File size: 11,757 Bytes
21ff762 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 | #!/usr/bin/env python3
"""
memory_anchor.py -- Diff-aware memory anchoring for the auto-memory store.
Auto-memory notes live under:
~/.claude/projects/<slug>/memory/*.md
They accumulate backtick-wrapped file references that rot as the codebase
moves. This module scans each memory file, pulls out references that look
like file paths, and verifies whether each resolves under a repository
root. It then renders a dashboard and, under ``check --strict``, exits
non-zero when any reference is dead.
Reference shapes we recognize (inside backtick code spans only, to avoid
false positives on prose):
`scan_repo.py` — bare file with known extension
`src/foo.py` — relative path with slash
`src/foo.py:42` — with optional line suffix
`~/.claude/skills/foo/SKILL.md` — tilde-expanded absolute path
No prose is parsed. If a path isn't in backticks, it isn't a reference.
Usage:
python src/memory_anchor.py scan
python src/memory_anchor.py dashboard
python src/memory_anchor.py check --strict # exit 2 if dead refs
"""
from __future__ import annotations
import argparse
import json
import os
import re
import sys
import time
from dataclasses import asdict, dataclass
from pathlib import Path
from typing import Sequence
# ── Paths & config defaults ────────────────────────────────────────────────
DEFAULT_MEMORY_ROOT = Path(os.path.expanduser("~/.claude/projects"))
KNOWN_EXTENSIONS = frozenset({
".py", ".md", ".json", ".yaml", ".yml", ".toml",
".ts", ".tsx", ".js", ".jsx", ".mjs", ".cjs",
".rs", ".go", ".java", ".c", ".h", ".cpp", ".hpp",
".sh", ".bash", ".zsh", ".fish",
".html", ".css", ".scss", ".sass",
".sql", ".txt", ".rst", ".ini", ".cfg",
})
# Inline code spans: backtick-delimited, not crossing newlines.
_CODE_SPAN = re.compile(r"`([^`\n]+)`")
# Trailing ``:<line>`` to strip (keep drive letters like ``C:`` intact by
# requiring purely digit tail).
_LINE_TAIL = re.compile(r":(\d+)$")
# ── Data model ─────────────────────────────────────────────────────────────
@dataclass(frozen=True)
class AnchorRef:
raw: str # exactly what the memory note wrote between backticks
path: str # normalised path sans line suffix
line: int | None # optional line number if suffix was present
exists: bool
def to_dict(self) -> dict:
return asdict(self)
@dataclass(frozen=True)
class MemoryAnchorFile:
memory_path: str
refs: tuple[AnchorRef, ...]
@property
def live(self) -> tuple[AnchorRef, ...]:
return tuple(r for r in self.refs if r.exists)
@property
def dead(self) -> tuple[AnchorRef, ...]:
return tuple(r for r in self.refs if not r.exists)
def to_dict(self) -> dict:
return {
"memory_path": self.memory_path,
"refs": [r.to_dict() for r in self.refs],
"live": [r.to_dict() for r in self.live],
"dead": [r.to_dict() for r in self.dead],
}
@dataclass(frozen=True)
class AnchorReport:
generated_at: float
repo_root: str
memory_root: str
files: tuple[MemoryAnchorFile, ...]
@property
def all_refs(self) -> tuple[AnchorRef, ...]:
out: list[AnchorRef] = []
for f in self.files:
out.extend(f.refs)
return tuple(out)
@property
def dead_count(self) -> int:
return sum(1 for r in self.all_refs if not r.exists)
@property
def live_count(self) -> int:
return sum(1 for r in self.all_refs if r.exists)
@property
def has_dead(self) -> bool:
return self.dead_count > 0
def to_dict(self) -> dict:
return {
"generated_at": self.generated_at,
"repo_root": self.repo_root,
"memory_root": self.memory_root,
"files": [f.to_dict() for f in self.files],
"totals": {
"total": len(self.all_refs),
"live": self.live_count,
"dead": self.dead_count,
},
}
# ── Reference extraction ───────────────────────────────────────────────────
def _looks_like_path(token: str) -> bool:
"""
Heuristic: token looks like a file path if it has a known extension OR
contains a slash with a dotted last segment.
Rejects obvious non-paths: shell-style flags, bare function names like
``add_skill()``, URLs, commit hashes.
"""
t = token.strip()
if not t or any(ch.isspace() for ch in t):
return False
if t.startswith(("-", "http://", "https://")):
return False
if t.endswith("()"):
return False
stripped = _LINE_TAIL.sub("", t)
lower = stripped.lower()
for ext in KNOWN_EXTENSIONS:
if lower.endswith(ext):
return True
if "/" in stripped:
tail = stripped.rsplit("/", 1)[-1]
if "." in tail and not tail.startswith("."):
return True
return False
def _strip_line_suffix(token: str) -> tuple[str, int | None]:
m = _LINE_TAIL.search(token)
if not m:
return token, None
return token[: m.start()], int(m.group(1))
def extract_refs(text: str) -> list[tuple[str, str, int | None]]:
"""
Return a list of (raw, path, line) triples for every backtick span that
looks like a file path. Preserves document order; deduplicates within
one document (same raw token scanned once).
