""" A python class to encapsulate CoMets's data and file handling """ """ Copyright 2012 Anthony Beville Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ from datetime import datetime import zipfile from pprint import pprint import xml.etree.ElementTree as ET from genericmetadata import GenericMetadata import utils class CoMet: writer_synonyms = ['writer', 'plotter', 'scripter'] penciller_synonyms = [ 'artist', 'penciller', 'penciler', 'breakdowns' ] inker_synonyms = [ 'inker', 'artist', 'finishes' ] colorist_synonyms = [ 'colorist', 'colourist', 'colorer', 'colourer' ] letterer_synonyms = [ 'letterer'] cover_synonyms = [ 'cover', 'covers', 'coverartist', 'cover artist' ] editor_synonyms = [ 'editor'] def metadataFromString( self, string ): tree = ET.ElementTree(ET.fromstring( string )) return self.convertXMLToMetadata( tree ) def stringFromMetadata( self, metadata ): header = '\n' tree = self.convertMetadataToXML( self, metadata ) return header + ET.tostring(tree.getroot()) def indent( self, elem, level=0 ): # for making the XML output readable i = "\n" + level*" " if len(elem): if not elem.text or not elem.text.strip(): elem.text = i + " " if not elem.tail or not elem.tail.strip(): elem.tail = i for elem in elem: self.indent( elem, level+1 ) if not elem.tail or not elem.tail.strip(): elem.tail = i else: if level and (not elem.tail or not elem.tail.strip()): elem.tail = i def convertMetadataToXML( self, filename, metadata ): #shorthand for the metadata md = metadata # build a tree structure root = ET.Element("comet") root.attrib['xmlns:comet'] = "http://www.denvog.com/comet/" root.attrib['xmlns:xsi'] = "http://www.w3.org/2001/XMLSchema-instance" root.attrib['xsi:schemaLocation'] = "http://www.denvog.com http://www.denvog.com/comet/comet.xsd" #helper func def assign( comet_entry, md_entry): if md_entry is not None: ET.SubElement(root, comet_entry).text = u"{0}".format(md_entry) # title is manditory if md.title is None: md.title = "" assign( 'title', md.title ) assign( 'series', md.series ) assign( 'issue', md.issue ) #must be int... assign( 'volume', md.volume ) assign( 'description', md.comments ) #assign( 'date', md.year ) #need to make a date from month, year assign( 'publisher', md.publisher ) assign( 'genre', md.genre ) assign( 'pages', md.pageCount ) assign( 'format', md.format ) assign( 'language', md.language ) #assign( 'readingDirection', md.??? ) assign( 'character', md.characters ) assign( 'rating', md.maturityRating ) #assign( 'price', md.??? ) #assign( 'isVersionOf', md.??? ) #assign( 'rights', md.??? ) #assign( 'identifier', md.??? ) #assign( 'coverImage', md.??? ) #assign( 'lastMark', md.??? ) # need to specially process the credits, since they are structured differently than CIX credit_writer_list = list() credit_penciller_list = list() credit_inker_list = list() credit_colorist_list = list() credit_letterer_list = list() credit_cover_list = list() credit_editor_list = list() # loop thru credits, and build a list for each role that CoMet supports for credit in metadata.credits: if credit['role'].lower() in set( self.writer_synonyms ): ET.SubElement(root, 'writer').text = u"{0}".format(credit['person']) if credit['role'].lower() in set( self.penciller_synonyms ): ET.SubElement(root, 'penciller').text = u"{0}".format(credit['person']) if credit['role'].lower() in set( self.inker_synonyms ): ET.SubElement(root, 'inker').text = u"{0}".format(credit['person']) if credit['role'].lower() in set( self.colorist_synonyms ): ET.SubElement(root, 'inker').text = u"{0}".format(credit['person']) credit_colorist_list.append(credit['person'].replace(",","")) if credit['role'].lower() in set( self.letterer_synonyms ): ET.SubElement(root, 'letterer').text = u"{0}".format(credit['person']) if credit['role'].lower() in set( self.cover_synonyms ): ET.SubElement(root, 'coverDesigner').text = u"{0}".format(credit['person']) if credit['role'].lower() in set( self.editor_synonyms ): ET.SubElement(root, 'editor').text = u"{0}".format(credit['person']) # self pretty-print self.indent(root) # wrap it in an ElementTree instance, and save as XML tree = ET.ElementTree(root) return tree def convertXMLToMetadata( self, tree ): root = tree.getroot() if root.tag != 'comet': raise 1 return None metadata = GenericMetadata() md = metadata # Helper function def xlate( tag ): node = root.find( tag ) if node is not None: return node.text else: return None md.series = xlate( 'series' ) md.title = xlate( 'title' ) md.issue = xlate( 'issue' ) md.volume = xlate( 'volume' ) md.comments = xlate( 'description' ) #md.year = xlate( 'Year' ) #md.month = xlate( 'Month' ) md.publisher = xlate( 'publisher' ) md.genre = xlate( 'genre' ) md.language = xlate( 'language' ) md.format = xlate( 'format' ) #md.manga = xlate( 'readingDirection' ) #md.characters = xlate( 'character' ) md.pageCount = xlate( 'pages' ) md.maturityRating = xlate( 'rating' ) # Now extract the credit info for n in root: if ( n.tag == 'writer' or n.tag == 'penciller' or n.tag == 'inker' or n.tag == 'colorist' or n.tag == 'letterer' or n.tag == 'editor' ): metadata.addCredit( n.text.strip(), n.tag.title() ) if n.tag == 'coverDesigner': metadata.addCredit( n.text.strip(), "Cover" ) metadata.isEmpty = False return metadata def writeToExternalFile( self, filename, metadata ): tree = self.convertMetadataToXML( self, metadata ) #ET.dump(tree) tree.write(filename, encoding='utf-8') def readFromExternalFile( self, filename ): tree = ET.parse( filename ) return self.convertXMLToMetadata( tree )