Skip to content

Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.

[feat] Test/add resource ci validation and fix session api case 400 error (#1599)

kaisonglic7df2971dbce
diff --git a/.github/workflows/api_test.yml b/.github/workflows/api_test.yml
index 35ddd5eb9..aeedaf698 100644
--- a/.github/workflows/api_test.yml
+++ b/.github/workflows/api_test.yml
@@ -33,7 +33,7 @@ jobs:
fail-fast: false
max-parallel: 1
matrix:
- os: ${{ (github.event_name == 'push' && github.ref == 'refs/heads/main') && fromJSON('["ubuntu-24.04", "macos-14", "windows-latest"]') || fromJSON('["ubuntu-24.04"]') }}
+ os: ${{ ((github.event_name == 'push' && github.ref == 'refs/heads/main') || github.event_name == 'schedule' || github.event_name == 'workflow_dispatch') && fromJSON('["ubuntu-24.04", "macos-14", "windows-latest"]') || fromJSON('["ubuntu-24.04"]') }}
steps:
- uses: actions/checkout@v6
@@ -67,12 +67,12 @@ jobs:
- name: Install system dependencies (Ubuntu)
if: runner.os == 'Linux'
run: |
- sudo apt-get update -y && sudo apt-get install -y cmake build-essential --no-install-recommends
+ sudo apt-get update -y && sudo apt-get install -y cmake build-essential ffmpeg --no-install-recommends
- name: Install system dependencies (macOS)
if: runner.os == 'macOS'
run: |
- brew install cmake
+ brew install cmake ffmpeg
- name: Install system dependencies (Windows)
if: runner.os == 'Windows'
@@ -127,7 +127,7 @@ jobs:
shell: bash
run: |
RAGFS_LIB_DIR="openviking/lib"
- RAGFS_SO_COUNT=$(ls -1 "$RAGFS_LIB_DIR"/ragfs_python*.so "$RAGFS_LIB_DIR"/ragfs_python*.pyd "$RAGFS_LIB_DIR"/ragfs_python*.dylib 2>/dev/null | wc -l | xargs || echo "0")
+ RAGFS_SO_COUNT=$(ls -1 "$RAGFS_LIB_DIR"/ragfs_python*.so "$RAGFS_LIB_DIR"/ragfs_python*.pyd "$RAGFS_LIB_DIR"/ragfs_python*.dylib 2>/dev/null | wc -l | tr -d '[:space:]' || echo "0")
if [ "$RAGFS_SO_COUNT" -gt 0 ]; then
echo "✅ ragfs_python native extension already exists:"
@@ -150,13 +150,15 @@ jobs:
echo "Extracting ragfs_python from wheel: $WHL_FILE"
mkdir -p "$RAGFS_LIB_DIR"
+ WHL_FILE_PY=$(echo "$WHL_FILE" | sed 's/\\/\\\\/g')
+ RAGFS_LIB_DIR_PY=$(echo "$RAGFS_LIB_DIR" | sed 's/\\/\\\\/g')
uv run python -c "
import zipfile, os, sys, stat
- with zipfile.ZipFile('$WHL_FILE') as zf:
+ with zipfile.ZipFile('${WHL_FILE_PY}') as zf:
for name in zf.namelist():
bn = os.path.basename(name)
if bn.startswith('ragfs_python') and (bn.endswith('.so') or bn.endswith('.pyd') or bn.endswith('.dylib')):
- dst = os.path.join('$RAGFS_LIB_DIR', bn)
+ dst = os.path.join('${RAGFS_LIB_DIR_PY}', bn)
with zf.open(name) as src, open(dst, 'wb') as f:
f.write(src.read())
os.chmod(dst, 0o755)
@@ -443,13 +445,14 @@ jobs:
if [ "${{ env.HAS_SECRETS }}" = "true" ]; then
echo "Running full test suite with VLM/Embedding"
- uv run python -m pytest . -v --html=api-test-report.html --self-contained-html
+ uv run python -m pytest . -v --html=api-test-report.html --self-contained-html \
+ --ignore=scenarios/resources_retrieval_slow/
else
echo "Running basic tests only (no VLM/Embedding)"
uv run python -m pytest . -v --html=api-test-report.html --self-contained-html \
--ignore=retrieval/ --ignore=resources/test_pack.py --ignore=resources/test_wait_processed.py \
--ignore=admin/ --ignore=skills/ --ignore=system/test_system_status.py --ignore=system/test_is_healthy.py --ignore=system/test_system_wait.py \
- --ignore=scenarios/ --ignore=sessions/ -k "not test_observer"
+ --ignore=scenarios/ -k "not test_observer"
fi
continue-on-error: true
@@ -464,10 +467,10 @@ jobs:
if ($env:HAS_SECRETS -eq "true") {
Write-Host "Running full test suite with VLM/Embedding (Windows: skipping filesystem tests)"
- uv run python -m pytest . -v --html=api-test-report.html --self-contained-html --ignore=filesystem/
+ uv run python -m pytest . -v --html=api-test-report.html --self-contained-html --ignore=filesystem/ --ignore=scenarios/resources_retrieval_slow/
} else {
Write-Host "Running basic tests only (no VLM/Embedding, Windows: skipping filesystem tests)"
- uv run python -m pytest . -v --html=api-test-report.html --self-contained-html --ignore=retrieval/ --ignore=resources/test_pack.py --ignore=resources/test_wait_processed.py --ignore=admin/ --ignore=skills/ --ignore=system/test_system_status.py --ignore=system/test_is_healthy.py --ignore=system/test_system_wait.py --ignore=filesystem/ --ignore=scenarios/ --ignore=sessions/ -k "not test_observer"
+ uv run python -m pytest . -v --html=api-test-report.html --self-contained-html --ignore=retrieval/ --ignore=resources/test_pack.py --ignore=resources/test_wait_processed.py --ignore=admin/ --ignore=skills/ --ignore=system/test_system_status.py --ignore=system/test_is_healthy.py --ignore=system/test_system_wait.py --ignore=filesystem/ --ignore=scenarios/ -k "not test_observer"
}
continue-on-error: true
diff --git a/.github/workflows/api_test_effect.yml b/.github/workflows/api_test_effect.yml
new file mode 100644
index 000000000..c356a1c47
--- /dev/null
+++ b/.github/workflows/api_test_effect.yml
@@ -0,0 +1,222 @@
+name: 07. API Effect Tests
+
+on:
+ workflow_dispatch:
+ schedule:
+ - cron: '0 2 * * *'
+
+concurrency:
+ group: ${{ github.workflow }}-${{ github.ref }}
+ cancel-in-progress: true
+
+jobs:
+ effect-tests:
+ name: API Effect Tests (${{ matrix.os }})
+ runs-on: ${{ matrix.os }}
+ timeout-minutes: 120
+ strategy:
+ fail-fast: false
+ max-parallel: 1
+ matrix:
+ os: [ubuntu-24.04]
+
+ steps:
+ - uses: actions/checkout@v6
+ with:
+ submodules: recursive
+ fetch-depth: 100
+
+ - name: Set up Python 3.10
+ uses: actions/setup-python@v6
+ with:
+ python-version: '3.10'
+
+ - name: Cache Python dependencies
+ uses: actions/cache@v5
+ with:
+ path: ~/.cache/pip
+ key: ${{ runner.os }}-pip-effect-${{ hashFiles('**/requirements.txt', '**/pyproject.toml') }}
+ restore-keys: |
+ ${{ runner.os }}-pip-effect-
+
+ - name: Install system dependencies
+ run: |
+ sudo apt-get update -y && sudo apt-get install -y cmake build-essential ffmpeg --no-install-recommends
+
+ - name: Install uv
+ uses: astral-sh/setup-uv@v7
+ with:
+ enable-cache: true
+
+ - name: Install build dependencies (system pip)
+ run: |
+ python -m pip install --upgrade pip
+ pip install "setuptools>=70.1" setuptools_scm cmake wheel
+
+ - name: Install dependencies using uv sync
+ run: uv sync --frozen
+
+ - name: Install build dependencies (uv pip)
+ run: uv pip install setuptools setuptools_scm cmake wheel maturin
+
+ - name: Set up Rust toolchain
+ uses: dtolnay/rust-toolchain@v1
+ with:
+ toolchain: stable
+
+ - name: Build C++ extensions
+ shell: bash
+ run: |
+ export OV_SKIP_OV_BUILD=1
+ mkdir -p openviking/bin
+ touch openviking/bin/ov
+ chmod +x openviking/bin/ov
+ if [ -z "$SETUPTOOLS_SCM_PRETEND_VERSION_FOR_OPENVIKING" ]; then
+ PRETEND_VERSION=$(git describe --tags --always 2>/dev/null | sed -E 's/^v?([0-9]+\.[0-9]+\.[0-9]+).*/\1.dev0/' || echo "0.0.0.dev0")
+ echo "SETUPTOOLS_SCM_PRETEND_VERSION_FOR_OPENVIKING=$PRETEND_VERSION" >> $GITHUB_ENV
+ fi
+ uv run python setup.py build_ext --inplace
+
+ - name: Build ragfs-python native extension
+ shell: bash
+ run: |
+ RAGFS_LIB_DIR="openviking/lib"
+ RAGFS_SO_COUNT=$(ls -1 "$RAGFS_LIB_DIR"/ragfs_python*.so "$RAGFS_LIB_DIR"/ragfs_python*.pyd "$RAGFS_LIB_DIR"/ragfs_python*.dylib 2>/dev/null | wc -l | tr -d '[:space:]' || echo "0")
+ if [ "$RAGFS_SO_COUNT" -gt 0 ]; then exit 0; fi
+ TMPDIR_RAGFS=$(mktemp -d)
+ cd crates/ragfs-python
+ uv run --no-project maturin build --release --features s3 --out "$TMPDIR_RAGFS"
+ cd ../..
+ WHL_FILE=$(ls -1 "$TMPDIR_RAGFS"/ragfs_python-*.whl 2>/dev/null | head -1)
+ if [ -z "$WHL_FILE" ]; then exit 1; fi
+ mkdir -p "$RAGFS_LIB_DIR"
+ WHL_FILE_PY=$(echo "$WHL_FILE" | sed 's/\\/\\\\/g')
+ RAGFS_LIB_DIR_PY=$(echo "$RAGFS_LIB_DIR" | sed 's/\\/\\\\/g')
+ uv run python -c "
+ import zipfile, os, sys
+ with zipfile.ZipFile('${WHL_FILE_PY}') as zf:
+ for name in zf.namelist():
+ bn = os.path.basename(name)
+ if bn.startswith('ragfs_python') and (bn.endswith('.so') or bn.endswith('.pyd') or bn.endswith('.dylib')):
+ dst = os.path.join('${RAGFS_LIB_DIR_PY}', bn)
+ with zf.open(name) as src, open(dst, 'wb') as f:
+ f.write(src.read())
+ os.chmod(dst, 0o755)
+ sys.exit(0)
+ sys.exit(1)
+ "
+ rm -rf "$TMPDIR_RAGFS"
+
+ - name: Install API test dependencies
+ run: |
+ cd tests/api_test
+ uv pip install -r requirements.txt
+
+ - name: Create OpenViking config file
+ env:
+ VLM_API_KEY: ${{ secrets.VLM_API_KEY }}
+ EMBEDDING_API_KEY: ${{ secrets.EMBEDDING_API_KEY }}
+ run: |
+ mkdir -p $HOME/.openviking
+ if [ -n "$VLM_API_KEY" ] && [ -n "$EMBEDDING_API_KEY" ]; then
+ echo "HAS_SECRETS=true" >> $GITHUB_ENV
+ cat > $HOME/.openviking/ov.conf << EOF
+ {
+ "server": { "root_api_key": "test-root-api-key" },
+ "vlm": {
+ "provider": "volcengine",
+ "api_key": "$VLM_API_KEY",
+ "model": "doubao-seed-2-0-mini-260215",
+ "api_base": "https://ark.cn-beijing.volces.com/api/v3",
+ "temperature": 0.1, "max_retries": 3
+ },
+ "embedding": {
+ "dense": {
+ "provider": "volcengine",
+ "api_key": "$EMBEDDING_API_KEY",
+ "model": "doubao-embedding-vision-251215",
+ "api_base": "https://ark.cn-beijing.volces.com/api/v3",
+ "dimension": 1024, "input": "multimodal"
+ }
+ }
+ }
+ EOF
+ else
+ echo "HAS_SECRETS=false" >> $GITHUB_ENV
+ cat > $HOME/.openviking/ov.conf << EOF
+ {
+ "server": { "root_api_key": "test-root-api-key" },
+ "vlm": {
+ "provider": "volcengine",
+ "api_key": "dummy-vlm-api-key",
+ "model": "doubao-seed-2-0-mini-260215",
+ "api_base": "https://ark.cn-beijing.volces.com/api/v3",
+ "temperature": 0.1, "max_retries": 3
+ },
+ "embedding": {
+ "dense": {
+ "provider": "volcengine",
+ "api_key": "dummy-embedding-api-key",
+ "model": "doubao-embedding-vision-251215",
+ "api_base": "https://ark.cn-beijing.volces.com/api/v3",
+ "dimension": 1024, "input": "multimodal"
+ }
+ }
+ }
+ EOF
+ fi
+
+ - name: Start OpenViking Server
+ run: |
+ SERVER_PORT=1933
+ echo "SERVER_PORT=$SERVER_PORT" >> $GITHUB_ENV
+ export ROOT_API_KEY=test-root-api-key
+ export SERVER_PORT=$SERVER_PORT
+ nohup uv run python -m openviking.server.bootstrap > openviking-server.log 2>&1 &
+ echo $! > openviking-server.pid
+ echo "Waiting for server to start..."
+ for i in {1..30}; do
+ if curl -s http://127.0.0.1:$SERVER_PORT/health | grep -q '"healthy":true'; then
+ echo "Server is ready!"
+ break
+ fi
+ echo "Waiting... ($i/30)"
+ sleep 2
+ done
+ if ! curl -s http://127.0.0.1:$SERVER_PORT/health | grep -q '"healthy":true'; then
+ echo "Server failed to start!"
+ cat openviking-server.log
+ exit 1
+ fi
+
+ - name: Run Effect Tests
+ id: run-effect-tests
+ run: |
+ cd tests/api_test
+ export OPENVIKING_API_KEY=test-root-api-key
+ export SERVER_URL=http://127.0.0.1:${{ env.SERVER_PORT }}
+ uv run python -m pytest scenarios/resources_retrieval_slow/ -v --tb=short --durations=0 --html=api-effect-test-report.html --self-contained-html
+ continue-on-error: true
+
+ - name: Upload test reports
+ uses: actions/upload-artifact@v7
+ if: always()
+ with:
+ name: api-effect-test-reports-${{ github.run_id }}
+ path: |
+ tests/api_test/api-effect-test-report.html
+ openviking-server.log
+
+ - name: Stop OpenViking Server
+ if: always()
+ run: |
+ if [ -f openviking-server.pid ]; then
+ kill $(cat openviking-server.pid) 2>/dev/null || true
+ pkill -f "openviking.server.bootstrap" 2>/dev/null || true
+ fi
+
+ - name: Check test results
+ if: steps.run-effect-tests.outcome != 'success'
+ run: |
+ echo "Effect tests failed!"
+ exit 1
diff --git a/tests/api_test/api/client.py b/tests/api_test/api/client.py
index 30cf2ca5a..b716a456b 100644
--- a/tests/api_test/api/client.py
+++ b/tests/api_test/api/client.py
@@ -256,6 +256,7 @@ class OpenVikingAPIClient:
path: str,
to: Optional[str] = None,
reason: Optional[str] = None,
+ parent: Optional[str] = None,
wait: bool = False,
) -> requests.Response:
endpoint = "/api/v1/resources"
@@ -273,6 +274,8 @@ class OpenVikingAPIClient:
payload["to"] = to
if reason:
payload["reason"] = reason
+ if parent:
+ payload["parent"] = parent
if wait:
payload["wait"] = wait
try:
diff --git a/tests/api_test/conftest.py b/tests/api_test/conftest.py
index 1deb94bc7..56692f76b 100644
--- a/tests/api_test/conftest.py
+++ b/tests/api_test/conftest.py
@@ -58,6 +58,49 @@ TEST_CASE_DESCRIPTIONS = {
"test_concurrent_write.py::TestConcurrentWrite::test_concurrent_write_conflict": "并发写入冲突验证",
"test_account_isolation.py::TestAccountIsolation::test_processed_not_zero_after_resource_ops": "账户隔离完整性验证",
"test_account_isolation.py::TestAccountIsolation::test_consecutive_health_checks": "账户隔离连续健康检查",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_txt_file": "TC-B01 TXT文本构建",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_markdown_file": "TC-B02 Markdown构建",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_empty_file": "TC-B15 空文件构建",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_raw_content": "TC-B14 原始内容构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_pdf_file": "TC-B03 PDF构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_html_file": "TC-B04 HTML构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_docx_file": "TC-B05 DOCX构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_legacy_doc_file": "TC-B06 .doc构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_pptx_file": "TC-B07 PPTX构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_xlsx_file": "TC-B08 XLSX构建",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_epub_file": "TC-B09 EPUB构建",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_zip_file": "TC-B10 ZIP构建",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_directory": "TC-B11 目录构建",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_code_repository_url": "TC-B12 代码仓库URL构建",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_svg": "TC-B13 图片构建(SVG)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_jpg": "TC-B13 图片构建(JPG)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_png": "TC-B13 图片构建(PNG)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_webp": "TC-B13 图片构建(WebP)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_audio_mp3": "TC-B14 音频构建(MP3)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_audio_wav": "TC-B14 音频构建(WAV)",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_video_mp4": "TC-B15 视频构建(MP4)",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_repo": "TC-P01 GitHub仓库构建",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_repo_with_branch": "TC-P02 GitHub分支构建",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_raw_file": "TC-P03 GitHub原始文件构建",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_blob_page": "TC-P04 GitHub Blob页构建",
+ "test_build_platform_wikipedia.py::TestBuildPlatformWikipedia::test_build_wikipedia_page": "TC-P05 Wikipedia构建",
+ "test_build_platform_arxiv.py::TestBuildPlatformArxiv::test_build_arxiv_pdf": "TC-P06 arXiv PDF构建",
+ "test_build_platform_arxiv.py::TestBuildPlatformArxiv::test_build_arxiv_abstract_page": "TC-P07 arXiv摘要页构建",
+ "test_build_platform_general_web.py::TestBuildPlatformGeneralWeb::test_build_general_webpage": "TC-P10 通用网页构建",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_404": "TC-E01 远端404不存在",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_403": "TC-E02 远端403禁止访问",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_500": "TC-E03 远端500服务错误",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_http_to_https_redirect": "TC-E05 HTTP→HTTPS跳转",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_multi_redirect": "TC-E06 多重跳转",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_dns_resolve_failure": "TC-E08 DNS解析失败",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_ssh_url_invalid_format": "TC-E09 SSH URL格式错误",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_duplicate_resource_no_to": "TC-E12 同名资源二次添加",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_incremental_update_with_to": "TC-E13 同to增量更新",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_unsupported_file_type": "TC-E15 不支持的文件类型",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_corrupted_zip": "TC-E16 损坏的ZIP文件",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_with_to_param": "TC-E17 to参数指定URI",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_with_parent_param": "TC-E18 parent参数指定父目录",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_non_resources_scope_rejected": "TC-E11 scope参数非resources拒绝",
}
@@ -114,6 +157,49 @@ TEST_CASE_APIS = {
"test_concurrent_write.py::TestConcurrentWrite::test_concurrent_write_conflict": "/api/v1/resources (并发写入)",
"test_account_isolation.py::TestAccountIsolation::test_processed_not_zero_after_resource_ops": "/api/v1/resources,/api/v1/search,/api/v1/system/observer",
"test_account_isolation.py::TestAccountIsolation::test_consecutive_health_checks": "/api/v1/system/healthy,/api/v1/system/observer",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_txt_file": "/api/v1/resources",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_markdown_file": "/api/v1/resources",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_empty_file": "/api/v1/resources",
+ "test_build_text_resources_slow.py::TestBuildTextResourcesSlow::test_build_raw_content": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_pdf_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_html_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_docx_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_legacy_doc_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_pptx_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_xlsx_file": "/api/v1/resources",
+ "test_build_document_resources_slow.py::TestBuildDocumentResourcesSlow::test_build_epub_file": "/api/v1/resources",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_zip_file": "/api/v1/resources",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_directory": "/api/v1/resources",
+ "test_build_archive_resources.py::TestBuildArchiveResources::test_build_code_repository_url": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_svg": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_jpg": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_png": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_image_webp": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_audio_mp3": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_audio_wav": "/api/v1/resources",
+ "test_build_media_resources_slow.py::TestBuildMediaResourcesSlow::test_build_video_mp4": "/api/v1/resources",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_repo": "/api/v1/resources",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_repo_with_branch": "/api/v1/resources",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_raw_file": "/api/v1/resources",
+ "test_build_platform_github.py::TestBuildPlatformGithub::test_build_github_blob_page": "/api/v1/resources",
+ "test_build_platform_wikipedia.py::TestBuildPlatformWikipedia::test_build_wikipedia_page": "/api/v1/resources",
+ "test_build_platform_arxiv.py::TestBuildPlatformArxiv::test_build_arxiv_pdf": "/api/v1/resources",
+ "test_build_platform_arxiv.py::TestBuildPlatformArxiv::test_build_arxiv_abstract_page": "/api/v1/resources",
+ "test_build_platform_general_web.py::TestBuildPlatformGeneralWeb::test_build_general_webpage": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_404": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_403": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_remote_500": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_http_to_https_redirect": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_multi_redirect": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_dns_resolve_failure": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_ssh_url_invalid_format": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_duplicate_resource_no_to": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_incremental_update_with_to": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_unsupported_file_type": "/api/v1/resources",
+ "test_build_error_handling_slow.py::TestBuildErrorHandlingSlow::test_error_corrupted_zip": "/api/v1/resources",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_with_to_param": "/api/v1/resources",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_with_parent_param": "/api/v1/resources",
+ "test_build_uri_params_slow.py::TestBuildUriParamsSlow::test_build_non_resources_scope_rejected": "/api/v1/resources",
}
@@ -189,10 +275,10 @@ def format_memory_delta(delta_bytes):
def get_test_category(nodeid):
parts = nodeid.split(os.sep)
-
+
# 优先匹配更具体的路径(倒序匹配)
priority_categories = ["stability_error", "resources_retrieval", "filesystem", "sessions"]
-
+
for part in parts:
if part in priority_categories:
# 特殊处理:将子目录映射到正确的分类
@@ -204,18 +290,45 @@ def get_test_category(nodeid):
return "文件系统API"
elif part == "sessions":
return "会话管理API"
-
+
# 如果没有匹配到优先分类,则按原逻辑匹配
for part in parts:
if part in CATEGORY_NAMES:
return CATEGORY_NAMES[part]
-
+
return "其他"
@pytest.fixture(scope="session")
def api_client():
client = OpenVikingAPIClient(server_url=Config.SERVER_URL, api_key=Config.OPENVIKING_API_KEY)
+
+ account_id = Config.OPENVIKING_ACCOUNT
+ user_id = Config.OPENVIKING_USER
+
+ try:
+ resp = client.admin_list_users(account_id)
+ if resp.status_code == 200:
+ data = resp.json()
+ users = data.get("result", [])
+ user_exists = any(
+ (isinstance(u, dict) and u.get("user_id") == user_id)
+ or (isinstance(u, str) and u == user_id)
+ for u in users
+ )
+ if not user_exists:
+ reg_resp = client.admin_register_user(account_id, user_id, role="admin")
+ if reg_resp.status_code in (200, 201):
+ print(f"已注册用户 {account_id}/{user_id}")
+ else:
+ print(f"注册用户失败({reg_resp.status_code}): {reg_resp.text[:200]}")
+ else:
+ print(f"用户 {account_id}/{user_id} 已存在")
+ else:
+ print(f"无法查询用户列表({resp.status_code}), 跳过用户注册(可能为开发模式)")
+ except Exception as e:
+ print(f"用户注册检查异常, 跳过: {e}")
+
return client
@@ -263,7 +376,7 @@ def pytest_runtest_makereport(item, call):
curl = fixture_value.to_curl()
if curl:
report.sections.append(("cURL Command", curl))
-
+
# 添加 Response Body 显示
if hasattr(fixture_value, "last_response") and fixture_value.last_response:
response = fixture_value.last_response
@@ -272,9 +385,14 @@ def pytest_runtest_makereport(item, call):
if response_text:
try:
import json
+
response_json = json.loads(response_text)
- formatted_response = json.dumps(response_json, indent=2, ensure_ascii=False)
- report.sections.append(("Response Body", f"<pre>{formatted_response}</pre>"))
+ formatted_response = json.dumps(
+ response_json, indent=2, ensure_ascii=False
+ )
+ report.sections.append(
+ ("Response Body", f"<pre>{formatted_response}</pre>")
+ )
except Exception:
report.sections.append(("Response Body", f"<pre>{response_text}</pre>"))
diff --git a/tests/api_test/requirements.txt b/tests/api_test/requirements.txt
index e03da1947..656a41dbb 100644
