Fix ffmpeg c api example (#1185)

k2-fsa · Jul 29, 2024 · b1711ec · b1711ec
1 parent 646f99c
commit b1711ec
Show file tree

Hide file tree

Showing 4 changed files with 30 additions and 12 deletions.
diff --git a/.github/workflows/c-api.yaml b/.github/workflows/c-api.yaml
@@ -13,6 +13,7 @@ on:
       - 'sherpa-onnx/csrc/*'
       - 'sherpa-onnx/c-api/*'
       - 'c-api-examples/**'
+      - 'ffmpeg-examples/**'
   pull_request:
     branches:
       - master
@@ -23,6 +24,7 @@ on:
       - 'sherpa-onnx/csrc/*'
       - 'sherpa-onnx/c-api/*'
       - 'c-api-examples/**'
+      - 'ffmpeg-examples/**'
 
   workflow_dispatch:
 
@@ -81,6 +83,22 @@ jobs:
             otool -L ./install/lib/libsherpa-onnx-c-api.dylib
           fi
 
+      - name: Test ffmpeg
+        if: matrix.os == 'macos-latest'
+        shell: bash
+        run: |
+          brew install ffmpeg
+
+          cd ffmpeg-examples
+          curl -SL -O https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20.tar.bz2
+          tar xvf sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20.tar.bz2
+          rm sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20.tar.bz2
+
+          make
+          ls -lh
+          ./run.sh
+          rm -rf sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20
+
       - name: Test sense-voice
         shell: bash
         run: |

diff --git a/ffmpeg-examples/Makefile b/ffmpeg-examples/Makefile
@@ -14,19 +14,18 @@ ifeq ($(GDB), TRUE)
 	OPTFLAG += -g
 endif
 
-# CFLAGS := $(shell pkg-config --cflags $(SHARED_LIBS)) -I.. -Wall -std=c++11 -fopenmp ${OPTFLAG}
-CFLAGS := $(shell pkg-config --cflags $(SHARED_LIBS)) -I.. -Wall -std=c++11 ${OPTFLAG}
+# CFLAGS := $(shell pkg-config --cflags $(SHARED_LIBS)) -I.. -Wall -std=c++17 -fopenmp ${OPTFLAG}
+CFLAGS := $(shell pkg-config --cflags $(SHARED_LIBS)) -I.. -Wall -std=c++17  ${OPTFLAG}
 LDLIBS := $(shell pkg-config --libs $(SHARED_LIBS))
 
 CUR_DIR :=$(shell pwd)
 
 LDLIBS += -L ../build/lib
 LDLIBS += -L ../build/_deps/onnxruntime-src/lib
-LDLIBS += -lsherpa-onnx-c-api -lsherpa-onnx-core -lonnxruntime -lkaldi-native-fbank-core
+LDLIBS += -lsherpa-onnx-c-api -lonnxruntime
 LDLIBS += -Wl,-rpath,${CUR_DIR}/../build/lib
 LDLIBS += -Wl,-rpath,${CUR_DIR}/../build/_deps/onnxruntime-src/lib
 
-
 #Get libavutil version and extract major, minor and micro
 LIBAVUTIL_VERSION := $(shell pkg-config --modversion libavutil)
 LIBAVUTIL_MAJOR := $(shell echo "$(LIBAVUTIL_VERSION)" | awk -F. '{print $$1}')

diff --git a/ffmpeg-examples/run.sh b/ffmpeg-examples/run.sh
@@ -45,4 +45,4 @@ echo "Decoding a URL"
   ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/encoder-epoch-99-avg-1.onnx \
   ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/decoder-epoch-99-avg-1.onnx \
   ./sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/joiner-epoch-99-avg-1.onnx \
-  https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/resolve/main/test_wavs/4.wav
+  https://huggingface.co/csukuangfj/sherpa-onnx-streaming-zipformer-bilingual-zh-en-2023-02-20/resolve/main/test_wavs/3.wav
diff --git a/ffmpeg-examples/sherpa-onnx-ffmpeg.c b/ffmpeg-examples/sherpa-onnx-ffmpeg.c
@@ -216,7 +216,7 @@ static int init_filters(const char *filters_descr) {
 static void sherpa_decode_frame(const AVFrame *frame,
                                 SherpaOnnxOnlineRecognizer *recognizer,
                                 SherpaOnnxOnlineStream *stream,
-                                SherpaOnnxDisplay *display,
+                                const SherpaOnnxDisplay *display,
                                 int32_t *segment_id) {
 #define N 3200  // 100s. Sample rate is fixed to 16 kHz
   static float samples[N];
@@ -229,7 +229,7 @@ static void sherpa_decode_frame(const AVFrame *frame,
       SherpaOnnxDecodeOnlineStream(recognizer, stream);
     }
 
-    SherpaOnnxOnlineRecognizerResult *r =
+    const SherpaOnnxOnlineRecognizerResult *r =
         SherpaOnnxGetOnlineStreamResult(recognizer, stream);
     if (strlen(r->text)) {
       SherpaOnnxPrint(display, *segment_id, r->text);
@@ -290,10 +290,11 @@ int main(int argc, char **argv) {
   }
 
   SherpaOnnxOnlineRecognizerConfig config;
+	memset(&config, 0, sizeof(config));
   config.model_config.tokens = argv[1];
-  config.model_config.encoder = argv[2];
-  config.model_config.decoder = argv[3];
-  config.model_config.joiner = argv[4];
+  config.model_config.transducer.encoder = argv[2];
+  config.model_config.transducer.decoder = argv[3];
+  config.model_config.transducer.joiner = argv[4];
 
   if (argc == 7 && atoi(argv[6]) > 0) {
     num_threads = atoi(argv[6]);
@@ -320,7 +321,7 @@ int main(int argc, char **argv) {
   SherpaOnnxOnlineRecognizer *recognizer =
       SherpaOnnxCreateOnlineRecognizer(&config);
   SherpaOnnxOnlineStream *stream = SherpaOnnxCreateOnlineStream(recognizer);
-  SherpaOnnxDisplay *display = SherpaOnnxCreateDisplay(50);
+  const SherpaOnnxDisplay *display = SherpaOnnxCreateDisplay(50);
   int32_t segment_id = 0;
 
   if ((ret = open_input_file(argv[5])) < 0) exit(1);
@@ -383,7 +384,7 @@ int main(int argc, char **argv) {
     SherpaOnnxDecodeOnlineStream(recognizer, stream);
   }
 
-  SherpaOnnxOnlineRecognizerResult *r =
+  const SherpaOnnxOnlineRecognizerResult *r =
       SherpaOnnxGetOnlineStreamResult(recognizer, stream);
   if (strlen(r->text)) {
     SherpaOnnxPrint(display, segment_id, r->text);