Compare commits

..
Author SHA1 Message Date
Anthony Stirling e0dea581db Restrict tool upload picker to the formats each tool accepts
The "Files" step's Upload link on the left tool panel synthesises a
native file input — it now derives the `accept` attribute from the
active tool's `supportedFormats` instead of hard-coding PDF, so users
can no longer pick e.g. a .docx for /add-stamp. Convert and Automate
keep the broad CONVERT_SUPPORTED_FORMATS list, ScannerImageSplit gains
PDF + raw image formats to match the controller, everything else
defaults to PDF. Includes a stubbed Playwright spec covering 7 PDF-only
tools, Convert, and ScannerImageSplit.
2026-05-01 10:23:53 +01:00
710 changed files with 10602 additions and 50634 deletions
+1 -1
View File
@@ -22,7 +22,7 @@ runs:
steps:
- name: Generate a GitHub App Token
id: generate-token
uses: actions/create-github-app-token@bcd2ba49218906704ab6c1aa796996da409d3eb1 # v3.2.0
uses: actions/create-github-app-token@df432ceedc7162793a195dd1713ff69aefc7379e # v2.0.6
with:
app-id: ${{ inputs.app-id }}
private-key: ${{ inputs.private-key }}
+1 -1
View File
@@ -1,6 +1,6 @@
# Maintainer: Stirling PDF Inc <contact@stirlingpdf.com>
pkgname=stirling-pdf-desktop
pkgver=2.11.0
pkgver=2.10.0
pkgrel=1
pkgdesc="Locally hosted, web-based PDF manipulation tool (Tauri desktop app, official Stirling PDF Inc build)"
arch=('x86_64')
+1 -1
View File
@@ -1,6 +1,6 @@
# Maintainer: Stirling PDF Inc <contact@stirlingpdf.com>
pkgname=stirling-pdf-server-bin
pkgver=2.11.0
pkgver=2.10.0
pkgrel=1
pkgdesc="Locally hosted, web-based PDF manipulation tool (server JAR, prebuilt)"
arch=('any')
-93
View File
@@ -48,81 +48,6 @@ updates:
cooldown:
default-days: 7
rebase-strategy: "auto"
groups:
embedpdf:
patterns:
- "@embedpdf/*"
mantine:
patterns:
- "@mantine/*"
- "postcss-preset-mantine"
mui:
patterns:
- "@mui/*"
tauri-js:
patterns:
- "@tauri-apps/*"
emotion:
patterns:
- "@emotion/*"
react:
patterns:
- "react"
- "react-dom"
- "@types/react"
- "@types/react-dom"
typescript-eslint:
patterns:
- "@typescript-eslint/*"
- "typescript-eslint"
eslint:
patterns:
- "eslint"
- "@eslint/*"
vite:
patterns:
- "vite"
- "vite-*"
- "@vitejs/*"
vitest:
patterns:
- "vitest"
- "@vitest/*"
testing-library:
patterns:
- "@testing-library/*"
i18next:
patterns:
- "i18next"
- "i18next-*"
- "react-i18next"
iconify:
patterns:
- "@iconify/*"
- "@iconify-json/*"
stripe:
patterns:
- "@stripe/*"
posthog:
patterns:
- "@posthog/*"
- "posthog-js"
supabase:
patterns:
- "@supabase/*"
dnd-kit:
patterns:
- "@dnd-kit/*"
tailwind:
patterns:
- "tailwindcss"
- "@tailwindcss/*"
postcss:
patterns:
- "postcss"
- "postcss-*"
exclude-patterns:
- "postcss-preset-mantine"
- package-ecosystem: cargo
directories:
@@ -134,24 +59,6 @@ updates:
cooldown:
default-days: 7
rebase-strategy: "auto"
groups:
tauri:
patterns:
- "tauri"
- "tauri-build"
- "tauri-plugin-*"
serde:
patterns:
- "serde"
- "serde_*"
tracing:
patterns:
- "tracing"
- "tracing-*"
tokio:
patterns:
- "tokio"
- "tokio-*"
- package-ecosystem: pip
directory: /testing/cucumber
+7 -7
View File
@@ -35,13 +35,13 @@ jobs:
pr_ref: ${{ steps.resolve.outputs.ref }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Resolve PR info
id: resolve
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
script: |
const { owner, repo } = context.repo;
@@ -111,7 +111,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -132,7 +132,7 @@ jobs:
- name: Add deployment started comment
id: deployment-started
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -309,7 +309,7 @@ jobs:
- name: Post V2 deployment URL to PR
if: success()
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -359,7 +359,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -376,7 +376,7 @@ jobs:
private-key: ${{ secrets.GH_APP_PRIVATE_KEY }}
- name: Clean up V2 deployment comments
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -70,7 +70,7 @@ jobs:
enable_prototypes: ${{ steps.check-prototypes-flag.outputs.enable_prototypes }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -88,7 +88,7 @@ jobs:
- name: Get PR data
id: get-pr
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
script: |
const prNumber = context.eventName === 'workflow_dispatch'
@@ -154,7 +154,7 @@ jobs:
- name: Add 'in_progress' reaction to comment
if: github.event_name == 'issue_comment'
id: add-eyes-reaction
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -183,7 +183,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -212,7 +212,7 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -383,7 +383,7 @@ jobs:
- name: Add success reaction to comment
if: success() && github.event_name == 'issue_comment'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -418,7 +418,7 @@ jobs:
- name: Add failure reaction to comment
if: failure() && github.event_name == 'issue_comment'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -438,7 +438,7 @@ jobs:
- name: Post deployment URL to PR
if: success()
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -474,7 +474,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -489,7 +489,7 @@ jobs:
private-key: ${{ secrets.GH_APP_PRIVATE_KEY }}
- name: Apply label commands
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
+2 -2
View File
@@ -21,7 +21,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -39,7 +39,7 @@ jobs:
- name: Remove 'pr-deployed' label if present
id: remove-label-comment
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
+3 -3
View File
@@ -19,7 +19,7 @@ jobs:
pull-requests: write
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@8d3c67de8e2fe68ef647c8db1e6a09f647780f40 # v2.19.0
with:
egress-policy: audit
@@ -38,7 +38,7 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -71,7 +71,7 @@ jobs:
- name: Comment on fixer failures
if: steps.fixer_changes.outcome == 'failure' && github.event_name == 'pull_request'
continue-on-error: true
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@f28e40c7f34bde8b3046d885e986cb6290c5673b # v7.1.0
with:
script: |
const marker = '<!-- engine-check -->';
+2 -2
View File
@@ -19,7 +19,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -158,7 +158,7 @@ jobs:
- name: Post comment on PR if needed
if: steps.actor.outputs.is_repo_dev == 'true'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
continue-on-error: true
with:
github-token: ${{ steps.setup-bot.outputs.token }}
+3 -3
View File
@@ -26,7 +26,7 @@ jobs:
jar_sha256: ${{ steps.hashes.outputs.jar_sha256 }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@8d3c67de8e2fe68ef647c8db1e6a09f647780f40 # v2.19.0
with:
egress-policy: audit
@@ -70,7 +70,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@8d3c67de8e2fe68ef647c8db1e6a09f647780f40 # v2.19.0
with:
egress-policy: audit
@@ -106,7 +106,7 @@ jobs:
- name: Publish stirling-pdf-desktop to AUR
if: ${{ github.event_name == 'release' || inputs.dry_run == false }}
uses: KSXGitHub/github-actions-deploy-aur@v4.1.3
uses: KSXGitHub/github-actions-deploy-aur@v4.1.2
with:
pkgname: stirling-pdf-desktop
pkgbuild: .github/aur/stirling-pdf-desktop/PKGBUILD
+1 -1
View File
@@ -16,7 +16,7 @@ jobs:
pull-requests: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+8 -8
View File
@@ -1,7 +1,7 @@
name: Backend build, format check, and coverage
# Reusable workflow called from build.yml. Runs the backend build matrix
# (JDK 25 × spring-security on/off), Spotless formatting check, JUnit, and
# Reusable workflow called from build.yml. Runs the full backend build matrix
# (JDK 21/25 × spring-security on/off), Spotless formatting check, JUnit, and
# posts Jacoco coverage to PRs.
on:
workflow_call:
@@ -18,11 +18,11 @@ jobs:
strategy:
fail-fast: false
matrix:
jdk-version: [25]
jdk-version: [21, 25]
spring-security: [true, false]
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
@@ -35,7 +35,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependency artifacts
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/wrapper
@@ -44,7 +44,7 @@ jobs:
key: gradle-deps-${{ runner.os }}-jdk-${{ matrix.jdk-version }}-${{ hashFiles('**/gradle/wrapper/gradle-wrapper.properties', '**/*.gradle', '**/*.gradle.kts', 'settings.gradle', 'settings.gradle.kts', 'gradle/libs.versions.toml') }}
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
cache-disabled: true
@@ -66,7 +66,7 @@ jobs:
# issue/PR number, which doesn't exist on merge_group runs.
if: steps.spotless-check.outcome == 'failure' && github.event_name == 'pull_request'
continue-on-error: true
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
script: |
const marker = '<!-- java-formatting-check -->';
@@ -148,7 +148,7 @@ jobs:
- name: Upload Test Reports
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: test-reports-jdk-${{ matrix.jdk-version }}-spring-security-${{ matrix.spring-security }}
path: |
+7 -15
View File
@@ -44,7 +44,7 @@ jobs:
SYSTEM_ENABLEANALYTICS: "false"
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
@@ -55,7 +55,7 @@ jobs:
java-version: "25"
distribution: "temurin"
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -63,15 +63,7 @@ jobs:
- name: Install Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
- name: Install Playwright (chromium only)
run: task e2e:install -- chromium
- name: Build frontend (needed for playwright's vite preview webServer)
# Enterprise tests target :8080 (Spring Boot's bundled frontend), but
# playwright's webServer config still launches `vite preview --port 5173`
# before any test run, which needs dist/ to exist. VITE_BUILD_FOR_PREVIEW
# forces absolute asset paths so vite preview can serve deep SPA routes.
env:
VITE_BUILD_FOR_PREVIEW: "1"
run: task frontend:build
run: task frontend:test:e2e:install -- chromium
- name: Resolve kubernetes.docker.internal to localhost
# The compose stacks set KC_HOSTNAME=kubernetes.docker.internal so
@@ -148,7 +140,7 @@ jobs:
wait_for_backend
- name: Run enterprise OAuth Playwright tests
id: oauth-tests
run: task e2e:enterprise -- --grep "OAuth"
run: task frontend:test:e2e -- --project=enterprise --grep "OAuth"
- name: Stop backend + tear down OAuth Keycloak
if: always()
run: |
@@ -221,7 +213,7 @@ jobs:
wait_for_backend
- name: Run enterprise SAML Playwright tests
id: saml-tests
run: task e2e:enterprise -- --grep "SAML"
run: task frontend:test:e2e -- --project=enterprise --grep "SAML"
- name: Stop backend + tear down SAML Keycloak
if: always()
run: |
@@ -251,7 +243,7 @@ jobs:
wait_for_backend
- name: Run enterprise feature Playwright tests
id: feature-tests
run: task e2e:enterprise -- --grep "Enterprise license"
run: task frontend:test:e2e -- --project=enterprise --grep "Enterprise license"
- name: Print backend log on failure
if: failure()
run: |
@@ -265,7 +257,7 @@ jobs:
stop_backend
- name: Upload Playwright report
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: playwright-report-enterprise-${{ github.run_id }}
path: frontend/playwright-report/
+1 -1
View File
@@ -46,7 +46,7 @@ jobs:
proprietary: ${{ steps.changes.outputs.proprietary }}
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
+4 -4
View File
@@ -13,7 +13,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -27,7 +27,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependency artifacts
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/wrapper
@@ -36,7 +36,7 @@ jobs:
key: gradle-deps-${{ runner.os }}-jdk-25-${{ hashFiles('**/gradle/wrapper/gradle-wrapper.properties', '**/*.gradle', '**/*.gradle.kts', 'settings.gradle', 'settings.gradle.kts', 'gradle/libs.versions.toml') }}
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
cache-disabled: true
@@ -52,7 +52,7 @@ jobs:
- name: FAILED - check the licenses for compatibility
if: failure()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: dependencies-without-allowed-license.json
path: build/reports/dependency-license/dependencies-without-allowed-license.json
+4 -4
View File
@@ -14,7 +14,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -28,7 +28,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependency artifacts
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/wrapper
@@ -37,7 +37,7 @@ jobs:
key: gradle-deps-${{ runner.os }}-jdk-25-${{ hashFiles('**/gradle/wrapper/gradle-wrapper.properties', '**/*.gradle', '**/*.gradle.kts', 'settings.gradle', 'settings.gradle.kts', 'gradle/libs.versions.toml') }}
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
cache-disabled: true
@@ -53,7 +53,7 @@ jobs:
DISABLE_ADDITIONAL_FEATURES: true
- name: Upload OpenAPI Documentation
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: openapi-docs
path: ./SwaggerDoc.json
+4 -4
View File
@@ -27,7 +27,7 @@ jobs:
pull-requests: write # Allow writing to pull requests
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -43,7 +43,7 @@ jobs:
- name: Get PR data
id: get-pr-data
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -87,7 +87,7 @@ jobs:
- name: Determine reference file
id: determine-file
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -240,7 +240,7 @@ jobs:
- name: Post comment on PR
if: env.SCRIPT_OUTPUT != ''
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
+2 -2
View File
@@ -14,13 +14,13 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: "Checkout Repository"
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: "Dependency Review"
uses: actions/dependency-review-action@a1d282b36b6f3519aa1f3fc636f609c47dddb294 # v5.0.0
uses: actions/dependency-review-action@2031cfc080254a8a887f58cffee85186f0e49e48 # v4.9.0
with:
config-file: "./.github/config/dependency-review-config.yml"
+1 -1
View File
@@ -18,7 +18,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+5 -5
View File
@@ -25,7 +25,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -39,7 +39,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependency artifacts
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/wrapper
@@ -48,7 +48,7 @@ jobs:
key: gradle-deps-${{ runner.os }}-jdk-25-${{ hashFiles('**/gradle/wrapper/gradle-wrapper.properties', '**/*.gradle', '**/*.gradle.kts', 'settings.gradle', 'settings.gradle.kts', 'gradle/libs.versions.toml') }}
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
cache-disabled: true
@@ -90,7 +90,7 @@ jobs:
- name: Upload Cucumber Report
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: cucumber-report
path: testing/cucumber/report.html
@@ -99,7 +99,7 @@ jobs:
- name: Upload Test Reports
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: docker-compose-test-reports
path: testing/reports/
+43 -11
View File
@@ -15,7 +15,7 @@ jobs:
timeout-minutes: 30
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
@@ -26,7 +26,7 @@ jobs:
java-version: "25"
distribution: "temurin"
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -34,30 +34,62 @@ jobs:
- name: Install Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
- name: Install Playwright (chromium only)
run: task e2e:install -- chromium
- name: Build frontend (production bundle for vite preview)
run: task frontend:test:e2e:install -- chromium
- name: Start Spring Boot backend (background)
env:
VITE_BUILD_FOR_PREVIEW: "1"
run: task frontend:build
# Suppress the analytics opt-in modal that fires on first admin login when
# enableAnalytics is null (see Onboarding.tsx). The modal renders a Mantine
# overlay that intercepts pointer events on every tool page until dismissed,
# which causes every "click run button" assertion in the live suite to fail.
SYSTEM_ENABLEANALYTICS: "false"
# NOTE: SECURITY_INITIALLOGIN_USERNAME/PASSWORD are intentionally NOT set.
# The live-setup project's bootstrap spec performs the real first-login
# flow against the backend's default admin/stirling user, exercising the
# forced-password-change UI and leaving the DB at admin/adminadmin for
# the rest of the live suite. This is both real coverage of the first-
# login flow and a stronger seed than env-var-driven user creation.
run: |
nohup ./gradlew :stirling-pdf:bootRun > /tmp/backend.log 2>&1 &
echo $! > /tmp/backend.pid
- name: Wait for backend to become ready
run: |
start=$SECONDS
# 300 iterations × 2s = 10 minute ceiling
for i in $(seq 1 300); do
if curl -fsS http://localhost:8080/api/v1/info/status >/dev/null 2>&1; then
echo "Backend up after $((SECONDS - start))s"
exit 0
fi
sleep 2
done
echo "Backend did not become ready in $((SECONDS - start))s"
tail -200 /tmp/backend.log || true
exit 1
- name: Run live E2E tests (chromium)
id: live-tests
run: task e2e:live
run: task frontend:test:e2e -- --project=live
- name: Print backend log on failure
if: failure() && steps.live-tests.conclusion == 'failure'
run: |
echo "::group::Spring Boot backend log (last 500 lines)"
tail -500 .test-state/playwright/backend.log || echo "no backend log found"
tail -500 /tmp/backend.log || echo "no backend log found"
echo "::endgroup::"
- name: Stop backend
if: always()
run: |
if [ -f /tmp/backend.pid ]; then
kill "$(cat /tmp/backend.pid)" 2>/dev/null || true
fi
- name: Upload backend log
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: backend-log-live-${{ github.run_id }}
path: .test-state/playwright/backend.log
path: /tmp/backend.log
retention-days: 7
- name: Upload Playwright report
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: playwright-report-live-${{ github.run_id }}
path: frontend/playwright-report/
+5 -9
View File
@@ -14,13 +14,13 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -28,16 +28,12 @@ jobs:
- name: Install Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
- name: Install Playwright (chromium only)
run: task e2e:install -- chromium
- name: Build frontend (production bundle for vite preview)
env:
VITE_BUILD_FOR_PREVIEW: "1"
run: task frontend:build
run: task frontend:test:e2e:install -- chromium
- name: Run stubbed E2E tests (chromium)
run: task e2e:stubbed -- --workers=3
run: task frontend:test:e2e -- --project=stubbed --workers=3
- name: Upload Playwright report
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: playwright-report-stubbed-${{ github.run_id }}
path: frontend/playwright-report/
@@ -28,7 +28,7 @@ jobs:
licenses-backend: ${{ steps.changes.outputs.licenses-backend }}
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -52,7 +52,7 @@ jobs:
repository-projects: write # Required for enabling automerge
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -80,7 +80,7 @@ jobs:
persist-credentials: false
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -139,7 +139,7 @@ jobs:
# PR Event: Check licenses and comment on PR
- name: Delete previous license check comments
if: (github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == false) && github.actor != 'dependabot[bot]'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -192,7 +192,7 @@ jobs:
- name: Comment on PR - License Check Results
if: (github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == false) && github.actor != 'dependabot[bot]'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -316,7 +316,7 @@ jobs:
repository-projects: write # Required for enabling automerge
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -341,7 +341,7 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -369,7 +369,7 @@ jobs:
- name: Upload artifact on license issues
if: env.LICENSE_WARNINGS_EXIST == 'true'
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: backend-dependencies-without-allowed-license.json
path: build/reports/dependency-license/dependencies-without-allowed-license.json
@@ -382,7 +382,7 @@ jobs:
- name: Delete previous backend license check comments
if: (github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == false) && github.actor != 'dependabot[bot]'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
@@ -412,7 +412,7 @@ jobs:
- name: Comment on PR - Backend License Check Results
if: (github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == false) && github.actor != 'dependabot[bot]'
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
github-token: ${{ steps.setup-bot.outputs.token }}
script: |
+4 -24
View File
@@ -15,13 +15,13 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Checkout repository
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -37,7 +37,7 @@ jobs:
# issue/PR number, which doesn't exist on merge_group runs.
if: steps.frontend-check.outcome == 'failure' && github.event_name == 'pull_request'
continue-on-error: true
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
script: |
const marker = '<!-- frontend-check -->';
@@ -85,28 +85,8 @@ jobs:
echo "to see what still needs fixing manually."
echo "============================================"
exit 1
- name: Remove frontend check comment on success
if: steps.frontend-check.outcome == 'success' && github.event_name == 'pull_request'
continue-on-error: true
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
with:
script: |
const marker = '<!-- frontend-check -->';
const { data: comments } = await github.rest.issues.listComments({
owner: context.repo.owner,
repo: context.repo.repo,
issue_number: context.issue.number,
});
const existing = comments.find(c => c.body.includes(marker));
if (existing) {
await github.rest.issues.deleteComment({
owner: context.repo.owner,
repo: context.repo.repo,
comment_id: existing.id,
});
}
- name: Upload frontend build artifacts
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: frontend-build
path: frontend/dist/
+1 -1
View File
@@ -15,7 +15,7 @@ jobs:
issues: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+23 -49
View File
@@ -44,7 +44,7 @@ jobs:
version: ${{ steps.versionNumber.outputs.versionNumber }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -57,7 +57,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependencies
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/caches
@@ -67,7 +67,7 @@ jobs:
gradle-${{ runner.os }}-
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -90,21 +90,21 @@ jobs:
if [ "${{ github.event_name }}" = "workflow_dispatch" ]; then
case "${{ github.event.inputs.platform }}" in
"windows")
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64","jpdfium_platforms":"windows-x64"}]}' >> $GITHUB_OUTPUT
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64"}]}' >> $GITHUB_OUTPUT
;;
"macos")
echo 'matrix={"include":[{"platform":"macos-15","args":"--target universal-apple-darwin","name":"macos-universal","jpdfium_platforms":"darwin-arm64,darwin-x64"}]}' >> $GITHUB_OUTPUT
echo 'matrix={"include":[{"platform":"macos-15","args":"--target aarch64-apple-darwin","name":"macos-aarch64"},{"platform":"macos-15-intel","args":"--target x86_64-apple-darwin","name":"macos-x86_64"}]}' >> $GITHUB_OUTPUT
;;
"linux")
echo 'matrix={"include":[{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64","jpdfium_platforms":"linux-x64"}]}' >> $GITHUB_OUTPUT
echo 'matrix={"include":[{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64"}]}' >> $GITHUB_OUTPUT
;;
*)
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64","jpdfium_platforms":"windows-x64"},{"platform":"macos-15","args":"--target universal-apple-darwin","name":"macos-universal","jpdfium_platforms":"darwin-arm64,darwin-x64"},{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64","jpdfium_platforms":"linux-x64"}]}' >> $GITHUB_OUTPUT
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64"},{"platform":"macos-15","args":"--target aarch64-apple-darwin","name":"macos-aarch64"},{"platform":"macos-15-intel","args":"--target x86_64-apple-darwin","name":"macos-x86_64"},{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64"}]}' >> $GITHUB_OUTPUT
;;
esac
else
# For push/release events, build all platforms
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64","jpdfium_platforms":"windows-x64"},{"platform":"macos-15","args":"--target universal-apple-darwin","name":"macos-universal","jpdfium_platforms":"darwin-arm64,darwin-x64"},{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64","jpdfium_platforms":"linux-x64"}]}' >> $GITHUB_OUTPUT
echo 'matrix={"include":[{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64"},{"platform":"macos-15","args":"--target aarch64-apple-darwin","name":"macos-aarch64"},{"platform":"macos-15-intel","args":"--target x86_64-apple-darwin","name":"macos-x86_64"},{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64"}]}' >> $GITHUB_OUTPUT
fi
build-jars:
@@ -127,7 +127,7 @@ jobs:
file_suffix: "-server"
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -140,13 +140,13 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
- name: Setup Node.js
if: matrix.variant.build_frontend == true
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: 22
cache: "npm"
@@ -173,7 +173,7 @@ jobs:
cp app/core/build/libs/stirling-pdf-${{ needs.determine-matrix.outputs.version }}.jar ./jar-dist/Stirling-PDF${{ matrix.variant.file_suffix }}.jar
- name: Upload JAR artifacts
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: jar${{ matrix.variant.file_suffix }}
path: ./jar-dist/*.jar
@@ -191,7 +191,7 @@ jobs:
RELEASE_GPG_PRIVATE_KEY: ${{ secrets.RELEASE_GPG_PRIVATE_KEY }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
allowed-endpoints: >
@@ -208,7 +208,7 @@ jobs:
sudo apt-get install -y libgtk-3-dev libwebkit2gtk-4.0-dev libwebkit2gtk-4.1-dev libappindicator3-dev librsvg2-dev patchelf libjavascriptcoregtk-4.0-dev libsoup2.4-dev libjavascriptcoregtk-4.1-dev libsoup-3.0-dev
- name: Setup Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: 22
cache: "npm"
@@ -218,23 +218,7 @@ jobs:
uses: dtolnay/rust-toolchain@4be9e76fd7c4901c61fb841f559994984270fce7 # stable
with:
toolchain: stable
targets: ${{ matrix.platform == 'macos-15' && 'aarch64-apple-darwin,x86_64-apple-darwin' || '' }}
# x86_64 JDK is set up first so the aarch64 step below can leave its
# JAVA_HOME as the active one. The macOS universal JRE build needs
# jmods from both arches; the x64 path is captured into the env
# before the second setup-java overwrites JAVA_HOME.
- name: Set up x86_64 JDK 25 (macOS universal JRE)
if: matrix.platform == 'macos-15'
uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
with:
java-version: "25"
distribution: "temurin"
architecture: "x64"
- name: Capture x86_64 JAVA_HOME
if: matrix.platform == 'macos-15'
run: echo "X64_JAVA_HOME=$JAVA_HOME" >> "$GITHUB_ENV"
targets: ${{ (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && 'aarch64-apple-darwin,x86_64-apple-darwin' || '' }}
- name: Set up JDK 25
uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
@@ -243,30 +227,20 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
- name: Install Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
# Build the universal JRE before desktop:prepare so the jlink:runtime
# task short-circuits on its `test -d runtime/jre` status check.
- name: Build universal macOS JRE
if: matrix.platform == 'macos-15'
env:
AARCH64_JAVA_HOME: ${{ env.JAVA_HOME }}
JPDFIUM_PLATFORMS: ${{ matrix.jpdfium_platforms }}
run: task desktop:jlink:universal-mac
- name: Prepare desktop build
run: task desktop:prepare
env:
MAVEN_USER: ${{ secrets.MAVEN_USER }}
MAVEN_PASSWORD: ${{ secrets.MAVEN_PASSWORD }}
MAVEN_PUBLIC_URL: ${{ secrets.MAVEN_PUBLIC_URL }}
DISABLE_ADDITIONAL_FEATURES: true
JPDFIUM_PLATFORMS: ${{ matrix.jpdfium_platforms }}
run: task desktop:prepare
# DigiCert KeyLocker Setup (Cloud HSM)
- name: Setup DigiCert KeyLocker
@@ -347,7 +321,7 @@ jobs:
}
- name: Import Apple Developer Certificate
if: matrix.platform == 'macos-15' && (github.event_name == 'release' || (github.event_name == 'workflow_dispatch' && github.event.inputs.sign != 'false') || github.ref == 'refs/heads/V2-master')
if: (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && (github.event_name == 'release' || (github.event_name == 'workflow_dispatch' && github.event.inputs.sign != 'false') || github.ref == 'refs/heads/V2-master')
env:
APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
@@ -368,7 +342,7 @@ jobs:
rm certificate.p12
- name: Verify Certificate
if: matrix.platform == 'macos-15' && (github.event_name == 'release' || (github.event_name == 'workflow_dispatch' && github.event.inputs.sign != 'false') || github.ref == 'refs/heads/V2-master')
if: (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && (github.event_name == 'release' || (github.event_name == 'workflow_dispatch' && github.event.inputs.sign != 'false') || github.ref == 'refs/heads/V2-master')
run: |
echo "Verifying Apple Developer Certificate..."
