fix: robust changed file detection for PRs/MRs in GitHub, GitLab, and… (#106)

dacoburn · web-flow · commit 0546fab77d8b · 2025-07-28T21:09:49.000-07:00
* fix: robust changed file detection for PRs/MRs in GitHub, GitLab, and Bitbucket

- Use git diff with appropriate refs and environment variables to detect changed files in pull/merge requests across GitHub Actions, GitLab CI, and Bitbucket Pipelines.
- Fallback to git show for single commit detection.
- Ensures manifest and other file changes are correctly picked up in all major

* Fixing logic for Github changed files

* Add fetch of previous commit

* Fix file names for changed files to work with globbing

* Fixed detection for files in the commit files change
diff --git a/pyproject.toml b/pyproject.toml
@@ -6,7 +6,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "socketsecurity"
-version = "2.1.28"
+version = "2.1.33"
 requires-python = ">= 3.10"
 license = {"file" = "LICENSE"}
 dependencies = [
diff --git a/socketsecurity/__init__.py b/socketsecurity/__init__.py
@@ -1,2 +1,2 @@
 __author__ = 'socket.dev'
-__version__ = '2.1.28'
+__version__ = '2.1.33'
diff --git a/socketsecurity/core/__init__.py b/socketsecurity/core/__init__.py
@@ -389,15 +389,20 @@ def has_manifest_files(self, files: list) -> bool:
             from .utils import socket_globs as fallback_patterns
             patterns = fallback_patterns
 
+        # Normalize all file paths for matching
+        norm_files = [f.replace('\\', '/').lstrip('./') for f in files]
+
         for ecosystem in patterns:
             ecosystem_patterns = patterns[ecosystem]
             for file_name in ecosystem_patterns:
                 pattern_str = ecosystem_patterns[file_name]["pattern"]
-                for file in files:
-                    if "\\" in file:
-                        file = file.replace("\\", "/")
-                    if PurePath(file).match(pattern_str):
-                        return True
+                # Expand brace patterns for each manifest pattern
+                expanded_patterns = Core.expand_brace_pattern(pattern_str)
+                for exp_pat in expanded_patterns:
+                    for file in norm_files:
+                        # Use PurePath.match for glob-like matching
+                        if PurePath(file).match(exp_pat):
+                            return True
         return False
 
     def check_file_count_limit(self, file_count: int) -> dict:
diff --git a/socketsecurity/core/git_interface.py b/socketsecurity/core/git_interface.py
@@ -15,6 +15,13 @@ def __init__(self, path: str):
         self.repo = Repo(path)
         assert self.repo
         self.head = self.repo.head
+
+        # Always fetch all remote refs to ensure branches exist for diffing
+        try:
+            self.repo.git.fetch('--all')
+            log.debug("Fetched all remote refs for diffing.")
+        except Exception as fetch_error:
+            log.debug(f"Failed to fetch all remote refs: {fetch_error}")
         
