add_clinvar_data.py 文件源码

def _download_latest_clinvar_xml(self, dest_dir):
        ftp = FTP('ftp.ncbi.nlm.nih.gov')
        ftp.login()
        ftp.cwd(CV_XML_DIR)
        # sort just in case the ftp lists the files in random order
        cv_xml_w_date = sorted(
            [f for f in ftp.nlst() if re.match(CV_XML_REGEX, f)])
        if len(cv_xml_w_date) == 0:
            raise CommandError('ClinVar reporting zero XML matching' +
                               ' regex: \'{0}\' in directory {1}'.format(
                                   CV_XML_REGEX, CV_XML_DIR))
        ftp_xml_filename = cv_xml_w_date[-1]
        dest_filepath = os.path.join(dest_dir, ftp_xml_filename)
        with open(dest_filepath, 'w') as fh:
            ftp.retrbinary('RETR {0}'.format(ftp_xml_filename), fh.write)
        return dest_filepath, ftp_xml_filename