feat: add subtitle field to XMP metadata parser, refactor for modern Java

.github/workflows/ci.yml

@@ -100,7 +100,7 @@ jobs:
             platform: darwin-arm64
             arch: arm64
             skipTests: false
-          - os: windows-2022
+          - os: windows-2025
             platform: windows-x64
             arch: x64
             skipTests: true

.github/workflows/publish.yml

@@ -94,7 +94,7 @@ jobs:
           - os: macos-15
             platform: darwin-arm64
             arch: arm64
-          - os: windows-2022
+          - os: windows-2025
             platform: windows-x64
             arch: x64
 

build.gradle.kts

@@ -15,7 +15,7 @@ plugins {
 
 allprojects {
     group = "org.grimmory"
-    version = "1.1.4"
+    version = "1.2.0"
 
     repositories {
         mavenCentral()
@@ -114,7 +114,8 @@ if (!enableCorpusTools.get()) {
 
 tasks.withType<JavaCompile> {
     options.compilerArgs.addAll(listOf(
-        "--enable-preview"
+        "--enable-preview",
+        "-Xlint:-preview"
     ))
 }
 
@@ -137,8 +138,21 @@ val pdfiumPlatforms = mapOf(
 )
 
 val platformFilter = findProperty("pdfiumPlatformFilter")?.toString()
+val hostOs = System.getProperty("os.name").lowercase()
+val hostArch = System.getProperty("os.arch").lowercase()
+val hostPlatform = when {
+    hostOs.contains("mac") && (hostArch == "aarch64" || hostArch == "arm64") -> "darwin-arm64"
+    hostOs.contains("mac") && (hostArch == "x86_64" || hostArch == "amd64") -> "darwin-x64"
+    hostOs.contains("linux") && (hostArch == "aarch64" || hostArch == "arm64") -> "linux-arm64"
+    hostOs.contains("linux") && (hostArch == "x86_64" || hostArch == "amd64") -> "linux-x64"
+    hostOs.contains("windows") && (hostArch == "x86_64" || hostArch == "amd64") -> "windows-x64"
+    else -> null
+}
+
 val activePlatforms = if (platformFilter != null) {
     pdfiumPlatforms.filterKeys { it == platformFilter }
+} else if (System.getenv("CI") == null && hostPlatform != null) {
+    pdfiumPlatforms.filterKeys { it == hostPlatform }
 } else {
     pdfiumPlatforms
 }
@@ -150,17 +164,6 @@ val prebuiltShimsDir = findProperty("prebuiltShimsDir")?.toString()?.let { proje
 val pdfiumArchiveDir = layout.buildDirectory.dir("pdfium-archives")
 val pdfiumNativesDir = layout.buildDirectory.dir("generated-natives")
 
-val hostOs = System.getProperty("os.name").lowercase()
-val hostArch = System.getProperty("os.arch").lowercase()
-val hostPlatform = when {
-    hostOs.contains("mac") && (hostArch == "aarch64" || hostArch == "arm64") -> "darwin-arm64"
-    hostOs.contains("mac") && (hostArch == "x86_64" || hostArch == "amd64") -> "darwin-x64"
-    hostOs.contains("linux") && (hostArch == "aarch64" || hostArch == "arm64") -> "linux-arm64"
-    hostOs.contains("linux") && (hostArch == "x86_64" || hostArch == "amd64") -> "linux-x64"
-    hostOs.contains("windows") && (hostArch == "x86_64" || hostArch == "amd64") -> "windows-x64"
-    else -> null
-}
-
 val downloadPdfiumBinaries by tasks.registering {
     description = "Downloads prebuilt PDFium binaries for all supported platforms"
     outputs.dir(pdfiumArchiveDir)
@@ -549,6 +552,7 @@ tasks.assemble {
 
