sa6anw.se/clean_language_blocks.py

48 lines
1.4 KiB
Python

import os
import re
import shutil
SOURCE_DIR = "docs"
ENGLISH_DIR = "docs/en"
# Rensa tidigare engelska mappar
if os.path.exists(ENGLISH_DIR):
shutil.rmtree(ENGLISH_DIR)
shutil.copytree(SOURCE_DIR, ENGLISH_DIR)
# --- Inline-hantering ---
def remove_other_language_inline(content, keep_lang):
def replacer(match):
lang, text = match.group(1), match.group(2)
return text if lang == keep_lang else ""
return re.sub(r"\{(sv|en):(.*?)}", replacer, content)
# --- Block-hantering ---
def remove_other_language_blocks(content, keep_lang):
content = re.sub(
r":(sv|en)\n(.*?)\n:::",
lambda m: m.group(2) if m.group(1) == keep_lang else "",
content,
flags=re.DOTALL
)
return content
# --- Kombinerad rengöring ---
def clean_file(path, keep_lang):
with open(path, "r", encoding="utf-8") as f:
content = f.read()
content = remove_other_language_blocks(content, keep_lang)
content = remove_other_language_inline(content, keep_lang)
with open(path, "w", encoding="utf-8") as f:
f.write(content.strip() + "\n")
# --- Rensa svenska filer ---
for filename in os.listdir(SOURCE_DIR):
if filename.endswith(".md"):
clean_file(os.path.join(SOURCE_DIR, filename), keep_lang="sv")
# --- Rensa engelska filer ---
for filename in os.listdir(ENGLISH_DIR):
if filename.endswith(".md"):
clean_file(os.path.join(ENGLISH_DIR, filename), keep_lang="en")