fix(mcp): resolve generated file paths containing spaces (#5819)

* fix(mcp): resolve generated file paths containing spaces

* fix(mcp): avoid partial rewrites at whitespace suffixes

---------

Co-authored-by: xbzz1018 <186267668+xbzz1018@users.noreply.github.com>

backend/packages/harness/deerflow/mcp/tools.py

@@ -288,6 +288,50 @@ def _rewrite_unique_bare_filenames(
     return rewritten
 
 
+def _rewrite_changed_paths_with_spaces(
+    text: str,
+    *,
+    changed_files: Iterable[Path],
+    thread_id: str,
+    user_id: str,
+    source_base_dir: Path | None,
+) -> str:
+    """Rewrite changed paths with spaces only at unambiguous text boundaries."""
+    candidates: dict[str, set[str]] = {}
+    for path in changed_files:
+        spellings = [str(path)]
+        if path.anchor == "/":
+            # Some servers print file URIs without percent-encoding spaces.
+            spellings.extend((f"file:{path}", f"file://{path}"))
+        if source_base_dir is not None:
+            try:
+                relative = path.relative_to(source_base_dir).as_posix()
+            except ValueError:
+                pass
+            else:
+                spellings.extend((relative, f"./{relative}"))
+        matching_spellings = [spelling for spelling in spellings if any(char.isspace() for char in spelling) and spelling in text]
+        if not matching_spellings:
+            continue
+        virtual_path = _local_uri_to_virtual_path(str(path), thread_id=thread_id, user_id=user_id)
+        if virtual_path is None:
+            continue
+        for spelling in matching_spellings:
+            candidates.setdefault(spelling, set()).add(virtual_path)
+
+    rewritten = text
+    for spelling in sorted(candidates, key=len, reverse=True):
+        destinations = candidates[spelling]
+        if len(destinations) != 1 or spelling not in rewritten:
+            continue
+        # Whitespace after an unquoted spelling may continue a longer filename.
+        # Require punctuation or the end of the text instead of rewriting a prefix.
+        pattern = re.compile(rf"(?<![\w./\\-]){re.escape(spelling)}(?!(?:[\w\s/\\-]|\.[\w]))")
+        replacement = next(iter(destinations))
+        rewritten = pattern.sub(lambda _match: replacement, rewritten)
+    return rewritten
+
+
 def _rewrite_local_paths_in_text(
     text: str,
     *,
@@ -327,6 +371,16 @@ def _rewrite_local_paths_in_text(
             return token
         return f"{rewritten}{trailing}"
 
+    if changed_files is not None:
+        changed_files = list(changed_files)
+        text = _rewrite_changed_paths_with_spaces(
+            text,
+            changed_files=changed_files,
+            thread_id=thread_id,
+            user_id=user_id,
+            source_base_dir=source_base_dir,
+        )
+
     rewritten = _LOCAL_PATH_IN_TEXT_RE.sub(_replace, text)
     if changed_files is None:
         return rewritten

backend/tests/test_mcp_file_migration.py

@@ -212,6 +212,125 @@ class TestRewriteLocalPathsInText:
 
         assert result == f"Screenshot saved to {VIRTUAL_PATH_PREFIX}/workspace/absolute-output/page.png"
 
+    def test_new_absolute_path_with_spaces_is_rewritten(self, paths: Paths):
+        workspace = paths.sandbox_work_dir("t1", user_id="u1")
+        src = _workspace_file(paths, "reports/final report.txt")
+        text = f'Saved as "{src}" and ready.'
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                source_base_dir=workspace,
+                changed_files=[src],
+            )
+
+        assert result == f'Saved as "{VIRTUAL_PATH_PREFIX}/workspace/reports/final report.txt" and ready.'
+
+    def test_new_relative_path_with_spaces_is_rewritten(self, paths: Paths):
+        workspace = paths.sandbox_work_dir("t1", user_id="u1")
+        src = _workspace_file(paths, "final reports/result.txt")
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                "Saved as final reports/result.txt.",
+                thread_id="t1",
+                user_id="u1",
+                source_base_dir=workspace,
+                changed_files=[src],
+            )
+
+        assert result == f"Saved as {VIRTUAL_PATH_PREFIX}/workspace/final reports/result.txt."
+
+    def test_new_dot_relative_path_with_spaces_is_rewritten(self, paths: Paths):
+        workspace = paths.sandbox_work_dir("t1", user_id="u1")
+        src = _workspace_file(paths, "final reports/result.txt")
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                "Saved as ./final reports/result.txt.",
+                thread_id="t1",
+                user_id="u1",
+                source_base_dir=workspace,
+                changed_files=[src],
+            )
+
+        assert result == f"Saved as {VIRTUAL_PATH_PREFIX}/workspace/final reports/result.txt."
+
+    def test_changed_path_with_spaces_outside_user_data_is_untouched(self, tmp_path: Path, paths: Paths):
+        outside = tmp_path / "outside report.txt"
+        outside.write_text("outside", encoding="utf-8")
+        text = f"Saved as {outside}"
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                changed_files=[outside],
+            )
+
+        assert result == text
+
+    @pytest.mark.skipif(os.name == "nt", reason="exercises POSIX file URI spellings")
+    def test_new_file_uri_with_literal_spaces_is_rewritten(self, paths: Paths):
+        src = _workspace_file(paths, "final report.txt")
+        text = f"Saved as file://{src}"
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                changed_files=[src],
+            )
+
+        assert result == f"Saved as {VIRTUAL_PATH_PREFIX}/workspace/final report.txt"
+
+    def test_changed_path_with_spaces_does_not_rewrite_longer_name(self, paths: Paths):
+        src = _workspace_file(paths, "final report.txt")
+        text = f"Backup: {src}.bak"
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                changed_files=[src],
+            )
+
+        assert result == text
+
+    def test_changed_path_with_spaces_does_not_rewrite_whitespace_suffix(self, paths: Paths):
+        src = _workspace_file(paths, "final report.txt")
+        longer = _workspace_file(paths, "final report.txt copy")
+        text = f"Backup: {longer}"
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                changed_files=[src],
+            )
+
+        assert result == text
+
+    def test_changed_path_with_spaces_followed_by_prose_is_untouched(self, paths: Paths):
+        src = _workspace_file(paths, "final report.txt")
+        text = f"Saved as {src} and ready."
+
+        with _patch_paths(paths):
+            result = mcp_tools._rewrite_local_paths_in_text(
+                text,
+                thread_id="t1",
+                user_id="u1",
+                changed_files=[src],
+            )
+
+        assert result == text
+
     def test_tmpdir_output_under_workspace_is_rewritten(self, paths: Paths):
         src = _workspace_file(paths, ".mcp/tmp/page.png")
         text = f"Saved to {src}"
@@ -629,6 +748,22 @@ class TestConvertCallToolResultRewrites:
 
         assert content[0]["text"] == f"Saved as {VIRTUAL_PATH_PREFIX}/workspace/page-2026.yml"
 
+    def test_text_path_with_spaces_rewritten_from_changed_files(self, paths: Paths):
+        workspace = paths.sandbox_work_dir("t1", user_id="u1")
+        src = _workspace_file(paths, "report with spaces.txt")
+        result = CallToolResult(content=[TextContent(type="text", text=f"Saved as {src}")], isError=False)
+
+        with _patch_paths(paths):
+            content, _ = mcp_tools._convert_call_tool_result(
+                result,
+                thread_id="t1",
+                user_id="u1",
+                source_base_dir=workspace,
+                changed_files=[src],
+            )
+
+        assert content[0]["text"] == f"Saved as {VIRTUAL_PATH_PREFIX}/workspace/report with spaces.txt"
+
     def test_no_context_does_not_rewrite(self, paths: Paths):
         src = _workspace_file(paths, "x.png", content=b"png")
         uri = src.as_uri()