extract-bmg.yml

name: » 🗺️ • Extract RR Tracks to JSON

on:
  workflow_dispatch:

jobs:
  check-and-build:
    runs-on: ubuntu-latest
    steps:
      - name: 📥 Checkout Riivolution Branch
        uses: actions/checkout@v4
        with:
          ref: Riivolution

      - name: Setup local Wiimms SZS Tools
        run: |
          # Make local tools executable and add to path
          chmod +x wszst/*
          echo "$(pwd)/wszst" >> $GITHUB_PATH

      - name: Download Fordy-RR release
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          # Download latest zip and extract
          gh release download --pattern "Fordy-RR-*.zip"
          unzip Fordy-RR-*.zip -d rr_mod

      - name: Extract SZS and Decode BMG
        run: |
          # Extract all szs and decode all bmg recursively
          find rr_mod -type f -name "*.szs" -exec wszst extract {} -d {}_ext \;
          find rr_mod -type f -name "*.bmg" -exec wbmgt decode {} \;

      - name: Generate Python Parser
        run: |
          # Write python script to parse decoded bmg files
          cat << 'EOF' > parse_bmg.py
          import os, json, re, glob

          bmg_files = glob.glob('rr_mod/**/*.txt', recursive=True)
          id_regex = re.compile(r'^\[(?:0x)?([0-9a-fA-F]+)\]')

          lang_map = {
              'E': 'en', 'G': 'de', 'F': 'fr', 'S': 'es', 'I': 'it',
              'J': 'ja', 'K': 'ko', 'N': 'nl', 'M': 'nl'
          }

          def get_lang(filepath):
              name = os.path.basename(filepath)
              match = re.search(r'_([EGFSIJKNM])\.', name, re.IGNORECASE)
              if match:
                  return lang_map.get(match.group(1).upper(), 'en')
              
              lower_path = filepath.lower()
              if 'german' in lower_path or '/de/' in lower_path: return 'de'
              if 'french' in lower_path or '/fr/' in lower_path: return 'fr'
              if 'spanish' in lower_path or '/es/' in lower_path: return 'es'
              if 'italian' in lower_path or '/it/' in lower_path: return 'it'
              if 'dutch' in lower_path or '/nl/' in lower_path: return 'nl'
              return 'en'

          db = {}

          for f in bmg_files:
              lang = get_lang(f)
              try:
                  with open(f, 'r', encoding='utf-8') as txt:
                      lines = txt.readlines()
              except:
                  continue

              current_id = None
              current_text = []

              for line in lines:
                  if line.startswith('#'): continue
                  match = id_regex.match(line.strip())
                  if match:
                      if current_id and current_text:
                          msg = ''.join(current_text).strip()
                          if current_id not in db: db[current_id] = {'en': '', 'translations': {}}
                          if lang == 'en': db[current_id]['en'] = msg
                          else: db[current_id]['translations'][lang] = msg
                      current_id = match.group(1).upper()
                      current_text = []
                  elif current_id:
                      if line.strip() != '' or current_text:
                          current_text.append(line)

              if current_id and current_text:
                  msg = ''.join(current_text).strip()
                  if current_id not in db: db[current_id] = {'en': '', 'translations': {}}
                  if lang == 'en': db[current_id]['en'] = msg
                  else: db[current_id]['translations'][lang] = msg

          # Write parsed data to JSON
          with open('RR-Tracks.json', 'w', encoding='utf-8') as out:
              json.dump(db, out, indent=2, ensure_ascii=False)
          EOF

      - name: Parse BMGs and build JSON
        run: python3 parse_bmg.py

      - name: Upload Artifact
        uses: actions/upload-artifact@v4
        with:
          name: RR-Tracks
          path: RR-Tracks.json