KEYCHAIN_PATH=$RUNNER_TEMP/app-signing.keychain-db
@@ -583,7 +557,7 @@ jobs:
# Only ship the MSI installer on Windows. The loose exe and WiX toolset exes
# are not the user-facing installer - the MSI contains the signed inner exe.
find . -name "*.msi" -exec cp {} "../../../dist/Stirling-PDF-${{ matrix.name }}.msi" \;
elif [ "${{ matrix.platform }}" = "macos-15" ]; then
elif [ "${{ matrix.platform }}" = "macos-15" ] || [ "${{ matrix.platform }}" = "macos-15-intel" ]; then
find . -name "*.dmg" -exec cp {} "../../../dist/Stirling-PDF-${{ matrix.name }}.dmg" \;
find . -name "*.app" -exec cp -r {} "../../../dist/Stirling-PDF-${{ matrix.name }}.app" \;
else
@@ -594,7 +568,7 @@ jobs:
- name: Upload build artifacts
if: always() && steps.digicert-setup.conclusion != 'failure'
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: Stirling-PDF-${{ matrix.name }}
path: ./dist/*
@@ -608,7 +582,7 @@ jobs:
contents: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -640,7 +614,7 @@ jobs:
run: ls -R ./artifacts
- name: Upload binaries to Release
uses: softprops/action-gh-release@b4309332981a82ec1c5618f44dd2e27cc8bfbfda # v3.0.0
uses: softprops/action-gh-release@153bb8e04406b158c6c84fc1615b65b24149a1fe # v2.6.1
with:
tag_name: v${{ needs.determine-matrix.outputs.version }}
generate_release_notes: true
+5 -5
View File
@@ -18,7 +18,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -26,7 +26,7 @@ jobs:
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Set up Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: "22"
cache: "npm"
@@ -35,14 +35,14 @@ jobs:
- name: Install Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
- name: Install all Playwright browsers
run: task e2e:install
run: task frontend:test:e2e:install
- name: Run E2E tests (all browsers)
run: task e2e:cross-browser
run: task frontend:test:e2e
- name: Upload Playwright report
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: playwright-nightly-${{ github.run_id }}
path: frontend/playwright-report/
+24 -20
View File
@@ -22,13 +22,14 @@ jobs:
runs-on: ubuntu-latest
outputs:
version: ${{ steps.info.outputs.version }}
dmg_sha256: ${{ steps.hashes.outputs.dmg_sha256 }}
dmg_arm64_sha256: ${{ steps.hashes.outputs.dmg_arm64_sha256 }}
dmg_x86_64_sha256: ${{ steps.hashes.outputs.dmg_x86_64_sha256 }}
msi_sha256: ${{ steps.hashes.outputs.msi_sha256 }}
deb_sha256: ${{ steps.hashes.outputs.deb_sha256 }}
jar_sha256: ${{ steps.hashes.outputs.jar_sha256 }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@8d3c67de8e2fe68ef647c8db1e6a09f647780f40 # v2.19.0
with:
egress-policy: audit
@@ -62,15 +63,17 @@ jobs:
sha256sum "$file" | awk '{print $1}'
}
DMG_SHA=$(download_sha256 "${BASE}/Stirling-PDF-macos-universal.dmg")
DMG_ARM64_SHA=$(download_sha256 "${BASE}/Stirling-PDF-macos-aarch64.dmg")
DMG_X64_SHA=$(download_sha256 "${BASE}/Stirling-PDF-macos-x86_64.dmg")
MSI_SHA=$(download_sha256 "${BASE}/Stirling-PDF-windows-x86_64.msi")
DEB_SHA=$(download_sha256 "${BASE}/Stirling-PDF-linux-x86_64.deb")
JAR_SHA=$(download_sha256 "${BASE}/Stirling-PDF-with-login.jar")
echo "dmg_sha256=$DMG_SHA" >> "$GITHUB_OUTPUT"
echo "msi_sha256=$MSI_SHA" >> "$GITHUB_OUTPUT"
echo "deb_sha256=$DEB_SHA" >> "$GITHUB_OUTPUT"
echo "jar_sha256=$JAR_SHA" >> "$GITHUB_OUTPUT"
echo "dmg_arm64_sha256=$DMG_ARM64_SHA" >> "$GITHUB_OUTPUT"
echo "dmg_x86_64_sha256=$DMG_X64_SHA" >> "$GITHUB_OUTPUT"
echo "msi_sha256=$MSI_SHA" >> "$GITHUB_OUTPUT"
echo "deb_sha256=$DEB_SHA" >> "$GITHUB_OUTPUT"
echo "jar_sha256=$JAR_SHA" >> "$GITHUB_OUTPUT"
update-homebrew-and-scoop:
needs: get-release-info
@@ -79,7 +82,7 @@ jobs:
contents: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@8d3c67de8e2fe68ef647c8db1e6a09f647780f40 # v2.19.0
with:
egress-policy: audit
@@ -90,21 +93,22 @@ jobs:
token: ${{ secrets.HOMEBREW_TAP_TOKEN }}
path: tap
# Stirling-PDF now ships a single universal DMG, so the Cask should
# have one sha256 line rather than separate on_arm/on_intel blocks.
# We rewrite every `sha256 "..."` in the Cask to the same value so
# the workflow keeps working through the Cask migration: pre-migration
# both arch blocks get the universal SHA (still correct, since both
# would resolve to the same DMG), post-migration the lone sha256 line
# is updated.
- name: Update Homebrew cask (Casks/stirling-pdf.rb)
env:
VERSION: ${{ needs.get-release-info.outputs.version }}
DMG_SHA: ${{ needs.get-release-info.outputs.dmg_sha256 }}
ARM64_SHA: ${{ needs.get-release-info.outputs.dmg_arm64_sha256 }}
X64_SHA: ${{ needs.get-release-info.outputs.dmg_x86_64_sha256 }}
run: |
CASK="tap/Casks/stirling-pdf.rb"
sed -i "s/version \"[^\"]*\"/version \"${VERSION}\"/" "$CASK"
sed -i "s/sha256 \"[^\"]*\"/sha256 \"${DMG_SHA}\"/g" "$CASK"
sed -i "s/version \".*\"/version \"${VERSION}\"/" "$CASK"
awk -v arm="$ARM64_SHA" -v x64="$X64_SHA" '
/on_arm/ { in_arm=1 }
/on_intel/ { in_arm=0; in_intel=1 }
/end/ { in_arm=0; in_intel=0 }
in_arm && /sha256/ { sub(/sha256 ".*"/, "sha256 \"" arm "\"") }
in_intel && /sha256/ { sub(/sha256 ".*"/, "sha256 \"" x64 "\"") }
{ print }
' "$CASK" > tmp && mv tmp "$CASK"
- name: Update Homebrew formula (Formula/stirling-pdf-server.rb)
env:
@@ -112,8 +116,8 @@ jobs:
JAR_SHA: ${{ needs.get-release-info.outputs.jar_sha256 }}
run: |
FORMULA="tap/Formula/stirling-pdf-server.rb"
sed -i "s/version \"[^\"]*\"/version \"${VERSION}\"/" "$FORMULA"
sed -i "s/sha256 \"[^\"]*\"/sha256 \"${JAR_SHA}\"/" "$FORMULA"
sed -i "s/version \".*\"/version \"${VERSION}\"/" "$FORMULA"
sed -i "s/sha256 \".*\"/sha256 \"${JAR_SHA}\"/" "$FORMULA"
- name: Update Scoop stirling-pdf.json
env:
+1 -1
View File
@@ -19,7 +19,7 @@ jobs:
PIP_DISABLE_PIP_VERSION_CHECK: "1"
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+2 -2
View File
@@ -43,7 +43,7 @@ jobs:
echo "version=${VERSION}" >> $GITHUB_OUTPUT
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -100,7 +100,7 @@ jobs:
sbom: true
- name: Install cosign
uses: sigstore/cosign-installer@6f9f17788090df1f26f669e9d70d6ae9567deba6 # v4.1.2
uses: sigstore/cosign-installer@cad07c2e89fa2edd6e2d7bab4c1aa38e53f76003 # v4.1.1
with:
cosign-release: "v2.4.1"
+12 -168
View File
@@ -2,22 +2,6 @@ name: Push Docker Image with VersionNumber
on:
workflow_dispatch:
inputs:
build_main_app:
description: "Build & push the main Stirling-PDF image (latest, fat, ultra-lite)."
required: false
type: boolean
default: true
build_unoserver:
description: "Build & push the standalone stirling-unoserver image."
required: false
type: boolean
default: true
force_unoserver_rebuild:
description: "Rebuild stirling-unoserver even if its source hash is unchanged."
required: false
type: boolean
default: false
push:
branches:
- master
@@ -47,13 +31,9 @@ jobs:
permissions:
packages: write
id-token: write
# On push events these stay 'true'; on workflow_dispatch they follow the inputs.
env:
RUN_MAIN_APP: ${{ github.event_name != 'workflow_dispatch' || inputs.build_main_app }}
RUN_UNOSERVER: ${{ github.event_name != 'workflow_dispatch' || inputs.build_unoserver }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -66,7 +46,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependencies
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/caches
@@ -76,7 +56,7 @@ jobs:
gradle-${{ runner.os }}-
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -96,13 +76,13 @@ jobs:
- name: Install cosign
if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master'
uses: sigstore/cosign-installer@6f9f17788090df1f26f669e9d70d6ae9567deba6 # v4.1.2
uses: sigstore/cosign-installer@cad07c2e89fa2edd6e2d7bab4c1aa38e53f76003 # v4.1.1
with:
cosign-release: "v2.4.1"
- name: Install cosign
if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master'
uses: sigstore/cosign-installer@6f9f17788090df1f26f669e9d70d6ae9567deba6 # v4.1.2
uses: sigstore/cosign-installer@cad07c2e89fa2edd6e2d7bab4c1aa38e53f76003 # v4.1.1
with:
cosign-release: "v2.4.1"
@@ -128,7 +108,6 @@ jobs:
- name: Generate tags for latest
id: meta
if: env.RUN_MAIN_APP == 'true'
uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
with:
images: |
@@ -143,8 +122,6 @@ jobs:
- name: Build and push Unified Dockerfile (latest variant)
id: build-push-latest
# Empty-tag guard: build-push-action errors when asked to push with no tags.
if: env.RUN_MAIN_APP == 'true' && steps.meta.outputs.tags != ''
uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
with:
builder: ${{ steps.buildx.outputs.name }}
@@ -163,7 +140,7 @@ jobs:
sbom: true
- name: Sign regular images
if: env.RUN_MAIN_APP == 'true' && (github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master') && steps.build-push-latest.outputs.digest != ''
if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master'
env:
DIGEST: ${{ steps.build-push-latest.outputs.digest }}
TAGS: ${{ steps.meta.outputs.tags }}
@@ -179,7 +156,7 @@ jobs:
- name: Generate tags for latest-fat
id: meta-fat
uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
if: github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
with:
images: |
${{ secrets.DOCKER_HUB_USERNAME }}/s-pdf
@@ -193,7 +170,7 @@ jobs:
- name: Build and push Unified Dockerfile (fat variant)
id: build-push-fat
uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain' && steps.meta-fat.outputs.tags != ''
if: github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
with:
builder: ${{ steps.buildx.outputs.name }}
context: .
@@ -209,7 +186,7 @@ jobs:
sbom: true
- name: Sign fat images
if: env.RUN_MAIN_APP == 'true' && (github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master') && steps.build-push-fat.outputs.digest != ''
if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master'
env:
DIGEST: ${{ steps.build-push-fat.outputs.digest }}
TAGS: ${{ steps.meta-fat.outputs.tags }}
@@ -223,7 +200,7 @@ jobs:
- name: Generate tags for ultra-lite
id: meta-lite
uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
if: github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
with:
images: |
${{ secrets.DOCKER_HUB_USERNAME }}/s-pdf
@@ -237,7 +214,7 @@ jobs:
- name: Build and push Unified Dockerfile (ultra-lite variant)
id: build-push-lite
uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain' && steps.meta-lite.outputs.tags != ''
if: github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
with:
builder: ${{ steps.buildx.outputs.name }}
context: .
@@ -253,7 +230,7 @@ jobs:
sbom: true
- name: Sign ultra-lite images
if: env.RUN_MAIN_APP == 'true' && (github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master') && steps.build-push-lite.outputs.digest != ''
if: github.ref == 'refs/heads/master' || github.ref == 'refs/heads/V2-master'
env:
DIGEST: ${{ steps.build-push-lite.outputs.digest }}
TAGS: ${{ steps.meta-lite.outputs.tags }}
@@ -263,136 +240,3 @@ jobs:
echo "$TAGS" | tr ',' '\n' | while read -r tag; do
cosign sign --key env://COSIGN_PRIVATE_KEY --yes "${tag}@${DIGEST}"
done
# Standalone unoserver image — versioned independently via
# docker/unoserver/VERSION. master/V2-master: publish <version>+latest
# only when the version is new. main/testMain: republish :alpha only
# when the source hash differs from the published image's annotation.
- name: Read unoserver image version
id: unoserverVersion
if: env.RUN_UNOSERVER == 'true'
run: |
version=$(tr -d '[:space:]' < docker/unoserver/VERSION)
if [ -z "$version" ]; then
echo "docker/unoserver/VERSION is empty"; exit 1
fi
echo "version=${version}" >> "$GITHUB_OUTPUT"
echo "Unoserver image version (from file): ${version}"
- name: Compute unoserver image source hash
id: unoserverHash
if: env.RUN_UNOSERVER == 'true'
run: |
set -eu
hash=$(cat \
docker/unoserver/Dockerfile \
docker/unoserver/entrypoint.sh \
docker/unoserver/healthcheck.sh \
docker/unoserver/VERSION \
| sha256sum | cut -d' ' -f1)
echo "hash=${hash}" >> "$GITHUB_OUTPUT"
echo "Unoserver source hash: ${hash}"
- name: Decide whether to publish unoserver image
id: unoserverDecision
if: env.RUN_UNOSERVER == 'true'
env:
UNOSERVER_VERSION: ${{ steps.unoserverVersion.outputs.version }}
UNOSERVER_HASH: ${{ steps.unoserverHash.outputs.hash }}
UNOSERVER_IMAGE: ghcr.io/${{ steps.repoowner.outputs.lowercase }}/stirling-unoserver
UNOSERVER_HASH_ANNOTATION: org.stirlingpdf.unoserver-source-hash
FORCE_REBUILD: ${{ inputs.force_unoserver_rebuild }}
GH_REF: ${{ github.ref }}
EVENT_NAME: ${{ github.event_name }}
run: |
set -eu
mode="skip"
tags=""
read_published_hash() {
local ref="$1"
docker buildx imagetools inspect "$ref" --raw 2>/dev/null \
| jq -r --arg key "$UNOSERVER_HASH_ANNOTATION" \
'.annotations[$key] // empty' \
2>/dev/null || true
}
# Manual dispatch from any branch routes to the :alpha publish path.
EFFECTIVE_REF="$GH_REF"
if [ "$EVENT_NAME" = "workflow_dispatch" ]; then
EFFECTIVE_REF="refs/heads/testMain"
fi
case "$EFFECTIVE_REF" in
refs/heads/master|refs/heads/V2-master)
if [ "${FORCE_REBUILD}" = "true" ]; then
echo "force_unoserver_rebuild=true — building stable regardless"
mode="stable"
tags="${UNOSERVER_IMAGE}:${UNOSERVER_VERSION},${UNOSERVER_IMAGE}:latest"
elif docker manifest inspect "${UNOSERVER_IMAGE}:${UNOSERVER_VERSION}" >/dev/null 2>&1; then
echo "stirling-unoserver:${UNOSERVER_VERSION} already on GHCR — skipping"
else
echo "stirling-unoserver:${UNOSERVER_VERSION} is new — will publish"
mode="stable"
tags="${UNOSERVER_IMAGE}:${UNOSERVER_VERSION},${UNOSERVER_IMAGE}:latest"
fi
;;
refs/heads/main|refs/heads/testMain)
published_hash=$(read_published_hash "${UNOSERVER_IMAGE}:alpha")
if [ "${FORCE_REBUILD}" = "true" ]; then
echo "force_unoserver_rebuild=true — rebuilding :alpha regardless"
mode="alpha"
tags="${UNOSERVER_IMAGE}:alpha"
elif [ -n "$published_hash" ] && [ "$published_hash" = "$UNOSERVER_HASH" ]; then
echo "Published :alpha source hash matches (${published_hash}) — skipping"
else
if [ -z "$published_hash" ]; then
echo ":alpha has no source-hash annotation (first publish or pre-tracking image) — will publish"
else
echo "Source hash changed (was ${published_hash}, now ${UNOSERVER_HASH}) — will publish"
fi
mode="alpha"
tags="${UNOSERVER_IMAGE}:alpha"
fi
;;
*)
echo "Branch ${GH_REF} does not publish unoserver image"
;;
esac
echo "mode=${mode}" >> "$GITHUB_OUTPUT"
echo "tags=${tags}" >> "$GITHUB_OUTPUT"
- name: Build and push unoserver image
id: build-push-unoserver
if: env.RUN_UNOSERVER == 'true' && steps.unoserverDecision.outputs.mode != 'skip'
uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
with:
builder: ${{ steps.buildx.outputs.name }}
context: .
file: ./docker/unoserver/Dockerfile
push: true
cache-from: type=gha,scope=stirling-unoserver
cache-to: type=gha,mode=max,scope=stirling-unoserver
tags: ${{ steps.unoserverDecision.outputs.tags }}
# Manifest annotation read by the decision step above to detect drift.
annotations: |
index:org.stirlingpdf.unoserver-source-hash=${{ steps.unoserverHash.outputs.hash }}
platforms: linux/amd64,linux/arm64/v8
provenance: true
sbom: true
- name: Sign unoserver image
if: env.RUN_UNOSERVER == 'true' && steps.unoserverDecision.outputs.mode == 'stable'
env:
DIGEST: ${{ steps.build-push-unoserver.outputs.digest }}
TAGS: ${{ steps.unoserverDecision.outputs.tags }}
COSIGN_PRIVATE_KEY: ${{ secrets.COSIGN_PRIVATE_KEY }}
COSIGN_PASSWORD: ${{ secrets.COSIGN_PASSWORD }}
run: |
if [ -n "$COSIGN_PRIVATE_KEY" ]; then
echo "$TAGS" | tr ',' '\n' | while read -r tag; do
cosign sign --key env://COSIGN_PRIVATE_KEY --yes "${tag}@${DIGEST}"
done
else
echo "Warning: COSIGN_PRIVATE_KEY not set, skipping unoserver image signing"
fi
+1 -1
View File
@@ -18,7 +18,7 @@ jobs:
packages: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+3 -3
View File
@@ -35,7 +35,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -67,7 +67,7 @@ jobs:
# Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
# format to the repository Actions tab.
- name: "Upload artifact"
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: SARIF file
path: results.sarif
@@ -75,6 +75,6 @@ jobs:
# Upload the results to GitHub's code scanning dashboard.
- name: "Upload to code-scanning"
uses: github/codeql-action/upload-sarif@68bde559dea0fdcac2102bfdf6230c5f70eb485e # v3.29.5
uses: github/codeql-action/upload-sarif@95e58e9a2cdfd71adc6e0353d5c52f41a045d225 # v3.29.5
with:
sarif_file: results.sarif
+1 -1
View File
@@ -17,7 +17,7 @@ jobs:
pull-requests: write
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+2 -2
View File
@@ -27,7 +27,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -40,7 +40,7 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
+1 -1
View File
@@ -35,7 +35,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+27 -48
View File
@@ -1,6 +1,6 @@
name: Build Tauri Applications
# Multi-OS Tauri desktop bundle build matrix (Windows / macOS universal /
# Multi-OS Tauri desktop bundle build matrix (Windows / macOS arm+intel /
# Linux). Called from build.yml on PRs that touch desktop sources (gated
# via the `tauri` filter in .github/config/.files.yaml). Also runnable
# on demand via workflow_dispatch with a per-platform selector.
@@ -37,7 +37,7 @@ jobs:
matrix: ${{ steps.set-matrix.outputs.matrix }}
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -47,15 +47,18 @@ jobs:
APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
PLATFORM: ${{ inputs.platform }}
run: |
WINDOWS='{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64","jpdfium_platforms":"windows-x64"}'
MACOS='{"platform":"macos-15","args":"--target universal-apple-darwin","name":"macos-universal","jpdfium_platforms":"darwin-arm64,darwin-x64"}'
LINUX='{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64","jpdfium_platforms":"linux-x64"}'
WINDOWS='{"platform":"windows-latest","args":"--target x86_64-pc-windows-msvc","name":"windows-x86_64"}'
MACOS_ARM='{"platform":"macos-15","args":"--target aarch64-apple-darwin","name":"macos-aarch64"}'
MACOS_INTEL='{"platform":"macos-15-intel","args":"--target x86_64-apple-darwin","name":"macos-x86_64"}'
LINUX='{"platform":"ubuntu-22.04","args":"","name":"linux-x86_64"}'
# Resolve requested platform — populated by either workflow_dispatch
# or workflow_call inputs; both paths default to "all".
case "$PLATFORM" in
windows) ENTRIES=("$WINDOWS") ;;
macos) ENTRIES=("$MACOS") ;;
macos) ENTRIES=("$MACOS_ARM" "$MACOS_INTEL") ;;
linux) ENTRIES=("$LINUX") ;;
*) ENTRIES=("$WINDOWS" "$MACOS" "$LINUX") ;;
*) ENTRIES=("$WINDOWS" "$MACOS_ARM" "$MACOS_INTEL" "$LINUX") ;;
esac
# Drop macOS entries when Apple certificate secret is unavailable
@@ -84,7 +87,7 @@ jobs:
RELEASE_GPG_PRIVATE_KEY: ${{ secrets.RELEASE_GPG_PRIVATE_KEY }}
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -98,7 +101,7 @@ jobs:
sudo apt-get install -y libgtk-3-dev libwebkit2gtk-4.0-dev libwebkit2gtk-4.1-dev libappindicator3-dev librsvg2-dev patchelf libjavascriptcoregtk-4.0-dev libsoup2.4-dev libjavascriptcoregtk-4.1-dev libsoup-3.0-dev
- name: Setup Node.js
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
node-version: 22
cache: "npm"
@@ -108,19 +111,7 @@ jobs:
uses: dtolnay/rust-toolchain@4be9e76fd7c4901c61fb841f559994984270fce7 # stable
with:
toolchain: stable
targets: ${{ matrix.platform == 'macos-15' && 'aarch64-apple-darwin,x86_64-apple-darwin' || '' }}
- name: Set up x86_64 JDK 25 (macOS universal JRE)
if: matrix.platform == 'macos-15'
uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
with:
java-version: "25"
distribution: "temurin"
architecture: "x64"
- name: Capture x86_64 JAVA_HOME
if: matrix.platform == 'macos-15'
run: echo "X64_JAVA_HOME=$JAVA_HOME" >> "$GITHUB_ENV"
targets: ${{ (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && 'aarch64-apple-darwin,x86_64-apple-darwin' || '' }}
- name: Set up JDK 25
uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
@@ -129,28 +120,19 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
- name: Setup Task
uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
- name: Build universal macOS JRE
if: matrix.platform == 'macos-15'
env:
AARCH64_JAVA_HOME: ${{ env.JAVA_HOME }}
JPDFIUM_PLATFORMS: ${{ matrix.jpdfium_platforms }}
run: task desktop:jlink:universal-mac
- name: Prepare desktop build
run: task desktop:prepare
env:
MAVEN_USER: ${{ secrets.MAVEN_USER }}
MAVEN_PASSWORD: ${{ secrets.MAVEN_PASSWORD }}
MAVEN_PUBLIC_URL: ${{ secrets.MAVEN_PUBLIC_URL }}
DISABLE_ADDITIONAL_FEATURES: true
JPDFIUM_PLATFORMS: ${{ matrix.jpdfium_platforms }}
run: task desktop:prepare
# DigiCert KeyLocker Setup (Cloud HSM)
- name: Setup DigiCert KeyLocker
@@ -231,7 +213,7 @@ jobs:
}
- name: Import Apple Developer Certificate
if: matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
if: (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && env.APPLE_CERTIFICATE != ''
env:
APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
@@ -252,7 +234,7 @@ jobs:
rm certificate.p12
- name: Verify Certificate
if: matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
if: (matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel') && env.APPLE_CERTIFICATE != ''
run: |
echo "Verifying Apple Developer Certificate..."
KEYCHAIN_PATH=$RUNNER_TEMP/app-signing.keychain-db
@@ -263,12 +245,8 @@ jobs:
echo "APPLE_SIGNING_IDENTITY=$CERT_ID" >> $GITHUB_ENV
echo "Certificate imported successfully."
- name: Sign JPDFium dylibs inside bootJar (macOS only)
if: matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
run: bash frontend/scripts/sign-jpdfium-dylibs-in-bootjar.sh
- name: Check DMG creation dependencies (macOS only)
if: matrix.platform == 'macos-15'
if: matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel'
run: |
echo "🔍 Checking DMG creation dependencies on ${{ matrix.platform }}..."
echo "hdiutil version: $(hdiutil --version || echo 'NOT FOUND')"
@@ -395,7 +373,7 @@ jobs:
fi
- name: Verify notarization (macOS only)
if: matrix.platform == 'macos-15'
if: matrix.platform == 'macos-15' || matrix.platform == 'macos-15-intel'
run: |
echo "🔍 Verifying notarization status..."
cd ./frontend/src-tauri/target
@@ -420,7 +398,7 @@ jobs:
# Only ship the MSI installer. The loose exe and WiX toolset exes
# are not the user-facing installer - the MSI contains the signed inner exe.
find . -name "*.msi" -exec cp {} "../../../dist/Stirling-PDF-${{ matrix.name }}.msi" \;
elif [ "${{ matrix.platform }}" = "macos-15" ]; then
elif [ "${{ matrix.platform }}" = "macos-15" ] || [ "${{ matrix.platform }}" = "macos-15-intel" ]; then
find . -name "*.dmg" -exec cp {} "../../../dist/Stirling-PDF-${{ matrix.name }}.dmg" \;
else
find . -name "*.deb" -exec cp {} "../../../dist/Stirling-PDF-${{ matrix.name }}.deb" \;
@@ -494,7 +472,7 @@ jobs:
}
- name: Upload artifacts
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: Stirling-PDF-${{ matrix.name }}
path: ./dist/*
@@ -513,7 +491,7 @@ jobs:
echo "❌ No Windows executable found"
exit 1
fi
elif [ "${{ matrix.platform }}" = "macos-15" ]; then
elif [ "${{ matrix.platform }}" = "macos-15" ] || [ "${{ matrix.platform }}" = "macos-15-intel" ]; then
echo "Checking for macOS artifacts..."
find . -name "*.dmg" | head -5
if [ $(find . -name "*.dmg" | wc -l) -eq 0 ]; then
@@ -555,12 +533,12 @@ jobs:
pull-requests: write
steps:
- name: Harden the runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- name: Post/Update PR Comment with Download Links
uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
uses: actions/github-script@ed597411d8f924073f98dfc5c65a23a2325f34cd # v8.0.0
with:
script: |
const owner = context.repo.owner;
@@ -578,7 +556,8 @@ jobs:
// Map of expected artifact names to display info
const artifactMap = {
'Stirling-PDF-windows-x86_64': { icon: '🪟', platform: 'Windows x64', files: '.exe, .msi' },
'Stirling-PDF-macos-universal': { icon: '🍎', platform: 'macOS Universal', files: '.dmg' },
'Stirling-PDF-macos-aarch64': { icon: '🍎', platform: 'macOS ARM64', files: '.dmg' },
'Stirling-PDF-macos-x86_64': { icon: '🍎', platform: 'macOS Intel', files: '.dmg' },
'Stirling-PDF-linux-x86_64': { icon: '🐧', platform: 'Linux x64', files: '.deb, .rpm, .AppImage' }
};
@@ -646,7 +625,7 @@ jobs:
if: always()
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
+4 -59
View File
@@ -48,7 +48,7 @@ jobs:
cache-scope: stirling-pdf-fat
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -80,7 +80,7 @@ jobs:
distribution: "temurin"
- name: Cache Gradle dependency artifacts
uses: actions/cache@27d5ce7f107fe9357f9df03efb73ab90386fccae # v5.0.5
uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
with:
path: |
~/.gradle/wrapper
@@ -89,7 +89,7 @@ jobs:
key: gradle-deps-${{ runner.os }}-jdk-25-${{ hashFiles('**/gradle/wrapper/gradle-wrapper.properties', '**/*.gradle', '**/*.gradle.kts', 'settings.gradle', 'settings.gradle.kts', 'gradle/libs.versions.toml') }}
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
cache-disabled: true
@@ -151,7 +151,7 @@ jobs:
- name: Upload Reports
if: always()
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
with:
name: reports-docker-${{ matrix.artifact-suffix }}
path: |
@@ -160,58 +160,3 @@ jobs:
build/reports/problems/
retention-days: 3
if-no-files-found: warn
test-build-unoserver-image:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
with:
egress-policy: audit
- name: Checkout Repository
uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Set up QEMU
uses: docker/setup-qemu-action@ce360397dd3f832beb865e1373c09c0e9f86d70a # v4.0.0
- name: Set up Docker Buildx
id: buildx
uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4.0.0
- name: Build docker/unoserver/Dockerfile
uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
with:
builder: ${{ steps.buildx.outputs.name }}
context: .
