import os
import re

SOURCE_DIR = "docs"
OUTPUT_DIRS = {"sv": "docs_sv", "en": "docs_en"}

lang_block_re = re.compile(r":::\s*(sv|en)\n(.*?)\n:::", re.DOTALL | re.IGNORECASE)

def split_content_by_language(content):
    blocks = {"sv": "", "en": ""}
    neutral_parts = []
    pos = 0

    matches = list(lang_block_re.finditer(content))

    if not matches:
        return {"sv": content, "en": ""}

    for match in matches:
        start, end = match.span()
        lang, text = match.groups()
        neutral = content[pos:start]
        neutral_parts.append(neutral)
        blocks[lang.lower()] += text.strip() + "\n\n"
        pos = end

    # Innehåll efter sista språkblock
    neutral_parts.append(content[pos:])

    # Lägg neutral text i båda
    tail = ''.join(neutral_parts)
    for l in blocks:
        blocks[l] = tail + blocks[l]

    return blocks

for filename in os.listdir(SOURCE_DIR):
    if filename.endswith(".md"):
        filepath = os.path.join(SOURCE_DIR, filename)
        with open(filepath, "r", encoding="utf-8") as f:
            content = f.read()

        blocks = split_content_by_language(content)

        for lang, text in blocks.items():
            if text.strip():
                os.makedirs(OUTPUT_DIRS[lang], exist_ok=True)
                out_path = os.path.join(OUTPUT_DIRS[lang], filename)
                with open(out_path, "w", encoding="utf-8") as f:
                    f.write(text.strip())
                print(f"✔ Skrev {lang}/{filename}")
            else:
                print(f"⚠️  Skippade {lang}/{filename} (tom)")