def getPubmedDoi(pmid):
""" retrieve doi for pmid via http eutils"""
url = 'http://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&[email protected]&retmode=xml&id=%s' % pmid
xp = maxXml.XmlParser(url=url)
doi = xp.getTextFirst("PubmedArticle/PubmedData/ArticleIdList/ArticleId", reqAttrDict={'IdType':'doi'}, default=None)
logging.debug("Found DOI: %s" % doi)
return doi
def getPubmedOutlinks(pmid, preferPmc=True):
""" use eutils to get outlinks from pubmed """
logging.debug("%s: Getting outlink from pubmed" % (pmid))
url = "http://eutils.ncbi.nlm.nih.gov/entrez/eutils/elink.fcgi?dbfrom=pubmed&id=%s&retmode=llinks&cmd=llinks" % pmid
xp = maxXml.XmlParser(url=url)
#req.add_header('User-Agent', 'User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US) [email protected]')
outlinks = []
aggregatorOutlinks = []
for objUrl in xp.getXmlAll("LinkSet/IdUrlList/IdUrlSet/ObjUrl"):
url = objUrl.getTextFirst("Url")
SubjType = objUrl.getTextFirst("SubjectType")
if SubjType != "publishers/providers":
logging.log(
5,
"skipping url %s, is not a URL to a provider/publisher" % url)
if SubjType == "aggregators":
aggregatorOutlinks.append(url)
continue
attrList = list(objUrl.getTextAll("Attribute"))
if not "full-text online" in attrList and \
"full-text PDF" not in attrList:
logging.log(
5, "skipping url %s, does not seem to provide fulltext" % url)
continue
else:
outlinks.append(url)
logging.debug("Found %d outlinks" % len(outlinks))
logging.log(5, "Outlinks: %s" % str(outlinks))
if len(outlinks) == 0:
logging.debug("No Outlinks found, checking for PMC aggregator")
if len(aggregatorOutlinks)>1 and \
"ukpmc" in aggregatorOutlinks[0]: # let's get rid of UKPMC
aggregatorOutlinks.pop(0)
for outlink in aggregatorOutlinks:
if preferPmc and httpStartsWith("http://www.ncbi.nlm.nih.gov/pmc",
outlink):
logging.debug("Found PMC outlink")
return outlink
logging.debug("No PMC outlink")
return None
else:
if "ukpmc" in outlinks[0] and len(outlinks) > 1:
outlinks.pop(0)
if "swetswise" in outlinks[0] and len(outlinks) > 1:
outlinks.pop(0)
outlink = outlinks[0]
logging.debug("Using outlink: %s" % outlink)
return outlink
def parsePubmedMedlineIter(xml):
"""
Parse pubmed xml format and yield as dictionary, see parseMedline
records come either from Pubmed as a <PubmedArticleSet><PubmedArticle><MedlineCitation>...
or from Medline as <MedlineCitationSet><MedlineCitation>...</MedlineCitation>
"""
#if fromMedline:
#recordTag = "MedlineCitation"
#closeTag = "</MedlineArticleSet>"
#openTag = "<MedlineArticleSet"
#else:
recordTag = "PubmedArticle"
#recordTag = "PubmedArticle/MedlineCitation"
closeTag = "</PubmedArticleSet>"
openTag = "<PubmedArticleSet>"
# NCBI eutils sometimes "forgets" the opening/closing tags
if xml.strip()=="":
logging.error("Got empty XML file from NCBI")
raise PubmedError("Got empty XML from NCBI", "pubmedEmptyXml")
if not openTag in xml:
logging.warn("Addding opening tag")
xml = openTag + "\n" + xml
if not closeTag in xml:
logging.warn("Addding closing tag")
xml = xml+"\n"+closeTag
logging.debug("Parsing pubmed file")
try:
topEl = maxXml.XmlParser(string=xml)
except ParseError:
logging.debug("Error on parsing this XML: %s" % xml)
raise
except ParseError2:
logging.debug("Error on parsing this XML: %s" % xml)
raise
for artEl in topEl.getXmlAll(recordTag):
medlineCitEl = artEl.getXmlFirst("MedlineCitation")
dataDict = parseMedline(medlineCitEl)
pubmedCitEl = artEl.getXmlFirst("PubmedData")
dataDict = parsePubmedFields(pubmedCitEl, dataDict)
yield dataDict
