react-native-sherpa-onnx 0.3.4 → 0.3.6

package/README.md

@@ -96,7 +96,7 @@ To build the sherpa-onnx iOS XCFramework yourself (e.g. custom version or patche
 | Text-to-Speech | ✅ **Supported** | Multiple model types (VITS, Matcha, Kokoro, etc.). See [Supported Model Types](#supported-model-types) and [TTS documentation](./docs/tts.md). |
 | Streaming Text-to-Speech | ✅ **Supported** | Incremental speech generation for low time-to-first-byte and playback while generating. See [Streaming TTS](./docs/tts-streaming.md). |
 | Execution providers (CPU, NNAPI, XNNPACK, Core ML, QNN) | ✅ **Supported** | See [Execution provider support](./docs/execution-providers.md). |
-| Play Asset Delivery (PAD) | ✅ **Supported** | Android only. See [Model Setup](./docs/model-setup.md). |
+| Play Asset Delivery (PAD) | ✅ **Supported** | Android only. See [Model Setup](./docs/model-setup.md) & [Extraction API](./docs/extraction.md). |
 | Automatic Model type detection | ✅ **Supported** | `detectSttModel()` and `detectTtsModel()` for a path. See [Model Setup: Model type detection](./docs/model-setup.md#model-detection). |
 | Model quantization | ✅ **Supported** | Automatic detection and preference for quantized (int8) models. |
 | Flexible model loading | ✅ **Supported** | Asset models, file system models, or auto-detection. |
@@ -166,6 +166,7 @@ For **streaming TTS** (incremental generation, low latency), use `createStreamin
 - [Source Separation](./docs/separation.md)
 - [Model Setup](./docs/model-setup.md) – Bundled assets, Play Asset Delivery (PAD), model discovery APIs, and troubleshooting
 - [Model Download Manager](./docs/download-manager.md)
+- [Extraction API](./docs/extraction.md)
 - [Disable FFMPEG](./docs/disable-ffmpeg.md)
 - [Disable LIBARCHIVE](./docs/disable-libarchive.md)
 

package/SherpaOnnx.podspec

@@ -27,7 +27,13 @@ Pod::Spec.new do |s|
   s.source       = { :git => "https://github.com/XDcobra/react-native-sherpa-onnx.git", :tag => "#{s.version}" }
 
   s.source_files = ["ios/**/*.{h,m,mm,swift,cpp}"]
-  s.private_header_files = "ios/**/*.h"
+  # Exclude vendored framework headers from the compile/copy phases to avoid
+  # duplicate PrivateHeaders outputs when CocoaPods builds this pod as framework.
+  s.exclude_files = ["ios/Frameworks/**/*"]
+  private_headers = Dir.glob(File.join(pod_root, "ios", "**", "*.h")).reject do |path|
+    path.start_with?(File.join(pod_root, "ios", "Frameworks") + File::SEPARATOR)
+  end
+  s.private_header_files = private_headers.map { |path| path.sub("#{pod_root}/", "") }
 
   s.frameworks = "Foundation", "Accelerate", "CoreML", "AVFoundation", "AudioToolbox"
 
@@ -61,9 +67,13 @@ Pod::Spec.new do |s|
   libarchive_xcframework_root = File.join(pod_root, "ios", "Frameworks", "libarchive.xcframework")
   libarchive_simulator_headers = File.join(libarchive_xcframework_root, "ios-arm64_x86_64-simulator", "Headers")
   libarchive_device_headers = File.join(libarchive_xcframework_root, "ios-arm64", "Headers")
+  libarchive_simulator_slice = File.join(libarchive_xcframework_root, "ios-arm64_x86_64-simulator")
+  libarchive_device_slice = File.join(libarchive_xcframework_root, "ios-arm64")
 
   ffmpeg_simulator_headers = File.join(ffmpeg_xcframework, "ios-arm64_x86_64-simulator", "Headers")
   ffmpeg_device_headers = File.join(ffmpeg_xcframework, "ios-arm64", "Headers")
+  ffmpeg_simulator_slice = File.join(ffmpeg_xcframework, "ios-arm64_x86_64-simulator")
+  ffmpeg_device_slice = File.join(ffmpeg_xcframework, "ios-arm64")
 
   gcc_defs = '$(inherited) PLATFORM_CONFIG_H=\\"libarchive_darwin_config.h\\"'
   gcc_defs += ' HAVE_FFMPEG=1' if has_ffmpeg
@@ -97,22 +107,33 @@ Pod::Spec.new do |s|
     header_search_paths << "\"#{ffmpeg_simulator_headers}\""
   end
 
+  library_search_paths_ios = ["$(inherited)", "\"#{device_slice}\""]
+  library_search_paths_sim = ["$(inherited)", "\"#{simulator_slice}\""]
+  if has_ffmpeg
+    library_search_paths_ios << "\"#{ffmpeg_device_slice}\""
+    library_search_paths_sim << "\"#{ffmpeg_simulator_slice}\""
+  end
+  if has_libarchive
+    library_search_paths_ios << "\"#{libarchive_device_slice}\""
+    library_search_paths_sim << "\"#{libarchive_simulator_slice}\""
+  end
+
   s.pod_target_xcconfig = {
     "HEADER_SEARCH_PATHS" => header_search_paths.join(" "),
     "GCC_PREPROCESSOR_DEFINITIONS" => gcc_defs,
     "CLANG_CXX_LANGUAGE_STANDARD" => "c++17",
     "CLANG_CXX_LIBRARY" => "libc++",
     "OTHER_CPLUSPLUSFLAGS" => "$(inherited)",
-    "LIBRARY_SEARCH_PATHS[sdk=iphoneos*]" => "$(inherited) \"#{device_slice}\"",
-    "LIBRARY_SEARCH_PATHS[sdk=iphonesimulator*]" => "$(inherited) \"#{simulator_slice}\"",
+    "LIBRARY_SEARCH_PATHS[sdk=iphoneos*]" => library_search_paths_ios.join(" "),
+    "LIBRARY_SEARCH_PATHS[sdk=iphonesimulator*]" => library_search_paths_sim.join(" "),
     "OTHER_LDFLAGS" => ld_flags
   }
 
   s.user_target_xcconfig = {
     "CLANG_CXX_LANGUAGE_STANDARD" => "c++17",
     "CLANG_CXX_LIBRARY" => "libc++",
-    "LIBRARY_SEARCH_PATHS[sdk=iphoneos*]" => "$(inherited) \"#{device_slice}\"",
-    "LIBRARY_SEARCH_PATHS[sdk=iphonesimulator*]" => "$(inherited) \"#{simulator_slice}\"",
+    "LIBRARY_SEARCH_PATHS[sdk=iphoneos*]" => library_search_paths_ios.join(" "),
+    "LIBRARY_SEARCH_PATHS[sdk=iphonesimulator*]" => library_search_paths_sim.join(" "),
     "OTHER_LDFLAGS" => ld_flags
   }
 

package/android/prebuilt-download.gradle

@@ -14,7 +14,7 @@ def requiredFfmpegSoFiles = [
 def requiredSherpaOnnxSoFiles = [
   'libsherpa-onnx-jni.so', 'libsherpa-onnx-c-api.so', 'libsherpa-onnx-cxx-api.so', 'libonnxruntime.so'
 ]
-def requiredLibarchiveSoFiles = ['libarchive.so']
+def requiredLibarchiveSoFiles = ['libarchive.so', 'libzstd.so']
 def requiredOnnxruntimeJniSoFiles = ['libonnxruntime4j_jni.so']
 
 def jniLibsDir = file("${project.projectDir}/src/main/jniLibs")

package/android/prebuilt-versions.gradle

@@ -36,7 +36,7 @@ println "[react-native-sherpa-onnx] ffmpeg version (extracted/used): ${ffmpegVer
 def libarchiveVersion = System.getenv('LIBARCHIVE_VERSION')
 if (!libarchiveVersion) {
   def v = readVersionFromTagFile(new File(moduleRoot, 'third_party/libarchive_prebuilt/ANDROID_RELEASE_TAG'), 'libarchive-android-v')
-  libarchiveVersion = v ?: (project.hasProperty('libarchiveVersion') ? project.libarchiveVersion : '3.8.5')
+  libarchiveVersion = v ?: (project.hasProperty('libarchiveVersion') ? project.libarchiveVersion : '3.8.5-2')
 }
 project.ext.libarchiveVersion = libarchiveVersion
 println "[react-native-sherpa-onnx] libarchive version (extracted/used): ${libarchiveVersion}"

