doocus first ver

doocus/naming.py

@@ -0,0 +1,52 @@
+"""
+Run-folder naming + content hashing for doocus.
+
+Reuses meetus's `YYYYMMDD-NNN-<stem>` run-folder convention (see
+meetus/output_manager.py) so both trees look the same. Kept here to keep doocus
+self-contained; meetus could adopt this helper later.
+"""
+from pathlib import Path
+from datetime import datetime
+import hashlib
+
+
+def sha256_file(path: Path, chunk: int = 1 << 20) -> str:
+    """Streaming sha256 of a file — never loads the whole file into memory."""
+    h = hashlib.sha256()
+    with open(path, "rb") as f:
+        for block in iter(lambda: f.read(chunk), b""):
+            h.update(block)
+    return h.hexdigest()
+
+
+def resolve_run_dir(base_output_dir: Path, stem: str, use_cache: bool = True) -> Path:
+    """Return (creating if needed) the run folder for `stem` under `base_output_dir`.
+
+    Mirrors meetus/output_manager.py:37-84 — reuse the most recent
+    `*-<stem>` folder when caching, else create `YYYYMMDD-NNN-<stem>` with the
+    next run number for today.
+    """
+    base = Path(base_output_dir)
+
+    if use_cache and base.exists():
+        existing = sorted(
+            [d for d in base.iterdir() if d.is_dir() and d.name.endswith(f"-{stem}")],
+            reverse=True,
+        )
+        if existing:
+            return existing[0]
+
+    date_str = datetime.now().strftime("%Y%m%d")
+    next_run = 1
+    if base.exists():
+        nums = []
+        for d in base.iterdir():
+            if d.is_dir() and d.name.startswith(date_str) and d.name.endswith(f"-{stem}"):
+                parts = d.name.split("-")
+                if len(parts) >= 2 and parts[1].isdigit():
+                    nums.append(int(parts[1]))
+        next_run = max(nums) + 1 if nums else 1
+
+    run_dir = base / f"{date_str}-{next_run:03d}-{stem}"
+    run_dir.mkdir(parents=True, exist_ok=True)
+    return run_dir