Add new flag --select-detectors

davidkopp · davidkopp · commit b2be89922b97 · 2025-09-18T23:36:13.000+02:00
diff --git a/TODOS.md b/TODOS.md
@@ -3,7 +3,6 @@
 - Rename CLI flag '--only-container-info' to '--only-source-info'
 - Publish package to PyPI
 - Add host_info_detector (similar to docker_info_detector)
-- Add CLI flag to select a subset of detectors (see how Syft does it: <https://github.com/anchore/syft/wiki/package-cataloger-selection>)
 - Extend the set of supported package managers with the common ones for Go, PHP and Java
 - Add Dockerfile
 - Add support for Podman
diff --git a/dependency_resolver/__main__.py b/dependency_resolver/__main__.py
@@ -37,6 +37,7 @@ def parse_arguments() -> argparse.Namespace:
   %(prog)s --debug                                # Enable debug output
   %(prog)s --skip-system-scope                    # Skip system scope package managers
   %(prog)s --skip-hash-collection                 # Skip hash collection for improved performance
+  %(prog)s --select-detectors "pip,dpkg"          # Use only pip and dpkg detectors
         """,
     )
 
@@ -85,6 +86,12 @@ def parse_arguments() -> argparse.Namespace:
         help="Skip hash collection for packages and project locations to improve performance",
     )
 
+    parser.add_argument(
+        "--select-detectors",
+        type=str,
+        help="Comma-separated list of detectors to use (e.g., 'pip,dpkg'). Available: pip, npm, dpkg, apk, maven, docker-info",
+    )
+
     return parser.parse_args()
 
 
@@ -133,6 +140,7 @@ def main() -> None:
             skip_system_scope=args.skip_system_scope,
             venv_path=args.venv_path,
             skip_hash_collection=args.skip_hash_collection,
+            selected_detectors=args.select_detectors,
         )
         dependencies = orchestrator.resolve_dependencies(executor, args.working_dir, args.only_container_info)
         formatter = OutputFormatter(debug=args.debug)
diff --git a/dependency_resolver/core/orchestrator.py b/dependency_resolver/core/orchestrator.py
@@ -17,13 +17,14 @@ def __init__(
         skip_system_scope: bool = False,
         venv_path: str | None = None,
         skip_hash_collection: bool = False,
+        selected_detectors: str | None = None,
     ):
         self.debug = debug
         self.skip_system_scope = skip_system_scope
         self.skip_hash_collection = skip_hash_collection
 
-        # Create detector instances
-        self.detectors: list[PackageManagerDetector] = [
+        # Create all detector instances
+        all_detectors: list[PackageManagerDetector] = [
             DockerInfoDetector(),
             DpkgDetector(),
             ApkDetector(),
@@ -32,6 +33,26 @@ def __init__(
             NpmDetector(debug=debug),
         ]
 
+        # Filter detectors based on selection
+        if selected_detectors:
+            selected_names = [name.strip() for name in selected_detectors.split(",")]
+            available_names = {detector.NAME for detector in all_detectors}
+
+            # Validate detector names
+            invalid_names = [name for name in selected_names if name not in available_names]
+            if invalid_names:
+                raise ValueError(
+                    f"Invalid detector names: {', '.join(invalid_names)}. Available detectors: {', '.join(sorted(available_names))}"
+                )
+
+            # Filter to only selected detectors
+            self.detectors = [detector for detector in all_detectors if detector.NAME in selected_names]
+            if self.debug:
+                selected_detector_names = [detector.NAME for detector in self.detectors]
+                print(f"Selected detectors: {', '.join(selected_detector_names)}")
+        else:
+            self.detectors = all_detectors
+
     def resolve_dependencies(
         self, executor: EnvironmentExecutor, working_dir: Optional[str] = None, only_container_info: bool = False
     ) -> dict[str, Any]:
diff --git a/docs/guides/troubleshooting.md b/docs/guides/troubleshooting.md
@@ -171,9 +171,10 @@ npm --version
 
 1. Use `--skip-system-scope` to skip system package managers
 2. Use `--skip-hash-collection` to skip hash generation for packages and locations
-3. Use `--only-container-info` for Docker metadata only
-4. Specify working directory to limit scope
-5. Check if system has many installed packages
+3. Use `--select-detectors` to analyze only specific package managers (e.g., `--select-detectors "pip,npm"`)
+4. Use `--only-container-info` for Docker metadata only
+5. Specify working directory to limit scope
+6. Check if system has many installed packages
 
 ## Getting Help
 
diff --git a/docs/usage/cli-guide.md b/docs/usage/cli-guide.md
@@ -46,8 +46,35 @@ python3 -m dependency_resolver --skip-system-scope
 
 # Skip hash collection for improved performance
 python3 -m dependency_resolver --skip-hash-collection
+
+# Select specific detectors to run
+python3 -m dependency_resolver --select-detectors "pip,dpkg"
 ```
 
+## Detector Selection
+
+Control which package managers are analyzed with the `--select-detectors` flag:
+
+```bash
+# Use only pip and dpkg detectors
+python3 -m dependency_resolver --select-detectors "pip,dpkg"
+
+# Use only npm detector for Node.js projects
+python3 -m dependency_resolver --select-detectors "npm"
+
+# Use multiple detectors with debug output
+python3 -m dependency_resolver --select-detectors "pip,npm,maven" --debug
+```
+
+**Available detectors:**
+
+- `pip` - Python packages (pip/PyPI)
+- `npm` - Node.js packages (npm/yarn)
+- `dpkg` - Debian/Ubuntu system packages
+- `apk` - Alpine Linux system packages
+- `maven` - Java Maven dependencies
+- `docker-info` - Docker container metadata
+
 ## Common Usage Patterns
 
 ### Analyzing a Specific Project
@@ -60,6 +87,9 @@ python3 -m dependency_resolver --working-dir /path/to/python/project
 
 # Analyze Node.js project
 python3 -m dependency_resolver --working-dir /path/to/nodejs/project
+
+# Analyze only Python dependencies in a project
+python3 -m dependency_resolver --working-dir /path/to/python/project --select-detectors "pip"
 ```
 
 ### Comprehensive Docker Analysis
diff --git a/docs/usage/programmatic-api.md b/docs/usage/programmatic-api.md
@@ -68,7 +68,12 @@ from dependency_resolver import Orchestrator, HostExecutor, DockerExecutor, Outp
 
 # Host system analysis with custom settings
 executor = HostExecutor()
-orchestrator = Orchestrator(debug=True, skip_system_scope=True, skip_hash_collection=True)
+orchestrator = Orchestrator(
+    debug=True,
+    skip_system_scope=True,
+    skip_hash_collection=True,
+    selected_detectors="pip,npm"  # Only analyze Python and Node.js dependencies
+)
 dependencies = orchestrator.resolve_dependencies(executor)
 
 # Format output
@@ -84,7 +89,12 @@ from dependency_resolver import Orchestrator, DockerExecutor, OutputFormatter
 # Docker container analysis
 container_id = "nginx"
 executor = DockerExecutor(container_id)
-orchestrator = Orchestrator(debug=False, skip_system_scope=False, skip_hash_collection=False)
+orchestrator = Orchestrator(
+    debug=False,
+    skip_system_scope=False,
+    skip_hash_collection=False,
+    selected_detectors="dpkg,docker-info"  # Analyze system packages and container info only
+)
 
 dependencies = orchestrator.resolve_dependencies(executor, working_dir="/app")
 
@@ -119,10 +129,30 @@ orchestrator = Orchestrator(
     debug=True,                    # Enable debug output
     skip_system_scope=False,       # Skip system-wide package managers
     venv_path="/path/to/venv",     # Specify Python virtual environment path
-    skip_hash_collection=False     # Skip hash collection for improved performance
+    skip_hash_collection=False,    # Skip hash collection for improved performance
+    selected_detectors="pip,npm"   # Use only specific detectors
 )
 ```
 
+### Detector Selection
+
+Control which package managers are analyzed by specifying the `selected_detectors` parameter:
+
+```python
+from dependency_resolver import Orchestrator, HostExecutor
+
+# Use only Python pip detector
+orchestrator = Orchestrator(selected_detectors="pip")
+
+# Use multiple specific detectors
+orchestrator = Orchestrator(selected_detectors="pip,npm,maven")
+
+# Use all detectors (default behavior)
+orchestrator = Orchestrator(selected_detectors=None)
+```
+
+**Available detectors:** `pip`, `npm`, `dpkg`, `apk`, `maven`, `docker-info`
+
 ### Executor Options
 
 ```python
diff --git a/performance-analysis/detector-benchmarks.sh b/performance-analysis/detector-benchmarks.sh
@@ -89,6 +89,7 @@ show_usage() {
     echo "  --profiling    Enable detailed profiling (slower, for analysis)"
     echo "  --interface TYPE       Interface type: 'cli' or 'programmatic' (default: cli)"
     echo "  --scenario TYPE        Specific scenario: varies by detector type"
+    echo "  --select-detectors LIST Comma-separated list of detectors to use (e.g., 'pip,dpkg')"
     echo ""
     echo "Specific scenario types by detector:"
     echo "  host:  skip-system, full-system, all (default: all)"
@@ -105,6 +106,7 @@ show_usage() {
     echo "  $0 --interface programmatic host  # Run host detector with programmatic interface"
     echo "  $0 --scenario small npm  # Run only small NPM scenario"
     echo "  $0 --scenario skip-system host  # Run only skip-system host scenario"
+    echo "  $0 --select-detectors 'pip,dpkg' host  # Run host detector with only pip and dpkg"
 }
 
 # Parse command line arguments
@@ -115,6 +117,7 @@ SKIP_SYSTEM_CHECK=false
 SKIP_HASH_COLLECTION=false
 ENABLE_PROFILING=false
 INTERFACE="cli"
+SELECT_DETECTORS=""
 
 while [[ $# -gt 0 ]]; do
     case $1 in
@@ -158,6 +161,16 @@ while [[ $# -gt 0 ]]; do
                 exit 1
             fi
             ;;
+        --select-detectors)
+            if [[ -n $2 ]]; then
+                SELECT_DETECTORS="$2"
+                shift 2
+            else
+                print_error "--select-detectors requires a comma-separated list of detectors"
+                show_usage
+                exit 1
+            fi
+            ;;
         all|host|npm|pip)
             DETECTOR_TYPE="$1"
             shift
@@ -177,6 +190,7 @@ SESSION_ID=$(date +%Y-%m-%d_%H-%M-%S)
 export ENABLE_PROFILING
 export SKIP_HASH_COLLECTION
 export SESSION_ID
+export SELECT_DETECTORS
 
 # Function to display enhanced results with current vs historical separation
 display_enhanced_results() {
@@ -365,6 +379,7 @@ main() {
     echo "Detector type: $DETECTOR_TYPE"
     echo "Specific scenario: $SPECIFIC_SCENARIO"
     echo "Interface: $INTERFACE"
+    echo "Selected detectors: ${SELECT_DETECTORS:-all}"
     echo "Profiles directory: $PROFILES_DIR"
     echo "Profiling enabled: ${ENABLE_PROFILING}"
     echo ""
diff --git a/performance-analysis/detectors/common.sh b/performance-analysis/detectors/common.sh
@@ -322,14 +322,26 @@ run_docker_benchmark() {
         echo "Hash collection disabled for faster execution"
     fi
 
+    # Prepare select detectors flag
+    local select_flag=""
+    local select_arg=""
+    if [[ -n "${SELECT_DETECTORS:-}" ]]; then
+        select_flag="--select-detectors '$SELECT_DETECTORS'"
+        select_arg=", selected_detectors='$SELECT_DETECTORS'"
+        echo "Using selected detectors: $SELECT_DETECTORS"
+    fi
+
     echo "Measuring execution time..."
     start_time=$(date +%s.%N)
 
     local cmd_result=0
     if [[ "$interface" == "cli" ]]; then
-        result=$(cd "$PROJECT_ROOT" && python3 -m dependency_resolver docker "$container_name" $hash_flag) || cmd_result=$?
+        local cmd="cd '$PROJECT_ROOT' && python3 -m dependency_resolver docker '$container_name'"
+        [[ -n "$hash_flag" ]] && cmd="$cmd $hash_flag"
+        [[ -n "$select_flag" ]] && cmd="$cmd $select_flag"
+        result=$(eval "$cmd") || cmd_result=$?
     else
-        result=$(cd "$PROJECT_ROOT" && python3 -c "import dependency_resolver; print(dependency_resolver.resolve_docker_dependencies('$container_name'$hash_arg))") || cmd_result=$?
+        result=$(cd "$PROJECT_ROOT" && python3 -c "import dependency_resolver; print(dependency_resolver.resolve_docker_dependencies('$container_name'$hash_arg$select_arg))") || cmd_result=$?
     fi
 
     end_time=$(date +%s.%N)
diff --git a/performance-analysis/detectors/host-benchmarks.sh b/performance-analysis/detectors/host-benchmarks.sh
@@ -56,47 +56,68 @@ if [[ "${SKIP_HASH_COLLECTION:-false}" == "true" ]]; then
     echo "Hash collection disabled for faster execution"
 fi
 
+# Prepare select detectors flag
+SELECT_FLAG=""
+SELECT_ARG=""
+if [[ -n "${SELECT_DETECTORS:-}" ]]; then
+    SELECT_FLAG="--select-detectors '$SELECT_DETECTORS'"
+    SELECT_ARG=", selected_detectors='$SELECT_DETECTORS'"
+    echo "Using selected detectors: $SELECT_DETECTORS"
+fi
+
 # Run selected scenarios based on scenario type and interface
 case $SCENARIO_TYPE in
     "skip-system")
         if [[ "$INTERFACE" == "cli" ]]; then
+            cmd="cd '$PROJECT_ROOT' && python3 -m dependency_resolver host --skip-system-scope"
+            [[ -n "$HASH_FLAG" ]] && cmd="$cmd $HASH_FLAG"
+            [[ -n "$SELECT_FLAG" ]] && cmd="$cmd $SELECT_FLAG"
             host_benchmark_scenario "Host Skip System Scope ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -m dependency_resolver host --skip-system-scope $HASH_FLAG" \
+                "$cmd" \
                 30
         else
             host_benchmark_scenario "Host Skip System Scope ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies(skip_system_scope=True$HASH_ARG))'" \
+                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies(skip_system_scope=True$HASH_ARG$SELECT_ARG))'" \
                 30
         fi
         ;;
     "full-system")
         if [[ "$INTERFACE" == "cli" ]]; then
+            cmd="cd '$PROJECT_ROOT' && python3 -m dependency_resolver host"
+            [[ -n "$HASH_FLAG" ]] && cmd="$cmd $HASH_FLAG"
+            [[ -n "$SELECT_FLAG" ]] && cmd="$cmd $SELECT_FLAG"
             host_benchmark_scenario "Host Full System ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -m dependency_resolver host $HASH_FLAG" \
+                "$cmd" \
                 60
         else
             host_benchmark_scenario "Host Full System ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies($HASH_ARG))'" \
+                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies($HASH_ARG$SELECT_ARG))'" \
                 60
         fi
         ;;
     "all")
         if [[ "$INTERFACE" == "cli" ]]; then
+            cmd1="cd '$PROJECT_ROOT' && python3 -m dependency_resolver host --skip-system-scope"
+            [[ -n "$HASH_FLAG" ]] && cmd1="$cmd1 $HASH_FLAG"
+            [[ -n "$SELECT_FLAG" ]] && cmd1="$cmd1 $SELECT_FLAG"
             host_benchmark_scenario "Host Skip System Scope ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -m dependency_resolver host --skip-system-scope $HASH_FLAG" \
+                "$cmd1" \
                 30
 
+            cmd2="cd '$PROJECT_ROOT' && python3 -m dependency_resolver host"
+            [[ -n "$HASH_FLAG" ]] && cmd2="$cmd2 $HASH_FLAG"
+            [[ -n "$SELECT_FLAG" ]] && cmd2="$cmd2 $SELECT_FLAG"
             host_benchmark_scenario "Host Full System ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -m dependency_resolver host $HASH_FLAG" \
+                "$cmd2" \
                 60
         else
             # Programmatic interface
             host_benchmark_scenario "Host Skip System Scope ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies(skip_system_scope=True$HASH_ARG))'" \
+                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies(skip_system_scope=True$HASH_ARG$SELECT_ARG))'" \
                 30
 
             host_benchmark_scenario "Host Full System ($INTERFACE)" \
-                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies($HASH_ARG))'" \
+                "cd '$PROJECT_ROOT' && python3 -c 'import dependency_resolver; print(dependency_resolver.resolve_host_dependencies($HASH_ARG$SELECT_ARG))'" \
                 60
         fi
         ;;
diff --git a/tests/core/test_orchestrator.py b/tests/core/test_orchestrator.py
diff --git a/tests/integration/test_cli.py b/tests/integration/test_cli.py
