Fixed indentation in lark and updated comments.

.gitignore

@@ -129,3 +129,6 @@ dmypy.json
 # Pyre type checker
 .pyre/
 
+# Kate swap files
+*.kate-swp
+

dat.py

@@ -1,62 +0,0 @@
-import xml.etree.ElementTree
-import hashdb
-
-class DatImportError(Exception):
-    '''This error is raised when a DAT import fails.'''
-
-# TODO: l o g g i n g
-
-# TODO: Consider using a context object to avoid keeping large XML trees in memory.
-class Dat:
-    '''A Dat object processes DAT files into the data structures defined in hashdb. '''
-    def __init__(self, filename):
-        '''Open the given DAT file and gather metadata from it.'''
-
-        xml_tree = xml.etree.ElementTree.parse(filename)
-        self._xml_root = xml_tree.getroot()
-
-        dat_header = self._xml_root.find('header')
-        self.info = hashdb.DatInfo(name=dat_header.find('name').text,
-                                   description=dat_header.find('description').text,
-                                   platform=None,
-                                   version=dat_header.find('version').text)
-
-    def set_platform(self, platform_info):
-        '''
-        Set a platform for this DAT file.
-        DAT files don't include platform metadata, but are all platform-specific.
-        '''
-        new_info = hashdb.DatInfo(name=self.info.name,
-                                  description=self.info.description,
-                                  platform=platform_info.shortcode,
-                                  version=self.info.version)
-        self.info = new_info
-
-    def set_name(self, new_name):
-        '''
-        Override the DAT file's name.
-        DAT files often have less-than-helpful names.
-        '''
-        new_info = hashdb.DatInfo(name=new_name,
-                                  description=self.info.description,
-                                  platform=self.info.platform,
-                                  version=self.info.version)
-        self.info = new_info
-
-    def read_all_hashes(self):
-        '''Read every hash in the DAT file and return it as a large list of RomInfo tuples.'''
-        if self.info.platform is None:
-            raise DatImportError('DAT platform not set.')
-
-        rom_info_list = []
-        all_rom_entries = self._xml_root.findall('.//rom')
-
-        for rom in all_rom_entries:
-            rom_info = hashdb.RomInfo(sha1sum=rom.get('sha1'),
-                                      filename=rom.get('name'),
-                                      platform=self.info.platform,
-                                      datorigin=self.info.name)
-
-            rom_info_list.append(rom_info)
-
-        return rom_info_list

metadata.py

@@ -13,6 +13,9 @@ import sqlalchemy.orm
 
 # TODO: l o g g i n g
 HASH_CHUNK_SIZE = 10485760 # 10mb
+
+db_session = None
+
 _db_session_maker = sqlalchemy.orm.sessionmaker()
 _engine = None
 _configured = False
@@ -25,6 +28,11 @@ class MetadataDBSessionException(Exception):
 def _uuidgen():
 	return str(uuid.uuid4())
 
+'''
+Metadata ORM classes for SQLAlchemy. For a detailed description of each piece of data, refer to
+metadata/README.md
+'''
+
 class Platform(_SQLBase):
 	'''SQLAlchemy ORM class for platform metadata.'''
 	__tablename__ = 'platforms'
@@ -33,6 +41,7 @@ class Platform(_SQLBase):
 		)
 	fullname = sqlalchemy.Column(sqlalchemy.String, nullable=False)
 	shortcode = sqlalchemy.Column(sqlalchemy.String, unique=True, nullable=False)
+	regional_names = sqlalchemy.Column(sqlalchemy.String)
 
 	release_groups = sqlalchemy.orm.relationship(
 		'ReleaseGroup', order_by=ReleaseGroup.id, back_populates='platform'
@@ -53,7 +62,6 @@ class ReleaseGroup(_SQLBase):
 	name = sqlalchemy.Column(sqlalchemy.String, unique=True, nullable=False)
 	platform_id = sqlalchemy.Column(sqlalchemy.Integer, sqlalchemy.ForeignKey('platforms.id'))
 
-	platform = sqlalchemy.orm.relationship('Platform', back_populates='release_groups')
 	releases = sqlalchemy.orm.relationship('Release', back_populates='release_group')
 
 	def __repr__(self):
@@ -76,7 +84,7 @@ class Release(_SQLBase):
 	release_group = sqlalchemy.orm.relationship('ReleaseGroup', back_populates='releases')
 
 	def __repr__(self):
-		return ('ROM Image: id: %s, en_name: %s' % (self.id, self.en_name )
+		return ('Release: id: %s, en_name: %s' % (self.id, self.en_name )
 
 class Image(_SQLBase):
 	'''SQLAlchemy ORM class for ROM image metadata.'''
@@ -105,6 +113,11 @@ class Image(_SQLBase):
 			)
 
 
+'''
+# TODO: I was attempting to avoid writing a class to see if I could, but it is not worth the clunk.
+Rewrite these functions into a class.
+'''
+
 def configure(db_path):
 	'''
 	Configure and initialize the database for the entire module.
@@ -118,7 +131,6 @@ def configure(db_path):
 	_db_session_maker.configure(bind=_engine)
 	_configured = True
 
-# TODO: Passing the session object is a little clunky. Maybe there's a way to infer it somehow?
 def search(session, table_object, **constraints):
 	'''
 	Search the database for entries matching the given constraints.

readme.md

@@ -8,17 +8,11 @@ directory structure.
 
 ## Planned features
 * Validate ROM images.
-* Import DAT files
 * Rename/move ROM files
 * Maintain a database of present ROMs
 * A nice, Beets-like interface
 * Grouping ROMS in archive files
 
 ## Known issues
-* This probably isn't terribly efficient. It's Python parsing XML into an SQLite database and I only
-  know pretty basic database design.
-
-* Python's `xml.etree` module has a couple of known security issues[1]. Stick to importing DATs from
-known places and it shouldn't be an issue.
-
-[1] - https://docs.python.org/3/library/xml.html#xml-vulnerabilities
+* This probably isn't terribly efficient. It's Python parsing JSON into an SQLite database and I
+onlyknow pretty basic database design.