def remove_duplicates(by='pmid'):
"""Remove duplicate articles by field.
:param str by: Article field to identify duplicates
"""
counts = collections.defaultdict(int)
values = [
value[by]
for value in Article._storage[0].store.find({}, {by: 1})
]
for value in values:
counts[value] += 1
for value, count in counts.items():
if count == 1:
continue
articles = list(Article.find(Q(by, 'eq', value)))
for duplicate in articles[1:]:
logger.debug(
'Deleting duplicate record: {}'.format(
value
)
)
Article.remove_one(duplicate)
def add_missing(query, max_count, randomize=False):
"""Search PubMed for articles and scrape documents.
:param str query: PubMed query
:param int max_count: Maximum number of articles to process
:param bool randomize: Randomize list of articles to fetch
:return: Added article objects
"""
pmids = pubtools.search_pmids(query)
stored_pmids = [
article['pmid'] for article in mongo['article'].find({}, {'pmid': 1})
]
missing_pmids = set(pmids) - set(stored_pmids)
logger.warn('Found {0} articles to add.'.format(len(missing_pmids)))
pmids_to_add = list(missing_pmids)[:max_count]
if randomize:
random.shuffle(pmids_to_add)
records = pubtools.download_pmids(pmids_to_add)
scraper = SCRAPE_CLASS(**SCRAPE_KWARGS)
added = []
for pmid, record in zip(pmids_to_add, records):
logger.debug('Adding article {}'.format(pmid))
article = Article.from_record(record)
article.scrape(scraper)
article.tag()
added.append(article)
return added
def rescrape(article_id, **kwargs):
logger.info('Re-scraping article {0}'.format(article_id))
try:
article = Article.load(article_id)
article.scrape(**kwargs)
except Exception as error:
logger.error('Error scraping article {0}'.format(article_id))
logger.exception(error)
def retag(article_id, **kwargs):
logger.info('Re-tagging article {0}'.format(article_id))
try:
article = Article.load(article_id)
article.tag(**kwargs)
except Exception as error:
logger.error('Error tagging article {0}'.format(article_id))
logger.exception(error)
def rescrape(command):
logger.info('Re-scraping article {0}'.format(command.article_id))
try:
article = Article.load(command.article_id)
article.scrape(overwrite=command.overwrite)
except Exception as error:
logger.error('Error scraping article {0}'.format(command.article_id))
logger.exception(error)
def retag(command):
logger.info('Re-tagging article {0}'.format(command.article_id))
try:
article = Article.load(command.article_id)
article.tag(overwrite=command.overwrite)
except Exception as error:
logger.error('Error tagging article {0}'.format(command.article_id))
logger.exception(error)
def rescrape(article_id, **kwargs):
logger.info('Re-scraping article {0}'.format(article_id))
try:
article = Article.load(article_id)
article.scrape(**kwargs)
except Exception as error:
logger.error('Error scraping article {0}'.format(article_id))
logger.exception(error)
def retag(article_id, **kwargs):
logger.info('Re-tagging article {0}'.format(article_id))
try:
article = Article.load(article_id)
article.tag(**kwargs)
except Exception as error:
logger.error('Error tagging article {0}'.format(article_id))
logger.exception(error)
def batch_rescrape(processes, query=None, limit=None, **kwargs):
pool = multiprocessing.Pool(processes=processes)
articles = Article.find(query)
if limit:
articles = articles.limit(limit)
results = pool.map(
functools.partial(rescrape, **kwargs),
(article._id for article in articles),
)
def batch_rescrape(processes, query=None, limit=None, overwrite=False):
pool = multiprocessing.Pool(processes=processes)
articles = Article.find(query)
if limit:
articles = articles.limit(limit)
results = pool.map(
rescrape,
(RescrapeCommand(article._id, overwrite) for article in articles),
)
def batch_rescrape(processes, query=None, limit=None, **kwargs):
pool = multiprocessing.Pool(processes=processes)
articles = Article.find(query)
if limit:
articles = articles.limit(limit)
results = pool.map(
functools.partial(rescrape, **kwargs),
(article._id for article in articles),
)
def remove_duplicates(by='pmid'):
"""Remove duplicate articles by field.
:param str by: Article field to identify duplicates
"""
counts = collections.defaultdict(int)
values = [
value[by] for value in Article._storage[0].store.find({}, {by: 1})
]
for value in values:
counts[value] += 1
for value, count in counts.items():
if count == 1:
continue
articles = list(Article.find(Q(by, 'eq', value)))
for duplicate in articles[1:]:
logger.debug('Deleting duplicate record: {}'.format(value))
Article.remove_one(duplicate)
def count_verified(threshold=VERIFY_THRESHOLD):
"""Count the number of downloaded and verified documents across all
articles.
:param float threshold: Document verification threshold
:return: Tuple of total and verified dictionaries, each mapping document
types to counts
"""
count = defaultdict(int)
verified = defaultdict(int)
for article in Article.find():
for type_, field in DOCUMENT_TYPES_TO_FIELDS.iteritems():
value = getattr(article, field)
if value:
count[type_] += 1
if value.verification_score > threshold:
verified[type_] += 1
return count, verified
def count_verified(threshold=VERIFY_THRESHOLD):
"""Count the number of downloaded and verified documents across all
articles.
:param float threshold: Document verification threshold
:return: Tuple of total and verified dictionaries, each mapping document
types to counts
"""
count = defaultdict(int)
verified = defaultdict(int)
for article in Article.find():
for type_, field in DOCUMENT_TYPES_TO_FIELDS.iteritems():
value = getattr(article, field)
if value:
count[type_] += 1
if value.verification_score > threshold:
verified[type_] += 1
return count, verified
def add_missing(query, max_count, randomize=False):
"""Search PubMed for articles and scrape documents.
:param str query: PubMed query
:param int max_count: Maximum number of articles to process
:param bool randomize: Randomize list of articles to fetch
:return: Added article objects
"""
pmids = pubtools.search_pmids(query)
stored_pmids = [
article['pmid']
for article in mongo['article'].find(
{}, {'pmid': 1}
)
]
missing_pmids = set(pmids) - set(stored_pmids)
logger.warn('Found {0} articles to add.'.format(len(missing_pmids)))
pmids_to_add = list(missing_pmids)[:max_count]
if randomize:
random.shuffle(pmids_to_add)
records = pubtools.download_pmids(pmids_to_add)
scraper = SCRAPE_CLASS(**SCRAPE_KWARGS)
added = []
for pmid, record in zip(pmids_to_add, records):
logger.debug('Adding article {}'.format(pmid))
article = Article.from_record(record)
article.scrape(scraper)
article.tag()
added.append(article)
return added
def update_dates(overwrite=False):
query = None if overwrite else Q('date', 'eq', None)
articles = Article.find(query)
for article in articles:
article.update_date()
article.save()
