Nicolas Cantu
2b99e8ff02
**Motivations:** - Ajout skill pour amélioration de documents en background - Scripts et documentation Collatz **Evolutions:** - .cursor/skills/document-improvement/ (SKILL, reference, examples) - v0/collatz_k_scripts/ (core, fusion, pipeline, utils, reproduce) - v0/journal.md, v0/log.md, v0/README collatz **Pages affectées:** - .cursor/skills/document-improvement/ - v0/collatz_k_scripts/ - v0/journal.md, v0/log.md
45 lines
1.3 KiB
Python
45 lines
1.3 KiB
Python
# -*- coding: utf-8 -*-
|
|
"""
|
|
collatz_k_utils.py
|
|
|
|
Utilitaires: lecture de listes depuis des fichiers Markdown,
|
|
écriture de rapports, etc.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
from pathlib import Path
|
|
import re
|
|
from typing import List, Optional
|
|
|
|
|
|
def extract_ints_from_markdown(path: str, header_regex: Optional[str] = None) -> List[int]:
|
|
"""
|
|
Extrait tous les entiers d'un fichier Markdown.
|
|
Si header_regex est fourni, extrait uniquement la section à partir du header.
|
|
"""
|
|
text = Path(path).read_text(encoding="utf-8")
|
|
if header_regex is not None:
|
|
m = re.search(header_regex, text, flags=re.S)
|
|
if not m:
|
|
raise ValueError(f"Section introuvable via regex: {header_regex}")
|
|
text = m.group(1)
|
|
return list(map(int, re.findall(r"\b\d+\b", text)))
|
|
|
|
|
|
def parse_markdown_table_to_rows(path: str) -> List[List[str]]:
|
|
"""
|
|
Parse simple des lignes de table Markdown commençant par '|'.
|
|
Retourne les champs (sans les '|', strip).
|
|
"""
|
|
rows: List[List[str]] = []
|
|
for ln in Path(path).read_text(encoding="utf-8").splitlines():
|
|
if not ln.startswith("|"):
|
|
continue
|
|
parts = [p.strip() for p in ln.strip().strip("|").split("|")]
|
|
rows.append(parts)
|
|
return rows
|
|
|
|
|
|
def write_text(path: str, content: str) -> None:
|
|
Path(path).write_text(content, encoding="utf-8")
|