[lint] use pre-commit to auto check and lint (#183)

* [lint] use pre-commit to auto check and lint

* [lint] auto format all by pre-commit, including c++, python

* [lint] cpplint==1.6.1

* [lint] remove sphinx

Author: cdliang11

.github/workflows/doc.yml

@@ -8,13 +8,6 @@ jobs:
     steps:
     - uses: actions/checkout@v1
 
-    # install sphinx related package and build sphinx files
-    - uses: ammaraskar/sphinx-action@master
-      with:
-        docs-folder: "docs/"
-        pre-build-command: "pip install sphinx-markdown-tables nbsphinx jinja2 recommonmark sphinx_rtd_theme"
-
-
     # add .nojekyll to notice Pages use the _* dirs
     - name: copy the generated site
       if: ${{ github.event_name == 'push' && github.ref == 'refs/heads/main' }}

.github/workflows/lint.yml

@@ -35,7 +35,7 @@ jobs:
       - name: Setup Python
         uses: actions/setup-python@v1
         with:
-          python-version: 3.x
+          python-version: 3.9
           architecture: x64
       - name: Fetch Wenet
         uses: actions/checkout@v1
@@ -80,7 +80,7 @@ jobs:
       - name: Run cpplint
         run: |
           set -eux
-          pip install cpplint
+          pip install cpplint==1.6.1
           cpplint --version
           cpplint --recursive .
           if [ $? != 0 ]; then exit 1; fi

.pre-commit-config.yaml

@@ -0,0 +1,23 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.5.0
+    hooks:
+    - id: trailing-whitespace
+  - repo: https://github.com/pre-commit/mirrors-yapf
+    rev: 'v0.32.0'
+    hooks:
+    - id: yapf
+  - repo: https://github.com/pycqa/flake8
+    rev: '3.8.2'
+    hooks:
+    - id: flake8
+  - repo: https://github.com/pre-commit/mirrors-clang-format
+    rev: 'v17.0.6'
+    hooks:
+    - id: clang-format
+      args: ['--style=file']
+      exclude: 'runtime/android/app/src/.*\.(json|java|js|m|mm|proto)'
+  - repo: https://github.com/cpplint/cpplint
+    rev: '1.6.1'
+    hooks:
+    - id: cpplint

README.md

@@ -45,6 +45,7 @@ pip install torch torchaudio
 
 ``` sh
 pip install -r requirements.txt
+pre-commit install  # for clean and tidy code
 ```
 
 ## Dataset

docs/conf.py

@@ -14,14 +14,12 @@
 # import sys
 # sys.path.insert(0, os.path.abspath('.'))
 
-
 # -- Project information -----------------------------------------------------
 
 project = 'Wenet'
 copyright = '2020, wenet-team'
 author = 'wenet-team'
 
-
 # -- General configuration ---------------------------------------------------
 
 # Add any Sphinx extension module names here, as strings. They can be
@@ -43,7 +41,6 @@
 # Add any paths that contain templates here, relative to this directory.
 templates_path = ['_templates']
 
-
 # The suffix(es) of source filenames.
 # You can specify multiple suffix as a list of string:
 source_suffix = {
@@ -57,7 +54,6 @@
 # This pattern also affects html_static_path and html_extra_path.
 exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
 
-
 # -- Options for HTML output -------------------------------------------------
 
 # The theme to use for HTML and HTML Help pages.  See the documentation for

requirements.txt

@@ -15,3 +15,4 @@ pyflakes==2.2.0
 lmdb
 scipy
 tqdm
+pre-commit==3.5.0

runtime/android/app/src/main/cpp/wekws.cc

@@ -82,18 +82,18 @@ void set_input_finished() {
 // }
 
 void start_spot() {
-    std::vector<std::vector<float>> feats;
-    feature_pipeline->Read(80, &feats);
-    std::vector<std::vector<float>> prob;
-    spotter->Forward(feats, &prob);
-    float max_prob = 0.0;
-    for (int t = 0; t < prob.size(); t++) {
-      for (int j = 0; j < prob[t].size(); j++) {
-        max_prob = std::max(prob[t][j], max_prob);
-      }
+  std::vector<std::vector<float>> feats;
+  feature_pipeline->Read(80, &feats);
+  std::vector<std::vector<float>> prob;
+  spotter->Forward(feats, &prob);
+  float max_prob = 0.0;
+  for (int t = 0; t < prob.size(); t++) {
+    for (int j = 0; j < prob[t].size(); j++) {
+      max_prob = std::max(prob[t][j], max_prob);
     }
-    result = std::to_string(offset) + " prob: " + std::to_string(max_prob);
-    offset += prob.size();
+  }
+  result = std::to_string(offset) + " prob: " + std::to_string(max_prob);
+  offset += prob.size();
 }
 