         # Use CI environment SHA if available, otherwise fall back to current HEAD commit
         github_sha = os.getenv('GITHUB_SHA')
@@ -128,12 +135,95 @@ def __init__(self, path: str):
         self.commit_sha = self.commit.binsha
         self.commit_message = self.commit.message
         self.committer = self.commit.committer
-        self.show_files = self.repo.git.show(self.commit, name_only=True, format="%n").splitlines()
+        # Detect changed files in PR/MR context for GitHub, GitLab, Bitbucket; fallback to git show
+        self.show_files = []
+        detected = False
+        # GitHub Actions PR context
+        github_base_ref = os.getenv('GITHUB_BASE_REF')
+        github_head_ref = os.getenv('GITHUB_HEAD_REF')
+        github_event_name = os.getenv('GITHUB_EVENT_NAME')
+        github_before_sha = os.getenv('GITHUB_EVENT_BEFORE')  # previous commit for push
+        github_sha = os.getenv('GITHUB_SHA')  # current commit
+        if github_event_name == 'pull_request' and github_base_ref and github_head_ref:
+            try:
+                # Fetch both branches individually
+                self.repo.git.fetch('origin', github_base_ref)
+                self.repo.git.fetch('origin', github_head_ref)
+                # Try remote diff first
+                diff_range = f"origin/{github_base_ref}...origin/{github_head_ref}"
+                try:
+                    diff_files = self.repo.git.diff('--name-only', diff_range)
+                    self.show_files = diff_files.splitlines()
+                    log.debug(f"Changed files detected via git diff (GitHub PR remote): {self.show_files}")
+                    detected = True
+                except Exception as remote_error:
+                    log.debug(f"Remote diff failed: {remote_error}")
+                    # Try local branch diff
+                    local_diff_range = f"{github_base_ref}...{github_head_ref}"
+                    try:
+                        diff_files = self.repo.git.diff('--name-only', local_diff_range)
+                        self.show_files = diff_files.splitlines()
+                        log.debug(f"Changed files detected via git diff (GitHub PR local): {self.show_files}")
+                        detected = True
+                    except Exception as local_error:
+                        log.debug(f"Local diff failed: {local_error}")
+            except Exception as error:
+                log.debug(f"Failed to fetch branches or diff for GitHub PR: {error}")
+        # Commits to default branch (push events)
+        elif github_event_name == 'push' and github_before_sha and github_sha:
+            try:
+                diff_files = self.repo.git.diff('--name-only', f'{github_before_sha}..{github_sha}')
+                self.show_files = diff_files.splitlines()
+                log.debug(f"Changed files detected via git diff (GitHub push): {self.show_files}")
+                detected = True
+            except Exception as error:
+                log.debug(f"Failed to get changed files via git diff (GitHub push): {error}")
+        elif github_event_name == 'push':
+            try:
+                self.show_files = self.repo.git.show(self.commit, name_only=True, format="%n").splitlines()
+                log.debug(f"Changed files detected via git show (GitHub push fallback): {self.show_files}")
+                detected = True
+            except Exception as error:
+                log.debug(f"Failed to get changed files via git show (GitHub push fallback): {error}")
+        # GitLab CI Merge Request context
+        if not detected:
+            gitlab_target = os.getenv('CI_MERGE_REQUEST_TARGET_BRANCH_NAME')
+            gitlab_source = os.getenv('CI_MERGE_REQUEST_SOURCE_BRANCH_NAME')
+            if gitlab_target and gitlab_source:
+                try:
+                    self.repo.git.fetch('origin', gitlab_target, gitlab_source)
+                    diff_range = f"origin/{gitlab_target}...origin/{gitlab_source}"
+                    diff_files = self.repo.git.diff('--name-only', diff_range)
+                    self.show_files = diff_files.splitlines()
+                    log.debug(f"Changed files detected via git diff (GitLab): {self.show_files}")
+                    detected = True
+                except Exception as error:
+                    log.debug(f"Failed to get changed files via git diff (GitLab): {error}")
+        # Bitbucket Pipelines PR context
+        if not detected:
+            bitbucket_pr_id = os.getenv('BITBUCKET_PR_ID')
+            bitbucket_source = os.getenv('BITBUCKET_BRANCH')
+            bitbucket_dest = os.getenv('BITBUCKET_PR_DESTINATION_BRANCH')
+            # BITBUCKET_BRANCH is the source branch in PR builds
+            if bitbucket_pr_id and bitbucket_source and bitbucket_dest:
+                try:
+                    self.repo.git.fetch('origin', bitbucket_dest, bitbucket_source)
+                    diff_range = f"origin/{bitbucket_dest}...origin/{bitbucket_source}"
+                    diff_files = self.repo.git.diff('--name-only', diff_range)
+                    self.show_files = diff_files.splitlines()
+                    log.debug(f"Changed files detected via git diff (Bitbucket): {self.show_files}")
+                    detected = True
+                except Exception as error:
+                    log.debug(f"Failed to get changed files via git diff (Bitbucket): {error}")
+        # Fallback to git show for single commit
+        if not detected:
+            self.show_files = self.repo.git.show(self.commit, name_only=True, format="%n").splitlines()
+            log.debug(f"Changed files detected via git show: {self.show_files}")
         self.changed_files = []
         for item in self.show_files:
             if item != "":
-                full_path = f"{self.path}/{item}"
-                self.changed_files.append(full_path)
+                # Use relative path for glob matching
+                self.changed_files.append(item)
         
         # Determine if this commit is on the default branch
         # This considers both GitHub Actions detached HEAD and regular branch situations
diff --git a/workflows/github-actions.yml b/workflows/github-actions.yml
@@ -15,7 +15,7 @@ on:
 
 # Prevent concurrent runs for the same commit
 concurrency:
-  group: socket-scan-${{ github.ref }}-${{ github.sha }}
+  group: socket-scan-${{ github.sha }}
   cancel-in-progress: true
 
 jobs:
@@ -33,7 +33,7 @@ jobs:
       - uses: actions/checkout@v4
         with:
           # For PRs, fetch one additional commit for proper diff analysis
-          fetch-depth: ${{ github.event_name == 'pull_request' && 2 || 0 }}
+          fetch-depth: 0
       
       - name: Run Socket Security Scan
         env:

Original file line number	Diff line number	Diff line change
`@@ -1,2 +1,2 @@`
`1`	`1`	`__author__ = 'socket.dev'`
`2`		`-__version__ = '2.1.28'`
	`2`	`+__version__ = '2.1.33'`