--- a/tests/api_test/requirements.txt
+++ b/tests/api_test/requirements.txt
@@ -4,3 +4,10 @@ pytest-html>=4.0.0
requests>=2.28.0
psutil>=5.9.0
aiohttp>=3.8.0
+fpdf2>=2.7.0
+python-docx>=0.8.11
+python-pptx>=0.6.21
+openpyxl>=3.1.0
+ebooklib>=0.18
+Pillow>=9.0.0
+# ffmpeg is a system dependency, install via: apt-get install ffmpeg (Ubuntu) or brew install ffmpeg (macOS)
diff --git a/tests/api_test/scenarios/resources_retrieval/build_test_helpers.py b/tests/api_test/scenarios/resources_retrieval/build_test_helpers.py
new file mode 100644
index 000000000..9961708e0
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/build_test_helpers.py
@@ -0,0 +1,270 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+
+def create_test_file(content=None, suffix=".txt", filename=None):
+ if content is None:
+ content = (
+ f"测试文件内容 - {uuid.uuid4()}\n这是一个用于构建测试的临时文件。\n包含一些测试数据。"
+ )
+
+ temp_dir = tempfile.mkdtemp()
+ fname = filename or f"test_file_{str(uuid.uuid4())[:8]}{suffix}"
+ test_file_path = os.path.join(temp_dir, fname)
+
+ with open(test_file_path, "w", encoding="utf-8") as f:
+ f.write(content)
+
+ return test_file_path, temp_dir
+
+
+def create_test_directory(file_count=3, nested=True):
+ temp_dir = tempfile.mkdtemp()
+
+ for i in range(file_count):
+ file_path = os.path.join(temp_dir, f"file_{i}.txt")
+ with open(file_path, "w", encoding="utf-8") as f:
+ f.write(f"测试文件 {i} 的内容\n一些测试数据 {uuid.uuid4()}")
+
+ if nested:
+ subdir = os.path.join(temp_dir, "subdir")
+ os.makedirs(subdir)
+ with open(os.path.join(subdir, "nested_file.txt"), "w", encoding="utf-8") as f:
+ f.write("嵌套文件的内容")
+
+ return temp_dir
+
+
+def create_test_zip(file_count=3):
+ import zipfile
+
+ temp_dir = tempfile.mkdtemp()
+ zip_path = os.path.join(temp_dir, f"test_archive_{str(uuid.uuid4())[:8]}.zip")
+
+ with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zf:
+ for i in range(file_count):
+ content = f"压缩包内文件 {i} 的内容\n测试数据 {uuid.uuid4()}"
+ zf.writestr(f"file_{i}.txt", content)
+ zf.writestr("subdir/nested_file.txt", "压缩包内嵌套文件的内容")
+
+ return zip_path, temp_dir
+
+
+def cleanup_temp_dir(temp_dir):
+ if temp_dir and os.path.exists(temp_dir):
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+
+def assert_root_uri_valid(root_uri):
+ assert root_uri, "add_resource 必须返回 root_uri"
+ assert root_uri.startswith("viking://resources/"), (
+ f"root_uri 应以 viking://resources/ 开头, 实际: {root_uri}"
+ )
+
+
+def assert_resource_findable(api_client, root_uri, keyword, timeout=90, poll_interval=3):
+ import time
+
+ api_client.wait_processed()
+
+ start = time.time()
+ poll_count = 0
+ while time.time() - start < timeout:
+ poll_count += 1
+
+ if poll_count % 3 == 0:
+ api_client.wait_processed()
+
+ find_resp = api_client.find(keyword, target_uri=root_uri)
+ if find_resp.status_code == 200:
+ find_data = find_resp.json()
+ if find_data.get("status") == "ok":
+ search_result = find_data.get("result", {})
+ for field in ["resources", "memories", "matches"]:
+ items = search_result.get(field, [])
+ if not isinstance(items, list):
+ continue
+ for item in items:
+ if isinstance(item, dict) and "uri" in item and root_uri in item["uri"]:
+ return True
+ elif isinstance(item, str) and root_uri in item:
+ return True
+
+ if time.time() - start >= timeout:
+ break
+ time.sleep(poll_interval)
+
+ return False
+
+
+def assert_resource_indexed(api_client, root_uri, keyword, timeout=120):
+ found = assert_resource_findable(api_client, root_uri, keyword, timeout=timeout)
+ assert found, f"搜索应能检索到资源 {root_uri}, 关键词: {keyword}"
+ return found
+
+
+def _extract_error_message(data):
+ if not isinstance(data, dict):
+ return str(data)
+ error = data.get("error")
+ if isinstance(error, dict):
+ return error.get("message", "") + " " + error.get("code", "")
+ errors = data.get("errors", [])
+ if isinstance(errors, list):
+ return " ".join(str(e) for e in errors)
+ return str(data.get("message", data.get("msg", "")))
+
+
+def assert_source_format(api_client, root_uri, expected_format):
+ if isinstance(expected_format, str):
+ expected_format = [expected_format]
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200, f"fs_stat 应返回200, root_uri: {root_uri}"
+ stat_data = stat_resp.json()
+ stat_result = stat_data.get("result")
+
+ if isinstance(stat_result, dict):
+ actual_format = stat_result.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ tree_resp = api_client.fs_tree(root_uri)
+ if tree_resp.status_code == 200:
+ tree_data = tree_resp.json()
+ tree_result = tree_data.get("result")
+ if isinstance(tree_result, dict):
+ actual_format = tree_result.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+ children = tree_result.get("children", [])
+ elif isinstance(tree_result, list):
+ children = tree_result
+ else:
+ children = []
+
+ for child in children:
+ if isinstance(child, dict):
+ actual_format = child.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ ls_resp = api_client.fs_ls(root_uri, recursive=True)
+ if ls_resp.status_code == 200:
+ ls_data = ls_resp.json()
+ ls_result = ls_data.get("result")
+ items = ls_result if isinstance(ls_result, list) else []
+ for item in items:
+ if isinstance(item, dict):
+ actual_format = item.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ ext_format_map = {
+ ".txt": ["text", "markdown"],
+ ".text": ["text"],
+ ".md": ["markdown"],
+ ".markdown": ["markdown"],
+ ".pdf": ["pdf"],
+ ".html": ["html"],
+ ".htm": ["html"],
+ ".docx": ["docx"],
+ ".doc": ["doc"],
+ ".pptx": ["pptx"],
+ ".xlsx": ["xlsx"],
+ ".epub": ["epub"],
+ ".zip": ["zip"],
+ ".png": ["image"],
+ ".jpg": ["image"],
+ ".jpeg": ["image"],
+ ".mp3": ["audio"],
+ ".wav": ["audio"],
+ ".mp4": ["video"],
+ }
+ if ls_resp.status_code == 200:
+ ls_data = ls_resp.json()
+ ls_result = ls_data.get("result")
+ items = ls_result if isinstance(ls_result, list) else []
+ for item in items:
+ if isinstance(item, dict):
+ rel_path = item.get("rel_path", "")
+ uri = item.get("uri", "")
+ for ext, fmt_list in ext_format_map.items():
+ if rel_path.endswith(ext) or uri.endswith(ext):
+ assert any(f in expected_format for f in fmt_list), (
+ f"根据扩展名 {ext} 推断 source_format 应为 {fmt_list}, 期望: {expected_format}"
+ )
+ return
+
+
+def assert_tree_has_child_nodes(api_client, root_uri, min_nodes=1):
+ tree_resp = api_client.fs_tree(root_uri)
+ assert tree_resp.status_code == 200, f"fs_tree 应返回200, root_uri: {root_uri}"
+ tree_data = tree_resp.json()
+ tree_result = tree_data.get("result")
+
+ if isinstance(tree_result, list):
+ children = tree_result
+ elif isinstance(tree_result, dict):
+ children = tree_result.get("children", [])
+ else:
+ children = []
+
+ assert len(children) >= min_nodes, (
+ f"tree 应至少包含 {min_nodes} 个子节点, 实际: {len(children)}"
+ )
+
+
+def assert_content_no_html_tags(api_client, root_uri):
+ grep_resp = api_client.grep(root_uri, pattern="<[a-z][a-z0-9]*[^>]*>")
+ if grep_resp.status_code != 200:
+ return
+ grep_data = grep_resp.json()
+ grep_result = grep_data.get("result", {})
+ matches = grep_result.get("matches", [])
+ html_matches = [m for m in matches if isinstance(m, dict) and "<" in str(m.get("line", ""))]
+ assert len(html_matches) == 0, (
+ f"HTML内容应被剥离, 但仍包含HTML标签: {[m.get('line', '')[:80] for m in html_matches[:3]]}"
+ )
+
+
+def assert_overview_or_abstract_contains(
+ api_client, root_uri, keyword, timeout=60, poll_interval=3
+):
+ import time
+
+ start = time.time()
+ while time.time() - start < timeout:
+ overview_resp = api_client.get_overview(root_uri)
+ if overview_resp.status_code == 200:
+ overview_data = overview_resp.json()
+ overview_text = str(overview_data.get("result", ""))
+ if keyword.lower() in overview_text.lower():
+ return True
+
+ abstract_resp = api_client.get_abstract(root_uri)
+ if abstract_resp.status_code == 200:
+ abstract_data = abstract_resp.json()
+ abstract_text = str(abstract_data.get("result", ""))
+ if keyword.lower() in abstract_text.lower():
+ return True
+
+ if time.time() - start >= timeout:
+ break
+ time.sleep(poll_interval)
+
+ return False
diff --git a/tests/api_test/scenarios/resources_retrieval/conftest.py b/tests/api_test/scenarios/resources_retrieval/conftest.py
new file mode 100644
index 000000000..e669d95eb
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/conftest.py
@@ -0,0 +1,4 @@
+import os
+import sys
+
+sys.path.insert(0, os.path.dirname(__file__))
diff --git a/tests/api_test/scenarios/resources_retrieval/test_build_document_resources.py b/tests/api_test/scenarios/resources_retrieval/test_build_document_resources.py
new file mode 100644
index 000000000..59bcbf6b5
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/test_build_document_resources.py
@@ -0,0 +1,109 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+import pytest
+from build_test_helpers import (
+ assert_content_no_html_tags,
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+)
+
+
+def _create_html_file():
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"html_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ html_path = os.path.join(temp_dir, f"test_{random_id}.html")
+
+ content = (
+ f"<html><head><title>HTML Test {random_id}</title></head>"
+ f"<body><h1>HTML测试标题 {random_id}</h1>"
+ f"<p>包含唯一关键词:{unique_keyword}</p>"
+ f"<p>用于验证HTML文件构建产物。</p>"
+ f"</body></html>"
+ )
+ with open(html_path, "w", encoding="utf-8") as f:
+ f.write(content)
+
+ return html_path, temp_dir, unique_keyword
+
+
+def _create_pptx_file():
+ try:
+ from pptx import Presentation
+ except ImportError:
+ pytest.skip("python-pptx 未安装,跳过 PPTX 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"pptx_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ pptx_path = os.path.join(temp_dir, f"test_{random_id}.pptx")
+
+ prs = Presentation()
+ slide = prs.slides.add_slide(prs.slide_layouts[1])
+ title = slide.shapes.title
+ title.text = f"PPT测试标题 {random_id}"
+ body = slide.placeholders[1]
+ body.text = f"包含唯一关键词:{unique_keyword}\n用于验证PPT文件构建产物。"
+ prs.save(pptx_path)
+
+ return pptx_path, temp_dir, unique_keyword
+
+
+class TestBuildDocumentResources:
+ """TC-B04, B07 文档类资源构建测试(快速用例,≤20s)"""
+
+ def test_build_html_file(self, api_client):
+ """TC-B04 HTML文件构建:验证 .html 文件添加后 source_format=html 且标签被剥离"""
+ html_path, temp_dir, unique_keyword = _create_html_file()
+ try:
+ response = api_client.add_resource(path=html_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["html", "markdown"])
+
+ assert_content_no_html_tags(api_client, root_uri)
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B04 HTML文件构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_pptx_file(self, api_client):
+ """TC-B07 PowerPoint构建:验证 .pptx 文件添加后 source_format=pptx 且内容可检索"""
+ pptx_path, temp_dir, unique_keyword = _create_pptx_file()
+ try:
+ response = api_client.add_resource(path=pptx_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["pptx", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B07 PowerPoint构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval/test_build_error_handling.py b/tests/api_test/scenarios/resources_retrieval/test_build_error_handling.py
new file mode 100644
index 000000000..72a97d7d5
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/test_build_error_handling.py
@@ -0,0 +1,158 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+from build_test_helpers import (
+ _extract_error_message,
+ assert_root_uri_valid,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildErrorHandling:
+ """TC-E01, E08, E09, E11, E16 异常与边界测试(快速用例,≤20s)"""
+
+ def test_error_remote_404(self, api_client):
+ """TC-E01 远端404不存在:验证 404 URL 返回错误且不崩溃,错误信息应包含状态码"""
+ url_404 = "https://httpbin.org/status/404"
+
+ response = api_client.add_resource(path=url_404, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert "404" in error_msg or "not found" in error_msg or "error" in error_msg, (
+ f"404错误信息应包含 404/not found/error, 实际: {error_msg}"
+ )
+ print("✓ TC-E01 远端404不存在处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E01 远端404处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"404 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_dns_resolve_failure(self, api_client):
+ """TC-E08 DNS解析失败:验证不存在的域名返回错误且不挂起"""
+ bad_dns_url = "https://nonexistent.domain.invalid.for.test/page"
+
+ response = api_client.add_resource(path=bad_dns_url, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "resolve" in error_msg
+ or "hostname" in error_msg
+ or "dns" in error_msg
+ or "error" in error_msg
+ ), f"DNS失败错误信息应包含 resolve/hostname/dns/error, 实际: {error_msg}"
+ print("✓ TC-E08 DNS解析失败处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E08 DNS解析失败处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"DNS失败 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_ssh_url_invalid_format(self, api_client):
+ """TC-E09 SSH URL格式错误:验证 git@invalid (无冒号) 返回 InvalidArgumentError"""
+ invalid_ssh_url = "git@invalid"
+
+ response = api_client.add_resource(path=invalid_ssh_url, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "invalid" in error_msg
+ or "ssh" in error_msg
+ or "uri" in error_msg
+ or "colon" in error_msg
+ or "error" in error_msg
+ or "permission" in error_msg
+ ), f"SSH格式错误应包含 invalid/ssh/uri/colon/error/permission, 实际: {error_msg}"
+ print("✓ TC-E09 SSH URL格式错误处理通过")
+ return
+
+ if data.get("status") == "ok":
+ print("✓ TC-E09 SSH URL格式错误处理通过(服务端降级)")
+ return
+
+ raise AssertionError(f"SSH URL格式错误应返回 error 或降级, 实际: {data.get('status')}")
+
+ def test_error_non_resources_scope_rejected(self, api_client):
+ """TC-E11 非resources scope拒绝:验证 to=viking://sessions/xxx 返回错误"""
+ random_id = str(uuid.uuid4())[:8]
+ test_content = f"scope测试内容 {random_id}"
+ test_file_path, temp_dir = create_test_file(content=test_content, suffix=".txt")
+ try:
+ response = api_client.add_resource(
+ path=test_file_path,
+ to="viking://sessions/test_session",
+ wait=True,
+ )
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "scope" in error_msg
+ or "resources" in error_msg
+ or "invalid" in error_msg
+ or "permission" in error_msg
+ or "internal" in error_msg
+ ), f"scope拒绝应包含 scope/resources/invalid/permission/internal, 实际: {error_msg}"
+ print("✓ TC-E11 非resources scope拒绝通过")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri", "")
+ assert "sessions" not in root_uri, (
+ f"非resources scope不应成功写入sessions, root_uri: {root_uri}"
+ )
+ print("✓ TC-E11 非resources scope处理通过(服务端重定向)")
+ return
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_error_corrupted_zip(self, api_client):
+ """TC-E16 损坏的ZIP文件:验证伪造 .zip 文件回退或报错且不崩溃"""
+ random_id = str(uuid.uuid4())[:8]
+
+ temp_dir = tempfile.mkdtemp()
+ zip_path = os.path.join(temp_dir, f"corrupted_{random_id}.zip")
+ with open(zip_path, "w", encoding="utf-8") as f:
+ f.write("这不是一个真正的ZIP文件内容")
+
+ try:
+ response = api_client.add_resource(path=zip_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") in ("ok", "error"), (
+ f"损坏ZIP应返回 ok 或 error, 实际: {data.get('status')}"
+ )
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+
+ print("✓ TC-E16 损坏的ZIP文件处理通过")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval/test_build_media_resources.py b/tests/api_test/scenarios/resources_retrieval/test_build_media_resources.py
new file mode 100644
index 000000000..6473580ed
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/test_build_media_resources.py
@@ -0,0 +1,95 @@
+import os
+import shutil
+import struct
+import tempfile
+import uuid
+import wave
+
+import pytest
+from build_test_helpers import assert_root_uri_valid, assert_source_format
+
+
+def _create_image_file(ext=".png"):
+ try:
+ from PIL import Image
+ except ImportError:
+ pytest.skip("Pillow 未安装,跳过图片构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ img_path = os.path.join(temp_dir, f"test_{random_id}{ext}")
+
+ img = Image.new("RGB", (100, 100), color=(73, 109, 137))
+ img.save(img_path)
+
+ return img_path, temp_dir, random_id
+
+
+def _create_audio_file(ext=".wav"):
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ audio_path = os.path.join(temp_dir, f"test_{random_id}{ext}")
+
+ with wave.open(audio_path, "w") as wf:
+ wf.setnchannels(1)
+ wf.setsampwidth(2)
+ wf.setframerate(44100)
+ frames = b""
+ for _i in range(44100):
+ value = int(32767 * 0.5)
+ frames += struct.pack("<h", value)
+ wf.writeframes(frames)
+
+ return audio_path, temp_dir, random_id
+
+
+class TestBuildMediaResources:
+ """TC-B13(PNG), B14(WAV) 媒体类资源构建测试(快速用例,≤20s)"""
+
+ def test_build_image_png(self, api_client):
+ """TC-B13 图片构建(PNG):验证 .png 文件添加后产物路径含 /images/ 且 source_format=image"""
+ img_path, temp_dir, random_id = _create_image_file(".png")
+ try:
+ response = api_client.add_resource(path=img_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/images/" in root_uri, f"图片 root_uri 应含 /images/, 实际: {root_uri}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "image")
+
+ print(f"✓ TC-B13 图片构建(PNG)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_audio_wav(self, api_client):
+ """TC-B14 音频构建(WAV):验证 .wav 文件添加后产物路径含 /audio/ 且 source_format=audio"""
+ audio_path, temp_dir, random_id = _create_audio_file(".wav")
+ try:
+ response = api_client.add_resource(path=audio_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/audio/" in root_uri, f"音频 root_uri 应含 /audio/, 实际: {root_uri}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "audio")
+
+ print(f"✓ TC-B14 音频构建(WAV)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval/test_build_text_resources.py b/tests/api_test/scenarios/resources_retrieval/test_build_text_resources.py
new file mode 100644
index 000000000..8db0ab9fc
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/test_build_text_resources.py
@@ -0,0 +1,73 @@
+import uuid
+
+from build_test_helpers import (
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+ assert_tree_has_child_nodes,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildTextResources:
+ """TC-B02, B15 文本类资源构建测试(快速用例,≤20s)"""
+
+ def test_build_markdown_file(self, api_client):
+ """TC-B02 Markdown文件构建:验证 .md 文件添加后 heading 结构保留为子节点"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"md_keyword_{random_id}"
+ content = (
+ f"# Markdown测试标题 {random_id}\n\n"
+ f"包含唯一关键词:{unique_keyword}\n\n"
+ f"## 第一节\n\n第一段内容。\n\n"
+ f"## 第二节\n\n第二段内容。\n\n"
+ f"### 子节\n\n子节内容。\n"
+ )
+
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".md")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "markdown")
+
+ assert_tree_has_child_nodes(api_client, root_uri, min_nodes=1)
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B02 Markdown文件构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_empty_file(self, api_client):
+ """TC-B15 空文件构建:验证空 .txt 文件添加不崩溃并返回合理结果"""
+ test_file_path, temp_dir = create_test_file(content="", suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") in ("ok", "error"), (
+ f"空文件构建应返回 ok 或 error, 实际: {data.get('status')}"
+ )
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+
+ print("✓ TC-B15 空文件构建通过")
+ finally:
+ cleanup_temp_dir(temp_dir)
diff --git a/tests/api_test/scenarios/resources_retrieval/test_build_uri_params.py b/tests/api_test/scenarios/resources_retrieval/test_build_uri_params.py
new file mode 100644
index 000000000..6ea4fb596
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval/test_build_uri_params.py
@@ -0,0 +1,89 @@
+import uuid
+
+from build_test_helpers import (
+ _extract_error_message,
+ assert_root_uri_valid,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildUriParams:
+ """TC-E11(参数部分), TC-E18 URI 参数与边界测试(快速用例,≤20s)"""
+
+ def test_build_with_parent_param(self, api_client):
+ """TC-E18 指定parent参数构建:验证 root_uri 以 parent 为前缀"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"parent_keyword_{random_id}"
+ parent_uri = f"viking://resources/parent_test_{random_id}"
+
+ mkdir_resp = api_client.fs_mkdir(parent_uri)
+ assert mkdir_resp.status_code == 200, (
+ f"无法创建parent目录, fs_mkdir返回: {mkdir_resp.status_code}, body: {mkdir_resp.text}"
+ )
+
+ content = f"parent参数测试 {random_id}\n包含唯一关键词:{unique_keyword}"
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, parent=parent_uri, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data)
+ raise AssertionError(f"parent参数构建失败, 服务端不支持parent参数: {error_msg}")
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ raise AssertionError(f"parent参数构建失败(内层错误): {inner_msg}")
+
+ root_uri = result.get("root_uri") if isinstance(result, dict) else None
+ assert_root_uri_valid(root_uri)
+ assert root_uri.startswith(parent_uri), (
+ f"指定parent时 root_uri 应以 parent 为前缀, parent: {parent_uri}, 实际: {root_uri}"
+ )
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-E18 指定parent参数构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_non_resources_scope_rejected(self, api_client):
+ """TC-E11 非resources scope拒绝:验证 to=viking://sessions/xxx 返回错误"""
+ random_id = str(uuid.uuid4())[:8]
+ test_content = f"scope测试内容 {random_id}"
+ test_file_path, temp_dir = create_test_file(content=test_content, suffix=".txt")
+ try:
+ response = api_client.add_resource(
+ path=test_file_path,
+ to="viking://sessions/test_session",
+ wait=True,
+ )
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "scope" in error_msg
+ or "resources" in error_msg
+ or "invalid" in error_msg
+ or "permission" in error_msg
+ or "internal" in error_msg
+ ), f"scope拒绝应包含 scope/resources/invalid/permission/internal, 实际: {error_msg}"
+ print("✓ TC-E11 非resources scope拒绝通过")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri", "")
+ assert "sessions" not in root_uri, (
+ f"非resources scope不应成功写入sessions, root_uri: {root_uri}"
+ )
+ print("✓ TC-E11 非resources scope处理通过(服务端重定向)")
+ return
+ finally:
+ cleanup_temp_dir(temp_dir)
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/build_test_helpers.py b/tests/api_test/scenarios/resources_retrieval_slow/build_test_helpers.py
new file mode 100644
index 000000000..9961708e0
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/build_test_helpers.py
@@ -0,0 +1,270 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+
+def create_test_file(content=None, suffix=".txt", filename=None):
+ if content is None:
+ content = (
+ f"测试文件内容 - {uuid.uuid4()}\n这是一个用于构建测试的临时文件。\n包含一些测试数据。"
+ )
+
+ temp_dir = tempfile.mkdtemp()
+ fname = filename or f"test_file_{str(uuid.uuid4())[:8]}{suffix}"
+ test_file_path = os.path.join(temp_dir, fname)
+
+ with open(test_file_path, "w", encoding="utf-8") as f:
+ f.write(content)
+
+ return test_file_path, temp_dir
+
+
+def create_test_directory(file_count=3, nested=True):
+ temp_dir = tempfile.mkdtemp()
+
+ for i in range(file_count):
+ file_path = os.path.join(temp_dir, f"file_{i}.txt")
+ with open(file_path, "w", encoding="utf-8") as f:
+ f.write(f"测试文件 {i} 的内容\n一些测试数据 {uuid.uuid4()}")
+
+ if nested:
+ subdir = os.path.join(temp_dir, "subdir")
+ os.makedirs(subdir)
+ with open(os.path.join(subdir, "nested_file.txt"), "w", encoding="utf-8") as f:
+ f.write("嵌套文件的内容")
+
+ return temp_dir
+
+
+def create_test_zip(file_count=3):
+ import zipfile
+
+ temp_dir = tempfile.mkdtemp()
+ zip_path = os.path.join(temp_dir, f"test_archive_{str(uuid.uuid4())[:8]}.zip")
+
+ with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zf:
+ for i in range(file_count):
+ content = f"压缩包内文件 {i} 的内容\n测试数据 {uuid.uuid4()}"
+ zf.writestr(f"file_{i}.txt", content)
+ zf.writestr("subdir/nested_file.txt", "压缩包内嵌套文件的内容")
+
+ return zip_path, temp_dir
+
+
+def cleanup_temp_dir(temp_dir):
+ if temp_dir and os.path.exists(temp_dir):
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+
+def assert_root_uri_valid(root_uri):
+ assert root_uri, "add_resource 必须返回 root_uri"
+ assert root_uri.startswith("viking://resources/"), (
+ f"root_uri 应以 viking://resources/ 开头, 实际: {root_uri}"
+ )
+
+
+def assert_resource_findable(api_client, root_uri, keyword, timeout=90, poll_interval=3):
+ import time
+
+ api_client.wait_processed()
+
+ start = time.time()
+ poll_count = 0
+ while time.time() - start < timeout:
+ poll_count += 1
+
+ if poll_count % 3 == 0:
+ api_client.wait_processed()
+
+ find_resp = api_client.find(keyword, target_uri=root_uri)
+ if find_resp.status_code == 200:
+ find_data = find_resp.json()
+ if find_data.get("status") == "ok":
+ search_result = find_data.get("result", {})
+ for field in ["resources", "memories", "matches"]:
+ items = search_result.get(field, [])
+ if not isinstance(items, list):
+ continue
+ for item in items:
+ if isinstance(item, dict) and "uri" in item and root_uri in item["uri"]:
+ return True
+ elif isinstance(item, str) and root_uri in item:
+ return True
+
+ if time.time() - start >= timeout:
+ break
+ time.sleep(poll_interval)
+
+ return False
+
+
+def assert_resource_indexed(api_client, root_uri, keyword, timeout=120):
+ found = assert_resource_findable(api_client, root_uri, keyword, timeout=timeout)
+ assert found, f"搜索应能检索到资源 {root_uri}, 关键词: {keyword}"
+ return found
+
+
+def _extract_error_message(data):
+ if not isinstance(data, dict):
+ return str(data)
+ error = data.get("error")
+ if isinstance(error, dict):
+ return error.get("message", "") + " " + error.get("code", "")
+ errors = data.get("errors", [])
+ if isinstance(errors, list):
+ return " ".join(str(e) for e in errors)
+ return str(data.get("message", data.get("msg", "")))
+
+
+def assert_source_format(api_client, root_uri, expected_format):
+ if isinstance(expected_format, str):
+ expected_format = [expected_format]
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200, f"fs_stat 应返回200, root_uri: {root_uri}"
+ stat_data = stat_resp.json()
+ stat_result = stat_data.get("result")
+
+ if isinstance(stat_result, dict):
+ actual_format = stat_result.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ tree_resp = api_client.fs_tree(root_uri)
+ if tree_resp.status_code == 200:
+ tree_data = tree_resp.json()
+ tree_result = tree_data.get("result")
+ if isinstance(tree_result, dict):
+ actual_format = tree_result.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+ children = tree_result.get("children", [])
+ elif isinstance(tree_result, list):
+ children = tree_result
+ else:
+ children = []
+
+ for child in children:
+ if isinstance(child, dict):
+ actual_format = child.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ ls_resp = api_client.fs_ls(root_uri, recursive=True)
+ if ls_resp.status_code == 200:
+ ls_data = ls_resp.json()
+ ls_result = ls_data.get("result")
+ items = ls_result if isinstance(ls_result, list) else []
+ for item in items:
+ if isinstance(item, dict):
+ actual_format = item.get("source_format")
+ if actual_format:
+ assert actual_format in expected_format, (
+ f"source_format 应为 {expected_format}, 实际: {actual_format}"
+ )
+ return
+
+ ext_format_map = {
+ ".txt": ["text", "markdown"],
+ ".text": ["text"],
+ ".md": ["markdown"],
+ ".markdown": ["markdown"],
+ ".pdf": ["pdf"],
+ ".html": ["html"],
+ ".htm": ["html"],
+ ".docx": ["docx"],
+ ".doc": ["doc"],
+ ".pptx": ["pptx"],
+ ".xlsx": ["xlsx"],
+ ".epub": ["epub"],
+ ".zip": ["zip"],
+ ".png": ["image"],
+ ".jpg": ["image"],
+ ".jpeg": ["image"],
+ ".mp3": ["audio"],
+ ".wav": ["audio"],
+ ".mp4": ["video"],
+ }
+ if ls_resp.status_code == 200:
+ ls_data = ls_resp.json()
+ ls_result = ls_data.get("result")
+ items = ls_result if isinstance(ls_result, list) else []
+ for item in items:
+ if isinstance(item, dict):
+ rel_path = item.get("rel_path", "")
+ uri = item.get("uri", "")
+ for ext, fmt_list in ext_format_map.items():
+ if rel_path.endswith(ext) or uri.endswith(ext):
+ assert any(f in expected_format for f in fmt_list), (
+ f"根据扩展名 {ext} 推断 source_format 应为 {fmt_list}, 期望: {expected_format}"
+ )
+ return
+
+
+def assert_tree_has_child_nodes(api_client, root_uri, min_nodes=1):
+ tree_resp = api_client.fs_tree(root_uri)
+ assert tree_resp.status_code == 200, f"fs_tree 应返回200, root_uri: {root_uri}"
+ tree_data = tree_resp.json()
+ tree_result = tree_data.get("result")
+
+ if isinstance(tree_result, list):
+ children = tree_result
+ elif isinstance(tree_result, dict):
+ children = tree_result.get("children", [])
+ else:
+ children = []
+
+ assert len(children) >= min_nodes, (
+ f"tree 应至少包含 {min_nodes} 个子节点, 实际: {len(children)}"
+ )
+
+
+def assert_content_no_html_tags(api_client, root_uri):
+ grep_resp = api_client.grep(root_uri, pattern="<[a-z][a-z0-9]*[^>]*>")
+ if grep_resp.status_code != 200:
+ return
+ grep_data = grep_resp.json()
+ grep_result = grep_data.get("result", {})
+ matches = grep_result.get("matches", [])
+ html_matches = [m for m in matches if isinstance(m, dict) and "<" in str(m.get("line", ""))]
+ assert len(html_matches) == 0, (
+ f"HTML内容应被剥离, 但仍包含HTML标签: {[m.get('line', '')[:80] for m in html_matches[:3]]}"
+ )
+
+
+def assert_overview_or_abstract_contains(
+ api_client, root_uri, keyword, timeout=60, poll_interval=3
+):
+ import time
+
+ start = time.time()
+ while time.time() - start < timeout:
+ overview_resp = api_client.get_overview(root_uri)
+ if overview_resp.status_code == 200:
+ overview_data = overview_resp.json()
+ overview_text = str(overview_data.get("result", ""))
+ if keyword.lower() in overview_text.lower():
+ return True
+
+ abstract_resp = api_client.get_abstract(root_uri)
+ if abstract_resp.status_code == 200:
+ abstract_data = abstract_resp.json()
+ abstract_text = str(abstract_data.get("result", ""))
+ if keyword.lower() in abstract_text.lower():
+ return True
+
+ if time.time() - start >= timeout:
+ break
+ time.sleep(poll_interval)
+
+ return False
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/conftest.py b/tests/api_test/scenarios/resources_retrieval_slow/conftest.py
new file mode 100644
index 000000000..e669d95eb
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/conftest.py
@@ -0,0 +1,4 @@
+import os
+import sys
+
+sys.path.insert(0, os.path.dirname(__file__))
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_archive_resources.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_archive_resources.py
new file mode 100644
index 000000000..965631e17
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_archive_resources.py
@@ -0,0 +1,112 @@
+import os
+import shutil
+import tempfile
+import uuid
+import zipfile
+
+from build_test_helpers import assert_resource_indexed, assert_root_uri_valid, assert_source_format
+
+
+class TestBuildArchiveResources:
+ """TC-B10~B12 压缩包与目录类资源构建测试"""
+
+ def test_build_zip_file(self, api_client):
+ """TC-B10 ZIP压缩包构建:验证 .zip 文件添加后解压并按目录结构处理"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"zip_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ zip_path = os.path.join(temp_dir, f"test_{random_id}.zip")
+
+ with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zf:
+ zf.writestr("file_0.txt", f"压缩包内文件0\n包含唯一关键词:{unique_keyword}")
+ zf.writestr("file_1.txt", f"压缩包内文件1\n测试数据 {random_id}")
+ zf.writestr("subdir/nested.txt", "嵌套文件内容")
+
+ try:
+ response = api_client.add_resource(path=zip_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["zip", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B10 ZIP压缩包构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_directory(self, api_client):
+ """TC-B11 目录递归构建:验证目录添加后子文件全部被索引"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"dir_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+
+ for i in range(3):
+ file_path = os.path.join(temp_dir, f"file_{i}.txt")
+ with open(file_path, "w", encoding="utf-8") as f:
+ f.write(f"目录内文件 {i}\n测试数据 {random_id}")
+
+ subdir = os.path.join(temp_dir, "subdir")
+ os.makedirs(subdir)
+ with open(os.path.join(subdir, "nested_file.txt"), "w", encoding="utf-8") as f:
+ f.write(f"嵌套文件内容\n包含唯一关键词:{unique_keyword}")
+
+ try:
+ response = api_client.add_resource(path=temp_dir, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["directory", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B11 目录递归构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_code_repository_url(self, api_client):
+ """TC-B12 代码仓库URL构建:验证 GitHub 仓库 URL 添加后 source_format=repository 且 URI 含 org/repo"""
+ repo_url = "https://github.com/volcengine/OpenViking"
+
+ response = api_client.add_resource(path=repo_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "volcengine" in root_uri and "OpenViking" in root_uri, (
+ f"代码仓库 root_uri 应含 org/repo, 实际: {root_uri}"
+ )
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("code_repository", None), (
+ f"meta.url_type 应为 code_repository, 实际: {meta.get('url_type')}"
+ )
+
+ assert_source_format(api_client, root_uri, ["repository", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-B12 代码仓库URL构建通过, root_uri: {root_uri}")
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_document_resources_slow.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_document_resources_slow.py
new file mode 100644
index 000000000..db47459a1
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_document_resources_slow.py
@@ -0,0 +1,332 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+import pytest
+from build_test_helpers import (
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+)
+
+
+def _create_pdf_file():
+ try:
+ from fpdf import FPDF
+ except ImportError:
+ pytest.skip("fpdf 未安装,跳过 PDF 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"pdf_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ pdf_path = os.path.join(temp_dir, f"test_{random_id}.pdf")
+
+ pdf = FPDF()
+ pdf.add_page()
+ pdf.set_font("Helvetica", size=12)
+ pdf.cell(200, 10, text=f"PDF Test Document {random_id}", new_x="LMARGIN", new_y="NEXT")
+ pdf.cell(200, 10, text=f"Unique keyword: {unique_keyword}", new_x="LMARGIN", new_y="NEXT")
+ pdf.cell(
+ 200, 10, text="This is a test PDF file for build validation.", new_x="LMARGIN", new_y="NEXT"
+ )
+ pdf.output(pdf_path)
+
+ return pdf_path, temp_dir, unique_keyword
+
+
+def _create_docx_file():
+ try:
+ from docx import Document
+ except ImportError:
+ pytest.skip("python-docx 未安装,跳过 DOCX 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"docx_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ docx_path = os.path.join(temp_dir, f"test_{random_id}.docx")
+
+ doc = Document()
+ doc.add_heading(f"Word测试标题 {random_id}", level=1)
+ doc.add_paragraph(f"包含唯一关键词:{unique_keyword}")
+ doc.add_paragraph("用于验证Word文档构建产物。")
+ doc.save(docx_path)
+
+ return docx_path, temp_dir, unique_keyword
+
+
+def _create_xlsx_file():
+ try:
+ from openpyxl import Workbook
+ except ImportError:
+ pytest.skip("openpyxl 未安装,跳过 XLSX 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"xlsx_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ xlsx_path = os.path.join(temp_dir, f"test_{random_id}.xlsx")
+
+ wb = Workbook()
+ ws = wb.active
+ ws.title = "TestSheet"
+ ws.append(["Column A", "Column B", "Column C"])
+ ws.append([f"数据1 {random_id}", unique_keyword, "数据3"])
+ ws.append(["数据4", "数据5", "数据6"])
+ wb.save(xlsx_path)
+
+ return xlsx_path, temp_dir, unique_keyword
+
+
+class TestBuildDocumentResourcesSlow:
+ """TC-B03~B09 文档类资源构建测试"""
+
+ def test_build_pdf_file(self, api_client):
+ """TC-B03 PDF文件构建:验证 .pdf 文件添加后 source_format=pdf 且内容可检索"""
+ pdf_path, temp_dir, unique_keyword = _create_pdf_file()
+ try:
+ response = api_client.add_resource(path=pdf_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["pdf", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B03 PDF文件构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_docx_file(self, api_client):
+ """TC-B05 Word文档构建:验证 .docx 文件添加后 source_format=docx 且内容可检索"""
+ docx_path, temp_dir, unique_keyword = _create_docx_file()
+ try:
+ response = api_client.add_resource(path=docx_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["docx", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B05 Word文档构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_xlsx_file(self, api_client):
+ """TC-B08 Excel构建:验证 .xlsx 文件添加后 source_format=xlsx 且表格数据可检索"""
+ xlsx_path, temp_dir, unique_keyword = _create_xlsx_file()
+ try:
+ response = api_client.add_resource(path=xlsx_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["xlsx", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B08 Excel构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_html_file(self, api_client):
+ """TC-B04 HTML文件构建:验证 .html 文件添加后 source_format=html 且标签被剥离"""
+ from build_test_helpers import assert_content_no_html_tags
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"html_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ html_path = os.path.join(temp_dir, f"test_{random_id}.html")
+
+ content = (
+ f"<html><head><title>HTML Test {random_id}</title></head>"
+ f"<body><h1>HTML测试标题 {random_id}</h1>"
+ f"<p>包含唯一关键词:{unique_keyword}</p>"
+ f"<p>用于验证HTML文件构建产物。</p>"
+ f"</body></html>"
+ )
+ with open(html_path, "w", encoding="utf-8") as f:
+ f.write(content)
+
+ try:
+ response = api_client.add_resource(path=html_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["html", "markdown"])
+
+ assert_content_no_html_tags(api_client, root_uri)
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B04 HTML文件构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_legacy_doc_file(self, api_client):
+ """TC-B06 旧版Word(.doc)构建:验证 .doc 文件添加后 source_format=doc 且内容可检索"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"doc_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ doc_path = os.path.join(temp_dir, f"test_{random_id}.doc")
+
+ try:
+ from docx import Document
+
+ doc = Document()
+ doc.add_heading(f"旧版Word测试标题 {random_id}", level=1)
+ doc.add_paragraph(f"包含唯一关键词:{unique_keyword}")
+ doc.save(doc_path)
+ except ImportError:
+ pytest.skip("python-docx 未安装,跳过旧版Word构建测试")
+
+ try:
+ response = api_client.add_resource(path=doc_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["doc", "docx", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B06 旧版Word(.doc)构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_pptx_file(self, api_client):
+ """TC-B07 PowerPoint构建:验证 .pptx 文件添加后 source_format=pptx 且内容可检索"""
+ try:
+ from pptx import Presentation
+ except ImportError:
+ pytest.skip("python-pptx 未安装,跳过 PPTX 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"pptx_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ pptx_path = os.path.join(temp_dir, f"test_{random_id}.pptx")
+
+ prs = Presentation()
