]> git.madduck.net Git - etc/vim.git/blobdiff - black.py

madduck's git repository

Every one of the projects in this repository is available at the canonical URL git://git.madduck.net/madduck/pub/<projectpath> — see each project's metadata for the exact URL.

All patches and comments are welcome. Please squash your changes to logical commits before using git-format-patch and git-send-email to patches@git.madduck.net. If you'd read over the Git project's submission guidelines and adhered to them, I'd be especially grateful.

SSH access, as well as push access can be individually arranged.

If you use my repositories frequently, consider adding the following snippet to ~/.gitconfig and using the third clone URL listed for each project:

[url "git://git.madduck.net/madduck/"]
  insteadOf = madduck:

Simplify single-file vs. multi-file modes
[etc/vim.git] / black.py
index 0323f67c257c1cca2a8ac3655171f3b7d74ddecd..9f230e794c57f8e45781fa0566495f91a897d812 100644 (file)
--- a/black.py
+++ b/black.py
@@ -1,6 +1,7 @@
 #!/usr/bin/env python3
 
 import asyncio
+import pickle
 from asyncio.base_events import BaseEventLoop
 from concurrent.futures import Executor, ProcessPoolExecutor
 from enum import Enum
@@ -32,6 +33,7 @@ from typing import (
     Union,
 )
 
+from appdirs import user_cache_dir
 from attr import dataclass, Factory
 import click
 
@@ -54,6 +56,10 @@ Priority = int
 Index = int
 LN = Union[Leaf, Node]
 SplitFunc = Callable[["Line", bool], Iterator["Line"]]
+Timestamp = float
+FileSize = int
+CacheInfo = Tuple[Timestamp, FileSize]
+Cache = Dict[Path, CacheInfo]
 out = partial(click.secho, bold=True, err=True)
 err = partial(click.secho, fg="red", err=True)
 
@@ -104,6 +110,12 @@ class WriteBack(Enum):
     DIFF = 2
 
 
+class Changed(Enum):
+    NO = 0
+    CACHED = 1
+    YES = 2
+
+
 @click.command()
 @click.option(
     "-l",
@@ -185,22 +197,10 @@ def main(
         write_back = WriteBack.YES
     if len(sources) == 0:
         ctx.exit(0)
+        return
+
     elif len(sources) == 1:
-        p = sources[0]
-        report = Report(check=check, quiet=quiet)
-        try:
-            if not p.is_file() and str(p) == "-":
-                changed = format_stdin_to_stdout(
-                    line_length=line_length, fast=fast, write_back=write_back
-                )
-            else:
-                changed = format_file_in_place(
-                    p, line_length=line_length, fast=fast, write_back=write_back
-                )
-            report.done(p, changed)
-        except Exception as exc:
-            report.failed(p, str(exc))
-        ctx.exit(report.return_code)
+        return_code = reformat_one(sources[0], line_length, fast, quiet, write_back)
     else:
         loop = asyncio.get_event_loop()
         executor = ProcessPoolExecutor(max_workers=os.cpu_count())
@@ -213,7 +213,45 @@ def main(
             )
         finally:
             shutdown(loop)
-            ctx.exit(return_code)
+    ctx.exit(return_code)
+
+
+def reformat_one(
+    src: Path, line_length: int, fast: bool, quiet: bool, write_back: WriteBack
+) -> int:
+    """Reformat a single file under `src` without spawning child processes.
+
+    If `quiet` is True, non-error messages are not output. `line_length`,
+    `write_back`, and `fast` options are passed to :func:`format_file_in_place`.
+    """
+    report = Report(check=write_back is WriteBack.NO, quiet=quiet)
+    try:
+        changed = Changed.NO
+        if not src.is_file() and str(src) == "-":
+            if format_stdin_to_stdout(
+                line_length=line_length, fast=fast, write_back=write_back
+            ):
+                changed = Changed.YES
+        else:
+            cache: Cache = {}
+            if write_back != WriteBack.DIFF:
+                cache = read_cache()
+                src = src.resolve()
+                if src in cache and cache[src] == get_cache_info(src):
+                    changed = Changed.CACHED
+            if (
+                changed is not Changed.CACHED
+                and format_file_in_place(
+                    src, line_length=line_length, fast=fast, write_back=write_back
+                )
+            ):
+                changed = Changed.YES
+            if write_back != WriteBack.DIFF and changed is not Changed.NO:
+                write_cache(cache, [src])
+        report.done(src, changed)
+    except Exception as exc:
+        report.failed(src, str(exc))
+    return report.return_code
 
