| """Shared runtime for Hugging Face Space scoring.""" |
|
|
| from __future__ import annotations |
|
|
| import json |
| import re |
| from pathlib import Path |
| from typing import Any, Dict, List, Mapping, MutableMapping |
|
|
| DATA_DIR = Path(__file__).resolve().parent / "data" |
| DATASET_PATH = DATA_DIR / "pentadrive-v1.json" |
| I18N_PATH = DATA_DIR / "markers-i18n.json" |
| HAS_LATIN = re.compile(r"[a-zA-Z\u00C0-\u024F]") |
|
|
|
|
| def _load_json(path: Path) -> Dict[str, Any]: |
| with path.open("r", encoding="utf-8") as f: |
| return json.load(f) |
|
|
|
|
| DATASET = _load_json(DATASET_PATH) |
| I18N = _load_json(I18N_PATH) |
|
|
|
|
| def match_marker(marker: str, lowered: str) -> int: |
| m = marker.lower() |
| if not m: |
| return 0 |
| if " " not in m and HAS_LATIN.search(m): |
| escaped = re.escape(m) |
| return len(re.findall(rf"\b{escaped}\b", lowered, re.IGNORECASE)) |
| count = 0 |
| start = 0 |
| while start <= len(lowered): |
| idx = lowered.find(m, start) |
| if idx < 0: |
| break |
| count += 1 |
| start = idx + max(1, len(m)) |
| return count |
|
|
|
|
| def detect_drives(text: str) -> Dict[str, Any]: |
| lowered = (text or "").lower() |
| by_drive: MutableMapping[str, int] = {k: 0 for k in ("S", "K", "A", "M", "G")} |
| hits: List[Dict[str, str]] = [] |
|
|
| phases = DATASET.get("phases") or () |
| for node in DATASET.get("nodes") or (): |
| drive = node.get("drive") |
| node_phases = node.get("phases") |
| if not drive or not isinstance(node_phases, Mapping): |
| continue |
| for phase in phases: |
| phase_data = node_phases.get(phase) |
| markers = phase_data.get("markers") if isinstance(phase_data, Mapping) else None |
| if not isinstance(markers, list): |
| continue |
| for marker in markers: |
| count = match_marker(str(marker), lowered) |
| if count == 0: |
| continue |
| by_drive[str(drive)] = by_drive.get(str(drive), 0) + count |
| hits.append( |
| { |
| "drive": str(drive), |
| "node": str(node.get("node", "")), |
| "phase": str(phase), |
| "marker": str(marker), |
| } |
| ) |
|
|
| i18n_drives = I18N.get("drives") |
| if isinstance(i18n_drives, dict): |
| for drive_code, langs in i18n_drives.items(): |
| if not isinstance(langs, dict): |
| continue |
| for _, markers in langs.items(): |
| if not isinstance(markers, list): |
| continue |
| for marker in markers: |
| count = match_marker(str(marker), lowered) |
| if count == 0: |
| continue |
| by_drive[str(drive_code)] = by_drive.get(str(drive_code), 0) + count |
| hits.append( |
| { |
| "drive": str(drive_code), |
| "node": "i18n", |
| "phase": "all", |
| "marker": str(marker), |
| } |
| ) |
|
|
| ranked = [ |
| {"drive": d, "score": s} |
| for d, s in sorted(by_drive.items(), key=lambda x: (-x[1], x[0])) |
| if s > 0 |
| ] |
| return {"ranked": ranked, "by_drive": dict(by_drive), "matches": hits} |
|
|
|
|