+ slide = prs.slides.add_slide(prs.slide_layouts[1])
+ title = slide.shapes.title
+ title.text = f"PPT测试标题 {random_id}"
+ body = slide.placeholders[1]
+ body.text = f"包含唯一关键词:{unique_keyword}\n用于验证PPT文件构建产物。"
+ prs.save(pptx_path)
+
+ try:
+ response = api_client.add_resource(path=pptx_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["pptx", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B07 PowerPoint构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_epub_file(self, api_client):
+ """TC-B09 EPUB构建:验证 .epub 文件添加后 source_format=epub 且内容可检索"""
+ try:
+ from ebooklib import epub
+ except ImportError:
+ pytest.skip("ebooklib 未安装,跳过 EPUB 构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"epub_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ epub_path = os.path.join(temp_dir, f"test_{random_id}.epub")
+
+ book = epub.EpubBook()
+ book.set_identifier(f"test-epub-{random_id}")
+ book.set_title(f"EPUB测试 {random_id}")
+ book.set_language("zh")
+
+ chapter = epub.EpubHtml(title="Chapter 1", file_name="chap_01.xhtml", lang="zh")
+ chapter.content = (
+ f"<html><body><h1>EPUB测试章节 {random_id}</h1>"
+ f"<p>包含唯一关键词:{unique_keyword}</p>"
+ f"<p>用于验证EPUB文件构建产物。</p>"
+ f"</body></html>"
+ )
+ book.add_item(chapter)
+ book.toc = (chapter,)
+ book.add_item(epub.EpubNcx())
+ book.add_item(epub.EpubNav())
+ epub.write_epub(epub_path, book, {})
+
+ try:
+ response = api_client.add_resource(path=epub_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["epub", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B09 EPUB构建通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_error_handling_slow.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_error_handling_slow.py
new file mode 100644
index 000000000..257ca7b0f
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_error_handling_slow.py
@@ -0,0 +1,429 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+from build_test_helpers import (
+ _extract_error_message,
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildErrorHandlingSlow:
+ """TC-E01~E16 异常与边界测试"""
+
+ def test_error_remote_404(self, api_client):
+ """TC-E01 远端404不存在:验证 404 URL 返回错误含状态码信息且不崩溃"""
+ url_404 = "https://httpbin.org/status/404"
+
+ response = api_client.add_resource(path=url_404, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert "404" in error_msg or "not found" in error_msg or "error" in error_msg, (
+ f"404错误信息应包含 404/not found/error, 实际: {error_msg}"
+ )
+ print("✓ TC-E01 远端404不存在处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors).lower()
+ assert (
+ "404" in inner_msg
+ or "not found" in inner_msg
+ or "failed" in inner_msg
+ or "error" in inner_msg
+ ), f"404内层错误应包含 404/not found/failed/error, 实际: {inner_msg}"
+ print(f"✓ TC-E01 远端404不存在处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E01 远端404处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"404 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_http_to_https_redirect(self, api_client):
+ """TC-E05 HTTP→HTTPS跳转:验证 http URL 自动跟随跳转且 root_uri 正常"""
+ redirect_url = "http://github.com/volcengine/OpenViking"
+
+ response = api_client.add_resource(path=redirect_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "volcengine" in root_uri and "OpenViking" in root_uri, (
+ f"跳转后 root_uri 应含 volcengine/OpenViking, 实际: {root_uri}"
+ )
+
+ print(f"✓ TC-E05 HTTP→HTTPS跳转通过, root_uri: {root_uri}")
+
+ def test_error_multi_redirect(self, api_client):
+ """TC-E06 多重跳转:验证短链 URL 自动跟随跳转且内容可检索"""
+ redirect_url = (
+ "https://httpbin.org/redirect-to?url=https://httpbin.org/html&status_code=302"
+ )
+
+ response = api_client.add_resource(path=redirect_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ assert_resource_indexed(api_client, root_uri, "httpbin")
+
+ print(f"✓ TC-E06 多重跳转通过, root_uri: {root_uri}")
+
+ def test_error_duplicate_resource_no_to(self, api_client):
+ """TC-E12 同名资源二次添加(无to):验证两次添加的 root_uri 不同(URI 附加后缀)"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"dup_keyword_{random_id}"
+ content = f"重复添加测试 {random_id}\n包含唯一关键词:{unique_keyword}"
+
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".txt")
+ try:
+ resp1 = api_client.add_resource(path=test_file_path, wait=True)
+ assert resp1.status_code == 200
+ data1 = resp1.json()
+ assert data1.get("status") == "ok"
+ root_uri_1 = data1.get("result", {}).get("root_uri")
+ assert_root_uri_valid(root_uri_1)
+
+ resp2 = api_client.add_resource(path=test_file_path, wait=True)
+ assert resp2.status_code == 200
+ data2 = resp2.json()
+ assert data2.get("status") == "ok"
+ root_uri_2 = data2.get("result", {}).get("root_uri")
+ assert_root_uri_valid(root_uri_2)
+
+ assert root_uri_1 != root_uri_2, (
+ f"同名资源二次添加(无to) root_uri 应不同, uri1: {root_uri_1}, uri2: {root_uri_2}"
+ )
+
+ print(f"✓ TC-E12 同名资源二次添加通过, uri1: {root_uri_1}, uri2: {root_uri_2}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_error_incremental_update_with_to(self, api_client):
+ """TC-E13 同to增量更新:验证同一 to 二次添加后 root_uri 不变且不报错"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"incr_keyword_{random_id}"
+ target_uri = f"viking://resources/incr_test_{random_id}"
+
+ content1 = f"增量更新测试1 {random_id}\n包含唯一关键词:{unique_keyword}_v1"
+ content2 = f"增量更新测试2 {random_id}\n包含唯一关键词:{unique_keyword}_v2"
+
+ file1, temp_dir1 = create_test_file(content=content1, suffix=".txt")
+ file2, temp_dir2 = create_test_file(content=content2, suffix=".txt")
+ try:
+ resp1 = api_client.add_resource(path=file1, to=target_uri, wait=True)
+ assert resp1.status_code == 200
+ data1 = resp1.json()
+ assert data1.get("status") == "ok"
+ root_uri_1 = data1.get("result", {}).get("root_uri")
+
+ resp2 = api_client.add_resource(path=file2, to=target_uri, wait=True)
+ assert resp2.status_code == 200
+ data2 = resp2.json()
+ assert data2.get("status") == "ok"
+ root_uri_2 = data2.get("result", {}).get("root_uri")
+
+ assert root_uri_1 == root_uri_2, (
+ f"同to增量更新 root_uri 应不变, uri1: {root_uri_1}, uri2: {root_uri_2}"
+ )
+
+ print(f"✓ TC-E13 同to增量更新通过, root_uri: {root_uri_1}")
+ finally:
+ cleanup_temp_dir(temp_dir1)
+ cleanup_temp_dir(temp_dir2)
+
+ def test_error_unsupported_file_type(self, api_client):
+ """TC-E15 不支持的文件类型:验证 .xyz 文件回退到 TextParser 且 source_format=text、内容可检索"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"xyz_keyword_{random_id}"
+
+ temp_dir = tempfile.mkdtemp()
+ file_path = os.path.join(temp_dir, f"test_{random_id}.xyz")
+ with open(file_path, "w", encoding="utf-8") as f:
+ f.write(f"不支持的文件类型测试 {random_id}\n包含唯一关键词:{unique_keyword}")
+
+ try:
+ response = api_client.add_resource(path=file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert "unsupported" in error_msg or "error" in error_msg or "type" in error_msg, (
+ f"不支持文件类型错误应包含 unsupported/error/type, 实际: {error_msg}"
+ )
+ print(f"✓ TC-E15 不支持的文件类型处理通过(服务端拒绝): {error_msg[:80]}")
+ return
+
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ assert (
+ "unsupported" in inner_msg.lower()
+ or "error" in inner_msg.lower()
+ or "parse" in inner_msg.lower()
+ ), f"不支持文件类型内层错误应包含 unsupported/error/parse, 实际: {inner_msg}"
+ print(f"✓ TC-E15 不支持的文件类型处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200, (
+ f"不支持文件类型 fs_stat 应返回200, root_uri: {root_uri}"
+ )
+
+ assert_source_format(api_client, root_uri, ["text", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-E15 不支持的文件类型处理通过(回退TextParser), root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_error_remote_403(self, api_client):
+ """TC-E02 远端403禁止访问:验证 403 URL 返回错误含状态码信息且不崩溃"""
+ url_403 = "https://httpbin.org/status/403"
+
+ response = api_client.add_resource(path=url_403, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert "403" in error_msg or "forbidden" in error_msg or "error" in error_msg, (
+ f"403错误信息应包含 403/forbidden/error, 实际: {error_msg}"
+ )
+ print("✓ TC-E02 远端403禁止访问处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors).lower()
+ assert (
+ "403" in inner_msg
+ or "forbidden" in inner_msg
+ or "failed" in inner_msg
+ or "error" in inner_msg
+ ), f"403内层错误应包含 403/forbidden/failed/error, 实际: {inner_msg}"
+ print(f"✓ TC-E02 远端403禁止访问处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E02 远端403处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"403 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_remote_500(self, api_client):
+ """TC-E03 远端500服务错误:验证 500 URL 返回错误含状态码信息且不崩溃"""
+ url_500 = "https://httpbin.org/status/500"
+
+ response = api_client.add_resource(path=url_500, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert "500" in error_msg or "server" in error_msg or "error" in error_msg, (
+ f"500错误信息应包含 500/server/error, 实际: {error_msg}"
+ )
+ print("✓ TC-E03 远端500服务错误处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors).lower()
+ assert (
+ "500" in inner_msg
+ or "server" in inner_msg
+ or "internal" in inner_msg
+ or "failed" in inner_msg
+ or "error" in inner_msg
+ ), f"500内层错误应包含 500/server/internal/failed/error, 实际: {inner_msg}"
+ print(f"✓ TC-E03 远端500服务错误处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E03 远端500处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"500 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_dns_resolve_failure(self, api_client):
+ """TC-E08 DNS解析失败:验证不存在的域名返回错误含DNS相关信息且不挂起"""
+ bad_dns_url = "https://nonexistent.domain.invalid.for.test/page"
+
+ response = api_client.add_resource(path=bad_dns_url, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "resolve" in error_msg
+ or "hostname" in error_msg
+ or "dns" in error_msg
+ or "error" in error_msg
+ or "connect" in error_msg
+ ), f"DNS失败错误信息应包含 resolve/hostname/dns/error/connect, 实际: {error_msg}"
+ print("✓ TC-E08 DNS解析失败处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors).lower()
+ assert (
+ "resolve" in inner_msg
+ or "hostname" in inner_msg
+ or "dns" in inner_msg
+ or "connect" in inner_msg
+ or "failed" in inner_msg
+ or "error" in inner_msg
+ ), f"DNS内层错误应包含 resolve/hostname/dns/connect/failed/error, 实际: {inner_msg}"
+ print(f"✓ TC-E08 DNS解析失败处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+ print("✓ TC-E08 DNS解析失败处理通过(降级为空资源)")
+ return
+
+ raise AssertionError(f"DNS失败 URL 应返回 error 或 ok, 实际: {data.get('status')}")
+
+ def test_error_ssh_url_invalid_format(self, api_client):
+ """TC-E09 SSH URL格式错误:验证 git@invalid (无冒号) 返回错误含SSH/URI相关信息"""
+ invalid_ssh_url = "git@invalid"
+
+ response = api_client.add_resource(path=invalid_ssh_url, wait=True)
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "invalid" in error_msg
+ or "ssh" in error_msg
+ or "uri" in error_msg
+ or "colon" in error_msg
+ or "error" in error_msg
+ or "permission" in error_msg
+ ), f"SSH格式错误应包含 invalid/ssh/uri/colon/error/permission, 实际: {error_msg}"
+ print("✓ TC-E09 SSH URL格式错误处理通过(返回error)")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors).lower()
+ assert (
+ "invalid" in inner_msg
+ or "ssh" in inner_msg
+ or "uri" in inner_msg
+ or "failed" in inner_msg
+ or "error" in inner_msg
+ ), f"SSH内层错误应包含 invalid/ssh/uri/failed/error, 实际: {inner_msg}"
+ print(f"✓ TC-E09 SSH URL格式错误处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ print("✓ TC-E09 SSH URL格式错误处理通过(服务端降级)")
+ return
+
+ raise AssertionError(f"SSH URL格式错误应返回 error 或降级, 实际: {data.get('status')}")
+
+ def test_error_corrupted_zip(self, api_client):
+ """TC-E16 损坏的ZIP文件:验证伪造 .zip 文件报错含 zip/corrupt 或回退处理,不产生有效子节点"""
+ random_id = str(uuid.uuid4())[:8]
+
+ temp_dir = tempfile.mkdtemp()
+ zip_path = os.path.join(temp_dir, f"corrupted_{random_id}.zip")
+ with open(zip_path, "w", encoding="utf-8") as f:
+ f.write("这不是一个真正的ZIP文件内容")
+
+ try:
+ response = api_client.add_resource(path=zip_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "zip" in error_msg
+ or "corrupt" in error_msg
+ or "error" in error_msg
+ or "archive" in error_msg
+ ), f"损坏ZIP错误信息应包含 zip/corrupt/error/archive, 实际: {error_msg}"
+ print(f"✓ TC-E16 损坏的ZIP文件处理通过(返回error): {error_msg[:80]}")
+ return
+
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ assert (
+ "zip" in inner_msg.lower()
+ or "corrupt" in inner_msg.lower()
+ or "error" in inner_msg.lower()
+ or "bad" in inner_msg.lower()
+ ), f"损坏ZIP内层错误应包含 zip/corrupt/error/bad, 实际: {inner_msg}"
+ print(f"✓ TC-E16 损坏的ZIP文件处理通过(内层错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ if root_uri:
+ assert_root_uri_valid(root_uri)
+
+ tree_resp = api_client.fs_tree(root_uri)
+ if tree_resp.status_code == 200:
+ tree_data = tree_resp.json()
+ tree_result = tree_data.get("result")
+ if isinstance(tree_result, list):
+ children = tree_result
+ elif isinstance(tree_result, dict):
+ children = tree_result.get("children", [])
+ else:
+ children = []
+ assert len(children) == 0, (
+ f"损坏ZIP不应产生有效子节点, 实际子节点数: {len(children)}"
+ )
+
+ print(f"✓ TC-E16 损坏的ZIP文件处理通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_media_resources_slow.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_media_resources_slow.py
new file mode 100644
index 000000000..be29452c4
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_media_resources_slow.py
@@ -0,0 +1,322 @@
+import os
+import shutil
+import tempfile
+import uuid
+
+import pytest
+from build_test_helpers import (
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+)
+
+
+class TestBuildMediaResourcesSlow:
+ """TC-B13~B15 媒体类资源构建测试"""
+
+ def test_build_image_svg(self, api_client):
+ """TC-B13 图片构建(SVG):验证 .svg 文件解析失败时错误信息合理,成功时 source_format=image 且可检索"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"svg_keyword_{random_id}"
+ temp_dir = tempfile.mkdtemp()
+ svg_path = os.path.join(temp_dir, f"test_{random_id}.svg")
+
+ svg_content = f'<svg xmlns="http://www.w3.org/2000/svg" width="100" height="100"><rect width="100" height="100" fill="blue"/><text x="10" y="50">SVG test {unique_keyword}</text></svg>'
+ with open(svg_path, "w", encoding="utf-8") as f:
+ f.write(svg_content)
+
+ try:
+ response = api_client.add_resource(path=svg_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") in ("ok", "error"), (
+ f"SVG文件应返回 ok 或 error, 实际: {data.get('status')}"
+ )
+
+ if data.get("status") == "error":
+ error_msg = str(data.get("error", "")).lower()
+ assert (
+ "image" in error_msg
+ or "svg" in error_msg
+ or "parse" in error_msg
+ or "error" in error_msg
+ ), f"SVG外层错误应包含 image/svg/parse/error, 实际: {error_msg}"
+ print("✓ TC-B13 图片构建(SVG)通过(服务端返回error,SVG格式不被支持)")
+ return
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ assert (
+ "image" in inner_msg.lower()
+ or "parse" in inner_msg.lower()
+ or "identify" in inner_msg.lower()
+ or "svg" in inner_msg.lower()
+ ), f"SVG解析错误信息应包含 image/parse/identify/svg, 实际: {inner_msg}"
+ print(
+ f"✓ TC-B13 图片构建(SVG)通过(服务端内层解析错误,SVG格式不被支持): {inner_msg[:80]}"
+ )
+ return
+
+ root_uri = result.get("root_uri") if isinstance(result, dict) else None
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["image", "html", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B13 图片构建(SVG)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_audio_mp3(self, api_client):
+ """TC-B14 音频构建(MP3):验证 .mp3 文件添加后产物路径含 /audio/ 且 source_format=audio"""
+ import subprocess
+ import uuid
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ mp3_path = os.path.join(temp_dir, f"test_{random_id}.mp3")
+
+ try:
+ result = subprocess.run(
+ [
+ "ffmpeg",
+ "-f",
+ "lavfi",
+ "-i",
+ "sine=frequency=440:duration=1",
+ "-frames:a",
+ "44100",
+ "-y",
+ mp3_path,
+ ],
+ capture_output=True,
+ timeout=10,
+ )
+ if result.returncode != 0:
+ pytest.skip("ffmpeg 不可用,跳过MP3音频构建测试")
+ except (FileNotFoundError, subprocess.TimeoutExpired):
+ pytest.skip("ffmpeg 不可用,跳过MP3音频构建测试")
+
+ try:
+ response = api_client.add_resource(path=mp3_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/audio/" in root_uri, f"音频 root_uri 应含 /audio/, 实际: {root_uri}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "audio")
+
+ print(f"✓ TC-B14 音频构建(MP3)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_audio_wav(self, api_client):
+ """TC-B14 音频构建(WAV):验证 .wav 文件添加后产物路径含 /audio/ 且 source_format=audio"""
+ import subprocess
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ wav_path = os.path.join(temp_dir, f"test_{random_id}.wav")
+
+ try:
+ result = subprocess.run(
+ [
+ "ffmpeg",
+ "-f",
+ "lavfi",
+ "-i",
+ "sine=frequency=880:duration=1",
+ "-frames:a",
+ "44100",
+ "-y",
+ wav_path,
+ ],
+ capture_output=True,
+ timeout=10,
+ )
+ if result.returncode != 0:
+ pytest.skip("ffmpeg 不可用,跳过WAV音频构建测试")
+ except (FileNotFoundError, subprocess.TimeoutExpired):
+ pytest.skip("ffmpeg 不可用,跳过WAV音频构建测试")
+
+ try:
+ response = api_client.add_resource(path=wav_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/audio/" in root_uri, f"音频 root_uri 应含 /audio/, 实际: {root_uri}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "audio")
+
+ print(f"✓ TC-B14 音频构建(WAV)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_image_jpg(self, api_client):
+ """TC-B13 图片构建(JPG):验证 .jpg 文件添加后产物路径含 /images/ 且 source_format=image"""
+ try:
+ from PIL import Image
+ except ImportError:
+ pytest.skip("Pillow 未安装,跳过JPG图片构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ img_path = os.path.join(temp_dir, f"test_{random_id}.jpg")
+
+ img = Image.new("RGB", (100, 100), color=(73, 109, 137))
+ img.save(img_path, format="JPEG")
+
+ try:
+ response = api_client.add_resource(path=img_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/images/" in root_uri, f"图片 root_uri 应含 /images/, 实际: {root_uri}"
+
+ assert_source_format(api_client, root_uri, "image")
+
+ print(f"✓ TC-B13 图片构建(JPG)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_image_png(self, api_client):
+ """TC-B13 图片构建(PNG):验证 .png 文件添加后产物路径含 /images/ 且 source_format=image"""
+ try:
+ from PIL import Image
+ except ImportError:
+ pytest.skip("Pillow 未安装,跳过PNG图片构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ img_path = os.path.join(temp_dir, f"test_{random_id}.png")
+
+ img = Image.new("RGBA", (100, 100), color=(73, 109, 137, 255))
+ img.save(img_path, format="PNG")
+
+ try:
+ response = api_client.add_resource(path=img_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/images/" in root_uri, f"图片 root_uri 应含 /images/, 实际: {root_uri}"
+
+ assert_source_format(api_client, root_uri, "image")
+
+ print(f"✓ TC-B13 图片构建(PNG)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_image_webp(self, api_client):
+ """TC-B13 图片构建(WebP):验证 .webp 文件添加后产物路径含 /images/ 且 source_format=image"""
+ try:
+ from PIL import Image
+ except ImportError:
+ pytest.skip("Pillow 未安装,跳过WebP图片构建测试")
+
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ img_path = os.path.join(temp_dir, f"test_{random_id}.webp")
+
+ img = Image.new("RGB", (100, 100), color=(73, 109, 137))
+ img.save(img_path, format="WEBP")
+
+ try:
+ response = api_client.add_resource(path=img_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/images/" in root_uri, f"图片 root_uri 应含 /images/, 实际: {root_uri}"
+
+ assert_source_format(api_client, root_uri, "image")
+
+ print(f"✓ TC-B13 图片构建(WebP)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
+
+ def test_build_video_mp4(self, api_client):
+ """TC-B15 视频构建(MP4):验证 .mp4 文件添加后产物路径含 /video/ 且 source_format=video"""
+ random_id = str(uuid.uuid4())[:8]
+ temp_dir = tempfile.mkdtemp()
+ mp4_path = os.path.join(temp_dir, f"test_{random_id}.mp4")
+
+ try:
+ import subprocess
+
+ result = subprocess.run(
+ [
+ "ffmpeg",
+ "-f",
+ "lavfi",
+ "-i",
+ "color=c=blue:s=100x100:d=1",
+ "-frames:v",
+ "25",
+ "-y",
+ mp4_path,
+ ],
+ capture_output=True,
+ timeout=10,
+ )
+ if result.returncode != 0:
+ pytest.skip("ffmpeg 不可用,跳过MP4视频构建测试")
+ except (FileNotFoundError, subprocess.TimeoutExpired):
+ pytest.skip("ffmpeg 不可用,跳过MP4视频构建测试")
+
+ try:
+ response = api_client.add_resource(path=mp4_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "/video/" in root_uri, f"视频 root_uri 应含 /video/, 实际: {root_uri}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "video")
+
+ print(f"✓ TC-B15 视频构建(MP4)通过, root_uri: {root_uri}")
+ finally:
+ shutil.rmtree(temp_dir, ignore_errors=True)
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_arxiv.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_arxiv.py
new file mode 100644
index 000000000..ec0fe52a3
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_arxiv.py
@@ -0,0 +1,59 @@
+from build_test_helpers import assert_resource_indexed, assert_root_uri_valid, assert_source_format
+
+
+class TestBuildPlatformArxiv:
+ """TC-P06~P07 arXiv 平台 URL 构建测试"""
+
+ def test_build_arxiv_pdf(self, api_client):
+ """TC-P06 arXiv PDF构建:验证 arxiv.org/pdf/ URL 走 DOWNLOAD_PDF 路由且 source_format=pdf"""
+ arxiv_pdf_url = "https://arxiv.org/pdf/2301.00234"
+
+ response = api_client.add_resource(path=arxiv_pdf_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("download_pdf", "webpage", None), (
+ f"meta.url_type 应为 download_pdf, 实际: {meta.get('url_type')}"
+ )
+
+ assert_source_format(api_client, root_uri, ["pdf", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-P06 arXiv PDF构建通过, root_uri: {root_uri}")
+
+ def test_build_arxiv_abstract_page(self, api_client):
+ """TC-P07 arXiv HTML页面构建:验证 arxiv.org/abs/ URL 走 WEBPAGE 路由且摘要可检索"""
+ arxiv_abs_url = "https://arxiv.org/abs/2301.00234"
+
+ response = api_client.add_resource(path=arxiv_abs_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("webpage", None), (
+ f"meta.url_type 应为 webpage, 实际: {meta.get('url_type')}"
+ )
+
+ assert_source_format(api_client, root_uri, ["html", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_resource_indexed(api_client, root_uri, "arxiv")
+
+ print(f"✓ TC-P07 arXiv HTML页面构建通过, root_uri: {root_uri}")
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_general_web.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_general_web.py
new file mode 100644
index 000000000..6709eba89
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_general_web.py
@@ -0,0 +1,33 @@
+from build_test_helpers import assert_resource_indexed, assert_root_uri_valid, assert_source_format
+
+
+class TestBuildPlatformGeneralWeb:
+ """TC-P10 通用网页 URL 构建测试"""
+
+ def test_build_general_webpage(self, api_client):
+ """TC-P10 通用网页构建:验证任意 HTTP URL 走 WEBPAGE 路由且 source_format=html、内容可检索"""
+ web_url = "https://httpbin.org/html"
+
+ response = api_client.add_resource(path=web_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("webpage", None), (
+ f"meta.url_type 应为 webpage, 实际: {meta.get('url_type')}"
+ )
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["html", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, "httpbin")
+
+ print(f"✓ TC-P10 通用网页构建通过, root_uri: {root_uri}")
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_github.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_github.py
new file mode 100644
index 000000000..d1e31d535
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_github.py
@@ -0,0 +1,125 @@
+from build_test_helpers import assert_resource_indexed, assert_root_uri_valid, assert_source_format
+
+
+class TestBuildPlatformGithub:
+ """TC-P01~P04 GitHub 平台 URL 构建测试"""
+
+ def test_build_github_repo(self, api_client):
+ """TC-P01 GitHub仓库构建:验证 github.com/org/repo 走 CodeRepositoryParser 且 root_uri 含 org/repo"""
+ repo_url = "https://github.com/volcengine/OpenViking"
+
+ response = api_client.add_resource(path=repo_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "volcengine" in root_uri and "OpenViking" in root_uri, (
+ f"GitHub仓库 root_uri 应含 volcengine/OpenViking, 实际: {root_uri}"
+ )
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("code_repository", None), (
+ f"meta.url_type 应为 code_repository, 实际: {meta.get('url_type')}"
+ )
+
+ assert_source_format(api_client, root_uri, ["repository", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-P01 GitHub仓库构建通过, root_uri: {root_uri}")
+
+ def test_build_github_repo_with_branch(self, api_client):
+ """TC-P02 GitHub指定分支构建:验证 github.com/org/repo/tree/branch 走 ZIP API 且 meta 含 repo_ref"""
+ repo_url = "https://github.com/volcengine/OpenViking/tree/main"
+
+ response = api_client.add_resource(path=repo_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+ assert "volcengine" in root_uri and "OpenViking" in root_uri, (
+ f"GitHub仓库 root_uri 应含 volcengine/OpenViking, 实际: {root_uri}"
+ )
+
+ meta = result.get("meta", {})
+ if "repo_ref" in meta:
+ assert meta["repo_ref"] == "main", f"meta.repo_ref 应为 main, 实际: {meta['repo_ref']}"
+
+ assert_source_format(api_client, root_uri, ["repository", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-P02 GitHub指定分支构建通过, root_uri: {root_uri}")
+
+ def test_build_github_raw_file(self, api_client):
+ """TC-P03 GitHub原始文件下载:验证 raw.githubusercontent.com URL 走 download_markdown 路由且内容可检索"""
+ raw_url = "https://raw.githubusercontent.com/volcengine/OpenViking/main/README.md"
+
+ response = api_client.add_resource(path=raw_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in (
+ "download_md",
+ "download_markdown",
+ "download_txt",
+ "webpage",
+ None,
+ ), (
+ f"meta.url_type 应为 download_md/download_markdown/download_txt/webpage, 实际: {meta.get('url_type')}"
+ )
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_resource_indexed(api_client, root_uri, "OpenViking")
+
+ print(f"✓ TC-P03 GitHub原始文件下载通过, root_uri: {root_uri}")
+
+ def test_build_github_blob_page(self, api_client):
+ """TC-P04 GitHub Blob页面构建:验证 github.com/org/repo/blob/branch/file 被转为 raw URL 下载且内容可检索"""
+ blob_url = "https://github.com/volcengine/OpenViking/blob/main/README.md"
+
+ response = api_client.add_resource(path=blob_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in (