file: ./docker/unoserver/Dockerfile
push: false
load: true
cache-from: type=gha,scope=stirling-unoserver
cache-to: type=gha,mode=max,scope=stirling-unoserver
platforms: linux/amd64
tags: stirling-unoserver:pr-test
provenance: false
sbom: false
- name: Smoke test the built image
run: |
set -eu
docker run -d --name unoserver-smoke \
-e UNOSERVER_RECYCLE_INTERVAL_SECONDS=0 \
stirling-unoserver:pr-test
deadline=$((SECONDS + 60))
while [ $SECONDS -lt $deadline ]; do
status=$(docker inspect -f '{{.State.Health.Status}}' unoserver-smoke 2>/dev/null || echo "starting")
if [ "$status" = "healthy" ]; then
echo "unoserver became healthy"
docker logs unoserver-smoke | tail -30
docker rm -f unoserver-smoke
exit 0
fi
sleep 3
done
echo "unoserver did not become healthy in time"
docker logs unoserver-smoke || true
docker rm -f unoserver-smoke || true
exit 1
+6 -6
View File
@@ -25,7 +25,7 @@ jobs:
runs-on: ubuntu-latest
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -39,7 +39,7 @@ jobs:
distribution: "temurin"
- name: Setup Gradle
uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
uses: gradle/actions/setup-gradle@f29f5a9d7b09a7c6b29859002d29d24e1674c884 # v5.0.1
with:
gradle-version: 9.3.1
@@ -131,7 +131,7 @@ jobs:
frontend: ${{ steps.changes.outputs.frontend }}
steps:
- name: Harden the runner (Audit all outbound calls)
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
@@ -150,14 +150,14 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
- uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
- name: Set up Node
uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
uses: actions/setup-node@53b83947a5a98c8d113130e565377fae1a50d02f # v6.3.0
with:
cache: "npm"
cache-dependency-path: frontend/package-lock.json
@@ -186,7 +186,7 @@ jobs:
steps:
- name: Harden Runner
uses: step-security/harden-runner@a5ad31d6a139d249332a2605b85202e8c0b78450 # v2.19.1
uses: step-security/harden-runner@58077d3c7e43986b6b15fba718e8ea69e387dfcc # v2.15.1
with:
egress-policy: audit
-7
View File
@@ -32,7 +32,6 @@ exampleYmlFiles/stirling/
/testing/cucumber/junit/
/testing/cucumber/report.html
/testing/.failed_tests
/.test-state/
SwaggerDoc.json
# Runtime storage for uploaded files and user data (not Java source code)
@@ -265,11 +264,5 @@ docs/type3/signatures/
# Type3 sample PDFs (development only)
**/type3/samples/
**/application-dev-local.properties
# Claude
.claude/
# Playwright MCP screenshots / traces
.playwright-mcp/
*.playwright-mcp.png
+12 -36
View File
@@ -1,20 +1,5 @@
version: '3'
# Gradle invocation strategy:
# - Linux/macOS: `./gradlew` runs the POSIX shell wrapper natively.
# - Windows: `cmd /c ".\gradlew.bat ..."` invokes the .bat wrapper through
# cmd.exe so it inherits the user's Windows-side `JAVA_HOME`
# and PATH. Routing through `bash gradlew` on Windows ends up
# under WSL or Git-Bash, neither of which inherits
# Adoptium/Temurin's default Windows-only Java env - gradlew
# then errors with "JAVA_HOME is not set and no 'java' command
# could be found".
#
# The entire `.\gradlew.bat ...` payload is double-quoted so
# the leading `.\` survives mvdan/sh's POSIX backslash
# stripping; cmd.exe also requires `.\` (not bare `gradlew.bat`)
# because modern Windows excludes cwd from cmd's search path.
tasks:
dev:
desc: "Start backend dev server"
@@ -25,24 +10,15 @@ tasks:
env:
SERVER_PORT: '{{.PORT}}'
cmds:
- cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED=true {{end}}cmd /c ".\gradlew.bat :stirling-pdf:bootRun"'
- cmd: '{{if .AIENGINE_URL}}set "AIENGINE_URL={{.AIENGINE_URL}}" && set "AIENGINE_ENABLED=true" && {{end}}cmd /c gradlew.bat :stirling-pdf:bootRun'
platforms: [windows]
- cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED=true {{end}}./gradlew :stirling-pdf:bootRun'
platforms: [linux, darwin]
dev:bundled:
desc: "Clean + bootRun with frontend bundled into the backend (single :8080 server)"
ignore_error: true
cmds:
- cmd: cmd /c ".\gradlew.bat clean bootRun -PbuildWithFrontend=true"
platforms: [windows]
- cmd: ./gradlew clean bootRun -PbuildWithFrontend=true
platforms: [linux, darwin]
build:
desc: "Full backend build"
cmds:
- cmd: cmd /c ".\gradlew.bat clean build"
- cmd: cmd /c gradlew.bat clean build
platforms: [windows]
- cmd: ./gradlew clean build
platforms: [linux, darwin]
@@ -50,7 +26,7 @@ tasks:
build:fast:
desc: "Build without tests"
cmds:
- cmd: cmd /c ".\gradlew.bat clean build -x test"
- cmd: cmd /c gradlew.bat clean build -x test
platforms: [windows]
- cmd: ./gradlew clean build -x test
platforms: [linux, darwin]
@@ -58,7 +34,7 @@ tasks:
build:ci:
desc: "Build for CI (formatting checked separately)"
cmds:
- cmd: cmd /c ".\gradlew.bat build -PnoSpotless"
- cmd: cmd /c gradlew.bat build -PnoSpotless
platforms: [windows]
- cmd: ./gradlew build -PnoSpotless
platforms: [linux, darwin]
@@ -66,7 +42,7 @@ tasks:
test:
desc: "Run backend tests"
cmds:
- cmd: cmd /c ".\gradlew.bat test"
- cmd: cmd /c gradlew.bat test
platforms: [windows]
- cmd: ./gradlew test
platforms: [linux, darwin]
@@ -74,7 +50,7 @@ tasks:
format:
desc: "Auto-fix code formatting"
cmds:
- cmd: cmd /c ".\gradlew.bat spotlessApply"
- cmd: cmd /c gradlew.bat spotlessApply
platforms: [windows]
- cmd: ./gradlew spotlessApply
platforms: [linux, darwin]
@@ -82,7 +58,7 @@ tasks:
format:check:
desc: "Check code formatting"
cmds:
- cmd: cmd /c ".\gradlew.bat spotlessCheck"
- cmd: cmd /c gradlew.bat spotlessCheck
platforms: [windows]
- cmd: ./gradlew spotlessCheck
platforms: [linux, darwin]
@@ -95,7 +71,7 @@ tasks:
swagger:
desc: "Generate OpenAPI docs"
cmds:
- cmd: cmd /c ".\gradlew.bat :stirling-pdf:copySwaggerDoc"
- cmd: cmd /c gradlew.bat :stirling-pdf:copySwaggerDoc
platforms: [windows]
- cmd: ./gradlew :stirling-pdf:copySwaggerDoc
platforms: [linux, darwin]
@@ -116,7 +92,7 @@ tasks:
desc: "Print project version"
silent: true
cmds:
- cmd: cmd /c ".\gradlew.bat printVersion --quiet" | tail -1
- cmd: cmd /c gradlew.bat printVersion --quiet | tail -1
platforms: [windows]
- cmd: ./gradlew printVersion --quiet | tail -1
platforms: [linux, darwin]
@@ -124,7 +100,7 @@ tasks:
licenses:check:
desc: "Check dependency licenses"
cmds:
- cmd: cmd /c ".\gradlew.bat checkLicense --no-parallel"
- cmd: cmd /c gradlew.bat checkLicense --no-parallel
platforms: [windows]
- cmd: ./gradlew checkLicense --no-parallel
platforms: [linux, darwin]
@@ -132,7 +108,7 @@ tasks:
licenses:generate:
desc: "Check and generate dependency license report"
cmds:
- cmd: cmd /c ".\gradlew.bat checkLicense generateLicenseReport --no-parallel"
- cmd: cmd /c gradlew.bat checkLicense generateLicenseReport --no-parallel
platforms: [windows]
- cmd: ./gradlew checkLicense generateLicenseReport --no-parallel
platforms: [linux, darwin]
@@ -140,7 +116,7 @@ tasks:
clean:
desc: "Clean build artifacts"
cmds:
- cmd: cmd /c ".\gradlew.bat clean"
- cmd: cmd /c gradlew.bat clean
platforms: [windows]
- cmd: ./gradlew clean
platforms: [linux, darwin]
+4 -36
View File
@@ -3,22 +3,6 @@ version: '3'
vars:
JLINK_MODULES: "java.base,java.compiler,java.desktop,java.instrument,java.logging,java.management,java.naming,java.net.http,java.prefs,java.rmi,java.scripting,java.security.jgss,java.security.sasl,java.sql,java.transaction.xa,java.xml,java.xml.crypto,jdk.crypto.ec,jdk.crypto.cryptoki,jdk.unsupported"
# Override via JPDFIUM_PLATFORMS env (csv of platform keys, or 'all').
JPDFIUM_PLATFORMS:
sh: |
if [ -n "${JPDFIUM_PLATFORMS:-}" ]; then
echo "$JPDFIUM_PLATFORMS"
else
case "{{OS}}-{{ARCH}}" in
darwin-arm64) echo "darwin-arm64";;
darwin-amd64) echo "darwin-x64";;
linux-amd64) echo "linux-x64";;
linux-arm64) echo "linux-arm64";;
windows-amd64) echo "windows-x64";;
*) echo "all";;
esac
fi
tasks:
prepare:
desc: "Prepare desktop build dependencies"
@@ -87,16 +71,15 @@ tasks:
deps: [jlink:jar, jlink:runtime]
jlink:jar:
desc: "Build backend JAR for Tauri bundling (host-OS natives only by default)"
desc: "Build backend JAR for Tauri bundling"
run: once
dir: ..
env:
DISABLE_ADDITIONAL_FEATURES: "true"
cmds:
- echo "Building bootJar with JPDFium natives for {{.JPDFIUM_PLATFORMS}}"
- cmd: cmd /c gradlew.bat bootJar --no-daemon -PjpdfiumPlatforms={{.JPDFIUM_PLATFORMS}}
- cmd: cmd /c gradlew.bat bootJar --no-daemon
platforms: [windows]
- cmd: ./gradlew bootJar --no-daemon -PjpdfiumPlatforms={{.JPDFIUM_PLATFORMS}}
- cmd: ./gradlew bootJar --no-daemon
platforms: [linux, darwin]
- mkdir -p frontend/src-tauri/libs
- cp app/core/build/libs/stirling-pdf-*.jar frontend/src-tauri/libs/
@@ -113,7 +96,7 @@ tasks:
jlink
--add-modules {{.JLINK_MODULES}}
--strip-debug
--compress=zip-6
--compress=2
--no-header-files
--no-man-pages
--output src-tauri/runtime/jre
@@ -124,18 +107,3 @@ tasks:
desc: "Remove JLink runtime and bundled JARs"
cmds:
- rm -rf src-tauri/libs src-tauri/runtime
# macOS-only. Replaces jlink:runtime's single-arch JRE with a universal
# (arm64 + x86_64) one for the universal Tauri shell. Runs the x86_64
# jlink under Rosetta on Apple Silicon, so it is opt-in and not part of
# the default desktop:build flow. Requires AARCH64_JAVA_HOME and
# X64_JAVA_HOME to point at matching JDK installations with jmods/.
jlink:universal-mac:
desc: "Create universal (arm64+x86_64) JRE for the macOS Tauri build"
deps: [jlink:jar]
platforms: [darwin]
env:
JLINK_MODULES: "{{.JLINK_MODULES}}"
OUTPUT_DIR: src-tauri/runtime/jre
cmds:
- scripts/build-universal-mac-jre.sh
-6
View File
@@ -55,9 +55,3 @@ tasks:
desc: "Tail docker compose logs"
cmds:
- docker compose -f {{.COMPOSE_DIR}}/docker-compose.yml logs -f
test:
desc: "Run full Docker integration test suite (builds all variants and tests them)"
ignore_error: true
cmds:
- bash testing/test.sh {{.CLI_ARGS}}
-204
View File
@@ -1,204 +0,0 @@
version: '3'
tasks:
install:
desc: "Install Playwright browsers"
dir: frontend
deps: [ ':frontend:install' ]
cmds:
- npx playwright install {{.CLI_ARGS}} --with-deps
stubbed:
desc: "Run stubbed E2E tests"
dir: frontend
deps: [ ':frontend:prepare' ]
cmds:
- npx playwright test --project=stubbed {{.CLI_ARGS}}
live:
desc: "Run live E2E tests"
summary: |
Auto-spawns a Spring Boot backend with isolated state under
.test-state/playwright/ (purged on every run) so the suite doesn't
touch your local DB, settings.yml, backups or customFiles. The
backend is stopped automatically when the runner exits.
Pass extra Playwright flags via -- :
task e2e:live -- --headed --grep merge
deps:
- live:backend
- live:runner
live:backend:
internal: true
ignore_error: true
vars:
BASE_DIR: '{{.ROOT_DIR}}/.test-state/playwright'
env:
STIRLING_BASE_PATH: '{{.BASE_DIR}}'
# Suppress the analytics opt-in modal that fires on first admin login.
# It renders a Mantine overlay that intercepts pointer events and blocks
# FirstLoginSlide from rendering, so the bootstrap spec times out waiting
# for the password-change prompt.
SYSTEM_ENABLEANALYTICS: "false"
# NOTE: SECURITY_INITIALLOGIN_USERNAME/PASSWORD are intentionally NOT set.
# The live-setup project's bootstrap spec performs the real first-login
# flow against the backend's default admin/stirling user, exercising the
# forced-password-change UI and leaving the DB at admin/adminadmin for
# the rest of the live suite. This is both real coverage of the first-
# login flow and a stronger seed than env-var-driven user creation.
cmds:
# Wrapped in `bash -c` because Task's embedded shell (mvdan/sh) doesn't
# support `&` + `$!` + `wait` reliably (gradle was never actually
# backgrounded, so `$!` recorded a stale PID and the runner saw an
# immediate "backend died"). Real bash backgrounds gradle properly and
# gives us a stable PID to clean up.
- |
bash -c '
set -e
rm -rf "{{.BASE_DIR}}"
mkdir -p "{{.BASE_DIR}}"
# Background gradle and record its PID so the runner can clean up
# the exact process tree (wrapper + forked Spring Boot JVM) without
# resorting to fuzzy `pkill -f` patterns. `wait` keeps this script
# alive for the lifetime of gradle so Task'"'"'s parallel deps stay
# synchronised.
bash gradlew :stirling-pdf:bootRun > "{{.BASE_DIR}}/backend.log" 2>&1 &
GRADLE_PID=$!
echo $GRADLE_PID > "{{.BASE_DIR}}/backend.pid"
wait $GRADLE_PID
'
live:runner:
internal: true
deps: [ ':frontend:prepare' ]
dir: frontend
vars:
BASE_DIR: '{{.ROOT_DIR}}/.test-state/playwright'
cmds:
# Wrapped in `bash -c` because Task's embedded shell (mvdan/sh) only
# supports `EXIT`/`ERR` in `trap`, not `INT`/`TERM`. CLI_ARGS are
# passed positionally so quoted args like --grep "foo bar" survive.
- |
bash -c '
set +e
# bootRun forks a separate Spring Boot JVM as a child of the gradle
# wrapper; killing only the wrapper leaves that JVM orphaned holding
# :8080. Walk the recorded PID'"'"'s process tree via pgrep so we kill
# every descendant, then lsof as a last-resort backstop.
kill_tree() {
local pid=$1
for child in $(pgrep -P "$pid" 2>/dev/null); do
kill_tree "$child"
done
kill "$pid" 2>/dev/null || true
}
cleanup() {
if [ -f "{{.BASE_DIR}}/backend.pid" ]; then
echo "Stopping backend..."
kill_tree "$(cat "{{.BASE_DIR}}/backend.pid")"
sleep 1
lsof -ti:8080 2>/dev/null | xargs kill 2>/dev/null || true
echo "Backend stopped"
fi
}
# Trap so cleanup runs on every exit path: tests pass, tests fail,
# backend never starts, user Ctrl-Cs. Without this the gradle JVM
# is left orphaned holding :8080 whenever the wait loop times out.
trap cleanup EXIT INT TERM
echo "Waiting for backend on :8080 (up to 10min; first compile can be slow)..."
WAITED=0
while [ $WAITED -lt 600 ]; do
if curl -sf http://localhost:8080/api/v1/info/status 2>/dev/null | grep -q UP; then
echo "Backend ready, starting Playwright"
break
fi
# Bail early if the backend already died: no point waiting 10min.
if [ -f "{{.BASE_DIR}}/backend.pid" ]; then
BACKEND_PID=$(cat "{{.BASE_DIR}}/backend.pid")
if ! kill -0 "$BACKEND_PID" 2>/dev/null; then
echo "Backend process exited before becoming ready; last 100 log lines:"
tail -100 "{{.BASE_DIR}}/backend.log" 2>/dev/null || true
exit 1
fi
fi
sleep 2
WAITED=$((WAITED + 2))
done
if [ $WAITED -ge 600 ]; then
echo "Backend did not become ready within 10min, aborting"
exit 1
fi
npx playwright test --project=live "$@"
' bash {{.CLI_ARGS}}
enterprise:
desc: "Run enterprise E2E tests"
summary: |
Requires an already-running keycloak compose stack on :8080. Bring one
up first with:
task e2e:oauth:up
task e2e:saml:up
dir: frontend
deps: [ ':frontend:prepare' ]
cmds:
- npx playwright test --project=enterprise {{.CLI_ARGS}}
cross-browser:
desc: "Run stubbed E2E tests on Chromium, Firefox, and WebKit"
dir: frontend
deps: [ ':frontend:prepare' ]
cmds:
- npx playwright test --project=stubbed --project=stubbed-firefox --project=stubbed-webkit {{.CLI_ARGS}}
check:
desc: "Run all E2E tests not requiring Docker"
cmds:
- task: stubbed
- task: live
check:all:
desc: "Run all E2E tests"
summary: |
Includes the enterprise project, which requires a keycloak compose
stack on :8080. Bring one up first with:
task e2e:oauth:up
task e2e:saml:up
cmds:
- task: stubbed
- task: live
- task: enterprise
oauth:up:
desc: "Start the OAuth keycloak test environment"
summary: |
Set LICENSE_KEY=<KEY> to skip the interactive license prompt:
task e2e:oauth:up LICENSE_KEY=abc123
Pass extra flags via -- :
task e2e:oauth:up -- --auto --nobuild
ignore_error: true
cmds:
- bash testing/compose/start-oauth-test.sh {{if .LICENSE_KEY}}--license-key "{{.LICENSE_KEY}}"{{end}} {{.CLI_ARGS}}
oauth:down:
desc: "Stop the OAuth keycloak test environment"
cmds:
- docker compose -f testing/compose/docker-compose-keycloak-oauth.yml down -v
saml:up:
desc: "Start the SAML keycloak test environment"
summary: |
Set LICENSE_KEY=<KEY> to skip the interactive license prompt:
task e2e:saml:up LICENSE_KEY=abc123
Pass extra flags via -- :
task e2e:saml:up -- --auto --with-storage --nobuild --language sv-SE
ignore_error: true
cmds:
- bash testing/compose/start-saml-test.sh {{if .LICENSE_KEY}}--license-key "{{.LICENSE_KEY}}"{{end}} {{.CLI_ARGS}}
saml:down:
desc: "Stop the SAML keycloak test environment"
cmds:
- docker compose -f testing/compose/docker-compose-keycloak-saml.yml down -v
+1 -2
View File
@@ -86,9 +86,8 @@ tasks:
fix:
desc: "Auto-fix lint + format"
cmds:
- task: format # Can auto-fix some things that `lint:fix` can't like line length violations
- task: lint:fix
- task: format # Ensure that after lint fixing that the code is still formatted correctly
- task: format
check:
desc: "Full engine quality gate"
+12 -1
View File
@@ -239,7 +239,6 @@ tasks:
- task: typecheck:saas
- task: typecheck:desktop
- task: typecheck:scripts
- task: typecheck:prototypes
# ============================================================
# Quality Gate
@@ -284,6 +283,18 @@ tasks:
cmds:
- npx vitest --coverage
test:e2e:
desc: "Run E2E tests"
deps: [prepare]
cmds:
- npx playwright test {{.CLI_ARGS}}
test:e2e:install:
desc: "Install E2E test browsers"
deps: [install]
cmds:
- npx playwright install {{.CLI_ARGS}} --with-deps
# ============================================================
# Code Generation
# ============================================================
+11 -11
View File
@@ -183,26 +183,26 @@ Use this pattern for desktop-specific or proprietary-specific features WITHOUT r
**Example - Desktop-specific footer:**
```typescript
// core/components/workbenchBar/WorkbenchBarFooterExtensions.tsx (stub)
interface WorkbenchBarFooterExtensionsProps {
// core/components/rightRail/RightRailFooterExtensions.tsx (stub)
interface RightRailFooterExtensionsProps {
className?: string;
}
export function WorkbenchBarFooterExtensions(_props: WorkbenchBarFooterExtensionsProps) {
export function RightRailFooterExtensions(_props: RightRailFooterExtensionsProps) {
return null; // Stub - does nothing in web builds
}
```
```tsx
// desktop/components/workbenchBar/WorkbenchBarFooterExtensions.tsx (real implementation)
// desktop/components/rightRail/RightRailFooterExtensions.tsx (real implementation)
import { Box } from '@mantine/core';
import { BackendHealthIndicator } from '@app/components/BackendHealthIndicator';
interface WorkbenchBarFooterExtensionsProps {
interface RightRailFooterExtensionsProps {
className?: string;
}
export function WorkbenchBarFooterExtensions({ className }: WorkbenchBarFooterExtensionsProps) {
export function RightRailFooterExtensions({ className }: RightRailFooterExtensionsProps) {
return (
<Box className={className}>
<BackendHealthIndicator />
@@ -212,15 +212,15 @@ export function WorkbenchBarFooterExtensions({ className }: WorkbenchBarFooterEx
```
```tsx
// core/components/shared/WorkbenchBar.tsx (usage - works in ALL builds)
import { WorkbenchBarFooterExtensions } from '@app/components/workbenchBar/WorkbenchBarFooterExtensions';
// core/components/shared/RightRail.tsx (usage - works in ALL builds)
import { RightRailFooterExtensions } from '@app/components/rightRail/RightRailFooterExtensions';
export function WorkbenchBar() {
export function RightRail() {
return (
<div>
{/* In web builds: renders nothing (stub returns null) */}
{/* In desktop builds: renders BackendHealthIndicator */}
<WorkbenchBarFooterExtensions className="workbench-bar-footer" />
<RightRailFooterExtensions className="right-rail-footer" />
</div>
);
}
@@ -431,7 +431,7 @@ The frontend is organized with a clear separation of concerns:
## Important Notes
- **Java Version**: Requires JDK 25.
- **Java Version**: Minimum JDK 21, supports and recommends JDK 25
- **Lombok**: Used extensively - ensure IDE plugin is installed
- **File Persistence**:
- **Backend**: Designed to be stateless - files are processed in memory/temp locations only
-70
View File
@@ -1,70 +0,0 @@
# Audit: feat/jpdfium-overlay
Verifies whether merging `feat/jpdfium-overlay` over `feat/jpdfium-integration` actually wins anything.
## What the branch actually does
Sole code commit: `fbadcda74 impl migration to pdfium for overlay and watermark and pagenumbers`. Net diff vs `feat/jpdfium-integration` is 5 files, +212 / -2.
For each of the three controllers the change is identical in shape:
```java
// JPDFium pre-validate catches corrupt PDFs cheaply before PDFBox does ...
try (PdfDocument ignored = PdfDocument.open(pdfFile.getBytes())) {
} catch (Exception e) {
log.debug("JPDFium pre-validate failed; proceeding with PDFBox: {}", e.getMessage());
}
try (PDDocument document = pdfDocumentFactory.load(pdfFile)) {
// ... all real overlay work done by PDFBox, exactly as before
}
```
There is no JPDFium overlay being performed. PDFBox still does 100% of:
- watermark tiled overlay (rows x cols, rotated, with non-Latin TTFs)
- 9-position page numbers with `{n}` / `{total}` / `{filename}` templating
- multi-line rotated stamps with rich `@var` templating
## Does the probe earn its keep?
No, on three counts:
1. **It does not short-circuit.** Catch swallows everything to DEBUG and falls through to PDFBox. A corrupt PDF still hits the full PDFBox parse path. The probe only adds work; it never saves any.
2. **It costs more than it saves on the hot path.** `PdfDocument.open(byte[])` does `MultipartFile.getBytes()` (full heap copy) + `JpdfiumLib.docOpenBytes` which `Arena.allocateFrom(JAVA_BYTE, data)` copies the entire payload into native memory, parses xref, then frees. For files > 10 MB `CustomPDFDocumentFactory.load(MultipartFile)` is normally stream-to-temp; the probe forces a full byte[] read in addition. Net: one extra heap-sized copy, one native copy, one PDFium parse, all discarded.
3. **No new corrupt-input test exercises the path.** `WatermarkControllerTest.OracleTests` and `PageNumbersControllerTest` only assert structure preservation on valid input. Probe is dead instrumentation as committed.
## JPDFium 1.0.x feature gaps that blocked real migration
Confirmed against `release/1.0.1` of JPDFium-pub.
| # | Gap | Where | Blocks |
|---|-----|-------|--------|
| 1 | `WatermarkApplier` is single-placement only; no row x col tiling. | `WatermarkApplier.applyTextWatermark` / `applyImageWatermark` | WatermarkController tiled overlay |
| 2 | High-level `WatermarkApplier` / `HeaderFooterApplier` only accept the built-in `FontName` enum. `PdfPageEditor.loadFont(byte[], type, cid)` exists but is not plumbed through. | `WatermarkApplier` line 72, `HeaderFooterApplier` line 103 | Non-Latin scripts (Arabic, JP, KR, ZH, Thai) |
| 3 | `HeaderFooterApplier.apply` hard-codes `size.width() / 2f` for both header and footer x. No 9-position grid. | `HeaderFooterApplier.apply` lines 49-56 | PageNumbersController 1..9 layout |
| 4 | `expandTemplate` only knows `{page}`, `{pages}`, `{date}`. | `HeaderFooterApplier.expandTemplate` line 93 | PageNumbers `{filename}`, Stamp `@filename`/`@author`/`@title`/`@uuid`/`@date{fmt}` etc. |
| 5 | `PdfAnnotationBuilder` takes a single `contents` string and a `Rect`; no multi-line layout, no per-line newline split, no rotated text-object matrices on the page content stream. | `PdfAnnotationBuilder.build` lines 130-137 | StampController `addTextStamp` multi-line + rotation |
| 6 | No image stamp/watermark with explicit width/height + rotation matrix; `WatermarkApplier.applyImageWatermark` forces `targetW = pageW * 0.3f` and never rotates. | `WatermarkApplier` lines 113-136 | Stamp/Watermark image variants |
Each maps to a one-line upstream feature request below.
## Upstream feature requests for JPDFium 1.0.2
1. `WatermarkApplier`: add `tileRows`/`tileCols` plus `widthSpacer`/`heightSpacer` builder fields and emit a row x col grid in `applyToPage`.
2. `Watermark`/`HeaderFooter` builders: accept `byte[] ttfFontData` (forward to `PdfPageEditor.loadFont` + use the returned `FPDF_FONT` in `createTextObject`).
3. `HeaderFooterApplier`: add `Position` enum (9-grid) for header and footer independently; replace the hard-coded `width/2f` x with `WatermarkApplier.computePosition`.
4. `HeaderFooter.expandTemplate`: add `{filename}`, `{author}`, `{title}`, `{subject}`, `{uuid}`, `{date:format}` and an `@var` alias set.
5. `PdfAnnotationBuilder` (or a new `TextStampApplier`): accept `List<String> lines`, `float rotationDegrees`, line-height, and emit one transformed text object per line via `PdfPageEditor.createTextObject` + `transform`.
6. `WatermarkApplier.applyImageWatermark`: take explicit `(width, height, rotation)` instead of fixed 30% width with no rotation.
## VERDICT
| Controller | Verdict | Reason |
|------------|---------|--------|
| WatermarkController | DROP | Probe is pure tax. Real migration blocked by gaps 1 + 2. |
| PageNumbersController | DROP | Probe is pure tax. Real migration blocked by gaps 3 + 4. |
| StampController | DROP | Probe is pure tax. Real migration blocked by gaps 4 + 5 + 6. |
## Bottom line
Drop the branch. The probe is a no-op that adds a full byte-copy + native parse to every overlay request and never short-circuits. Re-attempt overlay migration once JPDFium 1.0.2 lands feature requests 1-6, at which point a real migration (not a probe) is feasible.
+3 -3
View File
@@ -11,7 +11,7 @@ This guide focuses on developing for Stirling 2.0, including both the React fron
**Stirling 2.0** is built using:
**Backend:**
- Spring Boot (requires JDK 25)
- Spring Boot (Java 21+, JDK 25 recommended)
- PDFBox for core PDF operations
- LibreOffice for document conversions
- qpdf for PDF optimization
@@ -45,7 +45,7 @@ This guide focuses on developing for Stirling 2.0, including both the React fron
- [Task](https://taskfile.dev/installation/) — unified command runner (recommended)
- Docker
- Git
- Java JDK 25
- Java JDK 21 or later (JDK 25 recommended)
- Node.js 18+ and npm (required for frontend development)
- Gradle 7.0 or later (Included within the repo)
- [uv](https://docs.astral.sh/uv/) — Python package manager (required for engine development)
@@ -61,7 +61,7 @@ This guide focuses on developing for Stirling 2.0, including both the React fron
cd Stirling-PDF
```
2. Install Docker and JDK 25 if not already installed.
2. Install Docker and JDK 21 (or JDK 25 recommended) if not already installed.
3. Install a recommended Java IDE such as Eclipse, IntelliJ, or VSCode
1. Only VSCode
-2
View File
@@ -6,8 +6,6 @@ Portions of this software are licensed as follows:
* All content that resides under the "app/proprietary/" directory of this repository,
if that directory exists, is licensed under the license defined in "app/proprietary/LICENSE".
* All content that resides under the "app/saas/" directory of this repository,
if that directory exists, is licensed under the license defined in "app/saas/LICENSE".
* All content that resides under the "engine/" directory of this repository,
if that directory exists, is licensed under the license defined in "engine/LICENSE".
* All content that resides under the "frontend/src/proprietary/" directory of this repository,
+3 -6
View File
@@ -3,7 +3,7 @@ version: '3'
output: prefixed
vars:
FIND_FREE_PORT_PS: powershell -NoProfile -File scripts/find-free-port.ps1
FIND_FREE_PORT_PS: powershell -NoProfile -File scripts\find-free-port.ps1 -Preferred
FIND_FREE_PORT_SH: bash scripts/find-free-port.sh
includes:
@@ -22,9 +22,6 @@ includes:
desktop:
taskfile: .taskfiles/desktop.yml
dir: frontend
e2e:
taskfile: .taskfiles/e2e.yml
dir: .
tasks:
# ============================================================
@@ -45,7 +42,7 @@ tasks:
desc: "Start backend + frontend concurrently on free ports"
vars:
PORTS:
sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080 5173{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173{{end}}'
sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080,5173{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173{{end}}'
BACKEND_PORT: '{{index (splitList "\n" .PORTS) 0}}'
FRONTEND_PORT: '{{index (splitList "\n" .PORTS) 1}}'
deps:
@@ -62,7 +59,7 @@ tasks:
desc: "Start backend + frontend + engine concurrently on free ports"
vars:
PORTS:
sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080 5173 5001{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173 5001{{end}}'
sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080,5173,5001{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173 5001{{end}}'
BACKEND_PORT: '{{index (splitList "\n" .PORTS) 0}}'
FRONTEND_PORT: '{{index (splitList "\n" .PORTS) 1}}'
ENGINE_PORT: '{{index (splitList "\n" .PORTS) 2}}'
+1 -30
View File
@@ -45,39 +45,10 @@ dependencies {
api 'com.github.junrar:junrar:7.5.10' // RAR archive support for CBR files
api 'jakarta.servlet:jakarta.servlet-api:6.1.0'
api 'org.snakeyaml:snakeyaml-engine:3.0.1'
api "org.springdoc:springdoc-openapi-starter-webmvc-ui:3.0.3"
api "org.springdoc:springdoc-openapi-starter-webmvc-ui:3.0.2"
// Simple Java Mail for EML/MSG parsing (replaces direct Angus Mail usage)
api 'org.simplejavamail:simple-java-mail:8.12.6'
api 'org.simplejavamail:outlook-module:8.12.6' // MSG file support
api 'jakarta.mail:jakarta.mail-api:2.1.5'
runtimeOnly 'org.eclipse.angus:angus-mail:2.0.5'
// Tabula table extraction — used by the shared PDF parser and directly by downstream modules.
// api-scoped so downstream modules (core, proprietary) retain it on their compile classpath.
api ('technology.tabula:tabula:1.0.5') {
exclude group: 'org.slf4j', module: 'slf4j-simple'
exclude group: 'org.bouncycastle', module: 'bcprov-jdk15on'
exclude group: 'com.google.code.gson', module: 'gson'
}
api 'com.stirling:jpdfium:1.0.0'
// -PjpdfiumPlatforms=all|<csv of linux-x64,linux-arm64,darwin-x64,darwin-arm64,windows-x64>
def jpdfiumPlatformsProp = (project.findProperty('jpdfiumPlatforms') ?: 'all').toString().trim()
def jpdfiumAllPlatforms = ['linux-x64', 'linux-arm64', 'darwin-x64', 'darwin-arm64', 'windows-x64']
def jpdfiumPlatforms = jpdfiumPlatformsProp == 'all'
? jpdfiumAllPlatforms
: jpdfiumPlatformsProp.split(',').collect { it.trim() }.findAll { it }
def jpdfiumInvalid = jpdfiumPlatforms.findAll { !jpdfiumAllPlatforms.contains(it) }
if (jpdfiumInvalid) {
throw new GradleException("Unknown jpdfiumPlatforms value(s): ${jpdfiumInvalid.join(', ')}. " +
"Valid: ${jpdfiumAllPlatforms.join(', ')} or 'all'.")
}
logger.lifecycle("JPDFium native platforms: ${jpdfiumPlatforms.join(', ')}")
jpdfiumPlatforms.each { platform ->
runtimeOnly "com.stirling:jpdfium-natives-${platform}:1.0.0"
}
// ArchUnit: enforces module dependency direction (see ArchitectureTest)
testImplementation 'com.tngtech.archunit:archunit-junit5:1.4.2'
}
@@ -68,36 +68,6 @@ public class EndpointConfiguration {
return endpoint.startsWith("/") ? endpoint.substring(1) : endpoint;
}
/**
* Translate a full request URI like {@code /api/v1/general/remove-pages} into the endpoint key
* used by this configuration ({@code remove-pages}). Convert endpoints are a special case -
* {@code /api/v1/convert/pdf/img} is registered as {@code pdf-to-img}. Returns {@code null} if
* the URI is not an {@code /api/v1/<group>/<endpoint>} path.
*/
public static String endpointKeyForUri(String uri) {
if (uri == null || !uri.contains("/api/v1")) {
return null;
}
String[] parts = uri.split("/");
if (parts.length <= 4) {
return null;
}
if ("convert".equals(parts[3]) && parts.length > 5) {
return parts[4] + "-to-" + parts[5];
}
return parts[4];
}
/**
* Convenience wrapper around {@link #isEndpointEnabled(String)} that accepts a full request URI
* and translates it to the endpoint key. Falls back to treating the URI itself as a key for
* non-{@code /api/v1/...} paths so callers can pass arbitrary URIs.
*/
public boolean isEndpointEnabledForUri(String uri) {
String key = endpointKeyForUri(uri);
return isEndpointEnabled(key != null ? key : uri);
}
public void enableEndpoint(String endpoint) {
String normalized = normalizeEndpoint(endpoint);
endpointStatuses.put(normalized, true);
@@ -1,73 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.io.IOException;
import java.util.List;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.springframework.context.annotation.Primary;
import org.springframework.stereotype.Service;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
/**
* Chains table parsers in priority order: Tabula lattice → Tabula stream → {@link
* LineAlignmentTableParser}. The first parser returning a result above {@link
* #TABULA_CONFIDENCE_THRESHOLD} wins; results from different parsers are never mixed on one page.
*/
@Service
@Primary
@RequiredArgsConstructor
@Slf4j
public class CompositeTableParser implements TableParser {
/** Min Tabula confidence to accept results; below this LineAlignment is tried instead. */
static final float TABULA_CONFIDENCE_THRESHOLD = 0.5f;
private final TabulaTableParser tabulaParser;
private final LineAlignmentTableParser lineAlignmentParser;
@Override
public List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException {
// Step 1: Tabula lattice mode (ruled/bordered tables).
List<TableFragment> latticeResults = filterConfident(tabulaParser.parse(document, rawPage));
if (!latticeResults.isEmpty()) {
log.debug(
"Page {}: using Tabula lattice ({} table(s))",
rawPage.pageNumber(),
latticeResults.size());
return latticeResults;
}
// Step 2: Tabula stream mode (borderless/whitespace-delimited tables).
// parseStream is not on the TableParser interface — this intentionally couples to the
// concrete TabulaTableParser since stream mode is a Tabula-specific concept.
List<TableFragment> streamResults =
filterConfident(tabulaParser.parseStream(document, rawPage));
if (!streamResults.isEmpty()) {
log.debug(
"Page {}: using Tabula stream ({} table(s))",
rawPage.pageNumber(),
streamResults.size());
return streamResults;
}
// Step 3: Geometry-based line-alignment fallback.
List<TableFragment> lineResults = lineAlignmentParser.parse(document, rawPage);
if (!lineResults.isEmpty()) {
log.debug(
"Page {}: using LineAlignment ({} table(s))",
rawPage.pageNumber(),
lineResults.size());
return lineResults;
}
return List.of();
}
private List<TableFragment> filterConfident(List<TableFragment> tables) {
return tables.stream().filter(t -> t.confidence() >= TABULA_CONFIDENCE_THRESHOLD).toList();
}
}
@@ -1,528 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.Comparator;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.TreeMap;
import java.util.regex.Pattern;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.springframework.stereotype.Service;
import lombok.extern.slf4j.Slf4j;
/**
* Fallback {@link TableParser} for borderless financial tables using text geometry.
*
* <p>Identifies "anchor lines" (≥2 numeric tokens), builds a column grid from their right-edge
* positions, groups vertically proximate anchor lines into table candidates, then scores each group
* on column consistency and anchor density (confidence ceiling 0.85).
*/
@Service
@Slf4j
public class LineAlignmentTableParser implements TableParser {
/** Width in points of each column position bucket. */
static final float COLUMN_BUCKET_PT = 5f;
/** Tolerance in buckets when matching a token's right-edge to a confirmed column position. */
private static final int COLUMN_MATCH_BUCKETS = 2;
/** Maximum gap (as a multiple of modal line spacing) before splitting a group. */
private static final float MAX_GAP_FACTOR = 2.5f;
/** Minimum anchor rows (numeric-heavy) to form a valid table. */
static final int MIN_TABLE_ROWS = 3;
/** Minimum confirmed column positions to form a valid table. */
static final int MIN_COLUMNS = 2;
/**
* Min fraction of anchor lines a column must appear on to be confirmed (permissive for N/A
* rows).
*/
private static final double COLUMN_MIN_FREQUENCY = 0.40;
/**
* Matches financial numeric tokens: integers, decimals, parenthetical negatives, currency,
* percent, nil dashes.
*/
private static final Pattern NUMERIC =
Pattern.compile("^[\\(\\-\\$£€¥]?\\d[\\d,\\.]*[\\)%]?$|^[-–—]$");
/**
* Lines within this y-distance are merged into one row (restores rows split by LineBuilder's
* column-gap logic).
*/
static final float ROW_MERGE_TOLERANCE_PT = 2f;
// ── public API ───────────────────────────────────────────────────────────────────────────────
@Override
public List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException {
List<RawLine> lines = rawPage.lines();
if (lines.size() < MIN_TABLE_ROWS) return List.of();
float modalSpacing = computeModalSpacing(lines);
List<TokenizedLine> tokenized =
mergeCoincidentLines(lines.stream().map(this::tokenize).toList());
List<TokenizedLine> anchors = tokenized.stream().filter(TokenizedLine::isAnchor).toList();
if (anchors.size() < MIN_TABLE_ROWS) return List.of();
List<Float> columnGrid = buildColumnGrid(anchors);
if (columnGrid.size() < MIN_COLUMNS) {
log.debug(
"Page {}: LineAlignment — fewer than {} confirmed columns, skipping",
rawPage.pageNumber(),
MIN_COLUMNS);
return List.of();
}
List<List<TokenizedLine>> groups = groupRows(tokenized, columnGrid, modalSpacing);
List<TableFragment> results = new ArrayList<>();
for (int i = 0; i < groups.size(); i++) {
buildFragment(groups.get(i), columnGrid, rawPage.pageNumber(), i)
.ifPresent(results::add);
}
log.debug(
"Page {}: LineAlignment detected {} table(s) ({} anchor lines, {} columns)",
rawPage.pageNumber(),
results.size(),
anchors.size(),
columnGrid.size());
return results;
}
// ── coincident-line merging ──────────────────────────────────────────────────────────────────
/**
* Merges tokenised lines sharing the same y-position into one row, rejoining label/value halves
* split by LineBuilder.
*/
List<TokenizedLine> mergeCoincidentLines(List<TokenizedLine> tokenized) {
if (tokenized.size() < 2) return tokenized;
List<TokenizedLine> result = new ArrayList<>();
int i = 0;
while (i < tokenized.size()) {
float baseY = tokenized.get(i).line().bounds().y();
int j = i + 1;
while (j < tokenized.size()
&& Math.abs(tokenized.get(j).line().bounds().y() - baseY)
<= ROW_MERGE_TOLERANCE_PT) {
j++;
}
if (j == i + 1) {
result.add(tokenized.get(i));
} else {
result.add(mergeGroup(tokenized.subList(i, j)));
}
i = j;
}
return result;
}
private TokenizedLine mergeGroup(List<TokenizedLine> group) {
List<TextFragment> mergedFragments =
group.stream()
.flatMap(tl -> tl.line().fragments().stream())
.sorted(Comparator.comparingDouble(f -> f.bounds().x()))
.toList();
Bounds mergedBounds =
group.stream()
.map(tl -> tl.line().bounds())
.reduce(Bounds::merge)
.orElse(group.get(0).line().bounds());
RawLine mergedLine =
new RawLine(
group.get(0).line().lineId(),
mergedFragments,
mergedBounds,
group.get(0).line().pageNumber());
return tokenize(mergedLine);
}
// ── tokenisation ─────────────────────────────────────────────────────────────────────────────
/**
* Splits fragments into word-level tokens; x-positions are estimated linearly within each
* fragment.
*/
TokenizedLine tokenize(RawLine line) {
List<LineToken> tokens = new ArrayList<>();
for (TextFragment frag : line.fragments()) {
tokens.addAll(tokensFromFragment(frag));
}
List<LineToken> numeric = tokens.stream().filter(LineToken::numeric).toList();
return new TokenizedLine(line, tokens, numeric);
}
private List<LineToken> tokensFromFragment(TextFragment frag) {
String raw = frag.text();
if (raw == null || raw.isBlank()) return List.of();
float fragX = frag.bounds().x();
float fragWidth = frag.bounds().width();
int rawLen = raw.length();
List<LineToken> result = new ArrayList<>();
int offset = 0;
for (String part : raw.split("\\s+")) {
if (part.isEmpty()) {
offset++;
continue;
}
int idx = raw.indexOf(part, offset);
if (idx < 0) idx = offset;
float tokenX = rawLen > 0 ? fragX + ((float) idx / rawLen) * fragWidth : fragX;
float tokenRight =
rawLen > 0
? fragX + ((float) (idx + part.length()) / rawLen) * fragWidth
: fragX + fragWidth;
result.add(new LineToken(part, tokenX, tokenRight, NUMERIC.matcher(part).matches()));
offset = idx + part.length();
}
return result;
}
// ── column grid ──────────────────────────────────────────────────────────────────────────────
/**
* Returns confirmed column right-edge positions — those appearing on ≥ {@value
* #COLUMN_MIN_FREQUENCY} × N anchor lines.
*/
private List<Float> buildColumnGrid(List<TokenizedLine> anchors) {
// bucket → set of line indices that contributed a numeric token to that bucket
Map<Integer, List<Integer>> bucketLines = new HashMap<>();
for (int i = 0; i < anchors.size(); i++) {
for (LineToken t : anchors.get(i).numeric()) {
int bucket = bucket(t.right());
bucketLines.computeIfAbsent(bucket, k -> new ArrayList<>()).add(i);
}
}
int minHits =
Math.max(MIN_TABLE_ROWS, (int) Math.ceil(anchors.size() * COLUMN_MIN_FREQUENCY));
// Confirmed buckets → average right-edge for that bucket
TreeMap<Integer, Float> confirmed = new TreeMap<>();
for (Map.Entry<Integer, List<Integer>> entry : bucketLines.entrySet()) {
// Count distinct lines
long distinctLines = entry.getValue().stream().distinct().count();
if (distinctLines >= minHits) {
double avg =
entry.getValue().stream()
.distinct() // weight each line equally regardless of token count
.mapToDouble(
lineIdx ->
avgRightEdgeForBucket(
anchors, lineIdx, entry.getKey()))
.average()
.orElse(entry.getKey() * (double) COLUMN_BUCKET_PT);
confirmed.put(entry.getKey(), (float) avg);
}
}
return new ArrayList<>(confirmed.values()); // already sorted by bucket (left to right)
}
/**
* Returns the average right-edge position of tokens in {@code line} whose bucket matches {@code
* targetBucket}, falling back to the bucket's nominal centre when no tokens match.
*/
private double avgRightEdgeForBucket(
List<TokenizedLine> anchors, int lineIdx, int targetBucket) {
return anchors.get(lineIdx).numeric().stream()
.filter(t -> bucket(t.right()) == targetBucket)
.mapToDouble(LineToken::right)
.average()
.orElse(targetBucket * (double) COLUMN_BUCKET_PT);
}
// ── grouping ─────────────────────────────────────────────────────────────────────────────────
/**
* Groups anchor lines into table candidates, including adjacent label rows; a gap &gt;
* MAX_GAP_FACTOR × modal spacing splits groups.
*/
private List<List<TokenizedLine>> groupRows(
List<TokenizedLine> all, List<Float> columnGrid, float modalSpacing) {
float maxGap = modalSpacing > 0 ? modalSpacing * MAX_GAP_FACTOR : 30f;
List<List<TokenizedLine>> groups = new ArrayList<>();
List<TokenizedLine> current = new ArrayList<>();
for (int i = 0; i < all.size(); i++) {
TokenizedLine tl = all.get(i);
boolean fits = tl.isAnchor() && matchesGrid(tl, columnGrid);
if (current.isEmpty()) {
if (fits) current.add(tl);
continue;
}
float gap =
tl.line().bounds().y()
- current.get(current.size() - 1).line().bounds().bottom();
if (gap > maxGap) {
groups.add(current);
current = new ArrayList<>();
if (fits) current.add(tl);
continue;
}
if (fits) {
current.add(tl);
} else if (!tl.line().text().isBlank()) {
// Include non-anchor lines (labels) only if they have text and are within
// proximity.
current.add(tl);
}
}
if (!current.isEmpty()) groups.add(current);
return groups.stream().filter(g -> hasEnoughAnchorRows(g, columnGrid)).toList();
}
private boolean hasEnoughAnchorRows(List<TokenizedLine> group, List<Float> columnGrid) {
return group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count()
>= MIN_TABLE_ROWS;
}
/** A line "matches" the grid when ≥ 60 % of its numeric tokens land in confirmed columns. */
private boolean matchesGrid(TokenizedLine tl, List<Float> columnGrid) {
if (tl.numeric().isEmpty()) return false;
long matches =
tl.numeric().stream()
.filter(t -> nearestColumnIndex(t.right(), columnGrid) >= 0)
.count();
return (double) matches / tl.numeric().size() >= 0.60;
}
private boolean hasInconsistentColumnMatch(TokenizedLine tl, List<Float> columnGrid) {
if (tl.numeric().isEmpty()) return false;
long hits =
tl.numeric().stream()
.filter(t -> nearestColumnIndex(t.right(), columnGrid) >= 0)
.count();
return (double) hits / tl.numeric().size() < 0.60;
}
// ── fragment assembly ────────────────────────────────────────────────────────────────────────
private Optional<TableFragment> buildFragment(
List<TokenizedLine> group, List<Float> columnGrid, int pageNumber, int tableIndex) {
long anchorCount =
group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count();
if (anchorCount < MIN_TABLE_ROWS) return Optional.empty();
List<String> warnings = new ArrayList<>();
List<List<String>> rawRows = new ArrayList<>();
List<TableRow> rows = new ArrayList<>();
for (int rowIdx = 0; rowIdx < group.size(); rowIdx++) {
TokenizedLine tl = group.get(rowIdx);
List<String> rawRow = buildRawRow(tl, columnGrid);
rawRows.add(Collections.unmodifiableList(rawRow));
rows.add(buildTableRow(rowIdx, tl, rawRow, columnGrid));
}
// Column count = 1 label column + confirmed numeric columns
int colCount = columnGrid.size() + 1;
Bounds bounds = computeGroupBounds(group);
float confidence = computeConfidence(group, columnGrid, warnings);
return Optional.of(
new TableFragment(
"tbl-la-p" + pageNumber + "-" + tableIndex,
pageNumber,
bounds,
List.of(),
Collections.unmodifiableList(rows),
Collections.unmodifiableList(rawRows),
colCount,
confidence,
Collections.unmodifiableList(warnings),
null));
}
/**
* Builds a raw row as a list of strings: index 0 = label text, indices 1..N = column values.
*/
private List<String> buildRawRow(TokenizedLine tl, List<Float> columnGrid) {
String[] cells = new String[columnGrid.size() + 1];
Arrays.fill(cells, "");
// Separate label tokens (those not landing in any confirmed column) from column tokens.
List<String> labelParts = new ArrayList<>();
for (LineToken token : tl.all()) {
int col = nearestColumnIndex(token.right(), columnGrid);
if (col >= 0 && token.numeric()) {
int cellIdx = col + 1;
cells[cellIdx] =
cells[cellIdx].isEmpty()
? token.text()
: cells[cellIdx] + " " + token.text();
} else {
labelParts.add(token.text());
}
}
cells[0] = String.join(" ", labelParts).trim();
return Arrays.asList(cells);
}
private TableRow buildTableRow(
int rowIdx, TokenizedLine tl, List<String> rawRow, List<Float> columnGrid) {
List<TableCell> cells = new ArrayList<>(rawRow.size());
// Label cell: use the line's full bounds as an approximation.
cells.add(TableCell.of(0, rawRow.get(0), tl.line().bounds()));
for (int col = 0; col < columnGrid.size(); col++) {
String text = col + 1 < rawRow.size() ? rawRow.get(col + 1) : "";
float right = columnGrid.get(col);
float left = col > 0 ? columnGrid.get(col - 1) : right - 50f;
Bounds cellBounds =
new Bounds(
left,
tl.line().bounds().y(),
right - left,
tl.line().bounds().height());
cells.add(TableCell.of(col + 1, text, cellBounds));
}
return new TableRow(rowIdx, Collections.unmodifiableList(cells));
}
// ── confidence scoring ───────────────────────────────────────────────────────────────────────
/**
* Heuristic score in [0.0, 0.85] (ceiling keeps results below Tabula lattice which starts at
* 1.0). Base 0.70; +0.05/col beyond 2 (max +0.10); +0.05 at ≥5 anchors, +0.05 at ≥8; −0.15 if
* &gt;30 % of anchors have inconsistent columns; −0.10 if non-anchors outnumber anchors.
*/
private float computeConfidence(
List<TokenizedLine> group, List<Float> columnGrid, List<String> warnings) {
float score = 0.70f;
long anchorCount =
group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count();
long totalRows = group.size();
// More columns
int extraCols = Math.min(columnGrid.size() - MIN_COLUMNS, 2);
score += extraCols * 0.05f;
// More anchor rows
if (anchorCount >= 5) score += 0.05f;
if (anchorCount >= 8) score += 0.05f;
// Inconsistent column matching
long inconsistent =
group.stream()
.filter(TokenizedLine::isAnchor)
.filter(tl -> hasInconsistentColumnMatch(tl, columnGrid))
.count();
if (inconsistent > anchorCount * 0.30) {
score -= 0.15f;
warnings.add(
"Column match inconsistent on "
+ inconsistent
+ "/"
+ anchorCount
+ " anchor rows");
}
// Label-heavy
long nonAnchor = totalRows - anchorCount;
if (nonAnchor > anchorCount) {
score -= 0.10f;
warnings.add(
"Non-anchor rows ("
+ nonAnchor
+ ") outnumber anchor rows ("
+ anchorCount
+ ")");
}
return Math.max(0f, Math.min(0.85f, score));
}
// ── utility ──────────────────────────────────────────────────────────────────────────────────
/**
* Returns the grid index nearest to {@code rightEdge}, or -1 if none is within {@value
* #COLUMN_MATCH_BUCKETS} buckets.
*/
private int nearestColumnIndex(float rightEdge, List<Float> grid) {
int nearest = -1;
float minDist = COLUMN_MATCH_BUCKETS * COLUMN_BUCKET_PT + 1f;
for (int i = 0; i < grid.size(); i++) {
float dist = Math.abs(rightEdge - grid.get(i));
if (dist < minDist) {
minDist = dist;
nearest = i;
}
}
return nearest;
}
private Bounds computeGroupBounds(List<TokenizedLine> group) {
return group.stream()
.map(tl -> tl.line().bounds())
.reduce(Bounds::merge)
.orElse(new Bounds(0, 0, 0, 0));
}
/** Modal gap between consecutive line edges, used to calibrate the group-split threshold. */
private float computeModalSpacing(List<RawLine> lines) {
if (lines.size() < 2) return 0f;
Map<Float, Long> freq = new HashMap<>();
for (int i = 1; i < lines.size(); i++) {
float gap = lines.get(i).bounds().y() - lines.get(i - 1).bounds().bottom();
if (gap > 0) freq.merge(Math.round(gap / 2f) * 2f, 1L, Long::sum);
}
return freq.entrySet().stream()
.max(Map.Entry.comparingByValue())
.map(Map.Entry::getKey)
.orElse(0f);
}
private static int bucket(float x) {
return Math.round(x / COLUMN_BUCKET_PT);
}
// ── private data types ───────────────────────────────────────────────────────────────────────
/** A word-level token with an approximate right-edge x-position. */
record LineToken(String text, float x, float right, boolean numeric) {}
/** A {@link RawLine} with tokens pre-computed; an "anchor" has ≥ 2 numeric tokens. */
record TokenizedLine(RawLine line, List<LineToken> all, List<LineToken> numeric) {
boolean isAnchor() {
return numeric.size() >= 2;
}
}
}
@@ -1,139 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.util.ArrayList;
import java.util.Comparator;
import java.util.List;
import org.springframework.stereotype.Service;
import lombok.extern.slf4j.Slf4j;
/**
* Groups {@link TextFragment} objects into visual {@link RawLine}s using baseline proximity.
*
* <p>Fragments are on the same line when their baselines are within a font-size-derived tolerance.
* A new line starts whenever the horizontal gap exceeds an adaptive column-gap threshold ({@code
* max(effectiveWidth * COLUMN_GAP_RATIO, COLUMN_GAP_MIN_PT)}), splitting two-column text.
*/
@Service
@Slf4j
public class LineBuilder {
/** Baseline tolerance as a fraction of font size; 0.5 keeps mixed-size text on one line. */
private static final float BASELINE_TOLERANCE_FACTOR = 0.5f;
/** Absolute minimum tolerance so tiny font sizes don't collapse multi-line content. */
private static final float MIN_BASELINE_TOLERANCE = 2f;
/**
* Column-gap threshold as a fraction of page width; 0.10 clears tab stops but stays below
* two-column gutters.
*/
static final float COLUMN_GAP_RATIO = 0.10f;
/** Floor for the column-gap threshold so narrow pages don't over-split lines. */
static final float COLUMN_GAP_MIN_PT = 40f;
public List<RawLine> build(List<TextFragment> fragments, int pageNumber) {
if (fragments.isEmpty()) return List.of();
float effectiveWidth = inferEffectiveWidth(fragments);
float columnGapThreshold = Math.max(effectiveWidth * COLUMN_GAP_RATIO, COLUMN_GAP_MIN_PT);
log.debug(
"LineBuilder page {}: effectiveWidth={:.1f}pt, columnGapThreshold={:.1f}pt",
pageNumber,
effectiveWidth,
columnGapThreshold);
// Sort top-to-bottom first, then left-to-right within the same baseline band.
List<TextFragment> sorted =
fragments.stream()
.sorted(
Comparator.comparingDouble(TextFragment::baseline)
.thenComparingDouble(f -> f.bounds().x()))
.toList();
List<List<TextFragment>> groups = groupByBaseline(sorted, columnGapThreshold);
List<RawLine> lines = new ArrayList<>(groups.size());
for (int i = 0; i < groups.size(); i++) {
List<TextFragment> group =
groups.get(i).stream()
.sorted(Comparator.comparingDouble(f -> f.bounds().x()))
.toList();
Bounds lineBounds =
group.stream()
.map(TextFragment::bounds)
.reduce(Bounds::merge)
.orElse(new Bounds(0, 0, 0, 0));
lines.add(new RawLine("ln-p" + pageNumber + "-" + i, group, lineBounds, pageNumber));
}
return lines;
}
private List<List<TextFragment>> groupByBaseline(
List<TextFragment> sorted, float columnGapThreshold) {
List<List<TextFragment>> groups = new ArrayList<>();
List<TextFragment> current = new ArrayList<>();
float currentBaseline = Float.NaN;
for (TextFragment fragment : sorted) {
if (current.isEmpty()) {
current.add(fragment);
currentBaseline = fragment.baseline();
continue;
}
float maxFontSize =
Math.max(
fragment.fontSize(),
(float)
current.stream()
.mapToDouble(TextFragment::fontSize)
.max()
.orElse(0));
float tolerance =
Math.max(maxFontSize * BASELINE_TOLERANCE_FACTOR, MIN_BASELINE_TOLERANCE);
boolean sameBaseline = Math.abs(fragment.baseline() - currentBaseline) <= tolerance;
boolean columnGap = sameBaseline && hasColumnGap(fragment, current, columnGapThreshold);
if (sameBaseline && !columnGap) {
current.add(fragment);
// Anchor to the weighted mean baseline so long lines stay stable.
currentBaseline =
(currentBaseline * (current.size() - 1) + fragment.baseline())
/ current.size();
} else {
groups.add(current);
current = new ArrayList<>();
current.add(fragment);
currentBaseline = fragment.baseline();
}
}
if (!current.isEmpty()) groups.add(current);
return groups;
}
/**
* True when the gap from the rightmost fragment in {@code group} to {@code next} exceeds {@code
* threshold}.
*/
private static boolean hasColumnGap(
TextFragment next, List<TextFragment> group, float threshold) {
float lastRight = group.get(group.size() - 1).bounds().right();
return next.bounds().x() - lastRight > threshold;
}
/** Infers effective page width from the rightmost fragment right-edge plus a 10 % margin. */
private static float inferEffectiveWidth(List<TextFragment> fragments) {
double maxRight =
fragments.stream().mapToDouble(f -> f.bounds().right()).max().orElse(500.0);
return (float) maxRight * 1.10f;
}
}
@@ -1,79 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.springframework.stereotype.Service;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
/**
* Runs the per-page ingestion pipeline: {@link WordExtractingStripper} → {@link LineBuilder} →
* {@link TableParser}, producing a {@link PdfModels.ParsedPage} per page. The caller owns the
* {@link PDDocument} lifecycle.
*/
@Service
@RequiredArgsConstructor
@Slf4j
public class PdfIngester {
private final LineBuilder lineBuilder;
private final TableParser tableParser;
public List<ParsedPage> parse(PDDocument document) throws IOException {
return parse(document, document.getNumberOfPages());
}
public List<ParsedPage> parse(PDDocument document, int maxPages) throws IOException {
int pageCount = Math.min(document.getNumberOfPages(), maxPages);
List<ParsedPage> pages = new ArrayList<>(pageCount);
long fragmentsMs = 0;
long tablesMs = 0;
long t0 = System.currentTimeMillis();
for (int p = 1; p <= pageCount; p++) {
long ft = System.currentTimeMillis();
List<TextFragment> fragments = extractFragments(document, p);
fragmentsMs += System.currentTimeMillis() - ft;
PDPage page = document.getPage(p - 1);
PDRectangle mediaBox = page.getMediaBox();
List<RawLine> lines = lineBuilder.build(fragments, p);
RawPage rawPage = new RawPage(p, mediaBox.getWidth(), mediaBox.getHeight(), lines);
long tt = System.currentTimeMillis();
List<TableFragment> tables = tableParser.parse(document, rawPage);
tablesMs += System.currentTimeMillis() - tt;
log.debug(
"Page {}: {} fragments → {} lines, {} table(s)",
p,
fragments.size(),
lines.size(),
tables.size());
pages.add(new ParsedPage(p, mediaBox.getWidth(), mediaBox.getHeight(), tables, lines));
}
log.info(
"[timing] parse pages={} total={}ms fragments={}ms tables={}ms",
pageCount,
System.currentTimeMillis() - t0,
fragmentsMs,
tablesMs);
return pages;
}
private List<TextFragment> extractFragments(PDDocument document, int pageNumber)
throws IOException {
WordExtractingStripper stripper = new WordExtractingStripper(pageNumber);
stripper.getText(document);
return stripper.getFragments();
}
}
@@ -1,148 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import java.io.IOException;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
import org.apache.pdfbox.pdmodel.PDDocument;
/**
* All PDF parser model types and the table-parser contract in one place.
*
* <p>Import as {@code import static stirling.software.SPDF.pdf.parser.PdfModels.*;} to use all
* nested types without qualification.
*/
public final class PdfModels {
private PdfModels() {}
// ── Geometry ──────────────────────────────────────────────────────────────
public record Bounds(float x, float y, float width, float height) {
public float right() {
return x + width;
}
public float bottom() {
return y + height;
}
public static Bounds merge(Bounds a, Bounds b) {
float x = Math.min(a.x, b.x);
float y = Math.min(a.y, b.y);
return new Bounds(
x, y, Math.max(a.right(), b.right()) - x, Math.max(a.bottom(), b.bottom()) - y);
}
}
// ── Text fragments and lines ──────────────────────────────────────────────
/**
* A contiguous run of text from a single PDF content-stream string operation. {@code baseline}
* is preserved separately from bounds for line-grouping — characters of different sizes on the
* same visual line share a baseline but differ in top Y.
*/
public record TextFragment(
String fragmentId,
String text,
Bounds bounds,
float baseline,
float fontSize,
String fontName,
boolean bold) {}
public record RawLine(
String lineId, List<TextFragment> fragments, Bounds bounds, int pageNumber) {
public String text() {
if (fragments.isEmpty()) return "";
StringBuilder sb = new StringBuilder();
TextFragment prev = null;
for (TextFragment f : fragments) {
if (prev != null) {
float gap = f.bounds().x() - prev.bounds().right();
float avgCharWidth = prev.bounds().width() / Math.max(prev.text().length(), 1);
if (gap > avgCharWidth * 0.5f) sb.append(' ');
}
sb.append(f.text());
prev = f;
}
return sb.toString();
}
public float dominantFontSize() {
return fragments.stream()
.collect(Collectors.groupingBy(TextFragment::fontSize, Collectors.counting()))
.entrySet()
.stream()
.max(Map.Entry.comparingByValue())
.map(Map.Entry::getKey)
.orElse(0f);
}
public boolean hasBold() {
return fragments.stream().anyMatch(TextFragment::bold);
}
}
public record RawPage(int pageNumber, float widthPt, float heightPt, List<RawLine> lines) {}
// ── Table model ───────────────────────────────────────────────────────────
/**
* A single cell within a {@link TableRow}. {@code colSpan} and {@code rowSpan} are always 1 in
* v1 — span detection is deferred but the fields keep the schema stable.
*/
public record TableCell(int colIndex, String text, Bounds bounds, int colSpan, int rowSpan) {
public static TableCell of(int colIndex, String text, Bounds bounds) {
return new TableCell(colIndex, text, bounds, 1, 1);
}
}
public record TableRow(int rowIndex, List<TableCell> cells) {}
/**
* A table as it appears on a single page. {@code headers} is empty in v1 — all rows are in
* {@code rows}. {@code rawRows} preserves exact Tabula text output for diagnostics. {@code
* confidence} is a heuristic score in [0.0, 1.0]. {@code continuedFromPage} is null in v1.
*/
public record TableFragment(
String tableId,
int pageNumber,
Bounds bounds,
List<TableRow> headers,
List<TableRow> rows,
List<List<String>> rawRows,
int columnCount,
float confidence,
List<String> warnings,
Integer continuedFromPage) {}
// ── Page output ───────────────────────────────────────────────────────────
public record ParsedPage(
int pageNumber,
float widthPt,
float heightPt,
List<TableFragment> tables,
List<RawLine> layoutLines) {}
// ── Parser contract ───────────────────────────────────────────────────────
/**
* Extracts tables from a single page of a PDF. The caller owns the document lifecycle;
* implementations must not close it.
*/
public interface TableParser {
/**
* @param document open PDF; must not be closed by the implementation
* @param rawPage page metadata and lines for the page to process
* @return zero or more table fragments found on the page, never null
*/
List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException;
}
}
@@ -1,256 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Collections;
import java.util.List;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.springframework.stereotype.Service;
import lombok.extern.slf4j.Slf4j;
import technology.tabula.ObjectExtractor;
import technology.tabula.Page;
import technology.tabula.RectangularTextContainer;
import technology.tabula.Table;
import technology.tabula.extractors.BasicExtractionAlgorithm;
import technology.tabula.extractors.ExtractionAlgorithm;
import technology.tabula.extractors.SpreadsheetExtractionAlgorithm;
/**
* Primary {@link TableParser} implementation using Tabula's lattice-mode extraction.
*
* <h3>Algorithm</h3>
*
* Uses {@link SpreadsheetExtractionAlgorithm} (lattice mode), which detects tables from ruled lines
* (horizontal and vertical PDF path operators). This is reliable for tables with visible borders.
* Borderless or whitespace-delimited tables will not be detected — that requires stream mode, which
* is deferred to a later iteration.
*
* <h3>Coordinate system</h3>
*
* Tabula normalises page coordinates so that (0,0) is the top-left corner and Y increases downward,
* matching the {@link Bounds} convention used throughout this parser. No coordinate transformation
* is needed when mapping Tabula cell bounds to {@link Bounds}.
*
* <h3>Known limitations</h3>
*
* <ul>
* <li>Borderless tables are not detected (use stream mode, deferred).
* <li>Colspan and rowspan are not detected; all cells report colSpan=1, rowSpan=1.
* <li>Header rows are not identified; all rows appear in {@code rows}, headers is empty.
* <li>Cross-page table linking is not performed; each page is independent.
* <li>Rotated tables (90°/270° pages) may produce incorrect bounds.
* </ul>
*/
@Service
@Slf4j
public class TabulaTableParser implements TableParser {
/** Lattice mode — reliable for tables with visible ruled borders. */
@Override
public List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException {
return parseWithAlgorithm(
document, rawPage, new SpreadsheetExtractionAlgorithm(), "lattice");
}
/**
* Convenience overload for callers that only have a page number, not a full {@link RawPage}.
*/
public List<TableFragment> parse(PDDocument document, int pageNumber) throws IOException {
return parse(document, new RawPage(pageNumber, 0f, 0f, List.of()));
}
/** Stream mode — whitespace-based column detection for borderless tables. */
public List<TableFragment> parseStream(PDDocument document, RawPage rawPage)
throws IOException {
return parseWithAlgorithm(document, rawPage, new BasicExtractionAlgorithm(), "stream");
}
private List<TableFragment> parseWithAlgorithm(
PDDocument document, RawPage rawPage, ExtractionAlgorithm algorithm, String modeName)
throws IOException {
int pageNumber = rawPage.pageNumber();
List<Table> tabulaTables;
try {
// Do NOT use try-with-resources: ObjectExtractor.close() closes the underlying
// PDDocument, which we don't own. The extractor holds no resources of its own.
ObjectExtractor extractor = new ObjectExtractor(document);
Page page = extractor.extract(pageNumber);
tabulaTables = new ArrayList<>(algorithm.extract(page));
} catch (Exception e) {
log.warn(
"Tabula {} extraction failed on page {}: {}",
modeName,
pageNumber,
e.getMessage());
return List.of();
}
if (tabulaTables.isEmpty()) {
log.debug("Page {}: no tables detected by Tabula ({})", pageNumber, modeName);
return List.of();
}
log.debug(
"Page {}: Tabula ({}) detected {} table(s)",
pageNumber,
modeName,
tabulaTables.size());
List<TableFragment> fragments = new ArrayList<>(tabulaTables.size());
for (int i = 0; i < tabulaTables.size(); i++) {
fragments.add(toFragment(tabulaTables.get(i), pageNumber, i));
}
return fragments;
}
// ── private helpers ──────────────────────────────────────────────────────────────────────────
private TableFragment toFragment(Table table, int pageNumber, int tableIndex) {
List<List<RectangularTextContainer>> tabulaRows = table.getRows();
List<String> warnings = new ArrayList<>();
List<List<String>> rawRows = buildRawRows(tabulaRows);
int colCount = inferColumnCount(rawRows, warnings);
List<TableRow> rows = buildRows(tabulaRows, colCount, warnings);
float confidence = computeConfidence(rawRows, colCount, warnings);
Bounds bounds = tableBounds(table);
String tableId = "tbl-p" + pageNumber + "-" + tableIndex;
if (!warnings.isEmpty()) {
log.warn("Page {}, table {}: {}", pageNumber, tableIndex, warnings);
}
return new TableFragment(
tableId,
pageNumber,
bounds,
List.of(), // headers: deferred to v2
rows,
rawRows,
colCount,
confidence,
Collections.unmodifiableList(warnings),
null); // continuedFromPage: deferred to v2
}
private List<List<String>> buildRawRows(List<List<RectangularTextContainer>> tabulaRows) {
List<List<String>> rawRows = new ArrayList<>(tabulaRows.size());
for (List<RectangularTextContainer> tabulaRow : tabulaRows) {
List<String> cells = new ArrayList<>(tabulaRow.size());
for (RectangularTextContainer cell : tabulaRow) {
cells.add(normaliseText(cell.getText()));
}
rawRows.add(Collections.unmodifiableList(cells));
}
return rawRows;
}
private List<TableRow> buildRows(
List<List<RectangularTextContainer>> tabulaRows, int colCount, List<String> warnings) {
List<TableRow> rows = new ArrayList<>(tabulaRows.size());
for (int rowIdx = 0; rowIdx < tabulaRows.size(); rowIdx++) {
List<RectangularTextContainer> tabulaRow = tabulaRows.get(rowIdx);
List<TableCell> cells = new ArrayList<>(tabulaRow.size());
for (int colIdx = 0; colIdx < tabulaRow.size(); colIdx++) {
RectangularTextContainer c = tabulaRow.get(colIdx);
Bounds cellBounds =
new Bounds(
(float) c.getX(),
(float) c.getY(),
(float) c.getWidth(),
(float) c.getHeight());
cells.add(TableCell.of(colIdx, normaliseText(c.getText()), cellBounds));
}
if (tabulaRow.size() != colCount) {
warnings.add(
"Row "
+ rowIdx
+ " has "
+ tabulaRow.size()
+ " cells; expected "
+ colCount);
}
rows.add(new TableRow(rowIdx, Collections.unmodifiableList(cells)));
}
return rows;
}
/**
* The canonical column count for a table is the size of the widest row. Tabula can produce
* uneven rows when a cell's ruling lines are partially missing.
*/
private int inferColumnCount(List<List<String>> rawRows, List<String> warnings) {
if (rawRows.isEmpty()) return 0;
int max = rawRows.stream().mapToInt(List::size).max().orElse(0);
int mode =
rawRows.stream()
.collect(
java.util.stream.Collectors.groupingBy(
List::size, java.util.stream.Collectors.counting()))
.entrySet()
.stream()
.max(java.util.Map.Entry.comparingByValue())
.map(java.util.Map.Entry::getKey)
.orElse(0);
if (max != mode) {
warnings.add("Inconsistent column count: modal=" + mode + " max=" + max);
}
return mode > 0 ? mode : max;
}
/**
* Heuristic confidence score in [0.0, 1.0].
*
* <ul>
* <li>Starts at 1.0.
* <li>-0.3 if only one column (single-column tables are usually not real tables).
* <li>-0.1 per row with an inconsistent column count, capped at -0.4.
* <li>-0.3 if the empty-cell ratio across all cells exceeds 80%.
* </ul>
*/
private float computeConfidence(
List<List<String>> rawRows, int colCount, List<String> warnings) {
if (rawRows.isEmpty() || colCount == 0) return 0f;
float score = 1.0f;
if (colCount == 1) score -= 0.3f;
long inconsistentRows = rawRows.stream().filter(r -> r.size() != colCount).count();
score -= Math.min(inconsistentRows * 0.1f, 0.4f);
long totalCells = rawRows.stream().mapToLong(List::size).sum();
long emptyCells =
rawRows.stream().flatMap(Collection::stream).filter(String::isBlank).count();
if (totalCells > 0 && (float) emptyCells / totalCells > 0.8f) {
score -= 0.3f;
}
return Math.max(0f, Math.min(1f, score));
}
private Bounds tableBounds(Table table) {
return new Bounds(
(float) table.getX(),
(float) table.getY(),
(float) table.getWidth(),
(float) table.getHeight());
}
private String normaliseText(String raw) {
if (raw == null) return "";
// Tabula wraps multi-line cell content with \r\n — collapse to a single space.
return raw.replace("\r\n", " ").replace("\n", " ").replace("\r", " ").trim();
}
}
@@ -1,113 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.font.PDFont;
import org.apache.pdfbox.text.PDFTextStripper;
import org.apache.pdfbox.text.TextPosition;
/**
* Extends {@link PDFTextStripper} to capture per-fragment geometry and font metadata.
*
* <p>Overrides {@link #writeString} to split each content-stream string into word-level {@link
* TextFragment}s with bounding boxes, baseline, font name, and bold flag. Coordinates are in
* PDFTextStripper space: (0,0) top-left, Y increases downward, {@code getY()} is the baseline.
*/
class WordExtractingStripper extends PDFTextStripper {
private final int targetPage;
private final List<TextFragment> fragments = new ArrayList<>();
private int fragmentIndex = 0;
WordExtractingStripper(int pageNumber) throws IOException {
this.targetPage = pageNumber;
setStartPage(pageNumber);
setEndPage(pageNumber);
setSortByPosition(true);
}
@Override
protected void startPage(PDPage page) throws IOException {
super.startPage(page);
fragments.clear();
fragmentIndex = 0;
}
@Override
protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
if (text == null || text.isBlank()) return;
// Fast path: no whitespace → emit one fragment (most financial PDFs have each
// number as its own string operation, so this is the common case).
if (text.indexOf(' ') < 0) {
emitFragment(text, textPositions);
return;
}
// Per-word splitting requires 1:1 text-char to TextPosition correspondence.
// Fall back to one fragment when sizes differ (ligatures, encoding edge cases).
if (textPositions.size() != text.length()) {
emitFragment(text, textPositions);
return;
}
// Emit one TextFragment per whitespace-delimited word with accurate per-word bounds.
int start = 0;
for (int i = 0; i <= text.length(); i++) {
if (i == text.length() || text.charAt(i) == ' ') {
if (start < i) {
emitFragment(text.substring(start, i), textPositions.subList(start, i));
}
start = i + 1;
}
}
}
private void emitFragment(String text, List<TextPosition> positions) {
if (positions.isEmpty()) return;
float minX = Float.MAX_VALUE;
float minY = Float.MAX_VALUE;
float maxRight = -Float.MAX_VALUE;
float maxBaseline = -Float.MAX_VALUE;
TextPosition first = null;
for (TextPosition tp : positions) {
if (tp == null) continue;
if (first == null) first = tp;
float x = tp.getX();
// getY() is the baseline; top of character = getY() - getHeight().
float top = tp.getY() - tp.getHeight();
float right = x + tp.getWidth();
float baseline = tp.getY();
minX = Math.min(minX, x);
minY = Math.min(minY, top);
maxRight = Math.max(maxRight, right);
maxBaseline = Math.max(maxBaseline, baseline);
}
if (first == null) return;
PDFont font = first.getFont();
String fontName = font != null ? font.getName() : "";
boolean bold = fontName != null && fontName.toLowerCase().contains("bold");
// getHeight() gives the rendered glyph height, which is the most reliable visual size.
float fontSize = first.getHeight();
Bounds bounds = new Bounds(minX, minY, maxRight - minX, maxBaseline - minY);
String id = "tf-p" + targetPage + "-" + fragmentIndex++;
fragments.add(new TextFragment(id, text, bounds, maxBaseline, fontSize, fontName, bold));
}
List<TextFragment> getFragments() {
return Collections.unmodifiableList(fragments);
}
}
@@ -75,8 +75,8 @@ public @interface AutoJobPostMapping {
boolean queueable() default false;
/**
* Relative resource weight (1-100). See {@link
* stirling.software.common.enumeration.ResourceWeight} for the standard tiers.
* Relative resource weight (1–100) used by the scheduler to prioritise / throttle jobs. Values
* below 1 are clamped to&nbsp;1, values above 100 to&nbsp;100.
*/
int resourceWeight() default 1;
int resourceWeight() default 50;
}
@@ -59,18 +59,6 @@ public class InstallationPathConfig {
}
private static String initializeBasePath() {
// Allow tests / harnesses to redirect the entire state tree (configs,
// backups, customFiles, pipeline, logs) to an isolated location via
// -Dstirling.base-path=... or STIRLING_BASE_PATH=... so a Playwright
// run never touches a developer's working state.
String override = System.getProperty("stirling.base-path");
if (override == null || override.isBlank()) {
override = System.getenv("STIRLING_BASE_PATH");
}
if (override != null && !override.isBlank()) {
boolean hasTrailingSeparator = override.endsWith("/") || override.endsWith("\\");
return hasTrailingSeparator ? override : override + File.separator;
}
return "." + File.separator;
}
@@ -1,22 +0,0 @@
package stirling.software.common.enumeration;
/**
* Standard resource-weight tiers for {@link
* stirling.software.common.annotations.AutoJobPostMapping#resourceWeight()}.
*/
public final class ResourceWeight {
/** Lightweight: rotate, page numbers, extract pages, permissions, etc. */
public static final int SMALL_WEIGHT = 1;
/** Medium: merge, split, multi-tool batch. */
public static final int MEDIUM_WEIGHT = 3;
/** Heavy: compress, OCR (small), conversions, raster. */
public static final int LARGE_WEIGHT = 5;
/** Extra heavy: OCR on large files, full-document re-render, AI-assisted edits. */
public static final int XLARGE_WEIGHT = 10;
private ResourceWeight() {}
}
@@ -76,7 +76,6 @@ public class ApplicationProperties {
private ProcessExecutor processExecutor = new ProcessExecutor();
private PdfEditor pdfEditor = new PdfEditor();
private AiEngine aiEngine = new AiEngine();
private InternalApi internalApi = new InternalApi();
@Bean
public PropertySource<?> dynamicYamlPropertySource(ConfigurableEnvironment environment)
@@ -97,14 +96,7 @@ public class ApplicationProperties {
EncodedResource encodedResource = new EncodedResource(resource);
PropertySource<?> propertySource =
new YamlPropertySourceFactory().createPropertySource(null, encodedResource);
boolean saasActive = Arrays.asList(environment.getActiveProfiles()).contains("saas");
if (saasActive) {
// Saas-pinned values in application-saas.properties must beat settings.yml.
environment.getPropertySources().addLast(propertySource);
} else {
environment.getPropertySources().addFirst(propertySource);
}
environment.getPropertySources().addFirst(propertySource);
log.debug("Loaded properties: {}", propertySource.getSource());
@@ -245,26 +237,6 @@ public class ApplicationProperties {
private boolean enabled = false;
private String url = "http://localhost:5001";
private int timeoutSeconds = 120;
/**
* Longer timeout for heavy operations like RAG ingestion, which embeds the whole document
* and can take multiple minutes for large books. Applied per-call when the caller
* explicitly requests it via {@code AiEngineClient.postWithTimeout}.
*/
private int longRunningTimeoutSeconds = 600;
}
/**
* HTTP timeouts for loopback calls to internal Stirling API endpoints, used by the AI workflow
* executor and the pipeline processor. A bounded read timeout prevents a hung tool (e.g. an
* infinite loop in a PDF processing service) from stalling the entire chat workflow forever.
* Tools that legitimately need longer than the read timeout should be invoked through the async
* job executor instead of synchronously.
*/
@Data
public static class InternalApi {
private int connectTimeoutSeconds = 10;
private int readTimeoutSeconds = 300;
}
@Data
@@ -14,6 +14,7 @@ import lombok.NoArgsConstructor;
@Data
@NoArgsConstructor
@EqualsAndHashCode
@Schema(description = "PDF file input - either upload a file or provide a server-side file ID")
public class PDFFile {
@Schema(
@@ -1,15 +0,0 @@
package stirling.software.common.model.api.comments;
/**
* Absolute position of a PDF annotation in the document.
*
* <p>Coordinates are in PDF user-space with the origin at the page's bottom-left, consistent with
* PDFBox's {@code PDRectangle} convention.
*
* @param pageIndex 0-indexed page number the annotation lives on.
* @param x bottom-left x coordinate of the annotation rectangle.
* @param y bottom-left y coordinate of the annotation rectangle.
* @param width width of the annotation rectangle, in user-space units.
* @param height height of the annotation rectangle, in user-space units.
*/
public record AnnotationLocation(int pageIndex, float x, float y, float width, float height) {}
@@ -1,15 +0,0 @@
package stirling.software.common.model.api.comments;
/**
* Description of a single sticky-note (PDF Text) annotation to place on a document.
*
* <p>{@code author} and {@code subject} are optional — callers that pass {@code null} get a default
* author/subject from {@code PdfAnnotationService}.
*
* @param location where to anchor the annotation icon, in PDF user-space.
* @param text the comment body shown in the popup (required, non-blank).
* @param author optional author label shown in the popup; {@code null} → service default.
* @param subject optional subject line shown in the popup; {@code null} → service default.
*/
public record StickyNoteSpec(
AnnotationLocation location, String text, String author, String subject) {}
@@ -1,19 +0,0 @@
package stirling.software.common.model.api.general;
import io.swagger.v3.oas.annotations.media.Schema;
import lombok.Data;
import lombok.EqualsAndHashCode;
@Data
@EqualsAndHashCode
public class EditTextOperation {
@Schema(description = "The literal text to find.", requiredMode = Schema.RequiredMode.REQUIRED)
private String find;
@Schema(
description = "The replacement text. May be empty to delete the matched text.",
requiredMode = Schema.RequiredMode.REQUIRED)
private String replace;
}
@@ -1,26 +0,0 @@
package stirling.software.common.model.enumeration;
import lombok.Getter;
import lombok.RequiredArgsConstructor;
/** Team invitation status */
@Getter
@RequiredArgsConstructor
public enum InvitationStatus {
PENDING("PENDING"),
ACCEPTED("ACCEPTED"),
REJECTED("REJECTED"),
CANCELLED("CANCELLED"),
EXPIRED("EXPIRED");
private final String statusName;
public static InvitationStatus fromString(String statusName) {
for (InvitationStatus status : InvitationStatus.values()) {
if (status.getStatusName().equalsIgnoreCase(statusName)) {
return status;
}
}
throw new IllegalArgumentException("No InvitationStatus defined for name: " + statusName);
}
}
@@ -16,9 +16,6 @@ public enum Role {
// Unlimited access
USER("ROLE_USER", Integer.MAX_VALUE, Integer.MAX_VALUE, "adminUserSettings.user"),
// Paid tier; set by Stripe webhooks under saas mode.
PRO_USER("ROLE_PRO_USER", Integer.MAX_VALUE, Integer.MAX_VALUE, "adminUserSettings.proUser"),
// 40 API calls Per Day, 40 web calls
LIMITED_API_USER("ROLE_LIMITED_API_USER", 40, 40, "adminUserSettings.apiUser"),
@@ -1,26 +0,0 @@
package stirling.software.common.model.enumeration;
import lombok.Getter;
import lombok.RequiredArgsConstructor;
/**
* Team membership roles LEADER: Can invite/remove members, manage settings, view usage, manage
* billing MEMBER: Regular team member with standard access
*/
@Getter
@RequiredArgsConstructor
public enum TeamRole {
LEADER("LEADER"),
MEMBER("MEMBER");
private final String roleName;
public static TeamRole fromString(String roleName) {
for (TeamRole role : TeamRole.values()) {
if (role.getRoleName().equalsIgnoreCase(roleName)) {
return role;
}
}
throw new IllegalArgumentException("No TeamRole defined for name: " + roleName);
}
}
@@ -5,7 +5,6 @@ import java.io.UncheckedIOException;
import java.net.URLDecoder;
import java.nio.charset.StandardCharsets;
import java.nio.file.Files;
import java.time.Duration;
import java.util.regex.Pattern;
import org.springframework.beans.factory.annotation.Autowired;
@@ -13,18 +12,15 @@ import org.springframework.core.env.Environment;
import org.springframework.core.io.FileSystemResource;
import org.springframework.core.io.Resource;
import org.springframework.http.*;
import org.springframework.http.client.SimpleClientHttpRequestFactory;
import org.springframework.stereotype.Service;
import org.springframework.util.MultiValueMap;
import org.springframework.web.client.RequestCallback;
import org.springframework.web.client.ResourceAccessException;
import org.springframework.web.client.RestTemplate;
import jakarta.servlet.ServletContext;
import lombok.extern.slf4j.Slf4j;
import stirling.software.common.model.ApplicationProperties;
import stirling.software.common.model.enumeration.Role;
import stirling.software.common.util.TempFile;
import stirling.software.common.util.TempFileManager;
@@ -38,45 +34,26 @@ import stirling.software.common.util.TempFileManager;
@Slf4j
public class InternalApiClient {
// Allowlist for internal dispatch. Matches fixed namespace prefixes,
// Allowlist for internal dispatch. Matches a fixed namespace prefix,
// but rejects traversal (..), URL-encoding (%), query/fragment, backslashes, and any other
// character that could alter the resolved endpoint on the local Spring server.
//
// The second alternation carves out `/api/v1/ai/tools/*` specifically — AI tools are
// dispatchable, but the broader `/api/v1/ai/` surface (orchestrate, health, etc.) is
// intentionally NOT permitted to avoid plan steps re-entering the orchestrator.
private static final Pattern ALLOWED_ENDPOINT_PATH =
Pattern.compile(
"^/api/v1/(general|misc|security|convert|filter)(/[A-Za-z0-9_-]+)+$"
+ "|^/api/v1/ai/tools(/[A-Za-z0-9_-]+)+$");
Pattern.compile("^/api/v1/(general|misc|security|convert|filter)(/[A-Za-z0-9_-]+)+$");
private final ServletContext servletContext;
private final UserServiceInterface userService;
private final TempFileManager tempFileManager;
private final Environment environment;
private final Duration readTimeout;
private final RestTemplate restTemplate;
public InternalApiClient(
ServletContext servletContext,
@Autowired(required = false) UserServiceInterface userService,
TempFileManager tempFileManager,
Environment environment,
ApplicationProperties applicationProperties) {
Environment environment) {
this.servletContext = servletContext;
this.userService = userService;
this.tempFileManager = tempFileManager;
this.environment = environment;
ApplicationProperties.InternalApi internalApi = applicationProperties.getInternalApi();
// A bounded read timeout is what protects the workflow when an internal tool hangs
// (e.g. an infinite loop in a PDF processing service). The connect timeout is short
// because this is a loopback call; if connecting takes longer than a few seconds the
// local server is itself unhealthy.
this.readTimeout = Duration.ofSeconds(internalApi.getReadTimeoutSeconds());
SimpleClientHttpRequestFactory factory = new SimpleClientHttpRequestFactory();
factory.setConnectTimeout(Duration.ofSeconds(internalApi.getConnectTimeoutSeconds()));
factory.setReadTimeout(readTimeout);
this.restTemplate = new RestTemplate(factory);
}
/**
@@ -91,6 +68,7 @@ public class InternalApiClient {
validateUrl(endpointPath);
String url = getBaseUrl() + endpointPath;
RestTemplate restTemplate = new RestTemplate();
HttpHeaders headers = new HttpHeaders();
String apiKey = getApiKeyForUser();
if (apiKey != null && !apiKey.isEmpty()) {
@@ -100,38 +78,26 @@ public class InternalApiClient {
HttpEntity<MultiValueMap<String, Object>> entity = new HttpEntity<>(body, headers);
RequestCallback requestCallback = restTemplate.httpEntityCallback(entity, Resource.class);
try {
return restTemplate.execute(
url,
HttpMethod.POST,
requestCallback,
response -> {
try {
TempFile tempFile =
tempFileManager.createManagedTempFile("internal-api");
Files.copy(
response.getBody(),
tempFile.getPath(),
java.nio.file.StandardCopyOption.REPLACE_EXISTING);
String filename = extractFilename(response.getHeaders());
TempFileResource resource = new TempFileResource(tempFile, filename);
return ResponseEntity.status(response.getStatusCode())
.headers(response.getHeaders())
.body(resource);
} catch (IOException e) {
throw new UncheckedIOException(e);
}
});
} catch (ResourceAccessException e) {
// RestTemplate wraps low-level I/O failures in ResourceAccessException. Only the
// SocketTimeoutException-rooted case is a real timeout; other I/O failures (connection
// refused, DNS, etc.) propagate as-is so the upstream generic handler can describe
// them accurately.
if (e.getCause() instanceof java.net.SocketTimeoutException) {
throw new InternalApiTimeoutException(endpointPath, readTimeout, e);
}
throw e;
}
return restTemplate.execute(
url,
HttpMethod.POST,
requestCallback,
response -> {
try {
TempFile tempFile = tempFileManager.createManagedTempFile("internal-api");
Files.copy(
response.getBody(),
tempFile.getPath(),
java.nio.file.StandardCopyOption.REPLACE_EXISTING);
String filename = extractFilename(response.getHeaders());
TempFileResource resource = new TempFileResource(tempFile, filename);
return ResponseEntity.status(response.getStatusCode())
.headers(response.getHeaders())
.body(resource);
} catch (IOException e) {
throw new UncheckedIOException(e);
}
});
}
/**
@@ -1,36 +0,0 @@
package stirling.software.common.service;
import java.time.Duration;
/**
* Thrown when an internal Stirling tool invocation exceeds its configured read timeout or otherwise
* fails at the transport layer. Distinguishes a hung/timed-out tool from a tool that returned a
* non-2xx HTTP response (which is reported via the response status itself), so callers can present
* a clear "the tool didn't respond in time" message to the user instead of a generic stack trace.
*/
public class InternalApiTimeoutException extends RuntimeException {
private final String endpointPath;
private final Duration readTimeout;
public InternalApiTimeoutException(String endpointPath, Duration readTimeout, Throwable cause) {
super(buildMessage(endpointPath, readTimeout, cause), cause);
this.endpointPath = endpointPath;
this.readTimeout = readTimeout;
}
public String getEndpointPath() {
return endpointPath;
}
public Duration getReadTimeout() {
return readTimeout;
}
private static String buildMessage(String endpointPath, Duration readTimeout, Throwable cause) {
String reason = cause != null && cause.getMessage() != null ? cause.getMessage() : "";
return String.format(
"Internal tool %s did not respond within %ds (%s)",
endpointPath, readTimeout.toSeconds(), reason);
}
}
@@ -1,157 +0,0 @@
package stirling.software.common.service;
import java.util.Calendar;
import java.util.List;
import org.apache.pdfbox.cos.COSName;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.pdmodel.graphics.color.PDColor;
import org.apache.pdfbox.pdmodel.graphics.color.PDDeviceRGB;
import org.apache.pdfbox.pdmodel.interactive.annotation.PDAnnotationText;
import org.springframework.stereotype.Service;
import lombok.extern.slf4j.Slf4j;
import stirling.software.common.model.api.comments.AnnotationLocation;
import stirling.software.common.model.api.comments.StickyNoteSpec;
/**
* Shared primitive for adding sticky-note (PDF Text) annotations to a document.
*
* <p>Used by:
*
* <ul>
* <li>{@code /api/v1/misc/add-comments} — a deterministic, reusable tool.
