ghostbuster-cli 0.2.2__tar.gz → 0.4.0__tar.gz

ghostbuster_cli-0.4.0/.github/FUNDING.yml

@@ -0,0 +1,4 @@
+# GitHub Sponsor configuration for ghostbuster-cli
+ko_fi: adewanggar
+custom:
+  - https://saweria.co/adewanggar

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/.github/workflows/publish.yml

@@ -3,9 +3,6 @@ name: Publish to PyPI
 on:
   release:
     types: [published]
-  push:
-    tags:
-      - "v*"
 
 jobs:
   pypi-publish:
@@ -33,3 +30,5 @@ jobs:
 
       - name: Publish package distributions to PyPI
         uses: pypa/gh-action-pypi-publish@release/v1
+        with:
+          skip-existing: true

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/CHANGELOG.md

@@ -7,6 +7,28 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.4.0] - 2026-08-29
+
+### Added
+
+- **Git Diff / Incremental Fast Scan Mode (`ghostbuster scan --diff` / `-d`)**:
+  - Scan only modified, staged, and untracked files instead of the entire project tree.
+  - Added `--diff-base` flag to compare working branch against a target ref (e.g. `ghostbuster scan --diff-base origin/main`).
+  - Ideal for monorepos and GitHub Actions PR validation.
+  - Zero false positives: global reference verification remains intact for zombie code.
+
+## [0.3.0] - 2026-08-29
+
+### Added
+
+- **Node.js, JavaScript, and TypeScript Support**:
+  - **Dead Imports**: Scans `package.json` (`dependencies`, `devDependencies`, `peerDependencies`) and verifies against imports in `.js`, `.mjs`, `.cjs`, `.jsx`, `.ts`, `.mts`, `.cts`, `.tsx`, `.vue`, and `.svelte` files.
+  - **Scoped Packages & Subpaths**: Supports `@scope/pkg` and `pkg/subpath` imports.
+  - **Phantom Env**: Scans `process.env.VAR`, `process.env['VAR']`, `import.meta.env.VITE_VAR`, and `NEXT_PUBLIC_VAR`.
+  - **Auto-Fix for `package.json`**: `ghostbuster bust --confirm` automatically removes dead dependencies from `package.json` with preserved JSON formatting.
+  - **Orphan Files**: Added detection for Node log files (`npm-debug.log`, `yarn-error.log`, `pnpm-debug.log`).
+  - **Full-Stack Scanning**: Simultaneously scans Python backends and JS/TS frontends in monorepos with a unified Ghost Score.
+
 ## [0.2.2] - 2026-08-29
 
 ### Fixed

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: ghostbuster-cli
-Version: 0.2.2
+Version: 0.4.0
 Summary: Find and bust the ghosts haunting your codebase - unused deps, dead code, orphan files, and phantom env vars.
 Project-URL: Homepage, https://github.com/adewanggar/ghostbuster-cli
 Project-URL: Documentation, https://github.com/adewanggar/ghostbuster-cli#readme
@@ -128,6 +128,18 @@ ghostbuster scan ./my-project
 ghostbuster scan -v
 ```
 
+### Incremental / Git Diff Scan
+
+Scan only files that were modified, staged, or untracked in your working tree:
+
+```bash
+# Scan modified files only (fast mode)
+ghostbuster scan --diff
+
+# Compare working branch against main/master
+ghostbuster scan --diff-base origin/main
+```
+
 ### Filtered Scan
 
 ```bash
@@ -225,11 +237,11 @@ Ghostbuster uses **pure AST analysis** - no runtime imports, no code execution,
 
 ## Roadmap
 
-- <img src="https://raw.githubusercontent.com/adewanggar/ghostbuster-cli/main/assets/npm_package_icon.png" width="18" height="18" valign="middle" /> Node.js / TypeScript support (scan `package.json`, dead exports)
-- Pre-commit hook integration
-- GitHub Actions reporter (comment Ghost Score on PRs)
-- Config inheritance for monorepos
-- Ghost Score history tracking & trend chart
+- [x] <img src="https://raw.githubusercontent.com/adewanggar/ghostbuster-cli/main/assets/npm_package_icon.png" width="18" height="18" valign="middle" /> Node.js / JavaScript / TypeScript support (`package.json`, `process.env`, `import.meta.env`)
+- [x] Pre-commit hook integration (`.pre-commit-hooks.yaml`)
+- [ ] GitHub Actions reporter (comment Ghost Score on PRs)
+- [ ] Config inheritance for monorepos
+- [ ] Ghost Score history tracking & trend chart
 