 jstring get_result(JNIEnv* env, jobject) {

runtime/core/bin/kws_main.cc

@@ -12,7 +12,6 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-
 #include <iostream>
 #include <string>
 

runtime/core/frontend/feature_pipeline.h

@@ -21,8 +21,8 @@
 #include <vector>
 
 #include "frontend/fbank.h"
-#include "utils/log.h"
 #include "utils/blocking_queue.h"
+#include "utils/log.h"
 
 namespace wenet {
 

runtime/core/kws/keyword_spotting.cc

@@ -12,7 +12,6 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-
 #include "kws/keyword_spotting.h"
 
 #include <iostream>
@@ -35,30 +34,27 @@ KeywordSpotting::KeywordSpotting(const std::string& model_path) {
   out_names_ = {"output", "r_cache"};
   auto metadata = session_->GetModelMetadata();
   Ort::AllocatorWithDefaultOptions allocator;
-  cache_dim_ = std::stoi(metadata.LookupCustomMetadataMap("cache_dim",
-                                                          allocator));
-  cache_len_ = std::stoi(metadata.LookupCustomMetadataMap("cache_len",
-                                                          allocator));
+  cache_dim_ =
+      std::stoi(metadata.LookupCustomMetadataMap("cache_dim", allocator));
+  cache_len_ =
+      std::stoi(metadata.LookupCustomMetadataMap("cache_len", allocator));
   std::cout << "Kws Model Info:" << std::endl
             << "\tcache_dim: " << cache_dim_ << std::endl
             << "\tcache_len: " << cache_len_ << std::endl;
   Reset();
 }
 
-
 void KeywordSpotting::Reset() {
   Ort::MemoryInfo memory_info =
-        Ort::MemoryInfo::CreateCpu(OrtDeviceAllocator, OrtMemTypeCPU);
+      Ort::MemoryInfo::CreateCpu(OrtDeviceAllocator, OrtMemTypeCPU);
   cache_.resize(cache_dim_ * cache_len_, 0.0);
   const int64_t cache_shape[] = {1, cache_dim_, cache_len_};
-  cache_ort_ = Ort::Value::CreateTensor<float>(
-      memory_info, cache_.data(), cache_.size(), cache_shape, 3);
+  cache_ort_ = Ort::Value::CreateTensor<float>(memory_info, cache_.data(),
+                                               cache_.size(), cache_shape, 3);
 }
 
-
-void KeywordSpotting::Forward(
-    const std::vector<std::vector<float>>& feats,
-    std::vector<std::vector<float>>* prob) {
+void KeywordSpotting::Forward(const std::vector<std::vector<float>>& feats,
+                              std::vector<std::vector<float>>* prob) {
   prob->clear();
   if (feats.size() == 0) return;
   Ort::MemoryInfo memory_info =
@@ -78,9 +74,9 @@ void KeywordSpotting::Forward(
   inputs.emplace_back(std::move(feats_ort));
   inputs.emplace_back(std::move(cache_ort_));
   // ort_outputs.size() == 2
-  std::vector<Ort::Value> ort_outputs = session_->Run(
-      Ort::RunOptions{nullptr}, in_names_.data(), inputs.data(),
-      inputs.size(), out_names_.data(), out_names_.size());
+  std::vector<Ort::Value> ort_outputs =
+      session_->Run(Ort::RunOptions{nullptr}, in_names_.data(), inputs.data(),
+                    inputs.size(), out_names_.data(), out_names_.size());
 
   // 3. Update cache
   cache_ort_ = std::move(ort_outputs[1]);
@@ -92,9 +88,9 @@ void KeywordSpotting::Forward(
   int output_dim = type_info.GetShape()[2];
   prob->resize(num_outputs);
   for (int i = 0; i < num_outputs; i++) {
-     (*prob)[i].resize(output_dim);
-     memcpy((*prob)[i].data(), data + i * output_dim,
-            sizeof(float) * output_dim);
+    (*prob)[i].resize(output_dim);
+    memcpy((*prob)[i].data(), data + i * output_dim,
+           sizeof(float) * output_dim);
   }
 }