CM-55207-Fix commit range parsing for empty remote (#371)

Author: mateusz-sterczewski

cycode/cli/consts.py

@@ -269,6 +269,7 @@
 
 # git consts
 COMMIT_DIFF_DELETED_FILE_CHANGE_TYPE = 'D'
+COMMIT_RANGE_ALL_COMMITS = '--all'
 GIT_HEAD_COMMIT_REV = 'HEAD'
 GIT_EMPTY_TREE_OBJECT = '4b825dc642cb6eb9a060e54bf8d69288fbee4904'
 EMPTY_COMMIT_SHA = '0000000000000000000000000000000000000000'

cycode/cli/files_collector/commit_range_documents.py

@@ -351,10 +351,10 @@ def calculate_pre_push_commit_range(push_update_details: str) -> Optional[str]:
                 return f'{merge_base}..{local_object_name}'
 
             logger.debug('Failed to find merge base with any default branch')
-            return '--all'
+            return consts.COMMIT_RANGE_ALL_COMMITS
         except Exception as e:
             logger.debug('Failed to get repo for pre-push commit range calculation: %s', exc_info=e)
-            return '--all'
+            return consts.COMMIT_RANGE_ALL_COMMITS
 
     # If deleting a branch (local_object_name is all zeros), no need to scan
     if local_object_name == consts.EMPTY_COMMIT_SHA:
@@ -448,9 +448,25 @@ def parse_commit_range(commit_range: str, path: str) -> tuple[Optional[str], Opt
     - 'commit' (interpreted as 'commit..HEAD')
     - '..to' (interpreted as 'HEAD..to')
     - 'from..' (interpreted as 'from..HEAD')
+    - '--all' (interpreted as 'first_commit..HEAD' to scan all commits)
     """
     repo = git_proxy.get_repo(path)
 
+    # Handle '--all' special case: scan all commits from first to HEAD
+    # Usually represents an empty remote repository
+    if commit_range == consts.COMMIT_RANGE_ALL_COMMITS:
+        try:
+            head_commit = repo.rev_parse(consts.GIT_HEAD_COMMIT_REV).hexsha
+            all_commits = repo.git.rev_list('--reverse', head_commit).strip()
+            if all_commits:
+                first_commit = all_commits.splitlines()[0]
+                return first_commit, head_commit, '..'
+            logger.warning("No commits found for range '%s'", commit_range)
+            return None, None, None
+        except Exception as e:
+            logger.warning("Failed to parse commit range '%s'", commit_range, exc_info=e)
+            return None, None, None
+
     separator = '..'
     if '...' in commit_range:
         from_spec, to_spec = commit_range.split('...', 1)

tests/cli/files_collector/test_commit_range_documents.py

@@ -882,6 +882,67 @@ def test_single_commit_spec(self) -> None:
             parsed_from, parsed_to, separator = parse_commit_range(a, temp_dir)
             assert (parsed_from, parsed_to, separator) == (a, c, '..')
 
+    def test_parse_all_for_empty_remote_scenario(self) -> None:
+        """Test that '--all' is parsed correctly for empty remote repository.
+        This repository has one commit locally.
+        """
+        with temporary_git_repository() as (temp_dir, repo):
+            # Create a local commit (simulating first commit to empty remote)
+            test_file = os.path.join(temp_dir, 'test.py')
+            with open(test_file, 'w') as f:
+                f.write("print('test')")
+
+            repo.index.add(['test.py'])
+            commit = repo.index.commit('Initial commit')
+
+            # Test that '--all' (returned by calculate_pre_push_commit_range for empty remote)
+            # can be parsed to a valid commit range
+            parsed_from, parsed_to, separator = parse_commit_range('--all', temp_dir)
+
+            # Should return first commit to HEAD (which is the only commit in this case)
+            assert parsed_from == commit.hexsha
+            assert parsed_to == commit.hexsha
+            assert separator == '..'
+
+    def test_parse_all_for_empty_remote_scenario_with_two_commits(self) -> None:
+        """Test that '--all' is parsed correctly for empty remote repository.
+        This repository has two commits locally.
+        """
+        with temporary_git_repository() as (temp_dir, repo):
+            # Create first commit
+            test_file = os.path.join(temp_dir, 'test.py')
+            with open(test_file, 'w') as f:
+                f.write("print('test')")
+
+            repo.index.add(['test.py'])
+            commit1 = repo.index.commit('First commit')
+
+            # Create second commit
+            test_file2 = os.path.join(temp_dir, 'test2.py')
+            with open(test_file2, 'w') as f:
+                f.write("print('test2')")
+
+            repo.index.add(['test2.py'])
+            commit2 = repo.index.commit('Second commit')
+
+            # Test that '--all' returns first commit to HEAD (second commit)
+            parsed_from, parsed_to, separator = parse_commit_range('--all', temp_dir)
+
+            # Should return first commit to HEAD (second commit)
+            assert parsed_from == commit1.hexsha  # First commit
+            assert parsed_to == commit2.hexsha  # HEAD (second commit)
+            assert separator == '..'
+
+    def test_parse_all_with_empty_repository_returns_none(self) -> None:
+        """Test that '--all' returns None when repository has no commits."""
+        with temporary_git_repository() as (temp_dir, repo):
+            # Empty repository with no commits
+            parsed_from, parsed_to, separator = parse_commit_range('--all', temp_dir)
+            # Should return None, None, None when HEAD doesn't exist
+            assert parsed_from is None
+            assert parsed_to is None
+            assert separator is None
+
 
 class TestParsePreReceiveInput:
     """Test the parse_pre_receive_input function with various pre-receive hook input scenarios."""