[update_lib] todo shows last updated date (RustPython#7053)

Author: youknowone

scripts/update_lib/cmd_todo.py

@@ -15,6 +15,9 @@
 
 from update_lib.deps import (
     count_test_todos,
+    get_module_diff_stat,
+    get_module_last_updated,
+    get_test_last_updated,
     is_test_tracked,
     is_test_up_to_date,
 )
@@ -368,6 +371,18 @@ def compute_test_todo_list(
     return result
 
 
+def _format_meta_suffix(item: dict) -> str:
+    """Format metadata suffix (last updated date and diff count)."""
+    parts = []
+    last_updated = item.get("last_updated")
+    diff_lines = item.get("diff_lines", 0)
+    if last_updated:
+        parts.append(last_updated)
+    if diff_lines > 0:
+        parts.append(f"Δ{diff_lines}")
+    return f" | {' '.join(parts)}" if parts else ""
+
+
 def _format_test_suffix(item: dict) -> str:
     """Format suffix for test item (TODO count or untracked)."""
     tracked = item.get("tracked", True)
@@ -410,13 +425,15 @@ def format_test_todo_list(
         primary = tests[0]
         done_mark = "[x]" if primary["up_to_date"] else "[ ]"
         suffix = _format_test_suffix(primary)
-        lines.append(f"- {done_mark} {primary['name']}{suffix}")
+        meta = _format_meta_suffix(primary)
+        lines.append(f"- {done_mark} {primary['name']}{suffix}{meta}")
 
         # Rest are indented
         for item in tests[1:]:
             done_mark = "[x]" if item["up_to_date"] else "[ ]"
             suffix = _format_test_suffix(item)
-            lines.append(f"  - {done_mark} {item['name']}{suffix}")
+            meta = _format_meta_suffix(item)
+            lines.append(f"  - {done_mark} {item['name']}{suffix}{meta}")
 
     return lines
 
@@ -462,7 +479,8 @@ def format_todo_list(
         if rev_str:
             parts.append(f"({rev_str})")
 
-        lines.append(" ".join(parts))
+        line = " ".join(parts) + _format_meta_suffix(item)
+        lines.append(line)
 
         # Show hard_deps:
         # - Normal mode: only show if lib is up-to-date but hard_deps are not
@@ -482,7 +500,8 @@ def format_todo_list(
             for test_info in test_by_lib[name]:
                 test_done_mark = "[x]" if test_info["up_to_date"] else "[ ]"
                 suffix = _format_test_suffix(test_info)
-                lines.append(f"  - {test_done_mark} {test_info['name']}{suffix}")
+                meta = _format_meta_suffix(test_info)
+                lines.append(f"  - {test_done_mark} {test_info['name']}{suffix}{meta}")
 
         # Verbose mode: show detailed dependency info
         if verbose:
@@ -556,6 +575,29 @@ def format_all_todo(
         if include_done or lib_not_done or has_pending_test:
             lib_todo.append(item)
 
+    # Add metadata (last updated date and diff stat) to lib items
+    for item in lib_todo:
+        item["last_updated"] = get_module_last_updated(
+            item["name"], cpython_prefix, lib_prefix
+        )
+        item["diff_lines"] = (
+            0
+            if item["up_to_date"]
+            else get_module_diff_stat(item["name"], cpython_prefix, lib_prefix)
+        )
+
+    # Add last_updated to displayed test items (verbose only - slow)
+    if verbose:
+        for tests in test_by_lib.values():
+            for test in tests:
+                test["last_updated"] = get_test_last_updated(
+                    test["name"], cpython_prefix, lib_prefix
+                )
+        for test in no_lib_tests:
+            test["last_updated"] = get_test_last_updated(
+                test["name"], cpython_prefix, lib_prefix
+            )
+
     # Format lib todo with embedded tests
     lines.extend(format_todo_list(lib_todo, test_by_lib, limit, verbose))
 

scripts/update_lib/deps.py

@@ -8,6 +8,7 @@
 """
 
 import ast
+import difflib
 import functools
 import pathlib
 import re
@@ -1011,6 +1012,119 @@ def is_up_to_date(name: str, cpython_prefix: str, lib_prefix: str) -> bool:
     return found_any
 
