fast discover was also slower than os.walk

backend/app/services/scanner.py

@@ -1,12 +1,10 @@
 import concurrent.futures
 import hashlib
 import os
-import shutil
-import subprocess
 import threading
 import time
 from datetime import datetime, timezone
-from typing import Any, Dict, List, Optional, Tuple
+from typing import Any, Dict, List, Optional
 
 import psutil
 from loguru import logger
@@ -16,161 +14,6 @@ from sqlalchemy.orm.exc import ObjectDeletedError, StaleDataError
 from app.db import models
 from app.db.database import SessionLocal
 
-# Fast file discovery via `find -printf` (GNU find or compatible).
-# Detected once at import time; falls back to os.walk if unavailable.
-_FAST_FIND_BINARY: Optional[str] = None
-
-
-def _detect_fast_find() -> Optional[str]:
-    """Check if a `find` binary with `-printf` support is available.
-
-    Tries `gfind` (GNU find via Homebrew on macOS) first, then `find`.
-    Returns the binary path if `-printf` works, otherwise ``None``.
-    """
-    for candidate in ("gfind", "find"):
-        binary = shutil.which(candidate)
-        if binary is None:
-            continue
-        try:
-            result = subprocess.run(
-                [binary, "/tmp", "-maxdepth", "0", "-printf", "%f\n"],
-                capture_output=True,
-                timeout=5,
-            )
-            if result.returncode == 0 and result.stdout.strip() == b"tmp":
-                return binary
-        except Exception:
-            continue
-    return None
-
-
-def _init_fast_features() -> Optional[str]:
-    global _FAST_FIND_BINARY
-    _FAST_FIND_BINARY = _detect_fast_find()
-
-    if _FAST_FIND_BINARY:
-        logger.info(f"Fast file discovery enabled: using {_FAST_FIND_BINARY} -printf")
-    else:
-        logger.info("Fast file discovery unavailable: falling back to os.walk")
-
-    return _FAST_FIND_BINARY
-
-
-_FAST_FIND_BINARY = _init_fast_features()
-
-
-def _discover_files_fast(
-    root_base: str,
-    job_id: Optional[int],
-    batch_size: int,
-    current_timestamp,
-    resolve_tracking,
-    sync_metadata_batch,
-    metrics_lock,
-    metrics,
-    db_session: Session,
-) -> Tuple[int, int]:
-    """Walk a tree using `find -printf` for fast metadata extraction.
-
-    Streams output line-by-line via subprocess.Popen so progress updates
-    appear as files are discovered instead of waiting for find to finish.
-
-    Returns (files_found, files_batched) counts.
-    """
-    total_files_found = 0
-    files_batched = 0
-    pending_metadata: List[Dict[str, Any]] = []
-
-    # -printf format: path\tsize\tmtime (tab-separated; split from right for safety)
-    find_binary = _FAST_FIND_BINARY
-    if find_binary is None:
-        logger.warning(
-            "Fast file discovery requested but no compatible `find` binary found"
-        )
-        return 0, 0
-    cmd = [
-        find_binary,
-        root_base,
-        "-type",
-        "f",
-        "-printf",
-        "%p\t%s\t%T@\n",
-    ]
-
-    try:
-        proc = subprocess.Popen(
-            cmd,
-            stdout=subprocess.PIPE,
-            stderr=subprocess.DEVNULL,
-        )
-        if proc.stdout is None:
-            logger.error(
-                f"Fast file discovery failed: could not open stdout for {root_base}"
-            )
-            return 0, 0
-    except Exception as e:
-        logger.error(f"Fast file discovery failed for {root_base}: {e}")
-        return 0, 0
-
-    # Stream output line by line (tab-separated: path\tsize\tmtime)
-    for line in iter(proc.stdout.readline, b""):
-        if job_id is not None and JobManager.is_cancelled(job_id):
-            break
-
-        if not line.strip():
-            continue
-
-        # Split from right: mtime and size are always numeric
-        parts = line.split(b"\t")
-        if len(parts) < 3:
-            continue
-
-        # First n-2 parts may be path components (tabs in filename are rare)
-        full_file_path = b"\t".join(parts[:-2]).decode("utf-8", errors="replace")
-        try:
-            file_size = int(parts[-2])
-            file_mtime = float(parts[-1])
-        except (ValueError, IndexError):
-            continue
-
-        total_files_found += 1
-        with metrics_lock:
-            metrics["total_files_found"] = total_files_found
-            metrics["current_path"] = os.path.dirname(full_file_path)
-
-        is_ignored = resolve_tracking(full_file_path)
-        pending_metadata.append(
-            {
-                "path": full_file_path,
-                "size": file_size,
-                "mtime": file_mtime,
-                "ignored": is_ignored,
-            }
-        )
-
-        if len(pending_metadata) >= batch_size:
-            sync_metadata_batch(db_session, pending_metadata, current_timestamp)
-            db_session.commit()
-            files_batched += len(pending_metadata)
-            pending_metadata = []
-            if job_id is not None:
-                JobManager.update_job(
-                    job_id,
-                    10.0,
-                    f"Discovered {total_files_found} items...",
-                )
-
-    proc.stdout.close()
-    proc.wait()
-
-    # Flush remaining batch
-    if pending_metadata:
-        sync_metadata_batch(db_session, pending_metadata, current_timestamp)
-        db_session.commit()
-        files_batched += len(pending_metadata)
-
-    return total_files_found, files_batched
-
 
 class JobManager:
     """Manages operational job states and persistence with high resilience for background threads."""
@@ -439,27 +282,6 @@ class ScannerService:
                 if not os.path.exists(root_base):
                     continue
 
-                if _FAST_FIND_BINARY:
-                    # Fast path: GNU find -printf (metadata extracted in C)
-                    metrics = {
-                        "total_files_found": 0,
-                        "current_path": root_base,
-                    }
-                    found, _ = _discover_files_fast(
-                        root_base,
-                        job_id,
-                        BATCH_SIZE,
-                        current_timestamp,
-                        resolve_tracking,
-                        self._sync_metadata_batch,
-                        self._metrics_lock,
-                        metrics,
-                        db_session,
-                    )
-                    with self._metrics_lock:
-                        self.total_files_found += found
-                else:
-                    # Compatibility path: Python os.walk + os.stat
                 for current_dir, _sub_dirs, file_names in os.walk(root_base):
                     if job_id is not None and JobManager.is_cancelled(job_id):
                         break

backend/tests/test_service_scanner.py

@@ -114,9 +114,6 @@ def test_scan_sources_mocked(db_session, mocker):
     """Tests the discovery scan with mocked filesystem."""
     scanner = ScannerService()
 
-    # Disable fast find so the test uses the os.walk fallback path
-    mocker.patch("app.services.scanner._FAST_FIND_BINARY", None)
-
     # Mock settings
     mocker.patch("app.api.common.get_source_roots", return_value=["/mock_source"])
     mocker.patch("app.api.common.get_exclusion_spec", return_value=None)
@@ -146,7 +143,6 @@ def test_missing_file_marked_deleted_at_end_of_scan(db_session, mocker):
     """Tests that files not seen during a scan are marked as deleted."""
     scanner = ScannerService()
 
-    mocker.patch("app.services.scanner._FAST_FIND_BINARY", None)
     mocker.patch("app.api.common.get_source_roots", return_value=["/mock_source"])
     mocker.patch("app.api.common.get_exclusion_spec", return_value=None)
     mocker.patch("os.walk", return_value=[])
@@ -181,7 +177,6 @@ def test_existing_file_not_marked_deleted(db_session, mocker):
     """Tests that files found during scan retain is_deleted=False."""
     scanner = ScannerService()
 
-    mocker.patch("app.services.scanner._FAST_FIND_BINARY", None)
     mocker.patch("app.api.common.get_source_roots", return_value=["/mock_source"])
     mocker.patch("app.api.common.get_exclusion_spec", return_value=None)
     mocker.patch("os.path.exists", return_value=True)