* <li>AI-agent flows that generate comment specs (e.g. PDF review agent, math auditor review
* mode) and hand them off to this service for deterministic placement.
* </ul>
*/
@Slf4j
@Service
public class PdfAnnotationService {
/** Yellow sticky-note fill colour (R, G, B in 0..1 range). */
private static final float[] STICKY_NOTE_COLOR_RGB = {1f, 0.95f, 0.4f};
/** Opacity for the sticky-note icon. */
private static final float ANNOTATION_OPACITY = 0.9f;
/** PDF Text-annotation icon name — {@code "Comment"} is one of the standard icons. */
private static final String ANNOTATION_ICON_NAME = "Comment";
/** Default subject shown in the annotation popup when a spec does not supply one. */
private static final String DEFAULT_SUBJECT = "Stirling AI Comment";
/** Default author label shown in the annotation popup when a spec does not supply one. */
private static final String DEFAULT_AUTHOR = "Stirling AI";
/**
* Cap on sticky-note text length. PDF annotation bodies can technically be much longer, but
* anything beyond this is almost certainly pathological (accidental document-dump or malicious
* payload) and would bloat the output file.
*/
private static final int MAX_COMMENT_TEXT_LENGTH = 100_000;
/**
* Add a list of sticky notes to {@code doc}. Specs that reference an out-of-range page or
* contain blank text are logged and skipped; this method never throws for a single bad spec.
*
* @return the number of annotations actually applied
*/
public int addStickyNotes(PDDocument doc, List<StickyNoteSpec> specs) {
if (specs == null || specs.isEmpty()) {
return 0;
}
int totalPages = doc.getNumberOfPages();
Calendar now = Calendar.getInstance();
int applied = 0;
for (int i = 0; i < specs.size(); i++) {
StickyNoteSpec spec = specs.get(i);
if (!isValid(spec, totalPages, i)) {
continue;
}
apply(doc, spec, now);
applied++;
}
if (applied < specs.size()) {
log.warn(
"Applied {}/{} sticky notes; {} skipped due to invalid specs.",
applied,
specs.size(),
specs.size() - applied);
}
return applied;
}
/**
* Add a single sticky note. Convenience wrapper; prefer {@link #addStickyNotes(PDDocument,
* List)} when placing multiple annotations so log output is batched.
*/
public void addStickyNote(PDDocument doc, StickyNoteSpec spec) {
addStickyNotes(doc, List.of(spec));
}
private boolean isValid(StickyNoteSpec spec, int totalPages, int index) {
if (spec == null || spec.location() == null) {
log.warn("Skipping sticky-note[{}]: spec or location is null.", index);
return false;
}
if (spec.text() == null || spec.text().isBlank()) {
log.warn("Skipping sticky-note[{}]: text is blank.", index);
return false;
}
if (spec.text().length() > MAX_COMMENT_TEXT_LENGTH) {
log.warn(
"Skipping sticky-note[{}]: text length {} exceeds limit {}.",
index,
spec.text().length(),
MAX_COMMENT_TEXT_LENGTH);
return false;
}
AnnotationLocation loc = spec.location();
if (loc.width() <= 0f || loc.height() <= 0f) {
log.warn(
"Skipping sticky-note[{}]: non-positive dimensions width={} height={}.",
index,
loc.width(),
loc.height());
return false;
}
int page = loc.pageIndex();
if (page < 0 || page >= totalPages) {
log.warn(
"Skipping sticky-note[{}]: pageIndex={} out of range [0, {}).",
index,
page,
totalPages);
return false;
}
return true;
}
private void apply(PDDocument doc, StickyNoteSpec spec, Calendar now) {
AnnotationLocation loc = spec.location();
PDAnnotationText annot = new PDAnnotationText();
annot.setContents(spec.text());
annot.setRectangle(new PDRectangle(loc.x(), loc.y(), loc.width(), loc.height()));
annot.setSubject(nonBlankOr(spec.subject(), DEFAULT_SUBJECT));
annot.setTitlePopup(nonBlankOr(spec.author(), DEFAULT_AUTHOR));
annot.setColor(new PDColor(STICKY_NOTE_COLOR_RGB, PDDeviceRGB.INSTANCE));
annot.setCreationDate(now);
annot.setConstantOpacity(ANNOTATION_OPACITY);
annot.getCOSObject().setName(COSName.NAME, ANNOTATION_ICON_NAME);
try {
doc.getPage(loc.pageIndex()).getAnnotations().add(annot);
} catch (java.io.IOException e) {
log.warn(
"Failed to attach sticky note to page {}: {}", loc.pageIndex(), e.getMessage());
}
}
private static String nonBlankOr(String value, String fallback) {
return value != null && !value.isBlank() ? value : fallback;
}
}
@@ -2289,85 +2289,6 @@ public class FormUtils {
acroForm.getFields().add(field);
}
/** Drops AcroForm fields whose widgets are no longer on any page of {@code document}. */
public void pruneOrphanedFormFields(PDDocument document) {
if (document == null) {
return;
}
PDDocumentCatalog catalog = document.getDocumentCatalog();
if (catalog == null) {
return;
}
PDAcroForm form = catalog.getAcroForm(null);
if (form == null) {
return;
}
List<PDField> fields = form.getFields();
if (fields.isEmpty()) {
return;
}
Set<COSDictionary> liveWidgets = collectLiveWidgetDictionaries(document);
List<PDField> kept = pruneFieldList(fields, liveWidgets);
if (kept.isEmpty()) {
catalog.setAcroForm(null);
} else if (kept.size() != fields.size()) {
form.setFields(kept);
}
}
private Set<COSDictionary> collectLiveWidgetDictionaries(PDDocument document) {
Set<COSDictionary> live = new HashSet<>();
int pageCount = document.getNumberOfPages();
for (int i = 0; i < pageCount; i++) {
try {
for (PDAnnotation annotation : document.getPage(i).getAnnotations()) {
if (annotation instanceof PDAnnotationWidget) {
live.add(annotation.getCOSObject());
}
}
} catch (IOException e) {
log.debug("Failed reading page {} annotations: {}", i, e.getMessage());
}
}
return live;
}
private List<PDField> pruneFieldList(List<PDField> fields, Set<COSDictionary> liveWidgets) {
List<PDField> kept = new ArrayList<>(fields.size());
for (PDField field : fields) {
if (field instanceof PDNonTerminalField nonTerminal) {
List<PDField> children = nonTerminal.getChildren();
List<PDField> remaining = pruneFieldList(children, liveWidgets);
if (remaining.isEmpty()) {
continue;
}
if (remaining.size() != children.size()) {
nonTerminal.setChildren(remaining);
}
kept.add(nonTerminal);
} else if (field instanceof PDTerminalField terminal) {
List<PDAnnotationWidget> widgets = terminal.getWidgets();
List<PDAnnotationWidget> liveOnes = new ArrayList<>(widgets.size());
for (PDAnnotationWidget widget : widgets) {
if (liveWidgets.contains(widget.getCOSObject())) {
liveOnes.add(widget);
}
}
if (liveOnes.isEmpty()) {
continue;
}
if (liveOnes.size() != widgets.size()) {
terminal.setWidgets(liveOnes);
}
kept.add(terminal);
} else {
kept.add(field);
}
}
return kept;
}
// Delegation methods to GeneralFormCopyUtils for form field transformation
public boolean hasAnyRotatedPage(PDDocument document) {
return stirling.software.common.util.GeneralFormCopyUtils.hasAnyRotatedPage(document);
@@ -1,139 +0,0 @@
package stirling.software.common.util;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.Locale;
import java.util.Optional;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.text.PDFTextStripper;
import org.apache.pdfbox.text.TextPosition;
import org.springframework.stereotype.Component;
import lombok.extern.slf4j.Slf4j;
/**
* Locate text on a specific PDF page and return its bounding box in PDF user-space (bottom-left
* origin). Used by tools that receive "anchor by text" hints — e.g. {@code
* /api/v1/misc/add-comments} when callers supply an {@code anchorText} instead of explicit
* coordinates.
*
* <p>Matching is tolerant: case-insensitive with punctuation/whitespace stripped on both sides, so
* a caller-supplied needle of {@code "215000"} matches page text {@code "$215,000"}, and {@code
* "Total Revenue"} matches {@code "Total Revenue."}.
*/
@Slf4j
@Component
public class PdfTextLocator {
/** One found line of text with its user-space bounding box. */
public record MatchedBox(float x, float y, float width, float height) {}
/**
* Find the first line on {@code pageIndex} (0-indexed) whose text contains {@code needle} under
* the tolerant match. Returns empty when no match, when the page index is out of range, or when
* the needle is blank.
*/
public Optional<MatchedBox> findOnPage(PDDocument doc, int pageIndex, String needle) {
if (doc == null
|| needle == null
|| needle.isBlank()
|| pageIndex < 0
|| pageIndex >= doc.getNumberOfPages()) {
return Optional.empty();
}
String normalizedNeedle = normalize(needle);
if (normalizedNeedle.isEmpty()) {
return Optional.empty();
}
List<CapturedLine> lines = new ArrayList<>();
LineCapturingStripper stripper;
try {
stripper = new LineCapturingStripper(lines);
stripper.setStartPage(pageIndex + 1);
stripper.setEndPage(pageIndex + 1);
stripper.setSortByPosition(true);
// Side effect: populates `lines`. We don't need the concatenated text.
stripper.getText(doc);
} catch (IOException e) {
log.warn(
"PdfTextLocator failed to extract text on page {}: {}",
pageIndex,
e.getMessage());
return Optional.empty();
}
PDRectangle mediaBox = doc.getPage(pageIndex).getMediaBox();
float pageHeight = mediaBox.getHeight();
for (CapturedLine line : lines) {
if (normalize(line.text).contains(normalizedNeedle)) {
// PDFBox's *DirAdj coords descend from the top of the page; convert to PDF
// user-space (origin = bottom-left) so the bbox can feed a PDRectangle directly.
float userSpaceY = pageHeight - line.yTopDown - line.height;
return Optional.of(new MatchedBox(line.x, userSpaceY, line.width, line.height));
}
}
return Optional.empty();
}
/** Strip everything non-alphanumeric and lowercase for tolerant matching. */
private static String normalize(String s) {
return s.replaceAll("[^A-Za-z0-9]", "").toLowerCase(Locale.ROOT);
}
private static final class CapturedLine {
String text;
float x;
float yTopDown;
float width;
float height;
}
private static final class LineCapturingStripper extends PDFTextStripper {
private final List<CapturedLine> lines;
LineCapturingStripper(List<CapturedLine> sink) throws IOException {
super();
this.lines = sink;
}
@Override
protected void writeString(String text, List<TextPosition> textPositions)
throws IOException {
if (textPositions != null && !textPositions.isEmpty()) {
CapturedLine line = new CapturedLine();
line.text = text;
float minX = Float.MAX_VALUE;
float maxRight = 0f;
float minY = Float.MAX_VALUE;
float maxHeight = 0f;
for (TextPosition p : textPositions) {
float x = p.getXDirAdj();
float y = p.getYDirAdj();
float w = p.getWidthDirAdj();
float h = p.getHeightDir();
if (h == 0f) {
// Workaround: some fonts report 0 height via TextPosition; fall back to
// the nominal font size so downstream bboxes are never zero-height.
h = p.getFontSizeInPt();
}
if (x < minX) minX = x;
if (x + w > maxRight) maxRight = x + w;
if (y < minY) minY = y;
if (h > maxHeight) maxHeight = h;
}
line.x = minX;
line.width = maxRight - minX;
line.yTopDown = minY;
line.height = maxHeight;
lines.add(line);
}
super.writeString(text, textPositions);
}
}
}
@@ -15,7 +15,6 @@ import java.util.Set;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.Semaphore;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import io.github.pixee.security.BoundedLineReader;
@@ -203,15 +202,7 @@ public class ProcessExecutor {
boolean useSemaphore = true;
List<String> commandToRun = command;
if (shouldUseUnoServerPool(command)) {
try {
unoLease = unoServerPool.acquireEndpoint(timeoutDuration, TimeUnit.MINUTES);
} catch (TimeoutException e) {
throw new IOException(
"All unoserver endpoints busy; request timed out after "
+ timeoutDuration
+ " minutes",
e);
}
unoLease = unoServerPool.acquireEndpoint();
commandToRun = applyUnoServerEndpoint(command, unoLease.getEndpoint());
useSemaphore = false;
}
@@ -3,11 +3,8 @@ package stirling.software.common.util;
import java.util.ArrayList;
import java.util.Collections;
import java.util.List;
import java.util.Locale;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.LinkedBlockingQueue;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import java.util.concurrent.atomic.AtomicBoolean;
import stirling.software.common.model.ApplicationProperties;
@@ -45,27 +42,6 @@ public class UnoServerPool {
return new UnoServerLease(endpoints.get(index), index, this);
}
/** Fail-fast variant; non-positive timeout falls back to unbounded acquire. */
public UnoServerLease acquireEndpoint(long timeout, TimeUnit unit)
throws InterruptedException, TimeoutException {
if (endpoints.isEmpty()) {
return new UnoServerLease(defaultEndpoint(), null, this);
}
if (timeout <= 0) {
return acquireEndpoint();
}
Integer index = availableIndices.poll(timeout, unit);
if (index == null) {
throw new TimeoutException(
"Timed out waiting for a free unoserver endpoint after "
+ timeout
+ " "
+ unit.name().toLowerCase(Locale.ROOT));
}
return new UnoServerLease(endpoints.get(index), index, this);
}
private void releaseEndpoint(Integer index) {
if (index != null) {
availableIndices.offer(index);
@@ -6,18 +6,15 @@ import java.util.List;
import lombok.extern.slf4j.Slf4j;
import stirling.software.common.model.api.security.RedactionArea;
import tools.jackson.core.type.TypeReference;
import tools.jackson.databind.DeserializationFeature;
import tools.jackson.databind.JavaType;
import tools.jackson.databind.ObjectMapper;
import tools.jackson.databind.json.JsonMapper;
/**
* Spring property editor that decodes a JSON string into a typed {@link ArrayList}. Used to bind
* complex list parameters (e.g. {@code List<RedactionArea>}, {@code List<EditTextOperation>}) from
* multipart form fields, where Spring's default binding cannot deserialize a JSON array.
*/
@Slf4j
public class StringToArrayListPropertyEditor<T> extends PropertyEditorSupport {
public class StringToArrayListPropertyEditor extends PropertyEditorSupport {
private final ObjectMapper objectMapper =
JsonMapper.builder()
@@ -25,12 +22,6 @@ public class StringToArrayListPropertyEditor<T> extends PropertyEditorSupport {
.disable(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES)
.build();
private final Class<T> elementType;
public StringToArrayListPropertyEditor(Class<T> elementType) {
this.elementType = elementType;
}
@Override
public void setAsText(String text) throws IllegalArgumentException {
if (text == null || text.trim().isEmpty()) {
@@ -38,11 +29,8 @@ public class StringToArrayListPropertyEditor<T> extends PropertyEditorSupport {
return;
}
try {
JavaType listType =
objectMapper
.getTypeFactory()
.constructCollectionType(ArrayList.class, elementType);
List<T> list = objectMapper.readValue(text, listType);
TypeReference<ArrayList<RedactionArea>> typeRef = new TypeReference<>() {};
List<RedactionArea> list = objectMapper.readValue(text, typeRef);
setValue(list);
} catch (Exception e) {
log.error("Exception while converting {}", e);
@@ -1,39 +0,0 @@
package stirling.software.SPDF.config;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertNull;
import org.junit.jupiter.api.Test;
class EndpointConfigurationTest {
@Test
void endpointKeyForUriExtractsSimpleKebab() {
assertEquals(
"remove-pages",
EndpointConfiguration.endpointKeyForUri("/api/v1/general/remove-pages"));
assertEquals(
"compress-pdf",
EndpointConfiguration.endpointKeyForUri("/api/v1/misc/compress-pdf"));
assertEquals(
"add-watermark",
EndpointConfiguration.endpointKeyForUri("/api/v1/security/add-watermark"));
}
@Test
void endpointKeyForUriComposesConvertEndpoints() {
assertEquals(
"pdf-to-img", EndpointConfiguration.endpointKeyForUri("/api/v1/convert/pdf/img"));
assertEquals(
"pdf-to-word", EndpointConfiguration.endpointKeyForUri("/api/v1/convert/pdf/word"));
assertEquals(
"html-to-pdf", EndpointConfiguration.endpointKeyForUri("/api/v1/convert/html/pdf"));
}
@Test
void endpointKeyForUriReturnsNullForNonApiPaths() {
assertNull(EndpointConfiguration.endpointKeyForUri(null));
assertNull(EndpointConfiguration.endpointKeyForUri("/some-page"));
assertNull(EndpointConfiguration.endpointKeyForUri("/api/v1/general"));
}
}
@@ -1,153 +0,0 @@
package stirling.software.SPDF.pdf.parser;
import static org.assertj.core.api.Assertions.assertThat;
import static stirling.software.SPDF.pdf.parser.PdfModels.*;
import java.util.List;
import org.junit.jupiter.api.Test;
/**
* Unit tests for {@link LineAlignmentTableParser}, focused on the coincident-line merge logic and
* column-grid construction.
*/
class LineAlignmentTableParserTest {
private final LineAlignmentTableParser parser = new LineAlignmentTableParser();
// ── mergeCoincidentLines ─────────────────────────────────────────────────────────────────────
@Test
void mergeCoincidentLines_singleLine_unchanged() {
var lines = List.of(tokenized(rawLine(10f, 100f, "Revenue")));
assertThat(parser.mergeCoincidentLines(lines)).hasSize(1);
}
@Test
void mergeCoincidentLines_distinctYLines_unchanged() {
// Two lines at different y positions — must NOT be merged.
var lines =
List.of(
tokenized(rawLine(10f, 100f, "Revenue")),
tokenized(rawLine(10f, 115f, "Cost")));
assertThat(parser.mergeCoincidentLines(lines)).hasSize(2);
}
@Test
void mergeCoincidentLines_sameY_merged() {
// Simulates a financial-table row split by LineBuilder at the column gap:
// label fragment at x=72 → "Revenue"
// value fragment at x=350 → "1,234"
// Both have y=100. After merge they should form one TokenizedLine.
var label = rawLine(72f, 100f, "Revenue");
var value = rawLine(350f, 100f, "1,234");
var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(value)));
assertThat(merged).hasSize(1);
// The merged line should contain tokens from both halves.
var tokens = merged.get(0).all();
assertThat(tokens.stream().map(t -> t.text()).toList())
.containsExactlyInAnyOrder("Revenue", "1,234");
}
@Test
void mergeCoincidentLines_sameY_mergedLineHasCorrectBounds() {
var label = rawLine(72f, 100f, "Revenue"); // 7 chars × 6pt = 42pt wide → right = 114
var value = rawLine(350f, 100f, "1,234"); // 5 chars × 6pt = 30pt wide → right = 380
var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(value)));
var bounds = merged.get(0).line().bounds();
assertThat(bounds.x()).isEqualTo(72f);
assertThat(bounds.right()).isEqualTo(380f);
}
@Test
void mergeCoincidentLines_withinTolerance_merged() {
// Lines 1.5pt apart (within ROW_MERGE_TOLERANCE_PT = 2pt) should merge.
var a = rawLine(10f, 100.0f, "Alpha");
var b = rawLine(200f, 101.5f, "99");
var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b)));
assertThat(merged).hasSize(1);
}
@Test
void mergeCoincidentLines_beyondTolerance_notMerged() {
// Lines 3pt apart (beyond ROW_MERGE_TOLERANCE_PT = 2pt) should NOT merge.
var a = rawLine(10f, 100.0f, "Alpha");
var b = rawLine(200f, 103.0f, "99");
var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b)));
assertThat(merged).hasSize(2);
}
@Test
void mergeCoincidentLines_threeCoincident_allMerged() {
// Three fragments at the same y (e.g. wide financial table with two value columns).
var a = rawLine(72f, 100f, "Revenue");
var b = rawLine(300f, 100f, "1,234");
var c = rawLine(400f, 100f, "5,678");
var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b), tokenized(c)));
assertThat(merged).hasSize(1);
assertThat(merged.get(0).all()).hasSize(3);
}
@Test
void mergeCoincidentLines_coincidentPairFollowedByDistinctLine_twoGroups() {
var a = rawLine(72f, 100f, "Revenue");
var b = rawLine(350f, 100f, "1,234"); // same y as a → merges with a
var c = rawLine(10f, 115f, "Expenses"); // different y → stays separate
var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b), tokenized(c)));
assertThat(merged).hasSize(2);
}
@Test
void mergeCoincidentLines_numericAnchorStatus_correctAfterMerge() {
// After merging, the combined line should be an anchor (≥2 numeric tokens).
// "Revenue" alone → not an anchor. "1,234 567" alone → anchor.
// Merged → anchor with at least 2 numerics.
var label = rawLine(72f, 100f, "Revenue");
var values = rawLineMultiWord(350f, 100f, "1,234", 30f, "567", 30f);
var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(values)));
assertThat(merged).hasSize(1);
assertThat(merged.get(0).isAnchor()).isTrue();
}
// ── helpers ──────────────────────────────────────────────────────────────────────────────────
/** Creates a RawLine with a single TextFragment of the given text at the given position. */
private static RawLine rawLine(float x, float y, String text) {
float width = text.length() * 6f; // ~6pt per char — rough but consistent
float height = 12f;
Bounds bounds = new Bounds(x, y, width, height);
TextFragment fragment =
new TextFragment("tf-test", text, bounds, y + height, 11f, "Helvetica", false);
return new RawLine("ln-test", List.of(fragment), bounds, 1);
}
/**
* Creates a RawLine with two TextFragments representing two words separated by a small gap.
* Used to simulate a values-only line with multiple numeric tokens.
*/
private static RawLine rawLineMultiWord(
float x, float y, String word1, float w1, String word2, float w2) {
float height = 12f;
Bounds b1 = new Bounds(x, y, w1, height);
Bounds b2 = new Bounds(x + w1 + 5f, y, w2, height);
TextFragment f1 = new TextFragment("tf-1", word1, b1, y + height, 11f, "Helvetica", false);
TextFragment f2 = new TextFragment("tf-2", word2, b2, y + height, 11f, "Helvetica", false);
Bounds lineBounds = new Bounds(x, y, x + w1 + 5f + w2 - x, height);
return new RawLine("ln-test", List.of(f1, f2), lineBounds, 1);
}
/** Tokenises a RawLine via the parser's own tokenise logic (package-private access). */
private LineAlignmentTableParser.TokenizedLine tokenized(RawLine line) {
return parser.tokenize(line);
}
}
@@ -1,59 +0,0 @@
package stirling.software.common.architecture;
import static com.tngtech.archunit.lang.syntax.ArchRuleDefinition.noClasses;
import org.junit.jupiter.api.Test;
import com.tngtech.archunit.core.domain.JavaClasses;
import com.tngtech.archunit.core.importer.ClassFileImporter;
import com.tngtech.archunit.core.importer.ImportOption;
import com.tngtech.archunit.lang.ArchRule;
/**
* Module dependency-direction guardrails. Allowed direction: {@code saas → proprietary → common}
* and {@code stirling-pdf → proprietary → common}.
*/
class ArchitectureTest {
private static final JavaClasses commonClasses =
new ClassFileImporter()
.withImportOption(ImportOption.Predefined.DO_NOT_INCLUDE_JARS)
.withImportOption(ImportOption.Predefined.DO_NOT_INCLUDE_TESTS)
.importPackages("stirling.software.common");
@Test
void commonDoesNotDependOnCore() {
ArchRule rule =
noClasses()
.that()
.resideInAPackage("stirling.software.common..")
.should()
.dependOnClassesThat()
.resideInAPackage("stirling.software.SPDF..");
rule.check(commonClasses);
}
@Test
void commonDoesNotDependOnProprietary() {
ArchRule rule =
noClasses()
.that()
.resideInAPackage("stirling.software.common..")
.should()
.dependOnClassesThat()
.resideInAPackage("stirling.software.proprietary..");
rule.check(commonClasses);
}
@Test
void commonDoesNotDependOnSaas() {
ArchRule rule =
noClasses()
.that()
.resideInAPackage("stirling.software.common..")
.should()
.dependOnClassesThat()
.resideInAPackage("stirling.software.saas..");
rule.check(commonClasses);
}
}
@@ -1,32 +0,0 @@
package stirling.software.common.jpdfium;
import static org.junit.jupiter.api.Assertions.assertNotNull;
import static org.junit.jupiter.api.Assertions.assertTrue;
import java.io.IOException;
import java.io.InputStream;
import java.nio.file.Files;
import java.nio.file.Path;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.io.TempDir;
import stirling.software.jpdfium.PdfDocument;
class JPDFiumSmokeTest {
@Test
void opensExamplePdfAndReadsPageCount(@TempDir Path tmp) throws IOException {
Path pdf = tmp.resolve("example.pdf");
try (InputStream in = getClass().getResourceAsStream("/example.pdf")) {
assertNotNull(in, "example.pdf must exist under src/test/resources");
Files.copy(in, pdf);
}
try (PdfDocument doc = PdfDocument.open(pdf)) {
assertTrue(
doc.pageCount() >= 1,
"PdfDocument should report at least one page for example.pdf");
}
}
}
@@ -22,13 +22,11 @@ import org.springframework.mock.env.MockEnvironment;
import org.springframework.util.LinkedMultiValueMap;
import org.springframework.util.MultiValueMap;
import org.springframework.web.client.RequestCallback;
import org.springframework.web.client.ResourceAccessException;
import org.springframework.web.client.ResponseExtractor;
import org.springframework.web.client.RestTemplate;
import jakarta.servlet.ServletContext;
import stirling.software.common.model.ApplicationProperties;
import stirling.software.common.util.TempFile;
import stirling.software.common.util.TempFileManager;
@@ -44,19 +42,8 @@ class InternalApiClientTest {
@BeforeEach
void setUp() {
lenient().when(servletContext.getContextPath()).thenReturn("");
client = newClient();
}
/**
* Build a fresh client. Tests that use {@link org.mockito.Mockito#mockConstruction} to
* intercept {@link RestTemplate} must call this from inside their {@code mockConstruction}
* block, since the client now caches one RestTemplate per instance at construction time.
*/
private InternalApiClient newClient() {
MockEnvironment environment = new MockEnvironment().withProperty("server.port", "8080");
ApplicationProperties applicationProperties = new ApplicationProperties();
return new InternalApiClient(
servletContext, userService, tempFileManager, environment, applicationProperties);
client = new InternalApiClient(servletContext, userService, tempFileManager, environment);
}
@Test
@@ -88,10 +75,7 @@ class InternalApiClientTest {
.thenAnswer(inv -> fakeOkResponse(inv.getArgument(3)));
})) {
// Reconstruct the client so its cached RestTemplate is the mocked one.
InternalApiClient mockedClient = newClient();
ResponseEntity<Resource> response =
mockedClient.post("/api/v1/general/merge-pdfs", body);
ResponseEntity<Resource> response = client.post("/api/v1/general/merge-pdfs", body);
assertNotNull(response);
assertEquals(HttpStatus.OK, response.getStatusCode());
@@ -102,123 +86,12 @@ class InternalApiClientTest {
}
}
@Test
void postWrapsSocketTimeoutAsInternalApiTimeoutException() {
// Simulates the read-timeout case: RestTemplate wraps SocketTimeoutException in
// ResourceAccessException when the underlying socket times out. The client must repackage
// that into a typed timeout exception that carries the failing endpoint and the configured
// read timeout, so the workflow layer can surface a clean "tool didn't respond" message
// to the user.
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
body.add("fileInput", namedResource("input.pdf", "data"));
try (var ignored =
mockConstruction(
RestTemplate.class,
(rt, ctx) -> {
when(rt.httpEntityCallback(any(), eq(Resource.class)))
.thenAnswer(inv -> (RequestCallback) req -> {});
when(rt.execute(anyString(), eq(HttpMethod.POST), any(), any()))
.thenThrow(
new ResourceAccessException(
"I/O error on POST request: Read timed out",
new java.net.SocketTimeoutException(
"Read timed out")));
})) {
InternalApiClient mockedClient = newClient();
InternalApiTimeoutException thrown =
assertThrows(
InternalApiTimeoutException.class,
() -> mockedClient.post("/api/v1/general/merge-pdfs", body));
assertEquals("/api/v1/general/merge-pdfs", thrown.getEndpointPath());
assertNotNull(thrown.getReadTimeout());
assertTrue(thrown.getMessage().contains("/api/v1/general/merge-pdfs"));
}
}
@Test
void postRethrowsNonTimeoutResourceAccessExceptionAsIs() {
// ResourceAccessException covers more than just timeouts (e.g. connection refused, DNS
// failure). Only SocketTimeoutException-rooted failures are timeouts; everything else
// must propagate so the upstream generic handler can label it correctly instead of lying
// about a "tool didn't respond" timeout.
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
body.add("fileInput", namedResource("input.pdf", "data"));
try (var ignored =
mockConstruction(
RestTemplate.class,
(rt, ctx) -> {
when(rt.httpEntityCallback(any(), eq(Resource.class)))
.thenAnswer(inv -> (RequestCallback) req -> {});
when(rt.execute(anyString(), eq(HttpMethod.POST), any(), any()))
.thenThrow(
new ResourceAccessException(
"I/O error on POST request: Connection refused",
new java.net.ConnectException(
"Connection refused")));
})) {
InternalApiClient mockedClient = newClient();
assertThrows(
ResourceAccessException.class,
() -> mockedClient.post("/api/v1/general/merge-pdfs", body));
}
}
@Test
void postRejectsDisallowedPath() {
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
assertThrows(SecurityException.class, () -> client.post("/api/v1/admin/settings", body));
}
@Test
void postRejectsAiEndpointsOutsideToolsSubnamespace() {
// /api/v1/ai/orchestrate and other non-tool AI endpoints are not internally
// dispatchable. Only /api/v1/ai/tools/* and the general/misc/security/convert/filter
// namespaces are on the allowlist — letting a plan step re-enter /orchestrate would
// introduce recursion risk.
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
assertThrows(SecurityException.class, () -> client.post("/api/v1/ai/orchestrate", body));
}
@Test
void postAcceptsAiToolsSubnamespace() throws Exception {
// Agent tool paths like /api/v1/ai/tools/pdf-comment-agent are on the allowlist and
// should be dispatchable by the orchestrator's plan executor.
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
body.add("fileInput", namedResource("input.pdf", "data"));
Path tempPath = Files.createTempFile("internal-api-ai-tools-test", ".tmp");
TempFile tempFile = mock(TempFile.class);
when(tempFile.getPath()).thenReturn(tempPath);
when(tempFile.getFile()).thenReturn(tempPath.toFile());
when(tempFileManager.createManagedTempFile("internal-api")).thenReturn(tempFile);
try (var ignored =
mockConstruction(
RestTemplate.class,
(rt, ctx) -> {
when(rt.httpEntityCallback(any(), eq(Resource.class)))
.thenReturn((RequestCallback) req -> {});
when(rt.execute(anyString(), eq(HttpMethod.POST), any(), any()))
.thenAnswer(inv -> fakeOkResponse(inv.getArgument(3)));
})) {
// Reconstruct the client so its cached RestTemplate is the mocked one.
InternalApiClient mockedClient = newClient();
ResponseEntity<Resource> response =
mockedClient.post("/api/v1/ai/tools/pdf-comment-agent", body);
assertNotNull(response);
assertEquals(HttpStatus.OK, response.getStatusCode());
} finally {
Files.deleteIfExists(tempPath);
}
}
@Test
void postRejectsPathTraversal() {
MultiValueMap<String, Object> body = new LinkedMultiValueMap<>();
@@ -1,191 +0,0 @@
package stirling.software.common.service;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertNotNull;
import static org.junit.jupiter.api.Assertions.assertTrue;
import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.pdmodel.interactive.annotation.PDAnnotation;
import org.apache.pdfbox.pdmodel.interactive.annotation.PDAnnotationText;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import stirling.software.common.model.api.comments.AnnotationLocation;
import stirling.software.common.model.api.comments.StickyNoteSpec;
class PdfAnnotationServiceTest {
private PdfAnnotationService service;
@BeforeEach
void setUp() {
service = new PdfAnnotationService();
}
@Test
void addStickyNotesPlacesOneAnnotationPerValidSpec() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
List<StickyNoteSpec> specs =
List.of(
spec(0, 72f, 700f, "First comment", "alice", null),
spec(1, 100f, 650f, "Second comment", null, "Second"));
int applied = service.addStickyNotes(doc, specs);
assertEquals(2, applied);
byte[] saved = save(doc);
try (PDDocument reloaded = Loader.loadPDF(saved)) {
assertEquals(1, textAnnotations(reloaded.getPage(0).getAnnotations()).size());
assertEquals(1, textAnnotations(reloaded.getPage(1).getAnnotations()).size());
PDAnnotationText first =
textAnnotations(reloaded.getPage(0).getAnnotations()).get(0);
assertEquals("First comment", first.getContents());
assertEquals("alice", first.getTitlePopup(), "author override propagates");
assertNotNull(first.getSubject(), "subject falls back to default when null");
}
}
}
@Test
void skipsSpecsWithBlankText() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
List<StickyNoteSpec> specs =
List.of(
spec(0, 72f, 700f, "Valid", null, null),
spec(0, 72f, 680f, " ", null, null));
int applied = service.addStickyNotes(doc, specs);
assertEquals(1, applied, "Blank-text spec must be skipped");
}
}
@Test
void skipsSpecsWithOutOfRangePageIndex() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
List<StickyNoteSpec> specs =
List.of(
spec(0, 72f, 700f, "OK", null, null),
spec(99, 72f, 700f, "Too far", null, null),
spec(-1, 72f, 700f, "Negative", null, null));
int applied = service.addStickyNotes(doc, specs);
assertEquals(1, applied, "Only the in-range spec should be applied");
}
}
@Test
void handlesNullAndEmptySpecList() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
assertEquals(0, service.addStickyNotes(doc, null));
assertEquals(0, service.addStickyNotes(doc, List.of()));
}
}
@Test
void skipsSpecsWithNonPositiveDimensions() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
StickyNoteSpec zeroWidth =
new StickyNoteSpec(
new AnnotationLocation(0, 72f, 700f, 0f, 20f),
"Zero width",
null,
null);
StickyNoteSpec negativeHeight =
new StickyNoteSpec(
new AnnotationLocation(0, 72f, 680f, 20f, -5f),
"Negative height",
null,
null);
List<StickyNoteSpec> specs =
List.of(spec(0, 72f, 660f, "OK", null, null), zeroWidth, negativeHeight);
int applied = service.addStickyNotes(doc, specs);
assertEquals(1, applied, "Only the positively-sized spec should be applied");
}
}
@Test
void skipsSpecsWithOverlongText() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
String overlong = "x".repeat(100_001);
List<StickyNoteSpec> specs =
List.of(
spec(0, 72f, 700f, "Short", null, null),
spec(0, 72f, 680f, overlong, null, null));
int applied = service.addStickyNotes(doc, specs);
assertEquals(1, applied, "Overlong-text spec must be skipped");
}
}
@Test
void appliesDefaultAuthorAndSubjectWhenAbsent() throws IOException {
byte[] bytes = twoPagePdfBytes();
try (PDDocument doc = Loader.loadPDF(bytes)) {
service.addStickyNote(doc, spec(0, 72f, 700f, "No author given", null, null));
byte[] saved = save(doc);
try (PDDocument reloaded = Loader.loadPDF(saved)) {
PDAnnotationText annot =
textAnnotations(reloaded.getPage(0).getAnnotations()).get(0);
assertTrue(
annot.getTitlePopup() != null && !annot.getTitlePopup().isBlank(),
"Default author should be applied");
assertTrue(
annot.getSubject() != null && !annot.getSubject().isBlank(),
"Default subject should be applied");
}
}
}
// --- helpers ---
private static StickyNoteSpec spec(
int page, float x, float y, String text, String author, String subject) {
return new StickyNoteSpec(
new AnnotationLocation(page, x, y, 20f, 20f), text, author, subject);
}
private static byte[] twoPagePdfBytes() throws IOException {
try (PDDocument doc = new PDDocument()) {
doc.addPage(new PDPage(PDRectangle.A4));
doc.addPage(new PDPage(PDRectangle.A4));
return save(doc);
}
}
private static byte[] save(PDDocument doc) throws IOException {
ByteArrayOutputStream baos = new ByteArrayOutputStream();
doc.save(baos);
return baos.toByteArray();
}
private static List<PDAnnotationText> textAnnotations(List<PDAnnotation> annotations) {
List<PDAnnotationText> out = new ArrayList<>();
for (PDAnnotation a : annotations) {
if (a instanceof PDAnnotationText t) {
out.add(t);
}
}
return out;
}
}
@@ -1,229 +0,0 @@
package stirling.software.common.util;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertNotNull;
import static org.junit.jupiter.api.Assertions.assertNull;
import static org.junit.jupiter.api.Assertions.assertTrue;
import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.pdmodel.interactive.annotation.PDAnnotationWidget;
import org.apache.pdfbox.pdmodel.interactive.form.PDAcroForm;
import org.apache.pdfbox.pdmodel.interactive.form.PDField;
import org.apache.pdfbox.pdmodel.interactive.form.PDNonTerminalField;
import org.apache.pdfbox.pdmodel.interactive.form.PDTextField;
import org.junit.jupiter.api.Test;
class FormUtilsPruneOrphanedFieldsTest {
@Test
void noAcroFormIsNoOp() throws IOException {
try (PDDocument document = new PDDocument()) {
document.addPage(new PDPage(PDRectangle.A4));
FormUtils.pruneOrphanedFormFields(document);
assertNull(document.getDocumentCatalog().getAcroForm(null));
}
}
@Test
void dropsFieldsWhoseWidgetsAreAllOnRemovedPages() throws IOException {
byte[] pdfBytes = buildPdfWithFieldPerPage(3);
try (PDDocument document = Loader.loadPDF(pdfBytes)) {
document.removePage(2);
document.removePage(0);
FormUtils.pruneOrphanedFormFields(document);
PDAcroForm form = document.getDocumentCatalog().getAcroForm(null);
assertNotNull(form);
List<String> remainingNames = new ArrayList<>();
for (PDField field : form.getFields()) {
remainingNames.add(field.getPartialName());
}
assertEquals(List.of("field_1"), remainingNames);
}
}
@Test
void dropsAcroFormEntirelyWhenNoFieldsSurvive() throws IOException {
byte[] pdfBytes = buildPdfWithFieldPerPage(2);
try (PDDocument document = Loader.loadPDF(pdfBytes)) {
document.removePage(1);
document.removePage(0);
document.addPage(new PDPage(PDRectangle.A4));
FormUtils.pruneOrphanedFormFields(document);
assertNull(document.getDocumentCatalog().getAcroForm(null));
}
}
@Test
void keepsLiveWidgetsAndDropsOrphanWidgetsFromMultiWidgetField() throws IOException {
byte[] pdfBytes = buildPdfWithMultiWidgetField();
try (PDDocument document = Loader.loadPDF(pdfBytes)) {
document.removePage(0);
FormUtils.pruneOrphanedFormFields(document);
PDAcroForm form = document.getDocumentCatalog().getAcroForm(null);
assertNotNull(form);
assertEquals(1, form.getFields().size());
PDField field = form.getFields().get(0);
assertEquals("multi", field.getPartialName());
assertEquals(2, field.getWidgets().size(), "two widgets remain after one is dropped");
}
}
@Test
void survivesRoundTripWithoutOrphanPagesInOutput() throws IOException {
byte[] pdfBytes = buildPdfWithFieldPerPage(3);
byte[] writtenBytes;
try (PDDocument document = Loader.loadPDF(pdfBytes)) {
document.removePage(2);
document.removePage(1);
FormUtils.pruneOrphanedFormFields(document);
try (ByteArrayOutputStream out = new ByteArrayOutputStream()) {
document.save(out);
writtenBytes = out.toByteArray();
}
}
try (PDDocument reloaded = Loader.loadPDF(writtenBytes)) {
assertEquals(1, reloaded.getNumberOfPages());
PDAcroForm form = reloaded.getDocumentCatalog().getAcroForm(null);
assertNotNull(form);
assertEquals(1, form.getFields().size());
assertEquals("field_0", form.getFields().get(0).getPartialName());
}
}
@Test
void prunesNestedNonTerminalFields() throws IOException {
byte[] pdfBytes = buildPdfWithNestedFields();
try (PDDocument document = Loader.loadPDF(pdfBytes)) {
document.removePage(1);
FormUtils.pruneOrphanedFormFields(document);
PDAcroForm form = document.getDocumentCatalog().getAcroForm(null);
assertNotNull(form);
assertEquals(1, form.getFields().size());
PDField group = form.getFields().get(0);
assertEquals("group", group.getPartialName());
assertTrue(group instanceof PDNonTerminalField);
PDNonTerminalField nonTerminal = (PDNonTerminalField) group;
assertEquals(1, nonTerminal.getChildren().size());
assertEquals("kept", nonTerminal.getChildren().get(0).getPartialName());
}
}
private static byte[] buildPdfWithFieldPerPage(int pageCount) throws IOException {
try (PDDocument document = new PDDocument()) {
PDAcroForm acroForm = new PDAcroForm(document);
acroForm.setDefaultResources(new PDResources());
document.getDocumentCatalog().setAcroForm(acroForm);
for (int i = 0; i < pageCount; i++) {
PDPage page = new PDPage(PDRectangle.A4);
document.addPage(page);
PDTextField field = new PDTextField(acroForm);
field.setPartialName("field_" + i);
PDAnnotationWidget widget = new PDAnnotationWidget();
widget.setRectangle(new PDRectangle(50, 50, 100, 20));
widget.setPage(page);
field.setWidgets(List.of(widget));
acroForm.getFields().add(field);
page.getAnnotations().add(widget);
}
try (ByteArrayOutputStream out = new ByteArrayOutputStream()) {
document.save(out);
return out.toByteArray();
}
}
}
private static byte[] buildPdfWithMultiWidgetField() throws IOException {
try (PDDocument document = new PDDocument()) {
PDAcroForm acroForm = new PDAcroForm(document);
acroForm.setDefaultResources(new PDResources());
document.getDocumentCatalog().setAcroForm(acroForm);
List<PDAnnotationWidget> widgets = new ArrayList<>();
for (int i = 0; i < 3; i++) {
PDPage page = new PDPage(PDRectangle.A4);
document.addPage(page);
PDAnnotationWidget widget = new PDAnnotationWidget();
widget.setRectangle(new PDRectangle(50, 50, 100, 20));
widget.setPage(page);
page.getAnnotations().add(widget);
widgets.add(widget);
}
PDTextField field = new PDTextField(acroForm);
field.setPartialName("multi");
field.setWidgets(widgets);
acroForm.getFields().add(field);
try (ByteArrayOutputStream out = new ByteArrayOutputStream()) {
document.save(out);
return out.toByteArray();
}
}
}
private static byte[] buildPdfWithNestedFields() throws IOException {
try (PDDocument document = new PDDocument()) {
PDAcroForm acroForm = new PDAcroForm(document);
acroForm.setDefaultResources(new PDResources());
document.getDocumentCatalog().setAcroForm(acroForm);
PDPage pageA = new PDPage(PDRectangle.A4);
PDPage pageB = new PDPage(PDRectangle.A4);
document.addPage(pageA);
document.addPage(pageB);
PDNonTerminalField group = new PDNonTerminalField(acroForm);
group.setPartialName("group");
PDTextField kept = new PDTextField(acroForm);
kept.setPartialName("kept");
PDAnnotationWidget keptWidget = new PDAnnotationWidget();
keptWidget.setRectangle(new PDRectangle(50, 50, 100, 20));
keptWidget.setPage(pageA);
kept.setWidgets(List.of(keptWidget));
pageA.getAnnotations().add(keptWidget);
PDTextField dropped = new PDTextField(acroForm);
dropped.setPartialName("dropped");
PDAnnotationWidget droppedWidget = new PDAnnotationWidget();
droppedWidget.setRectangle(new PDRectangle(50, 100, 100, 20));
droppedWidget.setPage(pageB);
dropped.setWidgets(List.of(droppedWidget));
pageB.getAnnotations().add(droppedWidget);
group.setChildren(List.of(kept, dropped));
acroForm.getFields().add(group);
try (ByteArrayOutputStream out = new ByteArrayOutputStream()) {
document.save(out);
return out.toByteArray();
}
}
}
}
@@ -1,97 +0,0 @@
package stirling.software.common.util;
import static org.assertj.core.api.Assertions.assertThat;
import java.awt.Color;
import java.io.ByteArrayOutputStream;
import java.util.Optional;
import org.apache.pdfbox.Loader;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDPage;
import org.apache.pdfbox.pdmodel.PDPageContentStream;
import org.apache.pdfbox.pdmodel.common.PDRectangle;
import org.apache.pdfbox.pdmodel.font.PDType1Font;
import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
import org.junit.jupiter.api.Test;
import stirling.software.common.util.PdfTextLocator.MatchedBox;
class PdfTextLocatorTest {
private final PdfTextLocator locator = new PdfTextLocator();
@Test
void findsLineContainingNeedleAndReturnsUserSpaceBox() throws Exception {
byte[] pdf = pdfWithLines(new String[] {"Revenue: $215,000", "Expenses: $120,000"});
try (PDDocument doc = Loader.loadPDF(pdf)) {
Optional<MatchedBox> match = locator.findOnPage(doc, 0, "215000");
assertThat(match).isPresent();
MatchedBox box = match.get();
// Line was drawn at y=720 in user-space (bottom-left origin); locator
// should return a bbox close to that height band with non-zero width.
assertThat(box.width()).isGreaterThan(0f);
assertThat(box.height()).isGreaterThan(0f);
assertThat(box.y()).isBetween(700f, 740f);
}
}
@Test
void matchIsCaseAndPunctuationInsensitive() throws Exception {
byte[] pdf = pdfWithLines(new String[] {"Total Revenue.", "Q4 summary"});
try (PDDocument doc = Loader.loadPDF(pdf)) {
Optional<MatchedBox> match = locator.findOnPage(doc, 0, "total revenue");
assertThat(match).isPresent();
}
}
@Test
void returnsEmptyWhenNeedleNotFound() throws Exception {
byte[] pdf = pdfWithLines(new String[] {"Nothing to see here"});
try (PDDocument doc = Loader.loadPDF(pdf)) {
Optional<MatchedBox> match = locator.findOnPage(doc, 0, "not-on-this-page");
assertThat(match).isEmpty();
}
}
@Test
void returnsEmptyForBlankNeedle() throws Exception {
byte[] pdf = pdfWithLines(new String[] {"Any text"});
try (PDDocument doc = Loader.loadPDF(pdf)) {
assertThat(locator.findOnPage(doc, 0, "")).isEmpty();
assertThat(locator.findOnPage(doc, 0, " ")).isEmpty();
assertThat(locator.findOnPage(doc, 0, null)).isEmpty();
}
}
@Test
void returnsEmptyForOutOfRangePage() throws Exception {
byte[] pdf = pdfWithLines(new String[] {"Single page"});
try (PDDocument doc = Loader.loadPDF(pdf)) {
assertThat(locator.findOnPage(doc, -1, "single")).isEmpty();
assertThat(locator.findOnPage(doc, 99, "single")).isEmpty();
}
}
private static byte[] pdfWithLines(String[] lines) throws Exception {
try (PDDocument doc = new PDDocument()) {
PDPage page = new PDPage(PDRectangle.A4);
doc.addPage(page);
try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
cs.setFont(new PDType1Font(Standard14Fonts.FontName.HELVETICA), 12);
cs.setNonStrokingColor(Color.BLACK);
float y = 720f;
for (String line : lines) {
cs.beginText();
cs.newLineAtOffset(72f, y);
cs.showText(line);
cs.endText();
y -= 20f;
}
}
ByteArrayOutputStream baos = new ByteArrayOutputStream();
doc.save(baos);
return baos.toByteArray();
}
}
}
@@ -9,7 +9,6 @@ import java.util.concurrent.CountDownLatch;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.TimeoutException;
import java.util.concurrent.atomic.AtomicInteger;
import org.junit.jupiter.api.Test;
@@ -174,70 +173,6 @@ public class UnoServerPoolTest {
assertEquals(port1, port2, "Should reuse the same endpoint after release");
}
@Test
void testAcquireWithTimeoutFailsFast() throws InterruptedException {
List<ApplicationProperties.ProcessExecutor.UnoServerEndpoint> endpoints =
createEndpoints(1);
UnoServerPool pool = new UnoServerPool(endpoints);
UnoServerPool.UnoServerLease held = pool.acquireEndpoint();
long start = System.nanoTime();
assertThrows(
TimeoutException.class,
() -> pool.acquireEndpoint(100, TimeUnit.MILLISECONDS),
"Should time out when no endpoint available");
long elapsedMs = (System.nanoTime() - start) / 1_000_000;
assertTrue(
elapsedMs >= 100 && elapsedMs < 1000,
"Should fail-fast after roughly the timeout, got " + elapsedMs + "ms");
held.close();
}
@Test
void testAcquireWithTimeoutSucceedsWhenAvailable()
throws InterruptedException, TimeoutException {
List<ApplicationProperties.ProcessExecutor.UnoServerEndpoint> endpoints =
createEndpoints(1);
UnoServerPool pool = new UnoServerPool(endpoints);
try (UnoServerPool.UnoServerLease lease = pool.acquireEndpoint(1, TimeUnit.SECONDS)) {
assertNotNull(lease);
assertEquals(2003, lease.getEndpoint().getPort());
}
}
@Test
void testAcquireWithZeroTimeoutBlocksUnbounded() throws InterruptedException {
List<ApplicationProperties.ProcessExecutor.UnoServerEndpoint> endpoints =
createEndpoints(1);
UnoServerPool pool = new UnoServerPool(endpoints);
UnoServerPool.UnoServerLease held = pool.acquireEndpoint();
AtomicInteger acquired = new AtomicInteger(0);
Thread t =
Thread.ofVirtual()
.start(
() -> {
try {
UnoServerPool.UnoServerLease lease =
pool.acquireEndpoint(0, TimeUnit.MILLISECONDS);
acquired.incrementAndGet();
lease.close();
} catch (Exception e) {
fail("unexpected: " + e);
}
});
Thread.sleep(150);
assertEquals(0, acquired.get(), "Should still be blocked");
held.close();
t.join(1000);
assertEquals(1, acquired.get(), "Should acquire after release");
}
@Test
void testHostLocationAndProtocol() throws InterruptedException {
List<ApplicationProperties.ProcessExecutor.UnoServerEndpoint> endpoints = new ArrayList<>();
@@ -11,11 +11,11 @@ import stirling.software.common.model.api.security.RedactionArea;
class StringToArrayListPropertyEditorTest {
private StringToArrayListPropertyEditor<RedactionArea> editor;
private StringToArrayListPropertyEditor editor;
@BeforeEach
void setUp() {
editor = new StringToArrayListPropertyEditor<>(RedactionArea.class);
editor = new StringToArrayListPropertyEditor();
}
@Test
+15 -35
View File
@@ -39,14 +39,12 @@ spotless {
}
dependencies {
if (!gradle.ext.disableAdditional) {
if (System.getenv('DISABLE_ADDITIONAL_FEATURES') != 'true'
|| (project.hasProperty('DISABLE_ADDITIONAL_FEATURES')
&& System.getProperty('DISABLE_ADDITIONAL_FEATURES') != 'true')) {
implementation project(':proprietary')
}
if (gradle.ext.enableSaas) {
implementation project(':saas')
}
implementation project(':common')
implementation 'org.springframework.boot:spring-boot-starter-jetty'
implementation 'org.eclipse.jetty.http2:jetty-http2-server'
@@ -67,7 +65,7 @@ dependencies {
exclude group: 'com.fasterxml.jackson.jaxrs'
exclude group: 'com.fasterxml.jackson.module', module: 'jackson-module-jaxb-annotations'
}
implementation 'commons-io:commons-io:2.22.0'
implementation 'commons-io:commons-io:2.21.0'
implementation "org.bouncycastle:bcprov-jdk18on:$bouncycastleVersion"
implementation "org.bouncycastle:bcpkix-jdk18on:$bouncycastleVersion"
implementation 'io.micrometer:micrometer-core'
@@ -89,6 +87,12 @@ dependencies {
implementation 'com.sun.xml.bind:jaxb-core:4.0.7'
implementation 'org.apache.poi:poi-ooxml:5.5.1'
// https://mvnrepository.com/artifact/technology.tabula/tabula
implementation ('technology.tabula:tabula:1.0.5') {
exclude group: 'org.slf4j', module: 'slf4j-simple'
exclude group: 'org.bouncycastle', module: 'bcprov-jdk15on'
exclude group: 'com.google.code.gson', module: 'gson'
}
// CVE-2022-25647: Explicit gson 2.13.2 to prevent unsafe deserialization (tabula would pull 2.8.7)
implementation 'com.google.code.gson:gson:2.13.2'
implementation 'org.apache.pdfbox:jbig2-imageio:3.0.4'
@@ -162,8 +166,7 @@ bootJar {
manifest {
attributes(
'Implementation-Title': 'Stirling-PDF',
'Implementation-Version': project.version,
'Enable-Native-Access': 'ALL-UNNAMED'
'Implementation-Version': project.version
)
}
}
@@ -176,23 +179,6 @@ springBoot {
// Frontend build tasks - only enabled with -PbuildWithFrontend=true
def buildWithFrontend = project.hasProperty('buildWithFrontend') && project.property('buildWithFrontend') == 'true'
def buildPrototypes = project.hasProperty('prototypesMode') && project.property('prototypesMode') == 'true'
// Vite mode: -PprototypesMode > -PfrontendMode > enableSaas > disableAdditional > proprietary.
def frontendModeOverride = project.findProperty('frontendMode')?.toString()?.toLowerCase()
def frontendMode
if (buildPrototypes) {
frontendMode = 'prototypes'
} else if (frontendModeOverride) {
frontendMode = frontendModeOverride
} else if (gradle.ext.enableSaas) {
frontendMode = 'saas'
} else if (gradle.ext.disableAdditional) {
frontendMode = 'core'
} else {
frontendMode = 'proprietary'
}
def frontendBuildTask = "frontend:build:${frontendMode}"
def frontendDir = file('../../frontend')
def frontendDistDir = file('../../frontend/dist')
def resourcesStaticDir = file('src/main/resources/static')
@@ -261,7 +247,7 @@ tasks.register('npmBuild', Exec) {
group = 'frontend'
description = 'Build frontend application'
workingDir file('../..')
commandLine = ['task', frontendBuildTask]
commandLine = buildPrototypes ? ['task', 'frontend:build:prototypes'] : ['task', 'frontend:build']
inputs.dir(new File(frontendDir, 'src'))
inputs.dir(new File(frontendDir, 'public'))
inputs.file(new File(frontendDir, 'package.json'))
@@ -276,7 +262,7 @@ tasks.register('npmBuild', Exec) {
environment 'VITE_API_BASE_URL', '/'
doFirst {
println "Building frontend application for production (mode=${frontendMode}, VITE_API_BASE_URL=/)"
println "Building frontend application for production (VITE_API_BASE_URL=/)"
}
}
@@ -285,7 +271,6 @@ tasks.register('copyFrontendAssets', Copy) {
group = 'frontend'
description = 'Copy frontend build to static resources'
dependsOn npmBuild
dependsOn cleanFrontendAssets
from(frontendDistDir) {
// Exclude files that conflict with backend static resources
exclude 'robots.txt' // Backend already has this
@@ -326,7 +311,7 @@ tasks.named('copyFrontendAssets').configure {
}
if (buildWithFrontend) {
println "Frontend build enabled - JAR will include React frontend (mode=${frontendMode})"
println "Frontend build enabled - JAR will include React frontend"
processResources.dependsOn copyFrontendAssets
} else {
println "Frontend build disabled - JAR will be backend-only with API landing page"
@@ -335,9 +320,4 @@ if (buildWithFrontend) {
}
bootJar.dependsOn ':common:jar'
if (!gradle.ext.disableAdditional) {
bootJar.dependsOn ':proprietary:jar'
}
if (gradle.ext.enableSaas) {
bootJar.dependsOn ':saas:jar'
}
bootJar.dependsOn ':proprietary:jar'
@@ -35,8 +35,7 @@ import stirling.software.common.model.ApplicationProperties;
scanBasePackages = {
"stirling.software.SPDF",
"stirling.software.common",
"stirling.software.proprietary",
"stirling.software.saas"
"stirling.software.proprietary"
})
public class SPDFApplication {
@@ -206,22 +205,15 @@ public class SPDFApplication {
}
}
// 2. Detect classpath shape and pick the matching profile chain.
boolean hasSaas = isClassPresent("stirling.software.saas.security.SupabaseSecurityConfig");
boolean hasSecurity =
isClassPresent(
"stirling.software.proprietary.security.configuration.SecurityConfiguration");
if (hasSaas) {
log.info("SaaS features in jar");
return new String[] {"security", "saas"};
}
if (hasSecurity) {
// 2. Detect if SecurityConfiguration is present on classpath
if (isClassPresent(
"stirling.software.proprietary.security.configuration.SecurityConfiguration")) {
log.info("Additional features in jar");
return new String[] {"security"};
} else {
log.info("Without additional features in jar");
return new String[] {"default"};
}
log.info("Without additional features in jar");
return new String[] {"default"};
}
private static boolean isClassPresent(String className) {
@@ -21,7 +21,28 @@ public class EndpointInterceptor implements HandlerInterceptor {
HttpServletRequest request, HttpServletResponse response, Object handler)
throws Exception {
String requestURI = request.getRequestURI();
boolean isEnabled = endpointConfiguration.isEndpointEnabledForUri(requestURI);
boolean isEnabled;
// Extract the specific endpoint name (e.g: /api/v1/general/remove-pages -> remove-pages)
if (requestURI.contains("/api/v1") && requestURI.split("/").length > 4) {
String[] requestURIParts = requestURI.split("/");
String requestEndpoint;
// Endpoint: /api/v1/convert/pdf/img becomes pdf-to-img
if ("convert".equals(requestURIParts[3]) && requestURIParts.length > 5) {
requestEndpoint = requestURIParts[4] + "-to-" + requestURIParts[5];
} else {
requestEndpoint = requestURIParts[4];
}
log.debug("Request endpoint: {}", requestEndpoint);
isEnabled = endpointConfiguration.isEndpointEnabled(requestEndpoint);
log.debug("Is endpoint enabled: {}", isEnabled);
} else {
isEnabled = endpointConfiguration.isEndpointEnabled(requestURI);
}
if (!isEnabled) {
response.sendError(HttpServletResponse.SC_FORBIDDEN, "This endpoint is disabled");
return false;
@@ -23,7 +23,6 @@ import lombok.RequiredArgsConstructor;
import stirling.software.SPDF.config.swagger.JsonDataResponse;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.AnalysisApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.model.api.PDFFile;
import stirling.software.common.service.CustomPDFDocumentFactory;
@@ -33,10 +32,7 @@ public class AnalysisController {
private final CustomPDFDocumentFactory pdfDocumentFactory;
@AutoJobPostMapping(
value = "/page-count",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/page-count", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get PDF page count",
@@ -47,10 +43,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/basic-info",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/basic-info", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get basic PDF information",
@@ -67,8 +60,7 @@ public class AnalysisController {
@AutoJobPostMapping(
value = "/document-properties",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get PDF document properties",
@@ -98,10 +90,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/page-dimensions",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/page-dimensions", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get page dimensions for all pages",
@@ -121,10 +110,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/form-fields",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/form-fields", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get form field information",
@@ -148,10 +134,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/annotation-info",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/annotation-info", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get annotation information",
@@ -176,10 +159,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/font-info",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/font-info", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get font information",
@@ -205,10 +185,7 @@ public class AnalysisController {
}
}
@AutoJobPostMapping(
value = "/security-info",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/security-info", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@JsonDataResponse
@Operation(
summary = "Get security information",
@@ -28,7 +28,6 @@ import lombok.RequiredArgsConstructor;
import stirling.software.SPDF.model.api.general.BookletImpositionRequest;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.GeneralUtils;
import stirling.software.common.util.TempFileManager;
@@ -45,8 +44,7 @@ public class BookletImpositionController {
@AutoJobPostMapping(
value = "/booklet-imposition",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.MEDIUM_WEIGHT)
consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@Operation(
summary = "Create a booklet with proper page imposition",
description =
@@ -26,7 +26,6 @@ import stirling.software.SPDF.config.EndpointConfiguration;
import stirling.software.SPDF.model.api.general.CropPdfForm;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.ExceptionUtils;
import stirling.software.common.util.GeneralUtils;
@@ -127,10 +126,7 @@ public class CropController {
return endpointConfiguration.isGroupEnabled("Ghostscript");
}
@AutoJobPostMapping(
value = "/crop",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
@AutoJobPostMapping(value = "/crop", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@Operation(
summary = "Crops a PDF document",
description =
@@ -25,7 +25,6 @@ import lombok.extern.slf4j.Slf4j;
import stirling.software.SPDF.model.api.EditTableOfContentsRequest;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.GeneralUtils;
import stirling.software.common.util.TempFileManager;
@@ -45,8 +44,7 @@ public class EditTableOfContentsController {
@AutoJobPostMapping(
value = "/extract-bookmarks",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@Operation(
summary = "Extract PDF Bookmarks",
description = "Extracts bookmarks/table of contents from a PDF document as JSON.")
@@ -149,8 +147,7 @@ public class EditTableOfContentsController {
@AutoJobPostMapping(
value = "/edit-table-of-contents",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.SMALL_WEIGHT)
consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@Operation(
summary = "Edit Table of Contents",
description = "Add or edit bookmarks/table of contents in a PDF document.")
@@ -1,333 +0,0 @@
package stirling.software.SPDF.controller.api;
import java.io.OutputStream;
import java.nio.file.Files;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashSet;
import java.util.List;
import java.util.Objects;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
import org.springframework.core.io.Resource;
import org.springframework.http.ResponseEntity;
import org.springframework.web.bind.WebDataBinder;
import org.springframework.web.bind.annotation.InitBinder;
import org.springframework.web.bind.annotation.ModelAttribute;
import org.springframework.web.multipart.MultipartFile;
import io.github.pixee.security.Filenames;
import io.swagger.v3.oas.annotations.Operation;
import lombok.RequiredArgsConstructor;
import lombok.extern.slf4j.Slf4j;
import stirling.software.SPDF.config.swagger.StandardPdfResponse;
import stirling.software.SPDF.model.api.general.EditTextRequest;
import stirling.software.SPDF.model.json.PdfJsonDocument;
import stirling.software.SPDF.model.json.PdfJsonPage;
import stirling.software.SPDF.model.json.PdfJsonTextElement;
import stirling.software.SPDF.service.PdfJsonConversionService;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.model.api.general.EditTextOperation;
import stirling.software.common.util.ExceptionUtils;
import stirling.software.common.util.GeneralUtils;
import stirling.software.common.util.TempFile;
import stirling.software.common.util.TempFileManager;
import stirling.software.common.util.WebResponseUtils;
import stirling.software.common.util.propertyeditor.StringToArrayListPropertyEditor;
/**
* Find/replace text editing for PDFs. Round-trips through {@link PdfJsonConversionService}: the
* input PDF is parsed into the editable JSON model, find/replace operations are applied to the
* {@code text} field of each text element, and the mutated model is rebuilt into a PDF.
*
* <p>Matching joins all text elements on a page into a single string before searching, so find
* strings can span multiple visual runs (titles split per word, kerning-broken phrases, etc.).
*
* <p>For cross-element matches the entire replacement is written into the first matched element,
* any intermediate elements are emptied, and the suffix of the last matched element is preserved.
* This anchors the new text at the leftmost match position and lets the font lay it out as one run,
* which is more visually reliable than redistributing words to the original per-element X positions
* (those positions are calibrated to the original word widths and rarely match the replacement
* widths). Centered or tracked layouts will become left-aligned at the original first-word position
* when their text is replaced - the trade-off is accepted in favour of avoiding glyph overlaps and
* gaps from variable-width word distribution.
*/
@Slf4j
@GeneralApi
@RequiredArgsConstructor
public class EditTextController {
private static final Pattern FILE_EXTENSION_PATTERN = Pattern.compile("[.][^.]+$");
private final PdfJsonConversionService pdfJsonConversionService;
private final TempFileManager tempFileManager;
@InitBinder
public void initBinder(WebDataBinder binder) {
binder.registerCustomEditor(
List.class,
"edits",
new StringToArrayListPropertyEditor<>(EditTextOperation.class));
}
@AutoJobPostMapping(consumes = "multipart/form-data", value = "/edit-text")
@StandardPdfResponse
@Operation(
summary = "Edit text in a PDF via find and replace",
description =
"Applies an ordered list of find/replace operations to the text in a PDF and"
+ " returns the edited PDF. Useful for find-and-replace, bulk renames"
+ " (e.g. updating a company name throughout a document), and copy"
+ " editing where the AI agent has identified specific replacements."
+ " Matching is performed against the joined text of each page, so"
+ " find strings can span multiple visual runs (titles split per word,"
+ " kerning-broken phrases). Cross-element matches are written as a"
+ " single replacement run anchored at the leftmost matched position;"
+ " centered or tracked text may shift left when its content changes."
+ " Input:PDF Output:PDF Type:SISO")
public ResponseEntity<Resource> editText(@ModelAttribute EditTextRequest request)
throws Exception {
MultipartFile inputFile = request.getFileInput();
if (inputFile == null) {
throw ExceptionUtils.createFileNullOrEmptyException();
}
List<EditTextOperation> edits = request.getEdits();
if (edits == null || edits.isEmpty()) {
throw ExceptionUtils.createIllegalArgumentException(
"error.editText.no.edits",
"No find/replace operations provided for text editing");
}
for (EditTextOperation edit : edits) {
if (edit == null || edit.getFind() == null || edit.getFind().isEmpty()) {
throw ExceptionUtils.createIllegalArgumentException(
"error.editText.empty.find", "Each edit must have a non-empty find string");
}
}
boolean wholeWordSearch = Boolean.TRUE.equals(request.getWholeWordSearch());
List<CompiledEdit> compiledEdits = compileEdits(edits, wholeWordSearch);
PdfJsonDocument document = pdfJsonConversionService.convertPdfToJsonDocument(inputFile);
Set<Integer> pageFilter = resolvePageFilter(request, document);
int modifiedSpans = applyEdits(document, compiledEdits, pageFilter);
log.info(
"edit-text: modified {} text span(s) using {} edit(s) on {} page(s)",
modifiedSpans,
compiledEdits.size(),
pageFilter == null
? document.getPages() == null ? 0 : document.getPages().size()
: pageFilter.size());
String docName = buildOutputFilename(inputFile);
TempFile tempOut = tempFileManager.createManagedTempFile(".pdf");
try (OutputStream os = Files.newOutputStream(tempOut.getPath())) {
pdfJsonConversionService.convertJsonToPdf(document, os);
} catch (Exception e) {
tempOut.close();
throw e;
}
return WebResponseUtils.pdfFileToWebResponse(tempOut, docName);
}
private List<CompiledEdit> compileEdits(
List<EditTextOperation> edits, boolean wholeWordSearch) {
return edits.stream().map(edit -> compileEdit(edit, wholeWordSearch)).toList();
}
private CompiledEdit compileEdit(EditTextOperation edit, boolean wholeWordSearch) {
// Always treat the user-supplied find string as a literal: Pattern.quote escapes any
// regex metacharacters, so the constructed pattern can only ever do a literal match
// (optionally bounded by our own word-boundary anchors). This rules out catastrophic
// backtracking from a malicious find string.
String findRaw = edit.getFind();
String replacement = Objects.toString(edit.getReplace(), "");
String regex = Pattern.quote(findRaw);
if (wholeWordSearch) {
// Use lookarounds rather than \b so the bound works even when the find string starts
// or ends with a non-word character (e.g. "-foo" or "$id"). \b only fires at a
// word/non-word transition, which would never trigger for a find that starts with a
// non-word char and was preceded by another non-word char in the source text.
regex = "(?<!\\w)(?:" + regex + ")(?!\\w)";
}
Pattern pattern = Pattern.compile(regex);
String safeReplacement = Matcher.quoteReplacement(replacement);
return new CompiledEdit(pattern, safeReplacement);
}
private Set<Integer> resolvePageFilter(EditTextRequest request, PdfJsonDocument document) {
String pageNumbers = request.getPageNumbers();
int totalPages = document.getPages() == null ? 0 : document.getPages().size();
if (totalPages == 0) {
return Collections.emptySet();
}
if (pageNumbers == null || pageNumbers.isBlank() || "all".equalsIgnoreCase(pageNumbers)) {
return null;
}
List<Integer> pages = GeneralUtils.parsePageList(pageNumbers, totalPages, true);
return new HashSet<>(pages);
}
private int applyEdits(
PdfJsonDocument document, List<CompiledEdit> edits, Set<Integer> pageFilter) {
List<PdfJsonPage> pages = document.getPages();
if (pages == null) {
return 0;
}
int modifiedSpans = 0;
// The page filter is built from parsePageList against pages.size(), so it returns 1-based
// positional page numbers. Match against the same positional numbering here rather than
// the per-page pageNumber field, which can become non-sequential after split/merge round
// trips and would misalign the filter.
for (int i = 0; i < pages.size(); i++) {
int positionalPageNumber = i + 1;
if (pageFilter != null && !pageFilter.contains(positionalPageNumber)) {
continue;
}
modifiedSpans += applyEditsToPage(pages.get(i), edits);
}
return modifiedSpans;
}
private int applyEditsToPage(PdfJsonPage page, List<CompiledEdit> edits) {
List<PdfJsonTextElement> elements = page.getTextElements();
if (elements == null || elements.isEmpty()) {
return 0;
}
Set<Integer> modifiedIndices = new HashSet<>();
for (CompiledEdit edit : edits) {
applyEditToPage(elements, edit, modifiedIndices);
}
for (Integer index : modifiedIndices) {
// Char codes were captured for the original glyph sequence; clear them so the rebuild
// re-encodes from the new text via the font.
elements.get(index).setCharCodes(null);
}
return modifiedIndices.size();
}
/**
* Apply a single edit across the page by matching against the concatenation of all element
* texts, then writing the replacement back into the originating element(s).
*/
private void applyEditToPage(
List<PdfJsonTextElement> elements, CompiledEdit edit, Set<Integer> modifiedIndices) {
StringBuilder joined = new StringBuilder();
int[] starts = new int[elements.size()];
int[] ends = new int[elements.size()];
for (int i = 0; i < elements.size(); i++) {
starts[i] = joined.length();
String text = elements.get(i).getText();
if (text != null) {
joined.append(text);
}
ends[i] = joined.length();
}
Matcher matcher = edit.pattern().matcher(joined);
List<MatchSpan> spans = new ArrayList<>();
StringBuffer interpolation = new StringBuffer();
int previousAppendPosition = 0;
while (matcher.find()) {
if (matcher.start() == matcher.end()) {
// Skip zero-length matches (e.g. /a*/ on empty input) — they cannot be applied.
continue;
}
int sizeBefore = interpolation.length();
matcher.appendReplacement(interpolation, edit.replacement());
int prefixLength = matcher.start() - previousAppendPosition;
String actualReplacement =
interpolation.substring(sizeBefore + prefixLength, interpolation.length());
spans.add(new MatchSpan(matcher.start(), matcher.end(), actualReplacement));
previousAppendPosition = matcher.end();
}
// Apply right-to-left so earlier match positions stay valid as we mutate elements.
for (int i = spans.size() - 1; i >= 0; i--) {
MatchSpan span = spans.get(i);
int firstElement = findElementForCharIndex(starts, ends, span.start());
int lastElement = findElementForCharIndex(starts, ends, span.end() - 1);
if (firstElement < 0 || lastElement < 0) {
continue;
}
applyMatchToElements(
elements, starts, span, firstElement, lastElement, modifiedIndices);
}
}
/**
* Find the element whose text covers the character at {@code charIndex} in the joined string.
* Returns -1 if no element covers that index (which should not happen for valid match spans).
*/
private static int findElementForCharIndex(int[] starts, int[] ends, int charIndex) {
for (int i = 0; i < starts.length; i++) {
if (starts[i] <= charIndex && charIndex < ends[i]) {
return i;
}
}
return -1;
}
private static void applyMatchToElements(
List<PdfJsonTextElement> elements,
int[] starts,
MatchSpan span,
int firstElement,
int lastElement,
Set<Integer> modifiedIndices) {
if (firstElement == lastElement) {
PdfJsonTextElement element = elements.get(firstElement);
String text = nullToEmpty(element.getText());
int matchStartInElement = span.start() - starts[firstElement];
int matchEndInElement = span.end() - starts[firstElement];
element.setText(
text.substring(0, matchStartInElement)
+ span.replacement()
+ text.substring(matchEndInElement));
modifiedIndices.add(firstElement);
return;
}
// Cross-element match: write the whole replacement into the first matched element, empty
// any intermediate elements, and keep only the suffix of the last matched element. The
// JSON->PDF rebuild concatenates per-token text, so the font lays out the replacement as
// one continuous run anchored at the first element's X position.
String firstText = nullToEmpty(elements.get(firstElement).getText());
int firstSplit = span.start() - starts[firstElement];
elements.get(firstElement).setText(firstText.substring(0, firstSplit) + span.replacement());
modifiedIndices.add(firstElement);
for (int mid = firstElement + 1; mid < lastElement; mid++) {
elements.get(mid).setText("");
modifiedIndices.add(mid);
}
String lastText = nullToEmpty(elements.get(lastElement).getText());
int lastSplit = span.end() - starts[lastElement];
elements.get(lastElement).setText(lastText.substring(lastSplit));
modifiedIndices.add(lastElement);
}
private static String nullToEmpty(String value) {
return value != null ? value : "";
}
private String buildOutputFilename(MultipartFile inputFile) {
String originalName = inputFile.getOriginalFilename();
String baseName =
(originalName != null && !originalName.isBlank())
? FILE_EXTENSION_PATTERN
.matcher(Filenames.toSimpleFileName(originalName))
.replaceFirst("")
: "document";
return baseName + "_edited.pdf";
}
private record CompiledEdit(Pattern pattern, String replacement) {}
private record MatchSpan(int start, int end, String replacement) {}
}
@@ -3,14 +3,13 @@ package stirling.software.SPDF.controller.api;
import java.io.File;
import java.io.IOException;
import java.io.InputStream;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Comparator;
import java.util.List;
import java.util.regex.Pattern;
import org.apache.pdfbox.multipdf.PDFMergerUtility;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDDocumentCatalog;
import org.apache.pdfbox.pdmodel.PDDocumentInformation;
@@ -40,7 +39,6 @@ import stirling.software.SPDF.config.swagger.StandardPdfResponse;
import stirling.software.SPDF.model.api.general.MergePdfsRequest;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.ExceptionUtils;
import stirling.software.common.util.GeneralUtils;
@@ -48,11 +46,6 @@ import stirling.software.common.util.PdfErrorUtils;
import stirling.software.common.util.TempFile;
import stirling.software.common.util.TempFileManager;
import stirling.software.common.util.WebResponseUtils;
import stirling.software.jpdfium.PdfDocument;
import stirling.software.jpdfium.PdfMerge;
import stirling.software.jpdfium.doc.Bookmark;
import stirling.software.jpdfium.doc.PdfBookmarkEditor;
import stirling.software.jpdfium.doc.PdfBookmarkEditor.BookmarkTree;
@GeneralApi
@Slf4j
@@ -63,6 +56,7 @@ public class MergeController {
private final CustomPDFDocumentFactory pdfDocumentFactory;
private final TempFileManager tempFileManager;
// Merges a list of PDDocument objects into a single PDDocument
public PDDocument mergeDocuments(List<PDDocument> documents) throws IOException {
PDDocument mergedDoc = pdfDocumentFactory.createNewDocument();
boolean success = false;
@@ -81,8 +75,11 @@ public class MergeController {
}
}
// Re-order files to match the explicit order provided by the front-end.
// fileOrder is newline-delimited original filenames in the desired order.
private static MultipartFile[] reorderFilesByProvidedOrder(
MultipartFile[] files, String fileOrder) {
// Split by various line endings and trim each entry
String[] desired =
stirling.software.common.util.RegexPatternUtils.getInstance()
.getNewlineSplitPattern()
@@ -109,6 +106,7 @@ public class MergeController {
return ordered.toArray(new MultipartFile[0]);
}
// Returns a comparator for sorting MultipartFile arrays based on the given sort type
private Comparator<MultipartFile> getSortComparator(String sortType) {
return switch (sortType) {
case "byFileName" ->
@@ -156,16 +154,18 @@ public class MergeController {
return 0;
}
};
case "orderProvided" -> (file1, file2) -> 0;
default -> (file1, file2) -> 0;
case "orderProvided" -> (file1, file2) -> 0; // Default is the order provided
default -> (file1, file2) -> 0; // Default is the order provided
};
}
// Parse client file IDs from JSON string
private String[] parseClientFileIds(String clientFileIds) {
if (clientFileIds == null || clientFileIds.trim().isEmpty()) {
return new String[0];
}
try {
// Simple JSON array parsing - remove brackets and split by comma
String trimmed = clientFileIds.trim();
if (trimmed.startsWith("[") && trimmed.endsWith("]")) {
String inside = trimmed.substring(1, trimmed.length() - 1).trim();
@@ -185,29 +185,39 @@ public class MergeController {
return new String[0];
}
// Adds a table of contents to the merged document using filenames as chapter titles
private void addTableOfContents(PDDocument mergedDocument, MultipartFile[] files) {
// Create the document outline
PDDocumentOutline outline = new PDDocumentOutline();
mergedDocument.getDocumentCatalog().setDocumentOutline(outline);
int pageIndex = 0;
int pageIndex = 0; // Current page index in the merged document
// Iterate through the original files
for (MultipartFile file : files) {
// Get the filename without extension to use as bookmark title
String filename = file.getOriginalFilename();
String title = GeneralUtils.removeExtension(filename);
// Create an outline item for this file
PDOutlineItem item = new PDOutlineItem();
item.setTitle(title);
// Set the destination to the first page of this file in the merged document
if (pageIndex < mergedDocument.getNumberOfPages()) {
PDPage page = mergedDocument.getPage(pageIndex);
item.setDestination(page);
}
// Add the item to the outline
outline.addLast(item);
// Increment page index for the next file
try (PDDocument doc = pdfDocumentFactory.load(file)) {
pageIndex += doc.getNumberOfPages();
} catch (IOException e) {
ExceptionUtils.logException("document loading for TOC generation", e);
pageIndex++;
pageIndex++; // Increment by at least one if we can't determine page count
}
}
}
@@ -225,6 +235,7 @@ public class MergeController {
}
}
// Fallback to XMP metadata if Info dates are missing
PDMetadata metadata = doc.getDocumentCatalog().getMetadata();
if (metadata != null) {
try (InputStream is = metadata.createInputStream()) {
@@ -260,10 +271,7 @@ public class MergeController {
return -1;
}
@AutoJobPostMapping(
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
value = "/merge-pdfs",
resourceWeight = ResourceWeight.MEDIUM_WEIGHT)
@AutoJobPostMapping(consumes = MediaType.MULTIPART_FORM_DATA_VALUE, value = "/merge-pdfs")
@StandardPdfResponse
@Operation(
summary = "Merge multiple PDF files into one",
@@ -275,7 +283,7 @@ public class MergeController {
@ModelAttribute MergePdfsRequest request,
@RequestParam(value = "fileOrder", required = false) String fileOrder)
throws IOException {
List<File> filesToDelete = new ArrayList<>();
List<File> filesToDelete = new ArrayList<>(); // List of temporary files to delete
TempFile outputTempFile = null;
boolean removeCertSign = Boolean.TRUE.equals(request.getRemoveCertSign());
@@ -286,35 +294,48 @@ public class MergeController {
files = new MultipartFile[0];
}
// If front-end provided explicit visible order, honor it and override backend sorting
if (fileOrder != null && !fileOrder.isBlank()) {
log.info("Reordering files based on fileOrder parameter");
files = reorderFilesByProvidedOrder(files, fileOrder);
} else {
log.info("Sorting files based on sortType: {}", request.getSortType());
Arrays.sort(files, getSortComparator(request.getSortType()));
Arrays.sort(
files,
getSortComparator(
request.getSortType())); // Sort files based on requested sort type
}
try (TempFile mt = new TempFile(tempFileManager, ".pdf")) {
List<Path> inputPaths = new ArrayList<>(files.length);
PDFMergerUtility mergerUtility = new PDFMergerUtility();
long totalSize = 0;
List<Integer> invalidIndexes = new ArrayList<>();
for (int index = 0; index < files.length; index++) {
MultipartFile multipartFile = files[index];
File tempFile = tempFileManager.convertMultipartFileToFile(multipartFile);
filesToDelete.add(tempFile);
inputPaths.add(tempFile.toPath());
totalSize += multipartFile.getSize();
File tempFile =
tempFileManager.convertMultipartFileToFile(
multipartFile); // Convert MultipartFile to File
filesToDelete.add(tempFile); // Add temp file to the list for later deletion
try (PdfDocument ignored = PdfDocument.open(tempFile.toPath())) {
} catch (Exception e) {
// Pre-validate each PDF so we can report which one(s) are broken
// Use the original MultipartFile to avoid deleting the tempFile during validation
try (PDDocument ignored = pdfDocumentFactory.load(multipartFile)) {
// OK
} catch (IOException e) {
ExceptionUtils.logException("PDF pre-validate", e);
invalidIndexes.add(index);
}
mergerUtility.addSource(tempFile); // Add source file to the merger utility
}
int[] pageCounts;
mergerUtility.setDestinationFileName(mt.getFile().getAbsolutePath());
try {
pageCounts =
mergeWithJpdfium(inputPaths, files, generateToc, mt.getFile().toPath());
mergerUtility.mergeDocuments(
pdfDocumentFactory.getStreamCacheFunction(
totalSize)); // Merge the documents
} catch (IOException e) {
ExceptionUtils.logException("PDF merge", e);
if (PdfErrorUtils.isCorruptedPdfError(e)) {
@@ -323,26 +344,10 @@ public class MergeController {
throw e;
}
boolean sigFlattenNeeded = false;
if (removeCertSign) {
try (PdfDocument check = PdfDocument.open(mt.getFile().toPath())) {
sigFlattenNeeded = !check.signatures().isEmpty();
} catch (Exception e) {
log.debug(
"JPDFium signature pre-check failed; falling back to PDFBox flatten:"
+ " {}",
e.getMessage());
sigFlattenNeeded = true;
}
if (!sigFlattenNeeded) {
log.info(
"removeCertSign requested but merged document has no signature"
+ " fields; skipping PDFBox flatten pass");
}
}
if (sigFlattenNeeded) {
try (PDDocument mergedDocument = pdfDocumentFactory.load(mt.getFile())) {
// Load the merged PDF document and operate on it inside try-with-resources
try (PDDocument mergedDocument = pdfDocumentFactory.load(mt.getFile())) {
// Remove signatures if removeCertSign is true
if (removeCertSign) {
PDDocumentCatalog catalog = mergedDocument.getDocumentCatalog();
PDAcroForm acroForm = catalog.getAcroForm();
if (acroForm != null) {
@@ -350,26 +355,24 @@ public class MergeController {
acroForm.getFields().stream()
.filter(PDSignatureField.class::isInstance)
.toList();
if (!fieldsToRemove.isEmpty()) {
acroForm.flatten(fieldsToRemove, false);
acroForm.flatten(
fieldsToRemove,
false); // Flatten the fields, effectively removing them
}
}
outputTempFile = new TempFile(tempFileManager, ".pdf");
try {
mergedDocument.save(outputTempFile.getFile());
} catch (Exception e) {
outputTempFile.close();
outputTempFile = null;
throw e;
}
}
} else {
// Add table of contents if generateToc is true
if (generateToc && files.length > 0) {
addTableOfContents(mergedDocument, files);
}
// Save the modified document to a temporary file
outputTempFile = new TempFile(tempFileManager, ".pdf");
try {
Files.copy(
mt.getFile().toPath(),
outputTempFile.getFile().toPath(),
java.nio.file.StandardCopyOption.REPLACE_EXISTING);
mergedDocument.save(outputTempFile.getFile());
} catch (Exception e) {
outputTempFile.close();
outputTempFile = null;
@@ -388,7 +391,7 @@ public class MergeController {
throw ex;
} finally {
for (File file : filesToDelete) {
tempFileManager.deleteTempFile(file);
tempFileManager.deleteTempFile(file); // Delete temporary files
}
}
@@ -398,113 +401,4 @@ public class MergeController {
return WebResponseUtils.pdfFileToWebResponse(outputTempFile, mergedFileName);
}
private int[] mergeWithJpdfium(
List<Path> inputPaths, MultipartFile[] files, boolean generateToc, Path outputPath)
throws IOException {
if (inputPaths.isEmpty()) {
try (PdfDocument empty = PdfDocument.open(new byte[0])) {
empty.save(outputPath);
} catch (Exception ignored) {
Files.write(outputPath, new byte[0]);
}
return new int[0];
}
List<PdfDocument> docs = new ArrayList<>(inputPaths.size());
int[] pageCounts = new int[inputPaths.size()];
int[] pageOffsets = new int[inputPaths.size()];
List<List<Bookmark>> sourceBookmarks = new ArrayList<>(inputPaths.size());
int runningOffset = 0;
try {
for (int i = 0; i < inputPaths.size(); i++) {
Path p = inputPaths.get(i);
PdfDocument doc = PdfDocument.open(p);
docs.add(doc);
pageCounts[i] = doc.pageCount();
pageOffsets[i] = runningOffset;
sourceBookmarks.add(doc.bookmarks());
runningOffset += pageCounts[i];
}
BookmarkTree combinedTree =
buildCombinedBookmarkTree(files, pageOffsets, sourceBookmarks, generateToc);
try (PdfDocument merged = PdfMerge.merge(docs)) {
if (combinedTree.entries().isEmpty()) {
merged.save(outputPath);
} else {
PdfBookmarkEditor.setBookmarks(merged, combinedTree, outputPath);
}
}
} catch (RuntimeException e) {
throw new IOException("JPDFium merge failed", e);
} finally {
for (PdfDocument doc : docs) {
try {
doc.close();
} catch (Exception ignored) {
}
}
}
return pageCounts;
}
private BookmarkTree buildCombinedBookmarkTree(
MultipartFile[] files,
int[] pageOffsets,
List<List<Bookmark>> sourceBookmarks,
boolean generateToc) {
BookmarkTree.Builder builder = BookmarkTree.builder();
if (generateToc) {
for (int i = 0; i < files.length; i++) {
String filename = files[i].getOriginalFilename();
String title = GeneralUtils.removeExtension(filename);
if (title == null || title.isBlank()) {
title = "Document " + (i + 1);
}
builder.add(title, pageOffsets[i]);
}
}
for (int i = 0; i < sourceBookmarks.size(); i++) {
int offset = pageOffsets[i];
for (Bookmark bm : sourceBookmarks.get(i)) {
addBookmarkFlat(builder, bm, offset);
}
}
return builder.build();
}
private void addBookmarkFlat(BookmarkTree.Builder builder, Bookmark root, int offset) {
final int maxNodes = 100_000;
java.util.Deque<Bookmark> stack = new java.util.ArrayDeque<>();
java.util.Set<Bookmark> visited =
java.util.Collections.newSetFromMap(new java.util.IdentityHashMap<>());
stack.push(root);
int processed = 0;
while (!stack.isEmpty() && processed < maxNodes) {
Bookmark bm = stack.pop();
if (!visited.add(bm)) {
continue;
}
processed++;
if (bm.isInternal() && bm.title() != null) {
builder.add(bm.title(), offset + bm.pageIndex());
}
if (bm.hasChildren()) {
List<Bookmark> children = bm.children();
for (int i = children.size() - 1; i >= 0; i--) {
stack.push(children.get(i));
}
}
}
if (processed >= maxNodes) {
log.warn(
"Source bookmark traversal hit {}-node cap; remaining bookmarks dropped",
maxNodes);
}
}
}
@@ -24,7 +24,6 @@ import lombok.extern.slf4j.Slf4j;
import stirling.software.SPDF.model.api.general.MergeMultiplePagesRequest;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.ExceptionUtils;
import stirling.software.common.util.GeneralFormCopyUtils;
@@ -42,8 +41,7 @@ public class MultiPageLayoutController {
@AutoJobPostMapping(
value = "/multi-page-layout",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.MEDIUM_WEIGHT)
consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@Operation(
summary = "Merge multiple pages of a PDF document into a single page",
description =
@@ -25,7 +25,6 @@ import stirling.software.SPDF.config.swagger.StandardPdfResponse;
import stirling.software.SPDF.model.api.general.OverlayPdfsRequest;
import stirling.software.common.annotations.AutoJobPostMapping;
import stirling.software.common.annotations.api.GeneralApi;
import stirling.software.common.enumeration.ResourceWeight;
import stirling.software.common.service.CustomPDFDocumentFactory;
import stirling.software.common.util.ExceptionUtils;
import stirling.software.common.util.GeneralUtils;
@@ -40,10 +39,7 @@ public class PdfOverlayController {
private final CustomPDFDocumentFactory pdfDocumentFactory;
private final TempFileManager tempFileManager;
@AutoJobPostMapping(
value = "/overlay-pdfs",
consumes = MediaType.MULTIPART_FORM_DATA_VALUE,
resourceWeight = ResourceWeight.MEDIUM_WEIGHT)
@AutoJobPostMapping(value = "/overlay-pdfs", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
@StandardPdfResponse
@Operation(
summary = "Overlay PDF files in various modes",

Some files were not shown because too many files have changed in this diff Show More