 
+def _count_file_diff(file_a: pathlib.Path, file_b: pathlib.Path) -> int:
+    """Count changed lines between two text files using difflib."""
+    a_content = safe_read_text(file_a)
+    b_content = safe_read_text(file_b)
+    if a_content is None or b_content is None:
+        return 0
+    if a_content == b_content:
+        return 0
+    a_lines = a_content.splitlines()
+    b_lines = b_content.splitlines()
+    count = 0
+    for line in difflib.unified_diff(a_lines, b_lines, lineterm=""):
+        if (line.startswith("+") and not line.startswith("+++")) or (
+            line.startswith("-") and not line.startswith("---")
+        ):
+            count += 1
+    return count
+
+
+def _count_path_diff(path_a: pathlib.Path, path_b: pathlib.Path) -> int:
+    """Count changed lines between two paths (file or directory, *.py only)."""
+    if path_a.is_file() and path_b.is_file():
+        return _count_file_diff(path_a, path_b)
+    if path_a.is_dir() and path_b.is_dir():
+        total = 0
+        a_files = {f.relative_to(path_a) for f in path_a.rglob("*.py")}
+        b_files = {f.relative_to(path_b) for f in path_b.rglob("*.py")}
+        for rel in a_files & b_files:
+            total += _count_file_diff(path_a / rel, path_b / rel)
+        for rel in a_files - b_files:
+            content = safe_read_text(path_a / rel)
+            if content:
+                total += len(content.splitlines())
+        for rel in b_files - a_files:
+            content = safe_read_text(path_b / rel)
+            if content:
+                total += len(content.splitlines())
+        return total
+    return 0
+
+
+def get_module_last_updated(
+    name: str, cpython_prefix: str, lib_prefix: str
+) -> str | None:
+    """Get the last git commit date for a module's Lib files."""
+    local_paths = []
+    for cpython_path in get_lib_paths(name, cpython_prefix):
+        if not cpython_path.exists():
+            continue
+        try:
+            rel_path = cpython_path.relative_to(cpython_prefix)
+            local_path = pathlib.Path(lib_prefix) / rel_path.relative_to("Lib")
+            if local_path.exists():
+                local_paths.append(str(local_path))
+        except ValueError:
+            continue
+    if not local_paths:
+        return None
+    try:
+        result = subprocess.run(
+            ["git", "log", "-1", "--format=%cd", "--date=short", "--"] + local_paths,
+            capture_output=True,
+            text=True,
+            timeout=10,
+        )
+        if result.returncode == 0 and result.stdout.strip():
+            return result.stdout.strip()
+    except Exception:
+        pass
+    return None
+
+
+def get_module_diff_stat(name: str, cpython_prefix: str, lib_prefix: str) -> int:
+    """Count differing lines between cpython and local Lib for a module."""
+    total = 0
+    for cpython_path in get_lib_paths(name, cpython_prefix):
+        if not cpython_path.exists():
+            continue
+        try:
+            rel_path = cpython_path.relative_to(cpython_prefix)
+            local_path = pathlib.Path(lib_prefix) / rel_path.relative_to("Lib")
+        except ValueError:
+            continue
+        if not local_path.exists():
+            continue
+        total += _count_path_diff(cpython_path, local_path)
+    return total
+
+
+def get_test_last_updated(
+    test_name: str, cpython_prefix: str, lib_prefix: str
+) -> str | None:
+    """Get the last git commit date for a test's files."""
+    cpython_path = _get_cpython_test_path(test_name, cpython_prefix)
+    if cpython_path is None:
+        return None
+    local_path = _get_local_test_path(cpython_path, lib_prefix)
+    if not local_path.exists():
+        return None
+    try:
+        result = subprocess.run(
+            ["git", "log", "-1", "--format=%cd", "--date=short", "--", str(local_path)],
+            capture_output=True,
+            text=True,
+            timeout=10,
+        )
+        if result.returncode == 0 and result.stdout.strip():
+            return result.stdout.strip()
+    except Exception:
+        pass
+    return None
+
+
 def get_test_dependencies(
     test_path: pathlib.Path,
 ) -> dict[str, list[pathlib.Path]]: