kodit 0.2.6__py3-none-any.whl → 0.2.7__py3-none-any.whl

kodit/_version.py

@@ -17,5 +17,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.2.6'
-__version_tuple__ = version_tuple = (0, 2, 6)
+__version__ = version = '0.2.7'
+__version_tuple__ = version_tuple = (0, 2, 7)

kodit/infrastructure/cloning/git/factory.py

@@ -17,6 +17,7 @@ from kodit.infrastructure.cloning.metadata import (
     GitAuthorExtractor,
     GitFileMetadataExtractor,
 )
+from kodit.infrastructure.git.git_utils import sanitize_git_url
 from kodit.infrastructure.ignore.ignore_pattern_provider import GitIgnorePatternProvider
 
 
@@ -46,28 +47,33 @@ class GitSourceFactory:
             progress_callback = NullProgressCallback()
 
         # Normalize the URI
-        self.log.debug("Normalising git uri", uri=uri)
+        # Never log the raw URI in production
+        self.log.debug("Normalising git uri", uri="[REDACTED]" + uri[-4:])
         with tempfile.TemporaryDirectory() as temp_dir:
             git.Repo.clone_from(uri, temp_dir)
             remote = git.Repo(temp_dir).remote()
             uri = remote.url
 
+        # Sanitize the URI to remove any credentials
+        sanitized_uri = sanitize_git_url(uri)
+        self.log.debug("Sanitized git uri", sanitized_uri=sanitized_uri)
+
         # Check if source already exists
-        self.log.debug("Checking if source already exists", uri=uri)
-        source = await self.repository.get_by_uri(uri)
+        self.log.debug("Checking if source already exists", uri=sanitized_uri)
+        source = await self.repository.get_by_uri(sanitized_uri)
 
         if source:
             self.log.info("Source already exists, reusing...", source_id=source.id)
             return source
 
-        # Prepare working copy
+        # Prepare working copy (use original URI for cloning, sanitized for storage)
         clone_path = await self.working_copy.prepare(uri)
 
         # Create source record
-        self.log.debug("Creating source", uri=uri, clone_path=str(clone_path))
+        self.log.debug("Creating source", uri=sanitized_uri, clone_path=str(clone_path))
         source = await self.repository.save(
             Source(
-                uri=uri,
+                uri=sanitized_uri,
                 cloned_path=str(clone_path),
                 source_type=SourceType.GIT,
             )

kodit/infrastructure/cloning/git/working_copy.py

@@ -5,6 +5,8 @@ from pathlib import Path
 import git
 import structlog
 
+from kodit.infrastructure.git.git_utils import sanitize_git_url
+
 
 class GitWorkingCopyProvider:
     """Working copy provider for git-based sources."""
@@ -16,17 +18,23 @@ class GitWorkingCopyProvider:
 
     async def prepare(self, uri: str) -> Path:
         """Prepare a Git working copy."""
-        # Create a unique directory name for the clone
-        clone_path = self.clone_dir / uri.replace("/", "_").replace(":", "_")
+        # Sanitize the URI for directory name to prevent credential leaks
+        sanitized_uri = sanitize_git_url(uri)
+
+        # Create a unique directory name for the clone using the sanitized URI
+        clone_path = self.clone_dir / sanitized_uri.replace("/", "_").replace(":", "_")
         clone_path.mkdir(parents=True, exist_ok=True)
 
         try:
-            self.log.info("Cloning repository", uri=uri, clone_path=str(clone_path))
+            self.log.info(
+                "Cloning repository", uri=sanitized_uri, clone_path=str(clone_path)
+            )
+            # Use the original URI for cloning (with credentials if present)
             git.Repo.clone_from(uri, clone_path)
         except git.GitCommandError as e:
             if "already exists and is not an empty directory" not in str(e):
                 msg = f"Failed to clone repository: {e}"
                 raise ValueError(msg) from e
-            self.log.info("Repository already exists, reusing...", uri=uri)
+            self.log.info("Repository already exists, reusing...", uri=sanitized_uri)
 
         return clone_path

kodit/infrastructure/git/git_utils.py

@@ -1,6 +1,7 @@
 """Git utilities for infrastructure operations."""
 
 import tempfile
+from urllib.parse import urlparse, urlunparse
 
 import git
 
@@ -22,3 +23,65 @@ def is_valid_clone_target(target: str) -> bool:
             return False
         else:
             return True
+
+
+def sanitize_git_url(url: str) -> str:
+    """Remove credentials from a git URL while preserving the rest of the URL structure.
+
+    This function handles various git URL formats:
+    - HTTPS URLs with username:password@host
+    - HTTPS URLs with username@host (no password)
+    - SSH URLs (left unchanged)
+    - File URLs (left unchanged)
+
+    Args:
+        url: The git URL that may contain credentials.
+
+    Returns:
+        The sanitized URL with credentials removed.
+
+    Examples:
+        >>> sanitize_git_url("https://phil:token@dev.azure.com/org/project/_git/repo")
+        "https://dev.azure.com/org/project/_git/repo"
+        >>> sanitize_git_url("https://username@github.com/user/repo.git")
+        "https://github.com/user/repo.git"
+        >>> sanitize_git_url("git@github.com:user/repo.git")
+        "git@github.com:user/repo.git"
+
+    """
+    # Handle SSH URLs (they don't have credentials in the URL format)
+    if url.startswith(("git@", "ssh://")):
+        return url
+
+    # Handle file URLs
+    if url.startswith("file://"):
+        return url
+
+    try:
+        # Parse the URL
+        parsed = urlparse(url)
+
+        # If there are no credentials, return the URL as-is
+        if not parsed.username:
+            return url
+
+        # Reconstruct the URL without credentials
+        # Keep scheme, netloc (without username/password), path, params, query, fragment
+        sanitized_netloc = parsed.hostname
+        if parsed.port:
+            sanitized_netloc = f"{parsed.hostname}:{parsed.port}"
+
+        return urlunparse(
+            (
+                parsed.scheme,
+                sanitized_netloc,
+                parsed.path,
+                parsed.params,
+                parsed.query,
+                parsed.fragment,
+            )
+        )
+
+    except Exception:  # noqa: BLE001
+        # If URL parsing fails, return the original URL
+        return url

{kodit-0.2.6.dist-info → kodit-0.2.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kodit
-Version: 0.2.6
+Version: 0.2.7
 Summary: Code indexing for better AI code generation
 Project-URL: Homepage, https://docs.helixml.tech/kodit/
 Project-URL: Documentation, https://docs.helixml.tech/kodit/

{kodit-0.2.6.dist-info → kodit-0.2.7.dist-info}/RECORD

@@ -1,6 +1,6 @@
 kodit/.gitignore,sha256=ztkjgRwL9Uud1OEi36hGQeDGk3OLK1NfDEO8YqGYy8o,11
 kodit/__init__.py,sha256=aEKHYninUq1yh6jaNfvJBYg-6fenpN132nJt1UU6Jxs,59
-kodit/_version.py,sha256=nObnONsicQ3YX6SG5MVBxmIp5dmRacXDauSqZijWQbY,511
+kodit/_version.py,sha256=Xk20v7uvkFqkpy9aLJzVngs1eKQn0FYUP2oyA1MEQUU,511
 kodit/app.py,sha256=qKBWJ0VNSY_M6G3VFfAQ0133q5bnS99cUFD0p396taw,1032
 kodit/cli.py,sha256=JnhTlG1s04O0m8AzsBdrwP8T_BqSZMPXnRLvI7T_Gxc,12004
 kodit/config.py,sha256=3yh7hfLSILjZK_qJMhcExwRcrWJ0b5Eb1JjjOvMPJZo,4146
@@ -41,8 +41,8 @@ kodit/infrastructure/cloning/folder/__init__.py,sha256=w6ykrVtbYJlUDEXAjqgf6w2rM
 kodit/infrastructure/cloning/folder/factory.py,sha256=vl1hwnYA7lczjotn2fahJQAt7IK96CSArx8cSaRFKeY,4242
 kodit/infrastructure/cloning/folder/working_copy.py,sha256=FPhwzuPj40yGoYvwcm9VG8mv8MbJxwfby_N5JS-_daA,1154
 kodit/infrastructure/cloning/git/__init__.py,sha256=20ePcp0qE6BuLsjsv4KYB1DzKhMIMsPXwEqIEZtjTJs,34
-kodit/infrastructure/cloning/git/factory.py,sha256=1f0TKM9R_65WUMEhoJwBKTeo9xNYhm6VtscY7SqD6yU,5012
-kodit/infrastructure/cloning/git/working_copy.py,sha256=DMW_p7WWGoSeyDI9g55ItwsRomZSotXWRrlopqwszaQ,1115
+kodit/infrastructure/cloning/git/factory.py,sha256=cY0cxapp0NCvjMRpzesW_qRzbWbh-tMKIeAj0Eodyhw,5409
+kodit/infrastructure/cloning/git/working_copy.py,sha256=r_Uu6NYhRQLAQce6k4KThXLtGPqkzr6CgWx7AJ89gN4,1462
 kodit/infrastructure/embedding/__init__.py,sha256=F-8nLlWAerYJ0MOIA4tbXHLan8bW5rRR84vzxx6tRKI,39
 kodit/infrastructure/embedding/embedding_factory.py,sha256=1AypjhWJGxvLnZt1SEH_FHPk9P0Vkt9fXdSGzFPp2ow,3432
 kodit/infrastructure/embedding/local_vector_search_repository.py,sha256=UO8A3Eb_djFVrWKKSukAo4u7k8djDD1SlOPHk2pP9ps,3921
@@ -59,7 +59,7 @@ kodit/infrastructure/enrichment/local_enrichment_provider.py,sha256=8CATNtgMHgBR
 kodit/infrastructure/enrichment/null_enrichment_provider.py,sha256=5Ksyxl3qDLxUjmOeIdHZ0UAIULy7RcbLXJoT7_CNXoQ,775
 kodit/infrastructure/enrichment/openai_enrichment_provider.py,sha256=fenq4HiJ2UkrzsE2D0A0qpmro38z9mKaIzKKU5v7hnY,3189
 kodit/infrastructure/git/__init__.py,sha256=0iMosFzudj4_xNIMe2SRbV6l5bWqkjnUsZoFsoZFuM8,33
-kodit/infrastructure/git/git_utils.py,sha256=lOujEx41UuWfYSnFWbY4HC2tK5utytyzNkW1e5IPCr0,543
+kodit/infrastructure/git/git_utils.py,sha256=2DH6cyTjDRwFfL5Bzt1y2w0DwHZNypbC6R0Gm_A3hhg,2476
 kodit/infrastructure/ignore/__init__.py,sha256=VzFv8XOzHmsu0MEGnWVSF6KsgqLBmvHlRqAkT1Xb1MY,36
 kodit/infrastructure/ignore/ignore_pattern_provider.py,sha256=9m2XCsgW87UBTfzHr6Z0Ns6WpzwkLir3zyBY3PwsgXk,2225
 kodit/infrastructure/indexing/__init__.py,sha256=7UPRa2jwCAsa0Orsp6PqXSF8iIXJVzXHMFmrKkI9yH8,38
@@ -93,8 +93,8 @@ kodit/migrations/versions/85155663351e_initial.py,sha256=Cg7zlF871o9ShV5rQMQ1v7h
 kodit/migrations/versions/9e53ea8bb3b0_add_authors.py,sha256=a32Zm8KUQyiiLkjKNPYdaJDgjW6VsV-GhaLnPnK_fpI,3884
 kodit/migrations/versions/__init__.py,sha256=9-lHzptItTzq_fomdIRBegQNm4Znx6pVjwD4MiqRIdo,36
 kodit/migrations/versions/c3f5137d30f5_index_all_the_things.py,sha256=rI8LmjF-I2OMxZ2nOIF_NRmqOLXe45hL_iz_nx97DTQ,1680
-kodit-0.2.6.dist-info/METADATA,sha256=q9yeLcPnH31fKZe5i5Bzw30NDuJge6NsCJDVToc0-OE,5867
-kodit-0.2.6.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-kodit-0.2.6.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
-kodit-0.2.6.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-kodit-0.2.6.dist-info/RECORD,,
+kodit-0.2.7.dist-info/METADATA,sha256=G5rdRgHtm6V-p1tt0h7g7pZJKnQ5mAnCgDXVFRJG7Fg,5867
+kodit-0.2.7.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+kodit-0.2.7.dist-info/entry_points.txt,sha256=hoTn-1aKyTItjnY91fnO-rV5uaWQLQ-Vi7V5et2IbHY,40
+kodit-0.2.7.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+kodit-0.2.7.dist-info/RECORD,,