 tasks.withType<Test> {
     useJUnitPlatform()
+    maxHeapSize = "2g"
     testLogging {
         showStandardStreams = true
         events("passed", "skipped", "failed")
@@ -645,7 +649,7 @@ tasks.register<JavaExec>("runCorpusProcessor") {
     group = "application"
     description = "Runs the CorpusProcessor to write metadata to PDFs"
     if (enableCorpusTools.get()) {
-        dependsOn("buildShim")
+        dependsOn("generateNativeIndex")
     }
     mainClass.set("org.grimmory.pdfium4j.CorpusProcessor")
     classpath = sourceSets["test"].runtimeClasspath
@@ -682,7 +686,7 @@ tasks.register<JavaExec>("runCorpusMetadataStress") {
     group = "application"
     description = "Runs metadata save stress validation against corpus PDFs"
     if (enableCorpusTools.get()) {
-        dependsOn("buildShim")
+        dependsOn("generateNativeIndex")
     }
     mainClass.set("org.grimmory.pdfium4j.CorpusMetadataStressRunner")
     classpath = sourceSets["test"].runtimeClasspath

shim/CMakeLists.txt

@@ -1,8 +1,8 @@
 # PDFium4j Native Shim Build
-cmake_minimum_required(VERSION 3.15)
+cmake_minimum_required(VERSION 3.20)
 project(pdfium4j_shim VERSION 1.0.1 LANGUAGES CXX)
 
-set(CMAKE_CXX_STANDARD 17)
+set(CMAKE_CXX_STANDARD 23)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)
 set(CMAKE_POSITION_INDEPENDENT_CODE ON)
 set(CMAKE_POSITION_INDEPENDENT_CODE ON CACHE BOOL "" FORCE)
@@ -14,6 +14,13 @@ endif()
 # Suppress QPDF PointerHolder transition warnings (and fix MSVC #warning error)
 add_definitions(-DPOINTERHOLDER_TRANSITION=0)
 
+if(WIN32)
+    # Prevent windows.h from defining min/max macros that collide with C++ stdlib
+    add_definitions(-DNOMINMAX)
+    # Reduce the amount of Windows header bloat
+    add_definitions(-DWIN32_LEAN_AND_MEAN)
+endif()
+
 # QPDF dependency hints
 if(DEFINED ENV{ZLIB_ROOT})
     set(ZLIB_ROOT "$ENV{ZLIB_ROOT}")

shim/include/pdfium4j_shim.h

@@ -22,6 +22,15 @@
 extern "C" {
 #endif
 
+typedef struct {
+    int charCode;
+    float left;
+    float bottom;
+    float right;
+    float top;
+    float fontSize;
+} pdfium4j_char_info_t;
+
 SHIM_EXPORT int FPDF_CALLCONV pdfium4j_page_count(FPDF_DOCUMENT doc);
 
 SHIM_EXPORT int FPDF_CALLCONV pdfium4j_get_meta_utf8(FPDF_DOCUMENT doc, const char* key, char* buf, int buf_len);
@@ -128,6 +137,16 @@ SHIM_EXPORT int FPDF_CALLCONV pdfium4j_save_with_metadata_mem_native(
     int metadata_count
 );
 
+SHIM_EXPORT int FPDF_CALLCONV pdfium4j_save_with_metadata_mem_to_file_native(
+    const void* src_buf,
+    size_t      src_len,
+    const char* dst_path,
+    const char* xmp_metadata,
+    int xmp_len,
+    const char** metadata_pairs,
+    int metadata_count
+);
+
 /**
  * QPDF-based metadata enumeration
  */
@@ -146,6 +165,21 @@ SHIM_EXPORT int FPDF_CALLCONV pdfium4j_read_info_dict_mem(
     void*                userdata
 );
 
+SHIM_EXPORT int FPDF_CALLCONV pdfium4j_get_xmp_qpdf(
+    const char* src_path,
+    char*       buf,
+    int         buf_len
+);
+
+SHIM_EXPORT int FPDF_CALLCONV pdfium4j_get_xmp_qpdf_mem(
+    const char* data,
+    size_t      len,
+    char*       buf,
+    int         buf_len
+);
+
+SHIM_EXPORT void FPDF_CALLCONV pdfium4j_resolve_optional_symbols(void);
+
 #ifdef __cplusplus
 }
 #endif