Coverage for src/crawler/base_crawler.py: 69%

1import logging 

2import time 

3from concurrent.futures import ( 

4 Executor, 

5 ThreadPoolExecutor, 

6) 

7from datetime import datetime, timedelta 

8from email.policy import EmailPolicy 

9 

10import regex 

11import requests 

12from bs4 import BeautifulSoup 

13from django.conf import settings 

14from django.contrib.auth.models import User 

15from django.utils import timezone 

16from langcodes import standardize_tag 

17from lingua import LanguageDetector, LanguageDetectorBuilder 

18from opentelemetry import trace 

19from ptf.cmds.xml.ckeditor.utils import ( 

20 build_jats_data_from_html_field, 

21) 

22from ptf.cmds.xml.jats.builder.references import ( 

23 get_article_title_xml, 

24 get_author_xml, 

25 get_fpage_xml, 

26 get_lpage_xml, 

27 get_source_xml, 

28 get_year_xml, 

29) 

30from ptf.cmds.xml.jats.jats_parser import JatsBase 

31from ptf.model_data import ( 

32 ArticleData, 

33 ContributorDict, 

34 IssueData, 

35 ResourceData, 

36 TitleDict, 

37 create_abstract, 

38 create_contributor, 

39 create_extlink, 

40 create_issuedata, 

41 create_publisherdata, 

42 create_titledata, 

43) 

44from ptf.model_data_converter import update_data_for_jats 

45from pylatexenc.latex2text import LatexNodes2Text 

46from pymongo.errors import DocumentTooLarge 

47from pysolr import SolrError 

48from requests.adapters import HTTPAdapter 

49from requests_cache import CachedSession, MongoCache 

50from urllib3 import Retry 

51 

52from crawler.cmds.xml_cmds import addOrUpdateGDMLIssueXmlCmd 

53from crawler.models import Source 

54from crawler.types import CitationLiteral 

55from crawler.utils import add_pdf_link_to_xarticle, cleanup_str, get_or_create_collection 

56 

57 

58class CrawlerTitleDict(TitleDict): 

59 title_tex: str | None 

60 

61 

62class BaseCollectionCrawler: 

63 """ 

64 Base collection for the crawlers. 

65 To create a crawler: 

66 1) derive a class from BaseCollectionCrawler and name it XXXCrawler 

67 2) override the functions parse_collection_content, parse_issue_content and parse_article_content 

68 3) update factory.py so that crawler_factory can return your new crawler 

69 """ 

70 

71 logger = logging.getLogger(__name__) 

72 tracer = trace.get_tracer(__name__) 

73 

74 source_name = "" 

75 source_domain = "" 

76 source_website = "" 

77 

78 issue_href = "" 

79 

80 collection = None 

81 source = None 

82 user = None 

83 session: requests.Session | CachedSession 

84 

85 verify = True 

86 headers = { 

87 "accept_encoding": "utf-8", 

88 "User-Agent": getattr(settings, "REQUESTS_USER_AGENT", "Mathdoc/1.0.0"), 

89 "From": getattr(settings, "REQUESTS_EMAIL", "accueil@listes.mathdoc.fr"), 

90 } 

91 

92 next_allowed_request: float = time.time() 

93 

94 # seconds to wait between two http requests 

95 requests_interval = getattr(settings, "REQUESTS_INTERVAL", 90) 

96 # seconds to wait before aborting the connection (if no bytes are recieved) 

97 requests_timeout = 60 

98 

99 latext_parser = LatexNodes2Text() 

100 

101 # Override the values in your concrete crawler if the formulas in text (titles, abstracts) 

102 # do not use the "$" to surround tex formulas 

103 delimiter_inline_formula = "$" 

104 delimiter_disp_formula = "$" 

105 

106 # HACK : Workaround for tests (monkeypatching) 

107 # We store the class here, so we can monkeypatch it when running tests 

108 # subCrawlers = { 

109 # LofplCrawler: None 

110 # } 

111 subCrawlers: dict[type["BaseCollectionCrawler"], "BaseCollectionCrawler | None"] = {} 

112 

113 _language_detector: LanguageDetector | None = None 

114 _language_detector_builder = LanguageDetectorBuilder.from_all_languages() 

115 

116 force_refresh = False 

117 

118 # Whereas to include headers in requests cache key 

119 match_headers = False 

120 orcid_re = r"https\:\/\/orcid\.org\/(?P<orcid>\d{4}-\d{4}-\d{4}-\d{4})" 

121 

122 # Set this to False on a Crawler-basis to allow inserting articles without PDFs 

123 ignore_missing_pdf = True 

124 

125 database_executor: Executor 

126 

127 @classmethod 

128 def get_view_id(cls): 

129 return cls.source_domain 

130 

131 @property 

132 def language_detector(self): 

133 """Crawler Instance singleton for language builder. 

134 Late init of LanguageDetector to save on memory""" 

135 if not self._language_detector: 

136 self._language_detector = self._language_detector_builder.build() 

137 return self._language_detector 

138 

139 def __init__( 

140 self, 

141 *args, 

142 username: str, 

143 collection_id: str, 

144 collection_url: str, 

145 test_mode: bool = False, 

146 publisher: str = "mathdoc", 

147 force_refresh=False, 

148 ): 

149 for CrawlerClass in self.subCrawlers: 149 ↛ 150line 149 didn't jump to line 150 because the loop on line 149 never started

150 self.subCrawlers[CrawlerClass] = CrawlerClass( 

151 *args, 

152 username=username, 

153 collection_id=collection_id, 

154 collection_url=collection_url, 

155 test_mode=test_mode, 

156 publisher=publisher, 

157 ) 

158 self.logger = logging.getLogger(__name__ + "." + self.source_domain) 

159 

160 self.username = username 

161 

162 self.collection_id = collection_id 

163 self.collection_url = ( 

164 collection_url # url of the collection. Ex: https://eudml.org/journal/10098 

165 ) 

166 

167 self.test_mode = test_mode 

168 self.publisher = publisher 

169 

170 self.session = requests.session() 

171 

172 # Skipped when running tests 

173 self.initialize() 

174 self.session.verify = self.verify 

175 self.force_refresh = force_refresh 

176 

177 # We implemented custom retry behaviour, so we don't want to make extra requests here 

178 retries = Retry( 

179 total=0, 

180 ) 

181 self.session.mount("https://", HTTPAdapter(max_retries=retries)) 

182 self.session.mount("http://", HTTPAdapter(max_retries=retries)) 

183 

184 self.database_executor = ThreadPoolExecutor( 

185 max_workers=2, thread_name_prefix="crawler_database_thread" 

186 ) 

187 

188 def initialize(self): 

189 """ 

190 Acts as a "second" init function to skip model accesses during test data generation 

191 """ 

192 self.collection = get_or_create_collection(self.collection_id) 

193 self.source = self.get_or_create_source() 

194 self.user = User.objects.get(username=self.username) 

195 self.session = CachedSession( 

196 match_headers=self.match_headers, 

197 headers=self.headers, 

198 backend=MongoCache( 

199 host=getattr(settings, "MONGO_HOSTNAME", "localhost"), 

200 ), 

201 expire_after=timedelta(days=30), 

202 ) 

203 

204 @classmethod 

205 def can_crawl(cls, pid: str) -> bool: 

206 return True 

207 

208 def parse_collection_content(self, content: str) -> list[IssueData]: 

209 """ 

210 Parse the HTML content with BeautifulSoup 

211 returns a list of xissue. 

212 Override this function in a derived class 

213 """ 

214 return [] 

215 

216 def parse_issue_content(self, content: str, xissue: IssueData): 

217 """ 

218 Parse the HTML content with BeautifulSoup 

219 Fills the xissue.articles 

220 Override this function in a derived class. 

221 

222 CAV : You are supposed to create articles there. Please assign a PID to each article. 

223 The PID can be `a + article_index`, like this : `a0` `a21` 

224 """ 

225 

226 def parse_article_content( 

227 self, content: str, xissue: IssueData, xarticle: ArticleData, url: str 

228 ) -> ArticleData | None: 

229 """ 

230 Parse the HTML content with BeautifulSoup 

231 returns the xarticle. 

232 Override this function in a derived class. 

233 The xissue is passed to the function in case the article page has issue information (ex: publisher) 

234 The article url is also passed as a parameter 

235 

236 CAV : You are supposed to assign articles pid again here 

237 """ 

238 return xarticle 

239 

240 @tracer.start_as_current_span("crawl_collection") 

241 def crawl_collection(self): 

242 # TODO: Comments, filter 

243 """ 

244 Crawl an entire collection. ptf.models.Container objects are created. 

245 - get the HTML content of the collection_url 

246 - parse the HTML content with beautifulsoup to extract the list of issues 

247 - merge the xissues (some Source can have multiple pages for 1 volume/issue. We create only 1 container) 

248 - crawl each issue if col_only is False 

249 - Returns the list of merged issues. 

250 It is an OrderedDict {pid: {"issues": xissues}} 

251 The key is the pid of the merged issues. 

252 Ex: The source may have Ex: Volume 6 (2000) and Volume 6 (1999) 

253 the pid is then made with 1999-2000__6_ 

254 """ 

255 

256 if self.source is None: 

257 raise RuntimeError("ERROR: the source is not set") 

258 

259 content = self.download_file(self.collection_url) 

260 xissues = self.parse_collection_content(content) 

261 

262 """ 

263 Some collections split the same volumes in different pages 

264 Ex: Volume 6 (2000) and Volume 6 (1999) 

265 We merge the 2 xissues with the same volume number => Volume 6 (1999-2000) 

266 """ 

267 # merged_xissues = self.merge_xissues(xissues) 

268 

269 xissues_dict = {str(i.pid): i for i in xissues} 

270 

271 return xissues_dict 

272 

273 @tracer.start_as_current_span("crawl_issue") 

274 def crawl_issue(self, xissue: IssueData): 

275 """ 

276 Crawl 1 wag page of an issue. 

277 - get the HTML content of the issue 

278 - parse the HTML content with beautifulsoup to extract the list of articles and/or the issue metadata 

279 - crawl each article 

280 """ 

281 

282 # Some source, like EuDML do not have a separate HTML pages for an issue's table of content. 

283 # The list of articles directly come from the collection HTML page: the xissue has no url attribute 

284 

285 issue_url = xissue.url 

286 if issue_url is not None: 

287 if issue_url.endswith(".pdf"): 

288 add_pdf_link_to_xarticle(xissue, issue_url) 

289 xissue.url = None 

290 else: 

291 content = self.download_file(issue_url) 

292 with self.tracer.start_as_current_span("parse_issue_content"): 

293 self.parse_issue_content(content, xissue) 

294 

295 xarticles = xissue.articles 

296 

297 parsed_xarticles = [] 

298 

299 for xarticle in xarticles: 

300 parsed_xarticle = self.crawl_article(xarticle, xissue) 

301 if parsed_xarticle is not None: 

302 parsed_xarticles.append(parsed_xarticle) 

303 

304 xissue.articles = parsed_xarticles 

305 

306 article_has_pdf = self.article_has_pdf(xissue) 

307 

308 if self.ignore_missing_pdf: 

309 xissue.articles = [a for a in xissue.articles if self.article_has_pdf(a)] 

310 

311 if not self.test_mode and (len(xissue.articles) > 0 or article_has_pdf): 

312 self.process_resource_metadata(xissue, resource_type="issue") 

313 self.database_executor.submit(self.add_xissue_into_database, xissue) 

314 

315 @staticmethod 

316 def article_has_source(art: ArticleData | IssueData): 

317 return ( 

318 next( 

319 (e_link for e_link in art.ext_links if e_link["rel"] == "source"), 

320 None, 

321 ) 

322 is not None 

323 ) 

324 

325 @staticmethod 

326 def article_has_pdf(art: ArticleData | IssueData): 

327 return ( 

328 next((link for link in art.ext_links if link["rel"] == "article-pdf"), None) 

329 is not None 

330 ) 

331 

332 def crawl_article(self, xarticle: ArticleData, xissue: IssueData): 

333 # ARTICLE URL as en ExtLink (to display the link in the article page) 

334 if xarticle.url is None: 

335 if not self.article_has_source(xarticle): 335 ↛ 345line 335 didn't jump to line 345 because the condition on line 335 was always true

336 if xissue.url: 

337 article_source = xissue.url 

338 else: 

339 article_source = self.collection_url 

340 ext_link = create_extlink() 

341 ext_link["rel"] = "source" 

342 ext_link["location"] = article_source 

343 ext_link["metadata"] = self.source_domain 

344 xarticle.ext_links.append(ext_link) 

345 return self.process_article_metadata(xarticle) 

346 

347 content = self.download_file(xarticle.url) 

348 xarticle.pid = f"{xissue.pid}_{xarticle.pid}" 

349 

350 try: 

351 with self.tracer.start_as_current_span("parse_article_content"): 

352 parsed_xarticle = self.parse_article_content( 

353 content, xissue, xarticle, xarticle.url 

354 ) 

355 except ValueError as e: 

356 self.logger.warning(e) 

357 self.logger.warning("Retrying in 5 mins while invalidating cache") 

358 time.sleep(5 * 60) 

359 content = self.download_file(xarticle.url, force_refresh=True) 

360 with self.tracer.start_as_current_span("parse_article_content"): 

361 parsed_xarticle = self.parse_article_content( 

362 content, xissue, xarticle, xarticle.url 

363 ) 

364 

365 if parsed_xarticle is None: 365 ↛ 366line 365 didn't jump to line 366 because the condition on line 365 was never true

366 return None 

367 

368 if parsed_xarticle.doi: 

369 parsed_xarticle.pid = ( 

370 parsed_xarticle.doi.replace("/", "_").replace(".", "_").replace("-", "_") 

371 ) 

372 

373 if not self.article_has_source(parsed_xarticle) and parsed_xarticle.url: 

374 ext_link = create_extlink() 

375 ext_link["rel"] = "source" 

376 ext_link["location"] = parsed_xarticle.url 

377 ext_link["metadata"] = self.source_domain 

378 parsed_xarticle.ext_links.append(ext_link) 

379 

380 # The article title may have formulas surrounded with '$' 

381 return self.process_article_metadata(parsed_xarticle) 

382 

383 def process_resource_metadata(self, xresource: ResourceData, resource_type="article"): 

384 tag = "article-title" if resource_type == "article" else "issue-title" 

385 

386 # Process title tex 

387 ckeditor_data = build_jats_data_from_html_field( 

388 xresource.title_tex, 

389 tag=tag, 

390 text_lang=xresource.lang, 

391 delimiter_inline=self.delimiter_inline_formula, 

392 delimiter_disp=self.delimiter_disp_formula, 

393 ) 

394 

395 xresource.title_html = ckeditor_data["value_html"] 

396 # xresource.title_tex = ckeditor_data["value_tex"] 

397 xresource.title_xml = ckeditor_data["value_xml"] 

398 

399 # Process trans_title tex 

400 if xresource.trans_title_tex: 400 ↛ 401line 400 didn't jump to line 401 because the condition on line 400 was never true

401 self.logger.warning( 

402 "Deprecation Notice : prefer using xresource.title directly instead of xresource.trans_title_tex" 

403 ) 

404 trans_title = self.create_trans_title( 

405 xresource_lang=xresource.lang, 

406 resource_type=resource_type, 

407 title_tex=xresource.trans_title_tex, 

408 lang=xresource.trans_lang, 

409 ) 

410 xresource.titles.append(trans_title) 

411 

412 abstracts_to_parse = [ 

413 xabstract for xabstract in xresource.abstracts if xabstract["tag"] == "abstract" 

414 ] 

415 # abstract may have formulas surrounded with '$' 

416 if len(abstracts_to_parse) > 0: 

417 for xabstract in abstracts_to_parse: 

418 ckeditor_data = build_jats_data_from_html_field( 

419 xabstract["value_tex"], 

420 tag="abstract", 

421 text_lang=xabstract["lang"], 

422 resource_lang=xresource.lang, 

423 field_type="abstract", 

424 delimiter_inline=self.delimiter_inline_formula, 

425 delimiter_disp=self.delimiter_disp_formula, 

426 ) 

427 

428 xabstract["value_html"] = ckeditor_data["value_html"] 

429 # xabstract["value_tex"] = ckeditor_data["value_tex"] 

430 xabstract["value_xml"] = ckeditor_data["value_xml"] 

431 

432 return xresource 

433 

434 def process_article_metadata(self, xarticle: ArticleData): 

435 self.process_resource_metadata(xarticle) 

436 for bibitem in xarticle.bibitems: 

437 bibitem.type = "unknown" 

438 update_data_for_jats(xarticle, with_label=False) 

439 

440 return xarticle 

441 

442 def _wait_download_delay(self, url: str, force_refresh=False): 

443 # If we already have a key, we can skip the timeout 

444 if isinstance(self.session, CachedSession): 

445 if self.session.cache.contains(url=url) and not force_refresh: 

446 return 

447 

448 delta = self.next_allowed_request - time.time() 

449 if delta > 0: 

450 self.logger.info(f"Waiting {int(delta)}s before making another request") 

451 time.sleep(delta) 

452 self.next_allowed_request = time.time() + self.requests_interval 

453 

454 def _get(self, url: str, force_refresh=False, headers={}) -> requests.Response: 

455 """ 

456 Wrapper around requests.get with delay based on the crawler class instance 

457 """ 

458 

459 self._wait_download_delay(url, force_refresh) 

460 

461 kwargs = {} 

462 # self.session.cache.delete(urls=[url]) 

463 if isinstance(self.session, CachedSession): 

464 kwargs["force_refresh"] = force_refresh 

465 

466 try: 

467 response = self.session.get( 

468 url, 

469 headers={**self.headers, **headers}, 

470 timeout=self.requests_timeout, 

471 **kwargs, 

472 ) 

473 except DocumentTooLarge as e: 

474 self.logger.error(e) 

475 response = requests.get( 

476 url, headers={**self.headers, **headers}, timeout=self.requests_timeout 

477 ) 

478 

479 if not response.ok: 

480 raise requests.exceptions.HTTPError( 

481 f"Endpoint answered with code {response.status_code} : {url}", 

482 response=response, 

483 ) 

484 

485 return response 

486 

487 def download_file(self, url: str, force_refresh=False, headers={}): 

488 """ 

489 Downloads a page and returns its content (decoded string). 

490 This function handles retries and decoding 

491 """ 

492 attempts = 0 

493 while True: 

494 try: 

495 if attempts > 0: 

496 force_refresh = True 

497 response = self._get( 

498 url, force_refresh=force_refresh or self.force_refresh, headers=headers 

499 ) 

500 

501 if getattr(response, "from_cache", False): 

502 return response.text 

503 

504 content = self.decode_response(response) 

505 if content == "" or not content: 

506 raise requests.exceptions.HTTPError(response) 

507 

508 if isinstance(self.session, CachedSession): 

509 if "Expires" in response.headers: 

510 del response.headers["Expires"] 

511 del response.headers["Cache-Control"] 

512 try: 

513 self.session.cache.save_response(response) 

514 except DocumentTooLarge as e: 

515 self.logger.warning(e) 

516 return content 

517 except ( 

518 requests.ConnectionError, 

519 requests.ConnectTimeout, 

520 requests.exceptions.HTTPError, 

521 ) as e: 

522 if attempts > 3: 

523 raise e 

524 self.logger.debug(f"Caught error : {e}", extra={"url": url}) 

525 attempts += 1 

526 # 15 mins, 30 mins, 45 mins 

527 delay_minutes = attempts * 15 

528 self.logger.debug( 

529 f"Retrying in {delay_minutes}mins ({(datetime.now() + timedelta(minutes=delay_minutes)).time()})", 

530 extra={"url": url}, 

531 ) 

532 time.sleep(delay_minutes * 60) 

533 

534 def decode_response(self, response: requests.Response, encoding: str | None = None): 

535 """Override this if the content-type headers from the sources are advertising something else than the actual content 

536 SASA needs this""" 

537 # Force 

538 if encoding: 

539 response.encoding = encoding 

540 return response.text 

541 

542 # Attempt to get encoding using HTTP headers 

543 content_type_tag = response.headers.get("Content-Type", None) 

544 

545 if content_type_tag: 545 ↛ 552line 545 didn't jump to line 552 because the condition on line 545 was always true

546 charset = self.parse_content_type_charset(content_type_tag) 

547 if charset: 547 ↛ 548line 547 didn't jump to line 548 because the condition on line 547 was never true

548 response.encoding = charset 

549 return response.text 

550 

551 # Attempt to get encoding using HTML meta charset tag 

552 soup = BeautifulSoup(response.text, "html5lib") 

553 charset = soup.select_one("meta[charset]") 

554 if charset: 

555 htmlencoding = charset.get("charset") 

556 if isinstance(htmlencoding, str): 556 ↛ 561line 556 didn't jump to line 561 because the condition on line 556 was always true

557 response.encoding = htmlencoding 

558 return response.text 

559 

560 # Attempt to get encoding using HTML meta content type tag 

561 content_type_tag = soup.select_one('meta[http-equiv="Content-Type"]') 

562 if content_type_tag: 

563 content_type = content_type_tag.get("content") 

564 if isinstance(content_type, str): 564 ↛ 570line 564 didn't jump to line 570 because the condition on line 564 was always true

565 charset = self.parse_content_type_charset(content_type) 

566 if charset: 566 ↛ 570line 566 didn't jump to line 570 because the condition on line 566 was always true

567 response.encoding = charset 

568 return response.text 

569 

570 return response.text 

571 

572 @staticmethod 

573 def parse_content_type_charset(content_type: str): 

574 header = EmailPolicy.header_factory("content-type", content_type) 

575 if "charset" in header.params: 

576 return header.params.get("charset") 

577 

578 @tracer.start_as_current_span("add_xissue_to_database") 

579 def add_xissue_into_database(self, xissue: IssueData): 

580 xissue.journal = self.collection 

581 xissue.source = self.source_domain 

582 

583 if xissue.year == "": 

584 raise ValueError("Failsafe : Cannot insert issue without a year") 

585 

586 xpub = create_publisherdata() 

587 xpub.name = self.publisher 

588 xissue.publisher = xpub 

589 xissue.last_modified_iso_8601_date_str = timezone.now().isoformat() 

590 

591 attempt = 1 

592 success = False 

593 

594 while not success and attempt < 4: 

595 try: 

596 params = {"xissue": xissue, "use_body": False} 

597 cmd = addOrUpdateGDMLIssueXmlCmd(params) 

598 cmd.do() 

599 success = True 

600 self.logger.debug(f"Issue {xissue.pid} inserted in database") 

601 except SolrError: 

602 self.logger.warning( 

603 f"Encoutered SolrError while inserting issue {xissue.pid} in database" 

604 ) 

605 attempt += 1 

606 self.logger.debug(f"Attempt {attempt}. sleeping 10 seconds.") 

607 time.sleep(10) 

608 

609 if success is False: 

610 raise ConnectionRefusedError("Cannot connect to SolR") 

611 

612 def get_metadata_using_citation_meta( 

613 self, 

614 xarticle: ArticleData, 

615 xissue: IssueData, 

616 soup: BeautifulSoup, 

617 what: list[CitationLiteral] = [], 

618 ): 

619 """ 

620 :param xarticle: the xarticle that will collect the metadata 

621 :param xissue: the xissue that will collect the publisher 

622 :param soup: the BeautifulSoup object of tha article page 

623 :param what: list of citation_ items to collect. 

624 :return: None. The given article is modified 

625 """ 

626 

627 if "title" in what: 

628 # TITLE 

629 citation_title_node = soup.select_one("meta[name='citation_title']") 

630 if citation_title_node: 630 ↛ 635line 630 didn't jump to line 635 because the condition on line 630 was always true

631 title = citation_title_node.get("content") 

632 if isinstance(title, str): 632 ↛ 635line 632 didn't jump to line 635 because the condition on line 632 was always true

633 xarticle.title_tex = title 

634 

635 if "author" in what: 635 ↛ 664line 635 didn't jump to line 664 because the condition on line 635 was always true

636 # AUTHORS 

637 citation_author_nodes = soup.select("meta[name^='citation_author']") 

638 current_author: ContributorDict | None = None 

639 for citation_author_node in citation_author_nodes: 

640 if citation_author_node.get("name") == "citation_author": 

641 text_author = citation_author_node.get("content") 

642 if not isinstance(text_author, str): 642 ↛ 643line 642 didn't jump to line 643 because the condition on line 642 was never true

643 raise ValueError("Cannot parse author") 

644 if text_author == "": 644 ↛ 645line 644 didn't jump to line 645 because the condition on line 644 was never true

645 current_author = None 

646 continue 

647 current_author = create_contributor(role="author", string_name=text_author) 

648 xarticle.contributors.append(current_author) 

649 continue 

650 if current_author is None: 650 ↛ 651line 650 didn't jump to line 651 because the condition on line 650 was never true

651 self.logger.warning("Couldn't parse citation author") 

652 continue 

653 if citation_author_node.get("name") == "citation_author_institution": 

654 text_institution = citation_author_node.get("content") 

655 if not isinstance(text_institution, str): 655 ↛ 656line 655 didn't jump to line 656 because the condition on line 655 was never true

656 continue 

657 current_author["addresses"].append(text_institution) 

658 if citation_author_node.get("name") == "citation_author_ocrid": 658 ↛ 659line 658 didn't jump to line 659 because the condition on line 658 was never true

659 text_orcid = citation_author_node.get("content") 

660 if not isinstance(text_orcid, str): 

661 continue 

662 current_author["orcid"] = text_orcid 

663 

664 if "pdf" in what: 

665 # PDF 

666 citation_pdf_node = soup.select_one('meta[name="citation_pdf_url"]') 

667 if citation_pdf_node: 

668 pdf_url = citation_pdf_node.get("content") 

669 if isinstance(pdf_url, str): 669 ↛ 672line 669 didn't jump to line 672 because the condition on line 669 was always true

670 add_pdf_link_to_xarticle(xarticle, pdf_url) 

671 

672 if "lang" in what: 

673 # LANG 

674 citation_lang_node = soup.select_one("meta[name='citation_language']") 

675 if citation_lang_node: 675 ↛ 681line 675 didn't jump to line 681 because the condition on line 675 was always true

676 # TODO: check other language code 

677 content_text = citation_lang_node.get("content") 

678 if isinstance(content_text, str): 678 ↛ 681line 678 didn't jump to line 681 because the condition on line 678 was always true

679 xarticle.lang = standardize_tag(content_text) 

680 

681 if "abstract" in what: 

682 # ABSTRACT 

683 abstract_node = soup.select_one("meta[name='citation_abstract']") 

684 if abstract_node is not None: 

685 abstract = abstract_node.get("content") 

686 if not isinstance(abstract, str): 686 ↛ 687line 686 didn't jump to line 687 because the condition on line 686 was never true

687 raise ValueError("Couldn't parse abstract from meta") 

688 abstract = BeautifulSoup(abstract, "html.parser").text 

689 lang = abstract_node.get("lang") 

690 if not isinstance(lang, str): 

691 lang = self.detect_language(abstract, xarticle) 

692 xarticle.abstracts.append(create_abstract(lang=lang, value_tex=abstract)) 

693 

694 if "page" in what: 

695 # PAGES 

696 citation_fpage_node = soup.select_one("meta[name='citation_firstpage']") 

697 if citation_fpage_node: 

698 page = citation_fpage_node.get("content") 

699 if isinstance(page, str): 699 ↛ 704line 699 didn't jump to line 704 because the condition on line 699 was always true

700 page = page.split("(")[0] 

701 if len(page) < 32: 701 ↛ 704line 701 didn't jump to line 704 because the condition on line 701 was always true

702 xarticle.fpage = page 

703 

704 citation_lpage_node = soup.select_one("meta[name='citation_lastpage']") 

705 if citation_lpage_node: 

706 page = citation_lpage_node.get("content") 

707 if isinstance(page, str): 707 ↛ 712line 707 didn't jump to line 712 because the condition on line 707 was always true

708 page = page.split("(")[0] 

709 if len(page) < 32: 709 ↛ 712line 709 didn't jump to line 712 because the condition on line 709 was always true

710 xarticle.lpage = page 

711 

712 if "doi" in what: 

713 # DOI 

714 citation_doi_node = soup.select_one("meta[name='citation_doi']") 

715 if citation_doi_node: 

716 doi = citation_doi_node.get("content") 

717 if isinstance(doi, str): 717 ↛ 724line 717 didn't jump to line 724 because the condition on line 717 was always true

718 doi = doi.strip() 

719 pos = doi.find("10.") 

720 if pos > 0: 

721 doi = doi[pos:] 

722 xarticle.doi = doi 

723 

724 if "mr" in what: 

725 # MR 

726 citation_mr_node = soup.select_one("meta[name='citation_mr']") 

727 if citation_mr_node: 

728 mr = citation_mr_node.get("content") 

729 if isinstance(mr, str): 729 ↛ 735line 729 didn't jump to line 735 because the condition on line 729 was always true

730 mr = mr.strip() 

731 if mr.find("MR") == 0: 731 ↛ 735line 731 didn't jump to line 735 because the condition on line 731 was always true

732 mr = mr[2:] 

733 xarticle.extids.append(("mr-item-id", mr)) 

734 

735 if "zbl" in what: 

736 # ZBL 

737 citation_zbl_node = soup.select_one("meta[name='citation_zbl']") 

738 if citation_zbl_node: 

739 zbl = citation_zbl_node.get("content") 

740 if isinstance(zbl, str): 740 ↛ 746line 740 didn't jump to line 746 because the condition on line 740 was always true

741 zbl = zbl.strip() 

742 if zbl.find("Zbl") == 0: 742 ↛ 746line 742 didn't jump to line 746 because the condition on line 742 was always true

743 zbl = zbl[3:].strip() 

744 xarticle.extids.append(("zbl-item-id", zbl)) 

745 

746 if "publisher" in what: 

747 # PUBLISHER 

748 citation_publisher_node = soup.select_one("meta[name='citation_publisher']") 

749 if citation_publisher_node: 

750 pub = citation_publisher_node.get("content") 

751 if isinstance(pub, str): 751 ↛ 758line 751 didn't jump to line 758 because the condition on line 751 was always true

752 pub = pub.strip() 

753 if pub != "": 753 ↛ 758line 753 didn't jump to line 758 because the condition on line 753 was always true

754 xpub = create_publisherdata() 

755 xpub.name = pub 

756 xissue.publisher = xpub 

757 

758 if "keywords" in what: 

759 # KEYWORDS 

760 citation_kwd_nodes = soup.select("meta[name='citation_keywords']") 

761 for kwd_node in citation_kwd_nodes: 

762 kwds = kwd_node.get("content") 

763 if isinstance(kwds, str): 763 ↛ 761line 763 didn't jump to line 761 because the condition on line 763 was always true

764 kwds = kwds.split(",") 

765 for kwd in kwds: 

766 if kwd == "": 

767 continue 

768 kwd = kwd.strip() 

769 xarticle.kwds.append({"type": "", "lang": xarticle.lang, "value": kwd}) 

770 

771 if "references" in what: 

772 citation_references = soup.select("meta[name='citation_reference']") 

773 for index, tag in enumerate(citation_references): 

774 content = tag.get("content") 

775 if not isinstance(content, str): 775 ↛ 776line 775 didn't jump to line 776 because the condition on line 775 was never true

776 raise ValueError("Cannot parse citation_reference meta") 

777 label = str(index + 1) 

778 if regex.match(r"^\[\d+\].*", content): 778 ↛ 779line 778 didn't jump to line 779 because the condition on line 778 was never true

779 label = None 

780 xarticle.bibitems.append(self.__parse_meta_citation_reference(content, label)) 

781 

782 def create_xissue( 

783 self, 

784 url: str | None, 

785 year: str, 

786 volume_number: str | None, 

787 issue_number: str | None = "1", 

788 vseries: str | None = None, 

789 ): 

790 if url is not None and url.endswith("/"): 

791 url = url[:-1] 

792 xissue = create_issuedata() 

793 xissue.url = url 

794 

795 xissue.pid = self.get_issue_pid( 

796 self.collection_id, year, volume_number, issue_number, vseries 

797 ) 

798 

799 xissue.year = year 

800 

801 if volume_number is not None: 

802 xissue.volume = regex.sub(r"[^a-zA-Z0-9-]+", "_", volume_number) 

803 

804 if issue_number is not None: 

805 xissue.number = issue_number.replace(",", "-") 

806 

807 if vseries is not None: 807 ↛ 808line 807 didn't jump to line 808 because the condition on line 807 was never true

808 xissue.vseries = vseries 

809 return xissue 

810 

811 def detect_language(self, text: str, article: ArticleData | None = None): 

812 if article and article.lang is not None and article.lang != "und": 

813 return article.lang 

814 

815 language = self.language_detector.detect_language_of(text) 

816 

817 if not language: 817 ↛ 818line 817 didn't jump to line 818 because the condition on line 817 was never true

818 return "und" 

819 return language.iso_code_639_1.name.lower() 

820 

821 def create_trans_title( 

822 self, 

823 resource_type: str, 

824 title_tex: str, 

825 lang: str, 

826 xresource_lang: str, 

827 title_type: str = "main", 

828 ):