Fix default language assignment once and for all.

[zzz-pokedex.git] / pokedex / db / __init__.py

diff --git a/pokedex/db/__init__.py b/pokedex/db/__init__.py
index ea103e7..464629d 100644 (file)
--- a/pokedex/db/__init__.py
+++ b/pokedex/db/__init__.py
@@ -1,9 +1,14 @@
+# encoding: utf-8
+import re
+
 from sqlalchemy import engine_from_config, orm
 
 from ..defaults import get_default_db_uri
 from .tables import Language, metadata
 from .multilang import MultilangSession, MultilangScopedSession
 
+ENGLISH_ID = 9
+
 
 def connect(uri=None, session_args={}, engine_args={}, engine_prefix=''):
     """Connects to the requested URI.  Returns a session object.
@@ -21,7 +26,7 @@ def connect(uri=None, session_args={}, engine_args={}, engine_prefix=''):
         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:
@@ -41,13 +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(class_=MultilangSession, language_class=Language,
-        **all_session_args)
+    sm = orm.sessionmaker(class_=MultilangSession,
+        default_language_id=ENGLISH_ID, **all_session_args)
     session = MultilangScopedSession(sm)
 
-    # Default to English
-    session.default_language = session.query(Language) \
-        .filter_by(identifier=u'en') \
-        .one()
-
     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