def repackage_archive_zip_to_pmc_zip(self, doi_id):
"repackage the zip file in the TMP_DIR to a PMC zip format"
# unzip contents
zip_input_dir = os.path.join(self.get_tmp_dir(), self.TMP_DIR)
zip_extracted_dir = os.path.join(self.get_tmp_dir(), self.JUNK_DIR)
zip_renamed_files_dir = os.path.join(self.get_tmp_dir(), self.RENAME_DIR)
pmc_zip_output_dir = os.path.join(self.get_tmp_dir(), self.INPUT_DIR)
archive_zip_name = glob.glob(zip_input_dir + "/*.zip")[0]
with zipfile.ZipFile(archive_zip_name, 'r') as myzip:
myzip.extractall(zip_extracted_dir)
# rename the files and profile the files
file_name_map = article_processing.rename_files_remove_version_number(
files_dir = zip_extracted_dir,
output_dir = zip_renamed_files_dir
)
if self.logger:
self.logger.info("FTPArticle running %s workflow for article %s, file_name_map"
% (self.workflow, self.doi_id))
self.logger.info(file_name_map)
# convert the XML
article_xml_file = glob.glob(zip_renamed_files_dir + "/*.xml")[0]
article_processing.convert_xml(xml_file=article_xml_file,
file_name_map=file_name_map)
# rezip the files into PMC zip format
soup = parser.parse_document(article_xml_file)
volume = parser.volume(soup)
pmc_zip_file_name = article_processing.new_pmc_zip_filename(self.journal, volume, doi_id)
with zipfile.ZipFile(os.path.join(pmc_zip_output_dir, pmc_zip_file_name), 'w',
zipfile.ZIP_DEFLATED, allowZip64=True) as new_zipfile:
dirfiles = article_processing.file_list(zip_renamed_files_dir)
for df in dirfiles:
filename = df.split(os.sep)[-1]
new_zipfile.write(df, filename)
return True
def profile_article(self, document):
"""
Temporary, profile the article by folder names in test data set
In real code we still want this to return the same values
"""
# Temporary setting of version values from directory names
soup = self.article_soup(self.article_xml_file())
# elife id / doi id / manuscript id
fid = parser.doi(soup).split('.')[-1]
# article status
if parser.is_poa(soup) is True:
status = 'poa'
else:
status = 'vor'
# version
version = self.version_number(document)
# volume
volume = parser.volume(soup)
return (fid, status, version, volume)
def profile_article(self, document):
"""
Temporary, profile the article by folder names in test data set
In real code we still want this to return the same values
"""
# Temporary setting of version values from directory names
soup = self.article_soup(self.article_xml_file())
# elife id / doi id / manuscript id
fid = parser.doi(soup).split('.')[-1]
# article status
if parser.is_poa(soup) is True:
status = 'poa'
else:
status = 'vor'
# version
version = self.version_number(document)
# volume
volume = parser.volume(soup)
return (fid, status, version, volume)
def convert_xml(self, doi_id, xml_file, filenames, new_filenames):
# Register namespaces
xmlio.register_xmlns()
root, doctype_dict = xmlio.parse(xml_file, return_doctype_dict=True)
soup = self.article_soup(xml_file)
if parser.is_poa(soup):
# Capitalise subject group values in article categories
root = self.subject_group_convert_in_xml(root)
pub_date = None
if parser.pub_date(soup) is None:
# add the published date to the XML
pub_date = self.get_pub_date_if_missing(doi_id)
root = self.add_pub_date_to_xml(doi_id, pub_date, root)
else:
pub_date = parser.pub_date(soup)
if parser.volume(soup) is None:
# Get the pub-date year to calculate the volume
year = pub_date[0]
volume = year - 2011
self.add_volume_to_xml(doi_id, volume, root)
# set the article-id, to overwrite the v2, v3 value if present
root = self.set_article_id_xml(doi_id, root)
# if pdf file then add self-uri tag
if parser.self_uri(soup) is not None and len(
parser.self_uri(soup)) == 0:
for filename in new_filenames:
if filename.endswith('.pdf'):
root = self.add_self_uri_to_xml(doi_id, filename, root)
# if ds.zip file is there, then add it to the xml
poa_ds_zip_file = None
for f in new_filenames:
if f.endswith('.zip'):
poa_ds_zip_file = f
if poa_ds_zip_file:
root = self.add_poa_ds_zip_to_xml(doi_id, poa_ds_zip_file,
root)
# Start the file output
reparsed_string = xmlio.output(root,
type=None,
doctype_dict=doctype_dict)
# Remove extra whitespace here for PoA articles to clean up and one VoR file too
reparsed_string = reparsed_string.replace("\n", '').replace("\t", '')
f = open(xml_file, 'wb')
f.write(reparsed_string)
f.close()
