Update in response to feedback

- one configuration option: ignore_paths
- use shlex to parse option to support whitespace and in-line comments
- change option type to Tuple[str, ...]
- remove ignored directories from dirnames to prevent scanning

kittens/diff/collect.py

@@ -7,6 +7,7 @@ from contextlib import suppress
 from fnmatch import fnmatch
 from functools import lru_cache
 from hashlib import md5
+from itertools import product
 from kitty.guess_mime_type import guess_type
 from typing import TYPE_CHECKING, Dict, List, Set, Optional, Iterator, Tuple, Union
 
@@ -38,8 +39,7 @@ class Segment:
 
 class Collection:
 
-    file_ignores = ['']
-    ignore_paths = ['']
+    ignore_paths: Tuple[str, ...] = ()
 
     def __init__(self) -> None:
         self.changes: Dict[str, str] = {}
@@ -109,26 +109,29 @@ def resolve_remote_name(path: str, default: str) -> str:
     return default
 
 
+def walk(base: str, names: Set[str], pmap: Dict[str, str], ignore_paths: Tuple[str, ...]) -> None:
+    for dirpath, dirnames, filenames in os.walk(base):
+        ignored = [_dir for _dir, pat in product(dirnames, ignore_paths) if fnmatch(_dir, pat)]
+        for _dir in ignored:
+            dirnames.pop(dirnames.index(_dir))
+        for filename in filenames:
+            if any(fnmatch(filename, pat) for pat in ignore_paths if pat):
+                continue
+            path = os.path.abspath(os.path.join(dirpath, filename))
+            path_name_map[path] = name = os.path.relpath(path, base)
+            names.add(name)
+            pmap[name] = path
+
+
 def collect_files(collection: Collection, left: str, right: str) -> None:
     left_names: Set[str] = set()
     right_names: Set[str] = set()
     left_path_map: Dict[str, str] = {}
     right_path_map: Dict[str, str] = {}
 
-    def walk(base: str, names: Set[str], pmap: Dict[str, str]) -> None:
-        for dirpath, dirnames, filenames in os.walk(base):
-            if any(fnmatch(dirpath, f"*/{pat}") for pat in collection.ignore_paths if pat):
-                continue
-            for filename in filenames:
-                if any(fnmatch(filename, f"{pat}") for pat in collection.file_ignores if pat):
-                    continue
-                path = os.path.abspath(os.path.join(dirpath, filename))
-                path_name_map[path] = name = os.path.relpath(path, base)
-                names.add(name)
-                pmap[name] = path
+    walk(left, left_names, left_path_map, collection.ignore_paths)
+    walk(right, right_names, right_path_map, collection.ignore_paths)
 
-    walk(left, left_names, left_path_map)
-    walk(right, right_names, right_path_map)
     common_names = left_names & right_names
     changed_names = {n for n in common_names if data_for_path(left_path_map[n]) != data_for_path(right_path_map[n])}
     for n in changed_names:

kittens/diff/main.py

@@ -655,7 +655,6 @@ def main(args: List[str]) -> None:
     opts = init_config(cli_opts)
     set_diff_command(opts.diff_cmd)
     lines_for_path.replace_tab_by = opts.replace_tab_by
-    Collection.file_ignores = opts.file_ignores
     Collection.ignore_paths = opts.ignore_paths
     left, right = map(get_remote_file, (left, right))
     if os.path.isdir(left) != os.path.isdir(right):

kittens/diff/options/definition.py

@@ -48,14 +48,11 @@ opt('replace_tab_by', '\\x20\\x20\\x20\\x20',
     long_text='The string to replace tabs with. Default is to use four spaces.'
     )
 
-opt('file_ignores', '',
-    option_type='pattern_list',
-    long_text='List of file patterns that are ignored when directories are compared'
-    )
-
 opt('ignore_paths', '',
     option_type='pattern_list',
-    long_text='List of directory patterns that are ignored when directories are compared'
+    long_text='''List of patterns that are matched against directory and file
+names and ignored when directories are compared.
+''',
     )
 
 egr()  # }}}

kittens/diff/options/parse.py

@@ -86,10 +86,7 @@ class Parser:
         for k in parse_map(val):
             ans['map'].append(k)
 
-    def file_ignores(self, val: str, ans: typing.Dict[str, typing.List[str]]):
-        ans['file_ignores'] = pattern_list(val)
-
-    def ignore_paths(self, val: str, ans: typing.Dict[str, typing.List[str]]):
+    def ignore_paths(self, val: str, ans: typing.Dict[str, typing.Any]):
         ans['ignore_paths'] = pattern_list(val)
 
 

kittens/diff/options/types.py

@@ -14,7 +14,6 @@ option_names = (  # {{{
  'added_margin_bg',
  'background',
  'diff_cmd',
- 'file_ignores',
  'filler_bg',
  'foreground',
  'highlight_added_bg',
@@ -45,14 +44,13 @@ class Options:
     added_margin_bg: Color = Color(205, 255, 216)
     background: Color = Color(255, 255, 255)
     diff_cmd: str = 'auto'
-    file_ignores: typing.List[str] = ['']
     filler_bg: Color = Color(250, 251, 252)
     foreground: Color = Color(0, 0, 0)
     highlight_added_bg: Color = Color(172, 242, 189)
     highlight_removed_bg: Color = Color(253, 184, 192)
     hunk_bg: Color = Color(241, 248, 255)
     hunk_margin_bg: Color = Color(219, 237, 255)
-    ignore_paths: typing.List[str] = ['']
+    ignore_paths: typing.Tuple[str, ...] = ()
     margin_bg: Color = Color(250, 251, 252)
     margin_fg: Color = Color(170, 170, 170)
     margin_filler_bg: typing.Optional[kitty.fast_data_types.Color] = None

kittens/diff/options/utils.py

@@ -3,6 +3,7 @@
 # License: GPLv3 Copyright: 2021, Kovid Goyal <kovid at kovidgoyal.net>
 
 
+import shlex
 from typing import Any, Dict, Iterable, List, Tuple, Union
 
 from kitty.conf.utils import (
@@ -58,8 +59,8 @@ def syntax_aliases(raw: str) -> Dict[str, str]:
     return ans
 
 
-def pattern_list(raw: str) -> List[str]:
-    return raw.split(' ')
+def pattern_list(raw: str) -> Tuple[str, ...]:
+    return tuple(shlex.split(raw, comments=True))
 
 
 def parse_map(val: str) -> Iterable[KittensKeyDefinition]: