[Tool] Add format tools and execute format

.github/workflows/actionlint.yml

@@ -0,0 +1,40 @@
+name: Lint GitHub Actions workflows
+on:
+  push:
+    branches:
+      - "main"
+    paths:
+      - '.github/workflows/*.ya?ml'
+      - '.github/workflows/actionlint.*'
+      - '.github/workflows/matchers/actionlint.json'
+  pull_request:
+    branches:
+      - "main"
+    paths:
+      - '.github/workflows/*.ya?ml'
+      - '.github/workflows/actionlint.*'
+      - '.github/workflows/matchers/actionlint.json'
+
+env:
+  LC_ALL: en_US.UTF-8
+
+defaults:
+  run:
+    shell: bash
+
+permissions:
+  contents: read
+
+jobs:
+  actionlint:
+    runs-on: ubuntu-latest
+    steps:
+      - name: "Checkout"
+        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
+        with:
+          fetch-depth: 0
+
+      - name: "Run actionlint"
+        run: |
+          echo "::add-matcher::.github/workflows/matchers/actionlint.json"
+          tools/actionlint.sh -color

.github/workflows/matchers/actionlint.json

@@ -0,0 +1,17 @@
+{
+  "problemMatcher": [
+    {
+      "owner": "actionlint",
+      "pattern": [
+        {
+          "regexp": "^(?:\\x1b\\[\\d+m)?(.+?)(?:\\x1b\\[\\d+m)*:(?:\\x1b\\[\\d+m)*(\\d+)(?:\\x1b\\[\\d+m)*:(?:\\x1b\\[\\d+m)*(\\d+)(?:\\x1b\\[\\d+m)*: (?:\\x1b\\[\\d+m)*(.+?)(?:\\x1b\\[\\d+m)* \\[(.+?)\\]$",
+          "file": 1,
+          "line": 2,
+          "column": 3,
+          "message": 4,
+          "code": 5
+        }
+      ]
+    }
+  ]
+}

.github/workflows/matchers/mypy.json

@@ -0,0 +1,16 @@
+{
+  "problemMatcher": [
+    {
+      "owner": "mypy",
+      "pattern": [
+        {
+          "regexp": "^(.+):(\\d+):\\s(error|warning):\\s(.+)$",
+          "file": 1,
+          "line": 2,
+          "severity": 3,
+          "message": 4
+        }
+      ]
+    }
+  ]
+}

.github/workflows/matchers/ruff.json

@@ -0,0 +1,17 @@
+{
+    "problemMatcher": [
+      {
+        "owner": "ruff",
+        "pattern": [
+          {
+            "regexp": "^(.+?):(\\d+):(\\d+): (\\w+): (.+)$",
+            "file": 1,
+            "line": 2,
+            "column": 3,
+            "code": 4,
+            "message": 5
+          }
+        ]
+      }
+    ]
+  }