package/android/src/main/cpp/jni/archive/sherpa-onnx-archive-helper.cpp

@@ -75,6 +75,38 @@ static void DrainRemainingAndClose(ArchiveReadContext* ctx) {
   fclose(ctx->file);
   ctx->file = nullptr;
 }
+
+struct StreamReadContext {
+  std::array<unsigned char, 64 * 1024> buffer{};
+  Sha256Context sha_ctx{};
+  long long bytes_read = 0;
+  ArchiveHelper::StreamReadCallback read_cb = nullptr;
+  void* user_data = nullptr;
+};
+
+static la_ssize_t ArchiveStreamReadCallback(struct archive* archive, void* client_data, const void** buff) {
+  auto* ctx = static_cast<StreamReadContext*>(client_data);
+  if (!ctx || !ctx->read_cb) {
+    archive_set_error(archive, EINVAL, "Invalid stream read context");
+    return -1;
+  }
+  std::ptrdiff_t n = ctx->read_cb(ctx->buffer.data(), ctx->buffer.size(), ctx->user_data);
+  if (n > 0) {
+    sha256_update(&ctx->sha_ctx, ctx->buffer.data(), static_cast<size_t>(n));
+    ctx->bytes_read += static_cast<long long>(n);
+    *buff = ctx->buffer.data();
+    return static_cast<la_ssize_t>(n);
+  }
+  if (n == 0) return 0;
+  archive_set_error(archive, EINVAL, "Stream read error");
+  return -1;
+}
+
+// No-op close callback for stream mode: the stream lifetime is managed by the
+// caller (e.g. a JNI InputStream), so libarchive must not close it.
+static int ArchiveStreamCloseCallback(struct archive* /* archive */, void* /* client_data */) {
+  return ARCHIVE_OK;
+}
 #endif  // HAVE_LIBARCHIVE
 
 static std::string ToHex(const unsigned char* data, size_t size) {
@@ -122,13 +154,15 @@ bool ArchiveHelper::ExtractTarBz2(
     return false;
   }
 
-  // Check target directory
+  // If target exists and is a directory, extract into it (merge). Otherwise require empty or force-remove.
   if (std::filesystem::exists(target_path)) {
-    if (force) {
+    if (std::filesystem::is_directory(target_path)) {
+      // Merge: extract into existing directory (e.g. multiple archives → same base path)
+    } else if (force) {
       std::error_code ec;
       std::filesystem::remove_all(target_path, ec);
       if (ec) {
-        if (out_error) *out_error = "Failed to remove target directory: " + ec.message();
+        if (out_error) *out_error = "Failed to remove target path: " + ec.message();
         return false;
       }
     } else {
@@ -137,7 +171,6 @@ bool ArchiveHelper::ExtractTarBz2(
     }
   }
 
-  // Create target directory
   std::error_code ec;
   std::filesystem::create_directories(target_path, ec);
   if (ec) {
@@ -167,11 +200,12 @@ bool ArchiveHelper::ExtractTarBz2(
     return false;
   }
 
-  // Configure archive to support tar and bzip2
+  // Configure archive to support tar and common compression (bzip2, gzip, xz, zstd)
   archive_read_support_format_tar(archive);
   archive_read_support_filter_bzip2(archive);
   archive_read_support_filter_gzip(archive);  // Also support gzip for compatibility
-  archive_read_support_filter_xz(archive);    // And xz
+  archive_read_support_filter_xz(archive);   // And xz
+  archive_read_support_filter_zstd(archive); // And zstd (.tar.zst)
 
   ArchiveReadContext read_ctx;
   read_ctx.file = fopen(source_path.c_str(), "rb");
@@ -354,6 +388,18 @@ bool ArchiveHelper::ExtractTarBz2(
       close_reader();
       return false;
     }
+
+    result = archive_write_finish_entry(disk);
+    if (result != ARCHIVE_OK && result != ARCHIVE_WARN) {
+      const char* err = archive_error_string(disk);
+      if (out_error) {
+        *out_error = err ? std::string("Failed to finish entry: ") + err : "Failed to finish entry";
+      }
+      archive_read_free(archive);
+      archive_write_free(disk);
+      close_reader();
+      return false;
+    }
   }
 
   archive_read_free(archive);
@@ -376,6 +422,260 @@ bool ArchiveHelper::ExtractTarBz2(
 #endif  // HAVE_LIBARCHIVE
 }
 
+bool ArchiveHelper::ExtractTarZst(
+    const std::string& source_path,
+    const std::string& target_path,
+    bool force,
+    std::function<void(long long, long long, double)> on_progress,
+    std::string* out_error,
+    std::string* out_sha256) {
+  return ExtractTarBz2(source_path, target_path, force, on_progress, out_error, out_sha256);
+}
+
+bool ArchiveHelper::ExtractFromStream(
+    StreamReadCallback read_cb,
+    void* read_user_data,
+    const std::string& target_path,
+    bool force,
+    std::function<void(long long, long long, double)> on_progress,
+    std::string* out_error,
+    std::string* out_sha256) {
+  cancel_requested_.store(false);
+
+#ifndef HAVE_LIBARCHIVE
+  (void)read_cb;
+  (void)read_user_data;
+  (void)target_path;
+  (void)force;
+  (void)on_progress;
+  (void)out_sha256;
+  if (out_error) *out_error = "libarchive not available. Build with libarchive or set sherpaOnnxDisableLibarchive=false in gradle.properties. See docs/disable-libarchive.md.";
+  return false;
+#else
+  if (!read_cb) {
+    if (out_error) *out_error = "Stream read callback is null";
+    return false;
+  }
+
+  if (std::filesystem::exists(target_path)) {
+    if (std::filesystem::is_directory(target_path)) {
+      // Merge: extract into existing directory (e.g. multiple archives → same base path)
+    } else if (force) {
+      std::error_code ec;
+      std::filesystem::remove_all(target_path, ec);
+      if (ec) {
+        if (out_error) *out_error = "Failed to remove target path: " + ec.message();
+        return false;
+      }
+    } else {
+      if (out_error) *out_error = "Target path already exists";
+      return false;
+    }
+  }
+
+  std::error_code ec;
+  std::filesystem::create_directories(target_path, ec);
+  if (ec) {
+    if (out_error) *out_error = "Failed to create target directory: " + ec.message();
+    return false;
+  }
+
+#ifndef NDEBUG
+  __android_log_print(ANDROID_LOG_INFO, "SherpaOnnx",
+                      "ExtractFromStream target_path=%s", target_path.c_str());
+#endif
+
+  std::string canonical_target = std::filesystem::canonical(target_path).string();
+  if (canonical_target.back() != '/') canonical_target += '/';
+
+  const long long total_bytes = 0;
+
+  struct archive* archive = archive_read_new();
+  if (!archive) {
+    if (out_error) *out_error = "Failed to create archive reader";
+    return false;
+  }
+
+  archive_read_support_format_tar(archive);
+  archive_read_support_filter_bzip2(archive);
+  archive_read_support_filter_gzip(archive);
+  archive_read_support_filter_xz(archive);
+  archive_read_support_filter_zstd(archive);
+
+  StreamReadContext stream_ctx;
+  stream_ctx.read_cb = read_cb;
+  stream_ctx.user_data = read_user_data;
+  sha256_init(&stream_ctx.sha_ctx);
+
+  if (archive_read_open(archive, &stream_ctx, nullptr, ArchiveStreamReadCallback, ArchiveStreamCloseCallback) != ARCHIVE_OK) {
+    const char* err = archive_error_string(archive);
+    if (out_error) *out_error = err ? std::string("Failed to open archive: ") + err : "Failed to open archive";
+    archive_read_free(archive);
+    return false;
+  }
+
+  struct archive* disk = archive_write_disk_new();
+  if (!disk) {
+    if (out_error) *out_error = "Failed to create disk writer";
+    archive_read_free(archive);
+    return false;
+  }
+
+  archive_write_disk_set_options(disk,
+                                  ARCHIVE_EXTRACT_TIME |
+                                  ARCHIVE_EXTRACT_PERM |
+                                  ARCHIVE_EXTRACT_ACL |
+                                  ARCHIVE_EXTRACT_FFLAGS);
+  archive_write_disk_set_standard_lookup(disk);
+
+  struct archive_entry* entry = nullptr;
+  int result = ARCHIVE_OK;
+  long long extracted_bytes = 0;
+  int last_percent = -1;
+  long long last_emit_bytes = 0;
+  int entry_index = 0;
+
+  while ((result = archive_read_next_header(archive, &entry)) == ARCHIVE_OK) {
+    if (cancel_requested_.load()) {
+      if (out_error) *out_error = "Extraction cancelled";
+      archive_read_free(archive);
+      archive_write_free(disk);
+      return false;
+    }
+
+    const char* current_path = archive_entry_pathname(entry);
+    if (!current_path) {
+      archive_read_free(archive);
+      archive_write_free(disk);
+      if (out_error) *out_error = "Invalid entry path";
+      return false;
+    }
+
+    std::string entry_path(current_path);
+    std::string full_path = target_path;
+    if (full_path.back() != '/') full_path += '/';
+    full_path += entry_path;
+
+    std::string canonical_entry;
+    try {
+      std::filesystem::path p(full_path);
+      std::filesystem::path parent = p.parent_path();
+      if (std::filesystem::exists(parent)) {
+        canonical_entry = std::filesystem::canonical(parent).string();
+      } else {
+        while (!std::filesystem::exists(parent) && parent != parent.parent_path()) {
+          parent = parent.parent_path();
+        }
+        if (std::filesystem::exists(parent)) {
+          canonical_entry = std::filesystem::canonical(parent).string();
+        } else {
+          canonical_entry = canonical_target;
+        }
+      }
+      canonical_entry += '/';
+      canonical_entry += p.filename().string();
+    } catch (const std::exception&) {
+      canonical_entry = full_path;
+    }
+
+    if (canonical_entry.find(canonical_target) != 0) {
+      archive_read_free(archive);
+      archive_write_free(disk);
+      if (out_error) *out_error = "Blocked path traversal: " + entry_path;
+      return false;
+    }
+
+    archive_entry_set_pathname(entry, full_path.c_str());
+
+    result = archive_write_header(disk, entry);
+    if (result != ARCHIVE_OK) {
+      const char* err = archive_error_string(disk);
+      if (out_error) *out_error = err ? std::string("Failed to write entry: ") + err : "Failed to write entry";
+      archive_read_free(archive);
+      archive_write_free(disk);
+      return false;
+    }
+
+    const void* buff = nullptr;
+    size_t size = 0;
+    la_int64_t offset = 0;
+
+    while ((result = archive_read_data_block(archive, &buff, &size, &offset)) == ARCHIVE_OK) {
+      if (cancel_requested_.load()) {
+        if (out_error) *out_error = "Extraction cancelled";
+        archive_read_free(archive);
+        archive_write_free(disk);
+        return false;
+      }
+
+      result = archive_write_data_block(disk, buff, size, offset);
+      if (result != ARCHIVE_OK) {
+        const char* err = archive_error_string(disk);
+        if (out_error) *out_error = err ? std::string("Failed to write data: ") + err : "Failed to write data";
+        archive_read_free(archive);
+        archive_write_free(disk);
+        return false;
+      }
+
+      extracted_bytes += static_cast<long long>(size);
+
+      if (on_progress) {
+        if (total_bytes > 0) {
+          long long compressed_bytes = archive_filter_bytes(archive, -1);
+          int percent = static_cast<int>(total_bytes > 0 ? (compressed_bytes * 100) / total_bytes : 0);
+          percent = (percent > 100) ? 100 : ((percent < 0) ? 0 : percent);
+          if (percent != last_percent) {
+            last_percent = percent;
+            on_progress(compressed_bytes, total_bytes, static_cast<double>(percent));
+          }
+        } else if (stream_ctx.bytes_read - last_emit_bytes >= 1024 * 1024) {
+          last_emit_bytes = stream_ctx.bytes_read;
+          on_progress(stream_ctx.bytes_read, 0, 0.0);
+        }
+      }
+    }
+
+    if (result != ARCHIVE_EOF && result != ARCHIVE_OK) {
+      const char* err = archive_error_string(archive);
+      if (out_error) *out_error = err ? std::string("Failed to read data: ") + err : "Failed to read data";
+      archive_read_free(archive);
+      archive_write_free(disk);
+      return false;
+    }
+
+    result = archive_write_finish_entry(disk);
+    if (result != ARCHIVE_OK && result != ARCHIVE_WARN) {
+      const char* err = archive_error_string(disk);
+      if (out_error) *out_error = err ? std::string("Failed to finish entry: ") + err : "Failed to finish entry";
+      archive_read_free(archive);
+      archive_write_free(disk);
+      return false;
+    }
+    entry_index++;
+  }
+
+#ifndef NDEBUG
+  __android_log_print(ANDROID_LOG_INFO, "SherpaOnnx",
+                      "ExtractFromStream done entries=%d", entry_index);
+#endif
+
+  archive_read_free(archive);
+  archive_write_free(disk);
+
+  if (out_sha256) {
+    unsigned char digest[32];
+    sha256_final(&stream_ctx.sha_ctx, digest);
+    *out_sha256 = ToHex(digest, sizeof(digest));
+  }
+
+  if (on_progress) {
+    on_progress(stream_ctx.bytes_read, stream_ctx.bytes_read, 100.0);
+  }
+
+  return true;
+#endif
+}
+
 bool ArchiveHelper::ComputeFileSha256(
     const std::string& file_path,
     std::string* out_error,

