Link articles in "code-civil"

link.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

import os
import re

def find_between (s, first, last):
    try:
        start = s.index(first) + len(first)
        end = s.index(last, start)
        return s[start:end]
    except:
        return -1

def link_or_self(matched_article, article_path_dict, current_article):
    if matched_article == current_article or matched_article in {"1", "2", "3", "4", "5", "6", "7", "8", "9"} or matched_article not in article_path_dict:
        return " " + matched_article

    return " [{}]({})".format(matched_article, article_path_dict[matched_article])

# Dictionary article number -> file path
article_path_dict = {}
for (dir, _, files) in os.walk("."):
    for filename in files:
        article = find_between(filename, 'Article ', '.md')

        if article == -1:
            continue

        print os.path.join(dir, filename)[1:]
        article_path_dict[article] = os.path.join(dir, filename)[1:]

for (dir, _, files) in os.walk("."):
    for filename in files:
        article = find_between(filename, 'Article ', '.md')

        if article == -1:
            continue

        path = os.path.join(dir, filename)

        with open(path) as f:
            content = f.read()

        content = re.sub(r' (\d+\-?\d*\-?\d*)', lambda m: link_or_self(m.group(1), article_path_dict, article), content)
        
        with open(path, 'w') as f:
            f.write(content)

Ouais ok, en regardant vite fait je vois qu'il y a pas mal de connecteurs possibles, ex. x à y et z.
Donc soit quelqu'un peut faire l'ensemble des possibilités soit en effet ta méthode (regexp simple + corrections) et la meilleure puisqu'on aura moins de faux-négatifs qu'avec la mienne.