.gitignore

@@ -11,3 +11,12 @@ __pycache__/
 
 # egg-info
 vllm_ascend.egg-info/
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Linting
+actionlint
+shellcheck*/

README.md

@@ -109,6 +109,3 @@ bash format.sh
 # 3. Commit changed files with message 'Run yapf and ruff'
 git commit -m "Run yapf and ruff"
 ```
-
-> [!NOTE]
-> The warning "F401 `torch_npu` imported but unused" doesn't matter because the api `torch.npu` will call this library indirectly while there are not `torch_npu` in the codes explicitly. 

examples/offline_inference_audio_language.py

@@ -7,7 +7,6 @@
 """
 
 from transformers import AutoTokenizer
-
 from vllm import LLM, SamplingParams
 from vllm.assets.audio import AudioAsset
 from vllm.utils import FlexibleArgumentParser

format.sh

@@ -43,7 +43,7 @@ ISORT_VERSION=$(isort --vn)
 CLANGFORMAT_VERSION=$(clang-format --version | awk '{print $3}')
 SPHINX_LINT_VERSION=$(sphinx-lint --version | awk '{print $2}')
 
-# # params: tool name, tool version, required version
+# params: tool name, tool version, required version
 tool_version_check() {
     expected=$(grep "$1" requirements-lint.txt | cut -d'=' -f3)
     if [[ "$2" != "$expected" ]]; then
@@ -112,9 +112,9 @@ fi
 echo 'vLLM yapf: Done'
 
 # Run mypy
-# echo 'vLLM mypy:'
-# tools/mypy.sh
-# echo 'vLLM mypy: Done'
+echo 'vLLM mypy:'
+tools/mypy.sh
+echo 'vLLM mypy: Done'
 
 
 # If git diff returns a file that is in the skip list, the file may be checked anyway:
@@ -316,6 +316,6 @@ else
     echo "✨🎉 Format check passed! Congratulations! 🎉✨"
 fi
 
-echo 'vLLM sphinx-lint:'
-tools/sphinx-lint.sh
-echo 'vLLM sphinx-lint: Done'
+# echo 'vLLM sphinx-lint:'
+# tools/sphinx-lint.sh
+# echo 'vLLM sphinx-lint: Done'

mypy.ini

@@ -0,0 +1,10 @@
+[mypy]
+; warn_return_any = True
+warn_unused_configs = True
+
+; Suppress all missing import errors from torch_npu for mypy.
+[mypy-torch_npu.*]
+ignore_missing_imports = True
+
+[mypy-transformers.*]
+ignore_missing_imports = True

setup.py

@@ -1,5 +1,6 @@
 import os
-from typing import Dict, List
+from typing import List
+
 from setuptools import setup
 
 ROOT_DIR = os.path.dirname(__file__)
@@ -8,6 +9,7 @@
 def get_path(*filepath) -> str:
     return os.path.join(ROOT_DIR, *filepath)
 
+
 def get_requirements() -> List[str]:
     """Get Python package dependencies from requirements.txt."""
 
@@ -23,25 +25,25 @@ def _read_requirements(filename: str) -> List[str]:
             else:
                 resolved_requirements.append(line)
         return resolved_requirements
-    
+
     try:
         requirements = _read_requirements("requirements.txt")
     except ValueError:
         print("Failed to read requirements.txt in vllm_ascend.")
     return requirements
 
 
-setup(name='vllm_ascend',
-      version='0.1',
-      packages=['vllm_ascend'],
-      install_requires=get_requirements(),
-      extras_require={
+setup(
+    name='vllm_ascend',
+    version='0.1',
+    packages=['vllm_ascend'],
+    install_requires=get_requirements(),
+    extras_require={
         "tensorizer": ["tensorizer>=2.9.0"],
         "runai": ["runai-model-streamer", "runai-model-streamer-s3", "boto3"],
         "audio": ["librosa", "soundfile"],  # Required for audio processing
         "video": ["decord"]  # Required for video processing
-      },
-      entry_points={
-          'vllm.platform_plugins':
-          ["ascend_plugin = vllm_ascend:register"]
-      })
+    },
+    entry_points={
+        'vllm.platform_plugins': ["ascend_plugin = vllm_ascend:register"]
+    })

tools/actionlint.sh

@@ -0,0 +1,13 @@
+#!/bin/bash
+
+if command -v actionlint &> /dev/null; then
+    actionlint "$@"
+    exit 0
+elif [ -x ./actionlint ]; then
+    ./actionlint "$@"
+    exit 0
+fi
+
+# download a binary to the current directory - v1.7.3
+bash <(curl https://raw.githubusercontent.com/rhysd/actionlint/aa0a7be8e566b096e64a5df8ff290ec24fa58fbc/scripts/download-actionlint.bash)
