fix(workspace-changes): count diff body lines starting with "-- "/"++ " (#4303)

`_count_diff_lines` skipped every line starting with "+++ "/"--- " to drop
difflib's two file headers. But a deleted hunk-body line whose content
begins with "-- " (e.g. a SQL comment `-- get users`) becomes the diff
line `--- get users`, and an added `++ ...` line becomes `+++ ...`, so
those were dropped too — undercounting the user-visible +N/-M change
summary (WorkspaceChangeSummary, per-file WorkspaceFileChange, and the
run-event string in recorder.py).

`difflib.unified_diff` always emits the two headers first, so skip them
by position (`lines[2:]`) instead. Hunk `@@` lines start with `@` and are
still ignored. Numeric/plain diffs are unaffected.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

backend/packages/harness/deerflow/workspace_changes/diff.py

@@ -198,12 +198,12 @@ def _snapshot_text(file: FileSnapshot | None) -> str | None:
 def _count_diff_lines(lines: list[str]) -> tuple[int, int]:
     additions = 0
     deletions = 0
-    for line in lines:
-        # Unified-diff file headers are "+++ " / "--- " with a trailing space;
-        # a bare "+++"/"---" prefix would also swallow real content lines whose
-        # text begins with those sequences (e.g. an added line "+++foo").
-        if line.startswith("+++ ") or line.startswith("--- "):
-            continue
+    # difflib.unified_diff always emits the two file headers ("--- a/…" then
+    # "+++ b/…") first; skip them by position. A prefix test can't tell them
+    # apart from a hunk-body line whose content starts with "-- "/"++ " (e.g. a
+    # deleted SQL comment "-- get users" becomes the diff line "--- get users"),
+    # which would then be dropped from the count.
+    for line in lines[2:]:
         if line.startswith("+"):
             additions += 1
         elif line.startswith("-"):

backend/tests/test_workspace_changes.py

@@ -182,6 +182,33 @@ def test_count_diff_lines_ignores_only_real_headers():
     assert deletions == 2
 
 
+def test_count_diff_lines_counts_content_starting_with_dashes_or_pluses():
+    """Hunk-body lines whose content starts with '-- '/'++ ' must be counted.
+
+    difflib prefixes a deleted line "-- get users" to "--- get users"; the old
+    prefix skip mistook that for a file header and dropped it, undercounting
+    deletions in the user-visible +N/-M summary.
+    """
+    import difflib
+
+    from deerflow.workspace_changes.diff import _count_diff_lines
+
+    lines = list(
+        difflib.unified_diff(
+            ["SELECT 1", "-- get users"],
+            ["SELECT 1", "SELECT 2"],
+            fromfile="a/x.sql",
+            tofile="b/x.sql",
+            lineterm="",
+        )
+    )
+
+    additions, deletions = _count_diff_lines(lines)
+
+    assert additions == 1
+    assert deletions == 1
+
+
 def test_scan_workspace_roots_skips_excluded_directories(tmp_path):
     roots = _roots(tmp_path)
     workspace = roots[0].host_path