perf(watcher): optimize wildcard matching

.jules/bolt.md

@@ -141,3 +141,11 @@ Inside the `_is_ignored_impl` hot path, `os.path.relpath` is computationally exp
 Action:
 In `watchdog` event path normalization, bypass the computationally expensive `os.path.relpath` for the common case where `base_path` is `.` and the path is already relative by adding a fast-path condition: `elif self.base_path == "." and not os.path.isabs(path) and not path.startswith(".."): pass`.
 To optimize ignore pattern matching in hot loops, pre-compute a flag during initialization (e.g., `self._has_compound_ignores = any('/' in p for p in self.ignore_patterns)`) and use it to short-circuit the evaluation of compound directory paths if no slash-based ignore patterns exist.
+
+## 2026-05-01 — Wildcard Regex Split Optimization
+
+Learning:
+Inside the file watcher's `_is_ignored_impl` hot path, applying a combined wildcard regex that includes both simple patterns (e.g. `*.tmp`) and compound patterns (e.g. `src/*.tmp`) to individual path segments (`parts`) and cumulative directory prefixes (`prefix`) is redundant and computationally wasteful. A simple wildcard pattern incorrectly evaluated against a cumulative prefix path loop wastes time, and a compound wildcard will never match a simple directory segment.
+
+Action:
+Split wildcard patterns into `simple_wildcards` (no slashes) and `compound_wildcards` (contains slashes), and compile them into separate regular expressions (`simple_wildcard_regex` and `compound_wildcard_regex`). Only apply the simple regex when iterating over individual parts, and apply the compound regex when accumulating the directory prefix. This optimization prevents unnecessary regex checks in the hot path.

src/echo/watcher.py

@@ -32,11 +32,18 @@ def __init__(self, command: str, base_path: str = ".", ignore_patterns: list[str
         # Pre-compute exact vs wildcard patterns for faster matching
         self.exact_ignores = {p for p in self.ignore_patterns if not any(c in p for c in ('*', '?', '['))}
         wildcard_ignores = [p for p in self.ignore_patterns if any(c in p for c in ('*', '?', '['))]
-        self.wildcard_regex = None
+
+        simple_wildcards = [p for p in wildcard_ignores if '/' not in p]
+        compound_wildcards = [p for p in wildcard_ignores if '/' in p]
+
+        self.simple_wildcard_regex = None
+        self.compound_wildcard_regex = None
         self._has_compound_ignores = any('/' in p for p in self.ignore_patterns)
-        if wildcard_ignores:
-            regex_str = "|".join(f"(?:{fnmatch.translate(p)})" for p in wildcard_ignores)
-            self.wildcard_regex = re.compile(regex_str)
+
+        if simple_wildcards:
+            self.simple_wildcard_regex = re.compile("|".join(f"(?:{fnmatch.translate(p)})" for p in simple_wildcards))
+        if compound_wildcards:
+            self.compound_wildcard_regex = re.compile("|".join(f"(?:{fnmatch.translate(p)})" for p in compound_wildcards))
 
         self.current_process = None
         self.process_lock = threading.Lock()
@@ -188,9 +195,9 @@ def _is_ignored_impl(self, path: str) -> bool:
         if not self.exact_ignores.isdisjoint(parts):
             return True
 
-        if self.wildcard_regex:
+        if self.simple_wildcard_regex:
             for part in parts:
-                if self.wildcard_regex.match(part):
+                if self.simple_wildcard_regex.match(part):
                     return True
 
         # Check for exact and wildcard ignore patterns matching cumulative prefix directories
@@ -203,7 +210,7 @@ def _is_ignored_impl(self, path: str) -> bool:
                 prefix = f"{prefix}/{part}"
                 if prefix in self.exact_ignores:
                     return True
-                if self.wildcard_regex and self.wildcard_regex.match(prefix):
+                if self.compound_wildcard_regex and self.compound_wildcard_regex.match(prefix):
                     return True
 
         return False

tests/test_ignore.py

@@ -126,7 +126,7 @@ def test_character_class_wildcard_match():
     handler = CommandRunnerHandler("echo 1", ignore_patterns=["[a-z].tmp"])
 
     # Must correctly categorize as wildcard and compile regex
-    assert handler.wildcard_regex is not None
+    assert handler.simple_wildcard_regex is not None
     assert "[a-z].tmp" not in handler.exact_ignores
 
     assert handler._is_ignored("a.tmp") is True