 ## Contributing
 
@@ -243,6 +255,13 @@ pip install pytest ruff mypy
 pytest tests/ -v
 ```
 
+## Support & Sponsor
+
+If Ghostbuster helped clean up your codebase or saved CI minutes, consider supporting the project:
+
+[![Ko-fi](https://img.shields.io/badge/Ko--fi-Support%20on%20Ko--fi-FF5E5B?style=for-the-badge&logo=ko-fi&logoColor=white)](https://ko-fi.com/adewanggar)
+[![Saweria](https://img.shields.io/badge/Saweria-Dukung%20lewat%20Saweria-FA8072?style=for-the-badge&logo=tether&logoColor=white)](https://saweria.co/adewanggar)
+
 ## License
 
 MIT (c) Ghostbuster Contributors

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/README.md

@@ -90,6 +90,18 @@ ghostbuster scan ./my-project
 ghostbuster scan -v
 ```
 
+### Incremental / Git Diff Scan
+
+Scan only files that were modified, staged, or untracked in your working tree:
+
+```bash
+# Scan modified files only (fast mode)
+ghostbuster scan --diff
+
+# Compare working branch against main/master
+ghostbuster scan --diff-base origin/main
+```
+
 ### Filtered Scan
 
 ```bash
@@ -187,11 +199,11 @@ Ghostbuster uses **pure AST analysis** - no runtime imports, no code execution,
 
 ## Roadmap
 
-- <img src="https://raw.githubusercontent.com/adewanggar/ghostbuster-cli/main/assets/npm_package_icon.png" width="18" height="18" valign="middle" /> Node.js / TypeScript support (scan `package.json`, dead exports)
-- Pre-commit hook integration
-- GitHub Actions reporter (comment Ghost Score on PRs)
-- Config inheritance for monorepos
-- Ghost Score history tracking & trend chart
+- [x] <img src="https://raw.githubusercontent.com/adewanggar/ghostbuster-cli/main/assets/npm_package_icon.png" width="18" height="18" valign="middle" /> Node.js / JavaScript / TypeScript support (`package.json`, `process.env`, `import.meta.env`)
+- [x] Pre-commit hook integration (`.pre-commit-hooks.yaml`)
+- [ ] GitHub Actions reporter (comment Ghost Score on PRs)
+- [ ] Config inheritance for monorepos
+- [ ] Ghost Score history tracking & trend chart
 
 ## Contributing
 
@@ -205,6 +217,13 @@ pip install pytest ruff mypy
 pytest tests/ -v
 ```
 
+## Support & Sponsor
+
+If Ghostbuster helped clean up your codebase or saved CI minutes, consider supporting the project:
+
+[![Ko-fi](https://img.shields.io/badge/Ko--fi-Support%20on%20Ko--fi-FF5E5B?style=for-the-badge&logo=ko-fi&logoColor=white)](https://ko-fi.com/adewanggar)
+[![Saweria](https://img.shields.io/badge/Saweria-Dukung%20lewat%20Saweria-FA8072?style=for-the-badge&logo=tether&logoColor=white)](https://saweria.co/adewanggar)
+
 ## License
 
 MIT (c) Ghostbuster Contributors

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "ghostbuster-cli"
-version = "0.2.2"
+version = "0.4.0"
 description = "Find and bust the ghosts haunting your codebase - unused deps, dead code, orphan files, and phantom env vars."
 readme = "README.md"
 license = "MIT"

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/src/ghostbuster/__init__.py

@@ -1,3 +1,3 @@
 """Ghostbuster — Find and bust the ghosts haunting your codebase."""
 
-__version__ = "0.2.2"
+__version__ = "0.4.0"

{ghostbuster_cli-0.2.2 → ghostbuster_cli-0.4.0}/src/ghostbuster/cli/scan.py

@@ -23,6 +23,7 @@ from ghostbuster.cli.display import (
     print_scan_result_markdown,
     print_scanning_start,
 )
+from ghostbuster.core.git_diff import get_changed_files
 from ghostbuster.core.models import GhostCategory
 from ghostbuster.core.scanner import create_default_orchestrator
 
@@ -55,6 +56,17 @@ def scan(
         "-c",
         help="Only scan specific categories (dead-import, orphan-file, zombie-code, phantom-env).",
     ),
+    diff: bool = typer.Option(
+        False,
+        "--diff",
+        "-d",
+        help="Incremental fast scan: only scan modified, staged, and untracked files.",
+    ),
+    diff_base: str | None = typer.Option(
+        None,
+        "--diff-base",
+        help="Compare changed files against a specific git ref/branch (e.g. main, origin/main, HEAD~1).",
+    ),
     verbose: bool = typer.Option(
         False,
         "--verbose",
@@ -74,6 +86,8 @@ def scan(
 
     Examples:
         ghostbuster scan
+        ghostbuster scan --diff
+        ghostbuster scan --diff-base origin/main
         ghostbuster scan ./my-project --verbose
         ghostbuster scan --format json --category dead-import
     """
@@ -91,10 +105,19 @@ def scan(
                 )
                 raise typer.Exit(1)
 
+    # Handle diff mode
+    changed_files: set[Path] | None = None
+    if diff or diff_base:
+        changed_files = get_changed_files(path, base_ref=diff_base)
+
     # Show banner for rich format
     if format == OutputFormat.rich:
         print_banner()
         print_scanning_start(str(path), _get_scanner_names(category))
+        if diff or diff_base:
+            console.print(
+                f"  [cyan]Diff mode active ({len(changed_files or set())} changed files analyzed)[/cyan]\n"
+            )
 
     # Run the scan
     try:
@@ -105,7 +128,11 @@ def scan(
             if format == OutputFormat.rich
             else _nullcontext()
         ):
-            result = orchestrator.run(path, categories=category)
+            result = orchestrator.run(
+                path,
+                categories=category,
+                changed_files=changed_files if (diff or diff_base) else None,
+            )
 
     except Exception as exc:
         print_error(f"Scan failed: {exc}", "Run with --debug flag for full traceback.")

ghostbuster_cli-0.4.0/src/ghostbuster/core/dead_imports.py

@@ -0,0 +1,416 @@
+"""Dead Import Scanner - detects dependencies declared but never imported.
+
+Checks requirements.txt, pyproject.toml, and package.json for declared
+dependencies, then walks Python, JavaScript, and TypeScript files to see which
+are actually imported.
+"""
+
+from __future__ import annotations
+
+import ast
+import json
+import re
+import sys
+from pathlib import Path
+
+from ghostbuster.core.models import Ghost, GhostCategory, Severity
+
+# Mapping from PyPI package names to their actual import names.
+PACKAGE_TO_IMPORT: dict[str, str | list[str]] = {
+    "pillow": "PIL",
+    "scikit-learn": "sklearn",
+    "scikit-image": "skimage",
+    "opencv-python": "cv2",
+    "opencv-python-headless": "cv2",
+    "python-dateutil": "dateutil",
+    "pyyaml": "yaml",
+    "beautifulsoup4": "bs4",
+    "python-dotenv": "dotenv",
+    "pymysql": "pymysql",
+    "attrs": "attr",
+    "google-cloud-storage": "google.cloud.storage",
+    "google-auth": "google.auth",
+    "python-jose": "jose",
+    "python-multipart": "multipart",
+    "ujson": "ujson",
+    "msgpack": "msgpack",
+    "setuptools": ["setuptools", "pkg_resources"],
+    "typing-extensions": "typing_extensions",
+    "importlib-metadata": "importlib_metadata",
+    "importlib-resources": "importlib_resources",
+}
+
+# Python tool packages that are run via CLI or plugins
+TOOL_PACKAGES: set[str] = {
+    "pip",
+    "setuptools",
+    "wheel",
+    "build",
+    "twine",
+    "ruff",
+    "black",
+    "isort",
+    "flake8",
+    "mypy",
+    "pytest",
+    "pytest-cov",
+    "pytest-xdist",
+    "pytest-asyncio",
+    "pre-commit",
+    "tox",
+    "nox",
+    "sphinx",
+    "ipython",
+    "ipdb",
+    "debugpy",
+    "hatchling",
+    "hatch",
+    "poetry",
+    "flit",
+    "pdm",
+    "autopep8",
+    "pyright",
+    "pylint",
+    "bandit",
+    "safety",
+    "coverage",
+}
+
+# Node.js tool/build packages that are run via CLI or config
+NODE_TOOL_PACKAGES: set[str] = {
+    "typescript",
+    "eslint",
+    "prettier",
+    "vite",
+    "webpack",
+    "rollup",
+    "tsup",
+    "esbuild",
+    "jest",
+    "vitest",
+    "mocha",
+    "chai",
+    "husky",
+    "lint-staged",
+    "rimraf",
+    "nodemon",
+    "concurrently",
+    "cross-env",
+    "tailwindcss",
+    "postcss",
+    "autoprefixer",
+    "dotenv-cli",
+}
+
+# Directories to skip when scanning source files
+SKIP_DIRS: set[str] = {
+    "venv",
+    ".venv",
+    "node_modules",
+    ".git",
+    "__pycache__",
+    ".tox",
+    ".nox",
+    "build",
+    "dist",
+    ".next",
+    ".nuxt",
+    ".turbo",
+    "coverage",
+}
+
+JS_TS_EXTENSIONS: set[str] = {
+    ".js",
+    ".mjs",
+    ".cjs",
+    ".jsx",
+    ".ts",
+    ".mts",
+    ".cts",
+    ".tsx",
+    ".vue",
+    ".svelte",
+}
+
+
+class DeadImportScanner:
+    """Detects dependencies declared in config but never imported in code."""
+
+    name = "dead-import"
+
+    def scan(self, path: Path) -> list[Ghost]:
+        """Scan a project directory for dead imports across Python and Node.js."""
+        declared = self._find_declared_dependencies(path)
+        if not declared:
+            return []
+
+        imported = self._find_all_imports(path)
+        ghosts: list[Ghost] = []
+
+        for dep_name, source_file in declared:
+            is_node = source_file.startswith("package.json")
+
+            if is_node:
+                # Node.js package handling
+                if self._is_node_tool_package(dep_name):
+                    continue
+
+                if not self._is_node_dep_used(dep_name, imported):
+                    ghosts.append(
+                        Ghost(
+                            category=GhostCategory.DEAD_IMPORT,
+                            name=dep_name,
+                            message=f"Package '{dep_name}' is declared in {source_file} but never imported",
+                            file_path=path / "package.json",
+                            severity=Severity.MEDIUM,
+                            fixable=True,
+                            suggestion=f"Remove '{dep_name}' from package.json",
+                        )
+                    )
+            else:
+                # Python package handling
+                normalized = self._normalize_package_name(dep_name)
+                if normalized in TOOL_PACKAGES:
+                    continue
+
+                import_names = self._get_import_names(normalized)
+                is_used = any(
+                    self._import_matches(imp, name) for imp in imported for name in import_names
+                )
+
+                if not is_used:
+                    ghosts.append(
+                        Ghost(
+                            category=GhostCategory.DEAD_IMPORT,
+                            name=dep_name,
+                            message=f"Package '{dep_name}' is declared in {source_file} but never imported",
