def single_input(args, config=None): """ Single Input Mode works to convert a single input XML file into EPUB. This is probably the most typical use case and is the most highly configurable, see the argument parser and oaepub --help """ if config is None: config = get_config_module() #Determination of input type and processing #Fetch by URL if 'http:' in args.input: raw_name = u_input.url_input(args.input) abs_input_path = os.path.join(LOCAL_DIR, raw_name+'.xml') parsed_article = Article(abs_input_path, validation=args.no_dtd_validation) #Fetch by DOI elif args.input[:4] == 'doi:': raw_name = u_input.doi_input(args.input) abs_input_path = os.path.join(LOCAL_DIR, raw_name+'.xml') parsed_article = Article(abs_input_path, validation=args.no_dtd_validation) #Local XML input else: abs_input_path = utils.get_absolute_path(args.input) raw_name = u_input.local_input(abs_input_path) parsed_article = Article(abs_input_path, validation=args.no_dtd_validation) #Generate the output path name, this will be the directory name for the #output. This output directory will later be zipped into an EPUB output_name = os.path.join(utils.get_output_directory(args), raw_name) #Make the EPUB make_epub(parsed_article, outdirect=output_name, explicit_images=args.images, # Explicit image path batch=False, config=config) #Cleanup removes the produced output directory, keeps the ePub file if args.clean: # Defaults to False, --clean or -c to toggle on shutil.rmtree(output_name) #Running epubcheck on the output verifies the validity of the ePub, #requires a local installation of java and epubcheck. if args.no_epubcheck: epubcheck('{0}.epub'.format(output_name), config)
def collection_input(args, config=None): """ Collection Input Mode works to compile multiple articles into a single composite ePub. This is akin to such formats as Collections, Issues, and Omnibus; it may also be useful for those interested in the simple distribution of a reading list, personal publications, or topic reference. Collection Input Mode produces output that is necessarily unlike the output generated by Single or Batch (which is just sequential Single) input modes. The primary difference in output lies with the ePub metadata; as applying metadata from any single article to the whole would be inappropriate. Unlike other input modes, Collection Mode is strictly dependent on the local directory of execution. If there is a file named "order.txt" in the local directory, this file should contain the name of one input XML file on each line; the files will be added to the ePub output by line-order. If there is "order.txt" file, Collection Mode will assume that all XML files are input and the article in order in the collection will be random. Collection Input Mode has default epubcheck behavior, it will place a system call to epubcheck unless specified otherwise (--no-epubcheck or -N flags). """ if config is None: config = get_config_module() try: order = open('order.txt', 'r') except IOError: # No order.txt xml_files = list_xml_files(dir=os.getcwd()) else: #Add all nonempty lines, in order, to the xml_files list xml_files = [i.strip() for i in order.readlines() if i.strip()] order.close() #The output name will be the same as the parent directory name #This will also serve as the dc:title output_name = os.path.split(os.getcwd())[1] #The standard make_epub() method will not work for Collection Mode #So the work done here is an adaptation of it print('Processing output to {0}.epub'.format(output_name)) #Copy files from base_epub to the new output if os.path.isdir(output_name): dir_exists(output_name) epub_base = os.path.join(CACHE_LOCATION, 'base_epub') shutil.copytree(epub_base, output_name) if args.collection is True: try: title_txt = open('title.txt', 'r') except IOError: # No title.txt title = output_name else: title = title_txt.readline().strip() title_txt.close() if not title: title = output_name print('title.txt was empty or title was not on first line!') print('Defaulting to name of parent directory. {0}'.format(title)) else: title = args.collection toc = ncx.NCX(oae_version=__version__, location=output_name, collection_mode=True) myopf = opf.OPF(location=output_name, collection_mode=True, title=title) #Now it is time to operate on each of the xml files for xml_file in xml_files: raw_name = u_input.local_input(xml_file) # is this used? parsed_article = Article(xml_file, validation=args.no_dtd_validation) toc.take_article(parsed_article) myopf.take_article(parsed_article) #Get the Digital Object Identifier doi = parsed_article.get_DOI() journal_doi, article_doi = doi.split('/') #Check for images img_dir = os.path.join(output_name, 'OPS', 'images-{0}'.format(article_doi)) expected_local = 'images-{0}'.format(raw_name) if os.path.isdir(expected_local): utils.images.local_images(expected_local, img_dir) else: article_cache = os.path.join(config.image_cache, journal_doi, article_doi) if os.path.isdir(article_cache): utils.images.image_cache(article_cache, img_dir) else: print('Images for {0} (DOI: {1}) could not be found!'.format(xml_file, doi)) r = input('Try to download them? [Y/n]') if r in ['y', 'Y', '']: os.mkdir(img_dir) utils.images.fetch_plos_images(article_doi, img_dir, parsed_article) if config.use_image_cache: utils.images.move_images_to_cache(img_dir, article_cache) else: sys.exit(1) #TODO: Content stuff if journal_doi == '10.1371': # PLoS's publisher DOI ops_doc = ops.OPSPLoS(parsed_article, output_name) #TODO: Workflow change, parse table of contents from OPS processed document toc.write() myopf.write() utils.epub_zip(output_name) #Running epubcheck on the output verifies the validity of the ePub, #requires a local installation of java and epubcheck. if args.no_epubcheck: epubcheck('{0}.epub'.format(output_name), config)
def batch_input(args, config=None): """ Batch Input Mode works to convert all of the article XML files in a specified directory into individual article EPUB files. Batch Input Mode is employed under a few simplifying assumptions: any pre-existing folder for article EPUB conversion will be eliminated without asking user permission, all output that except the .epub and .log files will be removed, and image files in a custom directory are not being used. Unlike the other input modes, Batch Input Mode output is always relative to the batch directory rather than the working directory of oaepub execution. Batch Input Mode has default epubcheck behavior, it will place a system call to epubcheck unless specified otherwise (--no-epubcheck or -N flags). """ if config is None: config = get_config_module() error_file = open('batch_tracebacks.txt', 'w') #Iterate over all listed files in the batch directory for item in os.listdir(args.batch): item_path = os.path.join(args.batch, item) #Skip directories and files without .xml extension _root, extension = os.path.splitext(item) if not os.path.isfile(item_path): continue if not extension == '.xml': continue print(item_path) #Parse the article try: raw_name = u_input.local_input(item_path) except: traceback.print_exc(file=error_file) else: parsed_article = Article(os.path.join(args.batch, raw_name+'.xml'), validation=args.no_dtd_validation) #Create the output name output_name = os.path.join(utils.get_output_directory(args), raw_name) #Make the EPUB try: make_epub(parsed_article, outdirect=output_name, explicit_images=None, # No explicit image path batch=True, config=config) except: error_file.write(item_path + '\n') traceback.print_exc(file=error_file) #Cleanup output directory, keeps EPUB and log shutil.rmtree(output_name) #Running epubcheck on the output verifies the validity of the ePub, #requires a local installation of java and epubcheck. if args.no_epubcheck: epubcheck('{0}.epub'.format(output_name), config) error_file.close()