"""
seen: set[str] = set()
out: list[tuple[str, str, int | None]] = []
for match in _CODE_SPAN.finditer(text):
raw = match.group(1).strip()
if raw in seen:
continue
if not _looks_like_path(raw):
continue
seen.add(raw)
path, line = _strip_line_suffix(raw)
out.append((raw, path, line))
return out
# ── Resolution ─────────────────────────────────────────────────────────────
def _resolve(path_str: str, repo_root: Path) -> bool:
"""
A reference resolves if:
- absolute path (or tilde-expanded) exists, OR
- joining against repo_root exists, OR
- joining against repo_root/src exists (common Python convention).
"""
expanded = os.path.expanduser(path_str)
p = Path(expanded)
if p.is_absolute():
return p.exists()
candidates = [
repo_root / expanded,
repo_root / "src" / expanded,
]
return any(c.exists() for c in candidates)
def scan_memory_file(memory_path: Path,
repo_root: Path) -> MemoryAnchorFile:
try:
text = memory_path.read_text(encoding="utf-8", errors="replace")
except OSError:
return MemoryAnchorFile(memory_path=str(memory_path), refs=())
refs: list[AnchorRef] = []
for raw, path, line in extract_refs(text):
refs.append(AnchorRef(
raw=raw,
path=path,
line=line,
exists=_resolve(path, repo_root),
))
return MemoryAnchorFile(memory_path=str(memory_path), refs=tuple(refs))
def _iter_memory_files(memory_root: Path) -> list[Path]:
if not memory_root.exists():
return []
return sorted(memory_root.rglob("*.md"))
# ── Report ─────────────────────────────────────────────────────────────────
def build_report(repo_root: Path,
memory_root: Path = DEFAULT_MEMORY_ROOT,
now: float | None = None) -> AnchorReport:
files = tuple(
scan_memory_file(mf, repo_root)
for mf in _iter_memory_files(memory_root)
)
return AnchorReport(
generated_at=now if now is not None else time.time(),
repo_root=str(repo_root),
memory_root=str(memory_root),
files=files,
)
def format_dashboard(report: AnchorReport) -> str:
total = len(report.all_refs)
header = (
f"[memory-anchor] {len(report.files)} memory files "
f"refs={total} live={report.live_count} "
f"dead={report.dead_count}"
)
lines = [header]
dead_files = [f for f in report.files if f.dead]
if not dead_files:
lines.append("All references resolve.")
return "\n".join(lines)
lines.append("")
lines.append("Dead references:")
for f in dead_files:
lines.append(f" {f.memory_path}")
for ref in f.dead:
if ref.line is not None:
lines.append(f" - `{ref.raw}` (path={ref.path}, "
f"line={ref.line})")
else:
lines.append(f" - `{ref.raw}`")
return "\n".join(lines)
# ── Repo-root detection ────────────────────────────────────────────────────
def _detect_repo_root(start: Path) -> Path:
cur = start.resolve()
for parent in (cur, *cur.parents):
if (parent / ".git").exists():
return parent
return cur
# ── CLI ────────────────────────────────────────────────────────────────────
def _build_cli_report(args: argparse.Namespace) -> AnchorReport:
repo_root = Path(args.repo_root).resolve() if args.repo_root \
else _detect_repo_root(Path.cwd())
memory_root = Path(args.memory_root).expanduser() if args.memory_root \
else DEFAULT_MEMORY_ROOT
return build_report(repo_root=repo_root, memory_root=memory_root)
def cmd_scan(args: argparse.Namespace) -> int:
print(json.dumps(_build_cli_report(args).to_dict(), indent=2))
return 0
def cmd_dashboard(args: argparse.Namespace) -> int:
print(format_dashboard(_build_cli_report(args)))
return 0
def cmd_check(args: argparse.Namespace) -> int:
report = _build_cli_report(args)
print(format_dashboard(report))
if args.strict and report.has_dead:
return 2
return 0
def build_parser() -> argparse.ArgumentParser:
p = argparse.ArgumentParser(prog="memory_anchor")
sub = p.add_subparsers(dest="cmd", required=True)
def _add_common(sp: argparse.ArgumentParser) -> None:
sp.add_argument("--repo-root", default=None,
help="Repository root (default: auto-detect from cwd)")
sp.add_argument("--memory-root", default=None,
help="Memory directory "
"(default: ~/.claude/projects)")
sp = sub.add_parser("scan", help="Emit a JSON anchor report")
_add_common(sp)
sp.set_defaults(func=cmd_scan)
sp = sub.add_parser("dashboard", help="Pretty dashboard to stdout")
_add_common(sp)
sp.set_defaults(func=cmd_dashboard)
sp = sub.add_parser("check",
help="Dashboard + nonzero exit on dead refs in --strict")
_add_common(sp)
sp.add_argument("--strict", action="store_true")
sp.set_defaults(func=cmd_check)
return p
def main(argv: Sequence[str] | None = None) -> int:
args = build_parser().parse_args(list(argv) if argv is not None else None)
return args.func(args)
if __name__ == "__main__":
sys.exit(main())
|