+                            file_path=path / (source_file.split(" [")[0]),
+                            severity=Severity.MEDIUM,
+                            fixable=True,
+                            suggestion=f"Remove '{dep_name}' from {source_file.split(' [')[0]}",
+                        )
+                    )
+
+        return ghosts
+
+    def _is_node_tool_package(self, pkg_name: str) -> bool:
+        """Check if a Node package is a known CLI/type/build tool."""
+        if pkg_name.startswith("@types/"):
+            return True
+        if pkg_name.startswith("@vitejs/") or pkg_name.startswith("@babel/"):
+            return True
+        return pkg_name.lower() in NODE_TOOL_PACKAGES
+
+    def _is_node_dep_used(self, dep_name: str, imported: set[str]) -> bool:
+        """Check if a declared Node dependency is used in imports."""
+        dep_lower = dep_name.lower()
+        for imp in imported:
+            imp_lower = imp.lower()
+            if imp_lower == dep_lower or imp_lower.startswith(dep_lower + "/"):
+                return True
+        return False
+
+    def _normalize_package_name(self, name: str) -> str:
+        """Normalize a Python package name: lowercase, hyphens to underscores."""
+        name = re.split(r"[>=<!\[;@]", name)[0].strip()
+        return name.lower().replace("-", "_")
+
+    def _get_import_names(self, normalized: str) -> list[str]:
+        """Get all possible import names for a normalized package name."""
+        original = normalized.replace("_", "-")
+        if original in PACKAGE_TO_IMPORT:
+            mapping = PACKAGE_TO_IMPORT[original]
+            if isinstance(mapping, list):
+                return [m.lower().replace("-", "_") for m in mapping]
+            return [mapping.lower().replace("-", "_")]
+
+        if normalized in PACKAGE_TO_IMPORT:
+            mapping = PACKAGE_TO_IMPORT[normalized]
+            if isinstance(mapping, list):
+                return [m.lower().replace("-", "_") for m in mapping]
+            return [mapping.lower().replace("-", "_")]
+
+        return [normalized]
+
+    def _import_matches(self, imported: str, expected: str) -> bool:
+        """Check if an imported module matches an expected package name."""
+        imported_lower = imported.lower().replace("-", "_")
+        expected_lower = expected.lower().replace("-", "_")
+        return imported_lower == expected_lower or imported_lower.startswith(expected_lower + ".")
+
+    def _find_declared_dependencies(self, path: Path) -> list[tuple[str, str]]:
+        """Find all declared dependencies from Python and Node.js config files."""
+        deps: list[tuple[str, str]] = []
+
+        # Check requirements.txt
+        req_file = path / "requirements.txt"
+        if req_file.exists():
+            deps.extend(self._parse_requirements_txt(req_file))
+
+        # Check pyproject.toml
+        pyproject = path / "pyproject.toml"
+        if pyproject.exists():
+            deps.extend(self._parse_pyproject_toml(pyproject))
+
+        # Check package.json (Node.js)
+        package_json = path / "package.json"
+        if package_json.exists():
+            deps.extend(self._parse_package_json(package_json))
+
+        return deps
+
+    def _parse_requirements_txt(self, filepath: Path) -> list[tuple[str, str]]:
+        """Parse a requirements.txt file for package names."""
+        deps: list[tuple[str, str]] = []
+        try:
+            for line in filepath.read_text(encoding="utf-8").splitlines():
+                line = line.strip()
+                if not line or line.startswith("#") or line.startswith("-"):
+                    continue
+                name = re.split(r"[>=<!\[;@]", line)[0].strip()
+                if name:
+                    deps.append((name, "requirements.txt"))
+        except OSError:
+            pass
+        return deps
+
+    def _parse_pyproject_toml(self, filepath: Path) -> list[tuple[str, str]]:
+        """Parse pyproject.toml for declared dependencies."""
+        deps: list[tuple[str, str]] = []
+        try:
+            if sys.version_info >= (3, 11):
+                import tomllib
+            else:
+                import tomli as tomllib  # type: ignore[no-redef]
+
+            data = tomllib.loads(filepath.read_text(encoding="utf-8"))
+            project_deps = data.get("project", {}).get("dependencies", [])
+            for dep in project_deps:
+                name = re.split(r"[>=<!\[;@]", dep)[0].strip()
+                if name:
+                    deps.append((name, "pyproject.toml"))
+
+            # Also check optional dependencies
+            optional = data.get("project", {}).get("optional-dependencies", {})
+            for group_name, group_deps in optional.items():
+                for dep in group_deps:
+                    name = re.split(r"[>=<!\[;@]", dep)[0].strip()
+                    if name:
+                        deps.append((name, f"pyproject.toml [optional: {group_name}]"))
+
+            # Support Poetry: [tool.poetry.dependencies]
+            poetry_deps = data.get("tool", {}).get("poetry", {}).get("dependencies", {})
+            if isinstance(poetry_deps, dict):
+                for dep_name in poetry_deps:
+                    if dep_name.lower() != "python":
+                        deps.append((dep_name, "pyproject.toml [poetry.dependencies]"))
+
+            # Support Poetry groups: [tool.poetry.group.<group>.dependencies]
+            poetry_groups = data.get("tool", {}).get("poetry", {}).get("group", {})
+            if isinstance(poetry_groups, dict):
+                for group_name, group_data in poetry_groups.items():
+                    if isinstance(group_data, dict):
+                        g_deps = group_data.get("dependencies", {})
+                        if isinstance(g_deps, dict):
+                            for dep_name in g_deps:
+                                if dep_name.lower() != "python":
+                                    deps.append(
+                                        (dep_name, f"pyproject.toml [poetry.group.{group_name}]")
+                                    )
+
+        except (OSError, Exception):
+            pass
+        return deps
+
+    def _parse_package_json(self, filepath: Path) -> list[tuple[str, str]]:
+        """Parse package.json for declared dependencies."""
+        deps: list[tuple[str, str]] = []
+        try:
+            data = json.loads(filepath.read_text(encoding="utf-8"))
+            for section in ["dependencies", "devDependencies", "peerDependencies"]:
+                sec_deps = data.get(section, {})
+                if isinstance(sec_deps, dict):
+                    for dep_name in sec_deps:
+                        deps.append((dep_name, f"package.json [{section}]"))
+        except (OSError, ValueError):
+            pass
+        return deps
+
+    def _find_all_imports(self, path: Path) -> set[str]:
+        """Walk Python and JS/TS files and collect all imported module/package names."""
+        imports: set[str] = set()
+
+        for file_path in path.rglob("*"):
+            if not file_path.is_file():
+                continue
+
+            parts = file_path.relative_to(path).parts
+            if any(p in SKIP_DIRS for p in parts):
+                continue
+
+            suffix = file_path.suffix.lower()
+
+            # Python files
+            if suffix == ".py":
+                self._extract_python_imports(file_path, imports)
+            # JS/TS/Vue/Svelte files
+            elif suffix in JS_TS_EXTENSIONS:
+                self._extract_js_ts_imports(file_path, imports)
+
+        return imports
+
+    def _extract_python_imports(self, filepath: Path, imports: set[str]) -> None:
+        """Extract imports from a Python file."""
+        try:
+            source = filepath.read_text(encoding="utf-8", errors="ignore")
+            tree = ast.parse(source, filename=str(filepath))
+            for node in ast.walk(tree):
+                if isinstance(node, ast.Import):
+                    for alias in node.names:
+                        imports.add(alias.name)
+                elif isinstance(node, ast.ImportFrom) and node.module:
+                    imports.add(node.module)
+        except (SyntaxError, OSError):
+            pass
+
+    def _extract_js_ts_imports(self, filepath: Path, imports: set[str]) -> None:
+        """Extract imports/requires from a JavaScript or TypeScript file."""
+        try:
+            source = filepath.read_text(encoding="utf-8", errors="ignore")
+
+            # Patterns:
+            # 1. import ... from 'pkg'
+            # 2. import 'pkg'
+            # 3. export ... from 'pkg'
+            # 4. require('pkg')
+            # 5. import('pkg')
+            patterns = [
+                r"""(?:import|export)\s+(?:[\w\s{},*]+from\s+)?['"]([^'"]+)['"]""",
+                r"""(?:require|import)\s*\(\s*['"]([^'"]+)['"]\s*\)""",
+            ]
+
+            for pattern in patterns:
+                for match in re.finditer(pattern, source):
+                    raw_specifier = match.group(1).strip()
+                    # Skip relative or absolute paths
+                    if (
+                        raw_specifier.startswith(".")
+                        or raw_specifier.startswith("/")
+                        or raw_specifier.startswith("#")
+                    ):
+                        continue
+
+                    # Handle scoped package: @org/pkg/sub -> @org/pkg
+                    if raw_specifier.startswith("@"):
+                        parts = raw_specifier.split("/")
+                        if len(parts) >= 2:
+                            imports.add(f"{parts[0]}/{parts[1]}")
+                        else:
+                            imports.add(raw_specifier)
+                    else:
+                        # Unscoped package: pkg/sub -> pkg
+                        pkg_root = raw_specifier.split("/")[0]
+                        imports.add(pkg_root)
+
+        except OSError:
+            pass