+ "download_md",
+ "download_markdown",
+ "download_txt",
+ "download_html",
+ "webpage",
+ None,
+ ), f"meta.url_type 应为 download 类, 实际: {meta.get('url_type')}"
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_resource_indexed(api_client, root_uri, "OpenViking")
+
+ print(f"✓ TC-P04 GitHub Blob页面构建通过, root_uri: {root_uri}")
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_wikipedia.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_wikipedia.py
new file mode 100644
index 000000000..c339357ea
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_platform_wikipedia.py
@@ -0,0 +1,36 @@
+from build_test_helpers import assert_resource_indexed, assert_root_uri_valid, assert_source_format
+
+
+class TestBuildPlatformWikipedia:
+ """TC-P05 Wikipedia 平台 URL 构建测试"""
+
+ def test_build_wikipedia_page(self, api_client):
+ """TC-P05 Wikipedia页面构建:验证 wikipedia.org URL 走 WEBPAGE 路由且内容可检索"""
+ wiki_url = "https://en.wikipedia.org/wiki/Software_testing"
+
+ response = api_client.add_resource(path=wiki_url, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok", (
+ f"Wikipedia页面构建应返回ok, 实际: {data.get('status')}, error: {data.get('error')}"
+ )
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert root_uri, "Wikipedia页面构建应返回root_uri, 实际为空"
+ assert_root_uri_valid(root_uri)
+
+ meta = result.get("meta", {})
+ assert meta.get("url_type") in ("webpage", "download_text", "download_html", None), (
+ f"meta.url_type 应为 webpage/download_text/download_html, 实际: {meta.get('url_type')}"
+ )
+
+ assert_source_format(api_client, root_uri, ["html", "markdown"])
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_resource_indexed(api_client, root_uri, "software testing")
+
+ print(f"✓ TC-P05 Wikipedia页面构建通过, root_uri: {root_uri}")
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_text_resources_slow.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_text_resources_slow.py
new file mode 100644
index 000000000..57985dd39
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_text_resources_slow.py
@@ -0,0 +1,156 @@
+import uuid
+
+from build_test_helpers import (
+ assert_resource_indexed,
+ assert_root_uri_valid,
+ assert_source_format,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildTextResourcesSlow:
+ """TC-B01, B02, B14, B15 文本类资源构建测试"""
+
+ def test_build_txt_file(self, api_client):
+ """TC-B01 纯文本文件构建:验证 .txt 文件添加后 source_format=text 且内容可检索"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"txt_keyword_{random_id}"
+ content = f"纯文本测试文件 {random_id}\n包含唯一关键词:{unique_keyword}\n用于验证txt文件构建产物。"
+
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["text", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B01 纯文本文件构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_markdown_file(self, api_client):
+ """TC-B02 Markdown文件构建:验证 .md 文件添加后 heading 结构保留为子节点"""
+ from build_test_helpers import assert_tree_has_child_nodes
+
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"md_keyword_{random_id}"
+ content = (
+ f"# Markdown测试标题 {random_id}\n\n"
+ f"包含唯一关键词:{unique_keyword}\n\n"
+ f"## 第一节\n\n第一段内容。\n\n"
+ f"## 第二节\n\n第二段内容。\n\n"
+ f"### 子节\n\n子节内容。\n"
+ )
+
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".md")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, "markdown")
+
+ assert_tree_has_child_nodes(api_client, root_uri, min_nodes=1)
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B02 Markdown文件构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_empty_file(self, api_client):
+ """TC-B15 空文件构建:验证空 .txt 文件添加后 root_uri 有效、fs_stat 可查、source_format 合理"""
+ from build_test_helpers import assert_source_format
+
+ test_file_path, temp_dir = create_test_file(content="", suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = str(data.get("error", ""))
+ assert "empty" in error_msg.lower() or "error" in error_msg.lower(), (
+ f"空文件错误信息应包含 empty/error, 实际: {error_msg}"
+ )
+ print(f"✓ TC-B15 空文件构建通过(服务端拒绝空文件): {error_msg[:80]}")
+ return
+
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ assert (
+ "empty" in inner_msg.lower()
+ or "error" in inner_msg.lower()
+ or "parse" in inner_msg.lower()
+ ), f"空文件内层错误应包含 empty/error/parse, 实际: {inner_msg}"
+ print(f"✓ TC-B15 空文件构建通过(内层解析错误): {inner_msg[:80]}")
+ return
+
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200, f"空文件 fs_stat 应返回200, root_uri: {root_uri}"
+
+ assert_source_format(api_client, root_uri, ["text", "markdown", ""])
+
+ print(f"✓ TC-B15 空文件构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_raw_content(self, api_client):
+ """TC-B14 原始内容字符串构建:验证纯文本内容写入文件后添加可检索"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"raw_keyword_{random_id}"
+ raw_content = (
+ f"原始内容测试 {random_id}\n包含唯一关键词:{unique_keyword}\n用于验证字符串输入构建。"
+ )
+
+ test_file_path, temp_dir = create_test_file(content=raw_content, suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert_root_uri_valid(root_uri)
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ assert_source_format(api_client, root_uri, ["text", "markdown"])
+
+ assert_resource_indexed(api_client, root_uri, unique_keyword)
+
+ print(f"✓ TC-B14 原始内容字符串构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
diff --git a/tests/api_test/scenarios/resources_retrieval_slow/test_build_uri_params_slow.py b/tests/api_test/scenarios/resources_retrieval_slow/test_build_uri_params_slow.py
new file mode 100644
index 000000000..399ffdadc
--- /dev/null
+++ b/tests/api_test/scenarios/resources_retrieval_slow/test_build_uri_params_slow.py
@@ -0,0 +1,117 @@
+import uuid
+
+from build_test_helpers import (
+ _extract_error_message,
+ assert_root_uri_valid,
+ cleanup_temp_dir,
+ create_test_file,
+)
+
+
+class TestBuildUriParamsSlow:
+ """TC-E11, E17, E18 URI参数与边界测试"""
+
+ def test_build_with_to_param(self, api_client):
+ """TC-E17 指定to参数构建:验证 root_uri == to"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"to_keyword_{random_id}"
+ target_uri = f"viking://resources/to_test_{random_id}"
+
+ content = f"to参数测试 {random_id}\n包含唯一关键词:{unique_keyword}"
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, to=target_uri, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ assert data.get("status") == "ok"
+
+ result = data.get("result", {})
+ root_uri = result.get("root_uri")
+ assert root_uri == target_uri, (
+ f"指定to时 root_uri 应等于 to, 期望: {target_uri}, 实际: {root_uri}"
+ )
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-E17 指定to参数构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_with_parent_param(self, api_client):
+ """TC-E18 指定parent参数构建:验证 root_uri 以 parent 为前缀"""
+ random_id = str(uuid.uuid4())[:8]
+ unique_keyword = f"parent_keyword_{random_id}"
+ parent_uri = f"viking://resources/parent_test_{random_id}"
+
+ mkdir_resp = api_client.fs_mkdir(parent_uri)
+ assert mkdir_resp.status_code == 200, (
+ f"无法创建parent目录, fs_mkdir返回: {mkdir_resp.status_code}, body: {mkdir_resp.text}"
+ )
+
+ content = f"parent参数测试 {random_id}\n包含唯一关键词:{unique_keyword}"
+ test_file_path, temp_dir = create_test_file(content=content, suffix=".txt")
+ try:
+ response = api_client.add_resource(path=test_file_path, parent=parent_uri, wait=True)
+ assert response.status_code == 200
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data)
+ raise AssertionError(f"parent参数构建失败, 服务端不支持parent参数: {error_msg}")
+
+ result = data.get("result", {})
+ if isinstance(result, dict) and result.get("status") == "error":
+ inner_errors = result.get("errors", [])
+ inner_msg = " ".join(str(e) for e in inner_errors)
+ raise AssertionError(f"parent参数构建失败(内层错误): {inner_msg}")
+
+ root_uri = result.get("root_uri") if isinstance(result, dict) else None
+ assert_root_uri_valid(root_uri)
+ assert root_uri.startswith(parent_uri), (
+ f"指定parent时 root_uri 应以 parent 为前缀, parent: {parent_uri}, 实际: {root_uri}"
+ )
+
+ stat_resp = api_client.fs_stat(root_uri)
+ assert stat_resp.status_code == 200
+
+ print(f"✓ TC-E18 指定parent参数构建通过, root_uri: {root_uri}")
+ finally:
+ cleanup_temp_dir(temp_dir)
+
+ def test_build_non_resources_scope_rejected(self, api_client):
+ """TC-E11 非resources scope拒绝:验证 to=viking://sessions/xxx 返回错误"""
+ random_id = str(uuid.uuid4())[:8]
+ test_content = f"scope测试内容 {random_id}"
+ test_file_path, temp_dir = create_test_file(content=test_content, suffix=".txt")
+ try:
+ response = api_client.add_resource(
+ path=test_file_path,
+ to="viking://sessions/test_session",
+ wait=True,
+ )
+
+ data = response.json()
+ if data.get("status") == "error":
+ error_msg = _extract_error_message(data).lower()
+ assert (
+ "scope" in error_msg
+ or "resources" in error_msg
+ or "invalid" in error_msg
+ or "permission" in error_msg
+ or "internal" in error_msg
+ ), f"scope拒绝应包含 scope/resources/invalid/permission/internal, 实际: {error_msg}"
+ print("✓ TC-E11 非resources scope拒绝通过")
+ return
+
+ if data.get("status") == "ok":
+ result = data.get("result", {})
+ root_uri = result.get("root_uri", "")
+ assert "sessions" not in root_uri, (
+ f"非resources scope不应成功写入sessions, root_uri: {root_uri}"
+ )
+ print("✓ TC-E11 非resources scope处理通过(服务端重定向)")
+ return
+ finally:
+ cleanup_temp_dir(temp_dir)
diff --git a/tests/oc2ov_test/tests/base_cli_test.py b/tests/oc2ov_test/tests/base_cli_test.py
index 4899411a2..5291a08f7 100644
--- a/tests/oc2ov_test/tests/base_cli_test.py
+++ b/tests/oc2ov_test/tests/base_cli_test.py
@@ -220,6 +220,9 @@ class BaseOpenClawCLITest(unittest.TestCase):
"did not produce a response",
"LLM idle timeout",
"timed out",
+ "couldn't generate a response",
+ "couldn't generate",
+ "please try again",
]
return any(ind.lower() in text.lower() for ind in timeout_indicators)
@@ -241,6 +244,22 @@ class BaseOpenClawCLITest(unittest.TestCase):
]
if any(text.startswith(p) for p in tool_result_prefixes):
return True
+ import re
+
+ tool_result_pattern = r"^\[?\{[^}]*\"name\"\s*:\s*\"none\"[^}]*\}\]?[\s]*$"
+ if re.match(tool_result_pattern, text):
+ return True
+ if text.startswith("[{") and text.endswith("}]"):
+ try:
+ import json
+
+ parsed = json.loads(text)
+ if isinstance(parsed, list) and all(
+ isinstance(item, dict) and "name" in item for item in parsed
+ ):
+ return True
+ except (json.JSONDecodeError, ValueError):
+ pass
tool_result_patterns = [
'"name":"none"',
]
diff --git a/tests/oc2ov_test/utils/assertions.py b/tests/oc2ov_test/utils/assertions.py
index 69c24aedc..c1591b9e4 100644
--- a/tests/oc2ov_test/utils/assertions.py
+++ b/tests/oc2ov_test/utils/assertions.py
@@ -15,6 +15,23 @@ class AssertionHelper:
断言辅助类
"""
+ @staticmethod
+ def _strip_tool_result_prefix(text: str) -> str:
+ """
+ 去除 LLM 响应中的工具调用结果前缀
+ 例如: '[{"name":"none"}] 您的临时密码...' -> '您的临时密码...'
+ """
+ import re
+
+ stripped = text.strip()
+ tool_prefix_pattern = r"^\s*\[?\{[^}]*\}\]?\s*"
+ while re.match(tool_prefix_pattern, stripped):
+ remaining = re.sub(tool_prefix_pattern, "", stripped, count=1)
+ if remaining == stripped:
+ break
+ stripped = remaining.strip()
+ return stripped if stripped else text
+
@staticmethod
def extract_response_text(response: dict) -> str:
"""
@@ -22,17 +39,17 @@ class AssertionHelper:
支持多种响应格式
"""
if isinstance(response, str):
- return response
+ return AssertionHelper._strip_tool_result_prefix(response)
if isinstance(response, dict):
if "output" in response:
- return str(response["output"])
+ return AssertionHelper._strip_tool_result_prefix(str(response["output"]))
if "message" in response:
- return str(response["message"])
+ return AssertionHelper._strip_tool_result_prefix(str(response["message"]))
if "content" in response:
- return str(response["content"])
+ return AssertionHelper._strip_tool_result_prefix(str(response["content"]))
if "text" in response:
- return str(response["text"])
+ return AssertionHelper._strip_tool_result_prefix(str(response["text"]))
if "result" in response and isinstance(response["result"], dict):
result = response["result"]
payloads = result.get("payloads")
@@ -42,10 +59,10 @@ class AssertionHelper:
if isinstance(p, dict) and "text" in p:
texts.append(str(p["text"]))
if texts:
- return "\n".join(texts)
+ return AssertionHelper._strip_tool_result_prefix("\n".join(texts))
for key in ("output", "text", "content", "message", "response"):
if key in result and result[key]:
- return str(result[key])
+ return AssertionHelper._strip_tool_result_prefix(str(result[key]))
messages = result.get("messages")
if isinstance(messages, list) and len(messages) > 0:
for msg in reversed(messages):
@@ -54,7 +71,7 @@ class AssertionHelper:
if role == "assistant":
c = msg.get("content", "")
if c:
- return str(c)
+ return AssertionHelper._strip_tool_result_prefix(str(c))
if "choices" in response and len(response["choices"]) > 0:
choice = response["choices"][0]
if isinstance(choice, dict):