def convert_xml(self, doi_id, xml_file, filenames, new_filenames):
# Register namespaces
xmlio.register_xmlns()
root, doctype_dict = xmlio.parse(xml_file, return_doctype_dict=True)
soup = self.article_soup(xml_file)
if parser.is_poa(soup):
# Capitalise subject group values in article categories
root = self.subject_group_convert_in_xml(root)
pub_date = None
if parser.pub_date(soup) is None:
# add the published date to the XML
pub_date = self.get_pub_date_if_missing(doi_id)
root = self.add_pub_date_to_xml(doi_id, pub_date, root)
else:
pub_date = parser.pub_date(soup)
if parser.volume(soup) is None:
# Get the pub-date year to calculate the volume
year = pub_date[0]
volume = year - 2011
self.add_volume_to_xml(doi_id, volume, root)
# set the article-id, to overwrite the v2, v3 value if present
root = self.set_article_id_xml(doi_id, root)
# if pdf file then add self-uri tag
if parser.self_uri(soup) is not None and len(parser.self_uri(soup)) == 0:
for filename in new_filenames:
if filename.endswith('.pdf'):
root = self.add_self_uri_to_xml(doi_id, filename, root)
# if ds.zip file is there, then add it to the xml
poa_ds_zip_file = None
for f in new_filenames:
if f.endswith('.zip'):
poa_ds_zip_file = f
if poa_ds_zip_file:
root = self.add_poa_ds_zip_to_xml(doi_id, poa_ds_zip_file, root)
# Start the file output
reparsed_string = xmlio.output(root, type=None, doctype_dict=doctype_dict)
# Remove extra whitespace here for PoA articles to clean up and one VoR file too
reparsed_string = reparsed_string.replace("\n", '').replace("\t", '')
f = open(xml_file, 'wb')
f.write(reparsed_string)
f.close()
def build_article_from_xml(article_xml_filename, detail="brief"):
"""
Parse JATS XML with elifetools parser, and populate an
eLifePOA article object
Basic data crossref needs: article_id, doi, title, contributors with names set
detail="brief" is normally enough,
detail="full" will populate all the contributor affiliations that are linked by xref tags
"""
error_count = 0
soup = parser.parse_document(article_xml_filename)
# Get DOI
doi = parser.doi(soup)
# Create the article object
article = eLifePOA(doi, title=None)
# Related articles
article.related_articles = build_related_articles(parser.related_article(soup))
# Get publisher_id and set object manuscript value
publisher_id = parser.publisher_id(soup)
article.manuscript = publisher_id
# Set the articleType
article_type = parser.article_type(soup)
if article_type:
article.articleType = article_type
# title
article.title = parser.full_title(soup)
#print article.title
# abstract
article.abstract = clean_abstract(parser.full_abstract(soup))
# digest
article.digest = clean_abstract(parser.full_digest(soup))
# elocation-id
article.elocation_id = parser.elocation_id(soup)
# contributors
all_contributors = parser.contributors(soup, detail)
author_contributors = filter(lambda con: con.get('type')
in ['author', 'on-behalf-of'], all_contributors)
contrib_type = "author"
contributors = build_contributors(author_contributors, contrib_type)
contrib_type = "author non-byline"
authors = parser.authors_non_byline(soup, detail)
contributors_non_byline = build_contributors(authors, contrib_type)
article.contributors = contributors + contributors_non_byline
# license href
license = eLifeLicense()
license.href = parser.license_url(soup)
article.license = license
# article_category
article.article_categories = parser.category(soup)
# keywords
article.author_keywords = parser.keywords(soup)
# research organisms
article.research_organisms = parser.research_organism(soup)
# funding awards
article.funding_awards = build_funding(parser.full_award_groups(soup))
# references or citations
article.ref_list = build_ref_list(parser.refs(soup))
# components with component DOI
article.component_list = build_components(parser.components(soup))
# History dates
date_types = ["received", "accepted"]
for date_type in date_types:
history_date = parser.history_date(soup, date_type)
if history_date:
date_instance = eLifeDate(date_type, history_date)
article.add_date(date_instance)
# Pub date
pub_date = parser.pub_date(soup)
if pub_date:
date_instance = eLifeDate("pub", pub_date)
article.add_date(date_instance)
# Set the volume if present
volume = parser.volume(soup)
if volume:
article.volume = volume
article.is_poa = parser.is_poa(soup)
return article, error_count