+./actionlint "$@"

tools/check_repo.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+# Checks whether the repo is clean and whether tags are available (necessary to correctly produce vllm version at build time)
+
+if ! git diff --quiet; then
+	echo "Repo is dirty" >&2
+
+	exit 1
+fi
+
+if ! git describe --tags; then
+	echo "No tags are present. Is this a shallow clone? git fetch --unshallow --tags" >&2
+
+	exit 1
+fi

tools/mypy.sh

@@ -0,0 +1,21 @@
+#!/bin/bash
+
+CI=${1:-0}
+PYTHON_VERSION=${2:-3.9}
+
+if [ "$CI" -eq 1 ]; then
+    set -e
+fi
+
+run_mypy() {
+    echo "Running mypy on $1"
+    if [ "$CI" -eq 1 ] && [ -z "$1" ]; then
+        mypy --python-version "${PYTHON_VERSION}" "$@"
+        return
+    fi
+    mypy --follow-imports skip --python-version "${PYTHON_VERSION}" "$@"
+}
+
+run_mypy # Note that this is less strict than CI
+run_mypy vllm_ascend
+run_mypy examples

tools/png-lint.sh

@@ -0,0 +1,15 @@
+#!/bin/bash
+
+# Ensure that *.excalidraw.png files have the excalidraw metadata
+# embedded in them. This ensures they can be loaded back into
+# the tool and edited in the future.
+
+find . -iname '*.excalidraw.png' | while read -r file; do
+	if git check-ignore -q "$file"; then
+		continue
+	fi
+	if ! grep -q "excalidraw+json" "$file"; then
+		echo "$file was not exported from excalidraw with 'Embed Scene' enabled."
+		exit 1
+	fi
+done

tools/shellcheck.sh

@@ -0,0 +1,24 @@
+#!/bin/bash
+set -e
+
+scversion="stable"
+
+if [ -d "shellcheck-${scversion}" ]; then
+    PATH="$PATH:$(pwd)/shellcheck-${scversion}"
+    export PATH
+fi
+
+if ! [ -x "$(command -v shellcheck)" ]; then
+    if [ "$(uname -s)" != "Linux" ] || [ "$(uname -m)" != "x86_64" ]; then
+        echo "Please install shellcheck: https://github.com/koalaman/shellcheck?tab=readme-ov-file#installing"
+        exit 1
+    fi
+
+    # automatic local install if linux x86_64
+    wget -qO- "https://github.com/koalaman/shellcheck/releases/download/${scversion?}/shellcheck-${scversion?}.linux.x86_64.tar.xz" | tar -xJv
+    PATH="$PATH:$(pwd)/shellcheck-${scversion}"
+    export PATH
+fi
+
+# TODO - fix warnings in .buildkite/run-amd-test.sh
+find . -name "*.sh" -not -path "./.buildkite/run-amd-test.sh" -print0 | xargs -0 -I {} sh -c 'git check-ignore -q "{}" || shellcheck "{}"'

tools/sphinx-lint.sh

@@ -0,0 +1,3 @@
+#!/bin/bash
+
+sphinx-lint --disable trailing-whitespace,missing-final-newline docs

vllm_ascend/attention.py

@@ -5,9 +5,9 @@
 import torch
 
 try:
-    import torch_npu
+    import torch_npu  # noqa: F401
 except ImportError:
-    print("Failed to import torch_npu")
+    print("Failed to import torch_npu.")
 
 from vllm.attention.backends.abstract import (AttentionBackend, AttentionImpl,
                                               AttentionMetadata, AttentionType)
@@ -375,15 +375,16 @@ def _add_seq_group(
             # TODO(sang): Combine chunked prefill and prefix caching by
             # only allowing multiple of block_size chunk size.
             # NOTE: This only works for oooooooxxx style attention.
-            block_table = []
+            block_table: List[int] = []
             prefix_cache_hit = any([
                 inter_data.prefix_cache_hit
                 for inter_data in self.input_builder.inter_data_list
             ])
             if prefix_cache_hit:
                 # NOTE(woosuk): For flash-attn, the block table should
                 # include the entries for the incoming prefill tokens.
-                block_table = block_tables[seq_id]
+                if block_tables is not None:
+                    block_table = block_tables[seq_id]
             elif ((chunked_prefill_enabled or not is_prompt)
                   and block_tables is not None):
                 if curr_sliding_window_block == 0: