X-Git-Url: http://git.veekun.com/zzz-pokedex.git/blobdiff_plain/2eb2985a60e6cd216399ea66bd6a976f9abe8868..b11dddc0517b1c928695b71aa9b74c85485e9e6c:/pokedex/db/__init__.py?ds=sidebyside diff --git a/pokedex/db/__init__.py b/pokedex/db/__init__.py index 0da8713..464629d 100644 --- a/pokedex/db/__init__.py +++ b/pokedex/db/__init__.py @@ -1,7 +1,13 @@ -from sqlalchemy import MetaData, Table, engine_from_config, orm +# encoding: utf-8 +import re + +from sqlalchemy import engine_from_config, orm from ..defaults import get_default_db_uri -from .tables import metadata +from .tables import Language, metadata +from .multilang import MultilangSession, MultilangScopedSession + +ENGLISH_ID = 9 def connect(uri=None, session_args={}, engine_args={}, engine_prefix=''): @@ -15,12 +21,12 @@ def connect(uri=None, session_args={}, engine_args={}, engine_prefix=''): # If we didn't get a uri, fall back to the default if uri is None: - uri = engine_args[engine_prefix + 'url'] + uri = engine_args.get(engine_prefix + 'url', None) if uri is None: uri = get_default_db_uri() ### Do some fixery for MySQL - if uri[0:5] == 'mysql': + if uri.startswith('mysql:'): # MySQL uses latin1 for connections by default even if the server is # otherwise oozing with utf8; charset fixes this if 'charset' not in uri: @@ -40,7 +46,38 @@ def connect(uri=None, session_args={}, engine_args={}, engine_prefix=''): all_session_args = dict(autoflush=True, autocommit=False, bind=engine) all_session_args.update(session_args) - sm = orm.sessionmaker(**all_session_args) - session = orm.scoped_session(sm) + sm = orm.sessionmaker(class_=MultilangSession, + default_language_id=ENGLISH_ID, **all_session_args) + session = MultilangScopedSession(sm) return session + +def identifier_from_name(name): + """Make a string safe to use as an identifier. + + Valid characters are lowercase alphanumerics and "-". This function may + raise ValueError if it can't come up with a suitable identifier. + + This function is useful for scripts which add things with names. + """ + if isinstance(name, str): + identifier = name.decode('utf-8') + else: + identifier = name + identifier = identifier.lower() + identifier = identifier.replace(u'+', u' plus ') + identifier = re.sub(u'[ _–]+', u'-', identifier) + identifier = re.sub(u"['./;’(),:]", u'', identifier) + identifier = identifier.replace(u'é', u'e') + identifier = identifier.replace(u'♀', u'-f') + identifier = identifier.replace(u'♂', u'-m') + if identifier in (u'???', u'????'): + identifier = u'unknown' + elif identifier == u'!': + identifier = u'exclamation' + elif identifier == u'?': + identifier = u'question' + + if not identifier.replace(u"-", u"").isalnum(): + raise ValueError(identifier) + return identifier