product-vamp/generator/catalogue.py

from __future__ import annotations

import argparse
import json
import sqlite3
import tempfile
from pathlib import Path

APPLICATION_ID = 0x56414D50
SCHEMA_VERSION = 1
COMMON_COLUMNS = """
  id TEXT PRIMARY KEY NOT NULL CHECK(length(trim(id)) BETWEEN 1 AND 160),
  title TEXT NOT NULL CHECK(length(trim(title)) BETWEEN 1 AND 200),
  composer TEXT NOT NULL,
  key_signature TEXT NOT NULL,
  authored_key_signature TEXT NOT NULL,
  source_path TEXT NOT NULL UNIQUE CHECK(length(source_path) > 0),
  source_json TEXT NOT NULL CHECK(json_valid(source_json)),
  measure_count INTEGER NOT NULL CHECK(measure_count > 0),
  total_pages INTEGER NOT NULL CHECK(total_pages BETWEEN 1 AND 10000)
"""

def human_label(value: str) -> str:
  result, capitalize = '', True
  for character in value:
    if character in '-_':
      result += ' '
      capitalize = True
    else:
      result += character.upper() if capitalize and 'a' <= character <= 'z' else character
      capitalize = False
  return result or 'Score'

def metadata(path: Path, root: Path) -> dict:
  if path.is_symlink() or not path.resolve().is_relative_to(root.resolve()):
    raise ValueError('score escapes the asset root')
  text = path.read_text(encoding='utf-8')
  document = json.loads(text)
  if document.get('schema_version') != 3:
    raise ValueError('expected a schema-v3 score bundle')
  music, presentation = document['music'], document['presentation']
  if len(music['staves']) != 1 or (presentation['width'], presentation['height']) != (976, 484):
    raise ValueError('expected one staff and 976x484 pages')
  staff, meta = music['staves'][0], music['metadata']
  for name in ('title', 'composer', 'key_signature', 'time_signature'):
    if not isinstance(meta[name], str):
      raise ValueError(f'{name} must be a string')
  authored = meta.get('authored_key_signature', '')
  if not isinstance(authored, str) or ('authored_key_signature' in meta and not authored):
    raise ValueError('authored_key_signature must be a nonempty string')
  key = meta['key_signature']
  if staff.get('arrangement_part_type') == 'practice' and key == 'No key' and 'variant_tonic' in meta:
    if not isinstance(meta['variant_tonic'], str):
      raise ValueError('variant_tonic must be a string')
    key = meta['variant_tonic'] + ' (no key)'
  if not isinstance(staff['name'], str):
    raise ValueError('staff name must be a string')
  relative = path.relative_to(root)
  return dict(id=path.stem, title=meta['title'], composer=meta['composer'], key_signature=key,
    authored_key_signature=authored, source_path=relative.as_posix(), source_json=text,
    measure_count=len(staff['measures']), total_pages=len(presentation['pages']), part=staff['name'])

def populate(connection: sqlite3.Connection, root: Path) -> None:
  if not root.is_dir():
    raise ValueError(f'{root}: score asset directory not found')
  connection.executescript(f'''
    PRAGMA application_id = {APPLICATION_ID};
    PRAGMA user_version = {SCHEMA_VERSION};
    CREATE TABLE songs ({COMMON_COLUMNS}, part TEXT NOT NULL) STRICT;
    CREATE TABLE practice_exercises ({COMMON_COLUMNS},
      exercise_id TEXT NOT NULL,
      practice_type_id TEXT NOT NULL, practice_type_title TEXT NOT NULL,
      lesson_group_id TEXT NOT NULL, lesson_group_title TEXT NOT NULL) STRICT;
    CREATE INDEX songs_menu ON songs(title COLLATE NOCASE, composer COLLATE NOCASE, part COLLATE NOCASE);
    CREATE INDEX practice_menu ON practice_exercises(practice_type_id COLLATE NOCASE, lesson_group_id COLLATE NOCASE, source_path COLLATE NOCASE);
  ''')
  identities, authored_keys = {}, {}
  for category, table, depth in (('songs', 'songs', 1), ('practice', 'practice_exercises', 4)):
    folder = root / category
    if (folder.exists() or folder.is_symlink()) and not folder.is_dir():
      raise ValueError(f'{folder}: score asset directory not found')
    for path in sorted((root / category).rglob('*.json')):
      try:
        parts = path.relative_to(root / category).parts
        if len(parts) != depth:
          raise ValueError(f'unexpected {category} score path')
        row = metadata(path, root)
        if category == 'practice':
          identity = '-'.join(parts[:3])
          previous = identities.setdefault(identity, path.parent)
          if previous != path.parent:
            raise ValueError(f'practice exercise ID collision {identity!r}: {previous} and {path.parent}')
          if authored_keys.setdefault(identity, row['authored_key_signature']) != row['authored_key_signature']:
            raise ValueError(f'practice variants disagree on authored key: {identity}')
          row.pop('part')
          row.update(id=identity + '-' + path.stem, exercise_id=identity, practice_type_id=parts[0], practice_type_title=human_label(parts[0]) + ' Practice',
            lesson_group_id=parts[1], lesson_group_title=human_label(parts[1]))
        connection.execute(f'INSERT INTO {table} ({", ".join(row)}) VALUES ({", ".join("?" for _ in row)})', tuple(row.values()))
      except (ValueError, KeyError, TypeError, sqlite3.Error) as error:
        raise ValueError(f'{path}: {error}') from error
  connection.commit()
  if connection.execute('PRAGMA integrity_check').fetchall() != [('ok',)]:
    raise ValueError('catalogue integrity check failed')

def build_catalogue(root: Path, destination: Path | None = None) -> Path:
  if not root.is_dir():
    raise ValueError(f'{root}: score asset directory not found')
  destination = destination or root / 'catalogue.db'
  destination.parent.mkdir(parents=True, exist_ok=True)
  with tempfile.TemporaryDirectory(prefix='.vamp-catalogue-', dir=destination.parent) as temporary:
    staged = Path(temporary) / 'catalogue.db'
    connection = sqlite3.connect(staged)
    try:
      populate(connection, root)
    finally:
      connection.close()
    staged.replace(destination)
  return destination

if __name__ == '__main__':
  parser = argparse.ArgumentParser(description='Compile generated score bundles into a read-only runtime catalogue.')
  parser.add_argument('assets', type=Path)
  parser.add_argument('destination', type=Path, nargs='?')
  arguments = parser.parse_args()
  build_catalogue(arguments.assets, arguments.destination)