package/android/src/main/cpp/jni/archive/sherpa-onnx-archive-helper.h

@@ -1,5 +1,6 @@
 #pragma once
 
+#include <cstddef>
 #include <string>
 #include <functional>
 #include <atomic>
@@ -10,14 +11,17 @@
  */
 class ArchiveHelper {
  public:
+  /** Callback to read bytes from a stream (e.g. Java InputStream via JNI). Returns bytes read, 0 on EOF, -1 on error. */
+  using StreamReadCallback = std::ptrdiff_t (*)(void* buf, size_t len, void* user_data);
+
   /**
-   * Extract tar.bz2 file to target directory
+   * Extract tar.bz2 (or .tar.zst, .tar.gz, .tar.xz) file to target directory.
    *
-   * @param sourcePath Path to the .tar.bz2 file
+   * @param sourcePath Path to the archive file
    * @param targetPath Destination directory path
    * @param force Whether to overwrite existing target directory
    * @param onProgress Callback for progress updates (bytesExtracted, totalBytes, percent)
-    * @param outSha256 Optional output SHA-256 hex of the archive file
+   * @param outSha256 Optional output SHA-256 hex of the archive file
    * @return true if extraction succeeded, false otherwise
    */
   static bool ExtractTarBz2(
@@ -28,7 +32,32 @@ class ArchiveHelper {
       std::string* out_error = nullptr,
       std::string* out_sha256 = nullptr);
 
-    /**
+  /**
+   * Extract .tar.zst (or other supported tar compression) file to target directory.
+   * Uses the same implementation as ExtractTarBz2 (libarchive supports zstd when built with ENABLE_ZSTD).
+   */
+  static bool ExtractTarZst(
+      const std::string& source_path,
+      const std::string& target_path,
+      bool force,
+      std::function<void(long long, long long, double)> on_progress = nullptr,
+      std::string* out_error = nullptr,
+      std::string* out_sha256 = nullptr);
+
+  /**
+   * Extract a tar archive (tar.zst or tar.bz2) from a stream via read_cb.
+   * Used for Android AssetManager streams; total_bytes can be 0 (progress then uses compressed bytes or periodic emit).
+   */
+  static bool ExtractFromStream(
+      StreamReadCallback read_cb,
+      void* read_user_data,
+      const std::string& target_path,
+      bool force,
+      std::function<void(long long, long long, double)> on_progress = nullptr,
+      std::string* out_error = nullptr,
+      std::string* out_sha256 = nullptr);
+
+  /**
      * Compute SHA-256 of a file.
      *
      * @param file_path Path to the file