 
 async def schedule_formatting(
@@ -232,41 +270,55 @@ async def schedule_formatting(
     `line_length`, `write_back`, and `fast` options are passed to
     :func:`format_file_in_place`.
     """
-    lock = None
-    if write_back == WriteBack.DIFF:
-        # For diff output, we need locks to ensure we don't interleave output
-        # from different processes.
-        manager = Manager()
-        lock = manager.Lock()
-    tasks = {
-        src: loop.run_in_executor(
-            executor, format_file_in_place, src, line_length, fast, write_back, lock
-        )
-        for src in sources
-    }
-    _task_values = list(tasks.values())
-    loop.add_signal_handler(signal.SIGINT, cancel, _task_values)
-    loop.add_signal_handler(signal.SIGTERM, cancel, _task_values)
-    await asyncio.wait(tasks.values())
-    cancelled = []
     report = Report(check=write_back is WriteBack.NO, quiet=quiet)
-    for src, task in tasks.items():
-        if not task.done():
-            report.failed(src, "timed out, cancelling")
-            task.cancel()
-            cancelled.append(task)
-        elif task.cancelled():
-            cancelled.append(task)
-        elif task.exception():
-            report.failed(src, str(task.exception()))
-        else:
-            report.done(src, task.result())
+    cache: Cache = {}
+    if write_back != WriteBack.DIFF:
+        cache = read_cache()
+        sources, cached = filter_cached(cache, sources)
+        for src in cached:
+            report.done(src, Changed.CACHED)
+    cancelled = []
+    formatted = []
+    if sources:
+        lock = None
+        if write_back == WriteBack.DIFF:
+            # For diff output, we need locks to ensure we don't interleave output
+            # from different processes.
+            manager = Manager()
+            lock = manager.Lock()
+        tasks = {
+            src: loop.run_in_executor(
+                executor, format_file_in_place, src, line_length, fast, write_back, lock
+            )
+            for src in sources
+        }
+        _task_values = list(tasks.values())
+        loop.add_signal_handler(signal.SIGINT, cancel, _task_values)
+        loop.add_signal_handler(signal.SIGTERM, cancel, _task_values)
+        await asyncio.wait(_task_values)
+        for src, task in tasks.items():
+            if not task.done():
+                report.failed(src, "timed out, cancelling")
+                task.cancel()
+                cancelled.append(task)
+            elif task.cancelled():
+                cancelled.append(task)
+            elif task.exception():
+                report.failed(src, str(task.exception()))
+            else:
+                formatted.append(src)
+                report.done(src, Changed.YES if task.result() else Changed.NO)
+
     if cancelled:
         await asyncio.gather(*cancelled, loop=loop, return_exceptions=True)
     elif not quiet:
         out("All done! ✨ 🍰 ✨")
     if not quiet:
         click.echo(str(report))
+
+    if write_back != WriteBack.DIFF and formatted:
+        write_cache(cache, formatted)
+
     return report.return_code
 
 
@@ -282,6 +334,7 @@ def format_file_in_place(
     If `write_back` is True, write reformatted code back to stdout.
     `line_length` and `fast` options are passed to :func:`format_file_contents`.
     """
+
     with tokenize.open(src) as src_buffer:
         src_contents = src_buffer.read()
     try:
@@ -2201,16 +2254,20 @@ class Report:
     same_count: int = 0
     failure_count: int = 0
 
-    def done(self, src: Path, changed: bool) -> None:
+    def done(self, src: Path, changed: Changed) -> None:
         """Increment the counter for successful reformatting. Write out a message."""
-        if changed:
+        if changed is Changed.YES:
             reformatted = "would reformat" if self.check else "reformatted"
             if not self.quiet:
                 out(f"{reformatted} {src}")
             self.change_count += 1
         else:
             if not self.quiet:
-                out(f"{src} already well formatted, good job.", bold=False)
+                if changed is Changed.NO:
+                    msg = f"{src} already well formatted, good job."
+                else:
+                    msg = f"{src} wasn't modified on disk since last run."
+                out(msg, bold=False)
             self.same_count += 1
 
     def failed(self, src: Path, message: str) -> None:
@@ -2409,5 +2466,62 @@ def sub_twice(regex: Pattern[str], replacement: str, original: str) -> str:
     return regex.sub(replacement, regex.sub(replacement, original))
 
 
+CACHE_DIR = Path(user_cache_dir("black", version=__version__))
+CACHE_FILE = CACHE_DIR / "cache.pickle"
+
+
+def read_cache() -> Cache:
+    """Read the cache if it exists and is well formed.
+
+    If it is not well formed, the call to write_cache later should resolve the issue.
+    """
+    if not CACHE_FILE.exists():
+        return {}
+
+    with CACHE_FILE.open("rb") as fobj:
+        try:
+            cache: Cache = pickle.load(fobj)
+        except pickle.UnpicklingError:
+            return {}
+
+    return cache
+
+
+def get_cache_info(path: Path) -> CacheInfo:
+    """Return the information used to check if a file is already formatted or not."""
+    stat = path.stat()
+    return stat.st_mtime, stat.st_size
+
+
+def filter_cached(
+    cache: Cache, sources: Iterable[Path]
+) -> Tuple[List[Path], List[Path]]:
+    """Split a list of paths into two.
+
+    The first list contains paths of files that modified on disk or are not in the
+    cache. The other list contains paths to non-modified files.
+    """
+    todo, done = [], []
+    for src in sources:
+        src = src.resolve()
+        if cache.get(src) != get_cache_info(src):
+            todo.append(src)
+        else:
+            done.append(src)
+    return todo, done
+
+
+def write_cache(cache: Cache, sources: List[Path]) -> None:
+    """Update the cache file."""
+    try:
+        if not CACHE_DIR.exists():
+            CACHE_DIR.mkdir(parents=True)
+        new_cache = {**cache, **{src.resolve(): get_cache_info(src) for src in sources}}
+        with CACHE_FILE.open("wb") as fobj:
+            pickle.dump(new_cache, fobj, protocol=pickle.HIGHEST_PROTOCOL)
+    except OSError:
+        pass
+
+
 if __name__ == "__main__":
     main()