summaryrefslogtreecommitdiffstats
path: root/python/parse_pubmed_xml.py
blob: 9350e9a4bb6bc9647caf6daa38076dd30849d8a5 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370

import sys
import json
import datetime
from bs4 import BeautifulSoup
from bs4.element import NavigableString

# from: https://www.ncbi.nlm.nih.gov/books/NBK3827/table/pubmedhelp.T.publication_types/?report=objectonly
PUBMED_RELEASE_TYPE_MAP = {
    #Adaptive Clinical Trial
    "Address": "speech",
    "Autobiography": "book",
    #Bibliography
    "Biography": "book",
    #Case Reports
    "Classical Article": "article-journal",
    #Clinical Conference
    #Clinical Study
    #Clinical Trial
    #Clinical Trial, Phase I
    #Clinical Trial, Phase II
    #Clinical Trial, Phase III
    #Clinical Trial, Phase IV
    #Clinical Trial Protocol
    #Clinical Trial, Veterinary
    #Collected Works
    #Comparative Study
    #Congress
    #Consensus Development Conference
    #Consensus Development Conference, NIH
    #Controlled Clinical Trial
    "Dataset": "dataset",
    #Dictionary
    #Directory
    #Duplicate Publication
    "Editorial": "editorial",
    #English Abstract   # doesn't indicate that this is abstract-only
    #Equivalence Trial
    #Evaluation Studies
    #Expression of Concern
    #Festschrift
    #Government Document
    #Guideline
    "Historical Article": "article-journal",
    #Interactive Tutorial
    "Interview": "interview",
    "Introductory Journal Article": "article-journal",
    "Journal Article": "article-journal",
    "Lecture": "speech",
    "Legal Case": "legal_case",
    "Legislation": "legislation",
    "Letter": "letter",
    #Meta-Analysis
    #Multicenter Study
    #News
    "Newspaper Article": "article-newspaper",
    #Observational Study
    #Observational Study, Veterinary
    #Overall
    #Patient Education Handout
    #Periodical Index
    #Personal Narrative
    #Portrait
    #Practice Guideline
    #Pragmatic Clinical Trial
    #Publication Components
    #Publication Formats
    #Publication Type Category
    #Randomized Controlled Trial
    #Research Support, American Recovery and Reinvestment Act
    #Research Support, N.I.H., Extramural
    #Research Support, N.I.H., Intramural
    #Research Support, Non-U.S. Gov't Research Support, U.S. Gov't, Non-P.H.S.
    #Research Support, U.S. Gov't, P.H.S.
    #Review     # in the "literature review" sense, not "product review"
    #Scientific Integrity Review
    #Study Characteristics
    #Support of Research
    #Systematic Review
    "Technical Report": "report",
    #Twin Study
    #Validation Studies
    #Video-Audio Media
    #Webcasts
}

MONTH_ABBR_MAP = {
    "Jan":  1, "01":  1,
    "Feb":  2, "02":  2,
    "Mar":  3, "03":  3,
    "Apr":  4, "04":  4,
    "May":  5, "05":  5,
    "Jun":  6, "06":  6,
    "Jul":  7, "07":  7,
    "Aug":  8, "08":  8,
    "Sep":  9, "09":  9,
    "Oct": 10, "10": 10,
    "Nov": 11, "11": 11,
    "Dec": 12, "12": 12,
}

class PubMedParser():
    """
    Converts PubMed/MEDLINE XML into in release entity (which can dump as JSON)

    TODO: MEDLINE doesn't include PMC/OA license; could include in importer?
    TODO: clean (ftfy) title, original title, etc
    """

    def __init__(self):
        pass

    def parse_file(self, handle):

        # 1. open with beautiful soup
        soup = BeautifulSoup(handle, "xml")

        # 2. iterate over articles, call parse_article on each
        for article in soup.find_all("PubmedArticle"):
            resp = self.parse_article(article)
            print(json.dumps(resp))
            #sys.exit(-1)

    def parse_article(self, a):

        medline = a.MedlineCitation
        # PubmedData isn't required by DTD, but seems to always be present
        pubmed = a.PubmedData
        extra = dict()
        extra_pubmed = dict()

        identifiers = pubmed.ArticleIdList
        doi = identifiers.find("ArticleId", IdType="doi")
        if doi:
            doi = doi.string.lower()

        pmcid = identifiers.find("ArticleId", IdType="pmc")
        if pmcid:
            pmcid = pmcid.string

        release_type = None
        for pub_type in medline.Article.PublicationTypeList.find_all("PublicationType"):
            if pub_type.string in PUBMED_RELEASE_TYPE_MAP:
                release_type = PUBMED_RELEASE_TYPE_MAP[pub_type.string]
            break
        if medline.Article.PublicationTypeList.find(string="Retraction of Publication"):
            release_type = "retraction"
            retraction_of = medline.find("CommentsCorrections", RefType="RetractionOf")
            if retraction_of:
                extra_pubmed['retraction_of_raw'] = retraction_of.RefSource.string
                extra_pubmed['retraction_of_pmid'] = retraction_of.PMID.string

        # everything in medline is published
        release_status = "published"
        if medline.Article.PublicationTypeList.find(string="Corrected and Republished Article"):
            release_status = "updated"
        if medline.Article.PublicationTypeList.find(string="Retracted Publication"):
            release_status = "retracted"

        pages = medline.find('MedlinePgn')
        if pages:
            pages = pages.string

        title = medline.Article.ArticleTitle.string, # always present
        if type(title) is tuple:
            title = ': '.join(title)
        if title.endswith('.'):
            title = title[:-1]
        # this hides some "special" titles, but the vast majority are
        # translations; translations don't always include the original_title
        if title.startswith('[') and title.endswith(']'):
            title = title[1:-1]

        original_title = medline.Article.find("VernacularTitle", recurse=False)
        if original_title:
            original_title = original_title.string
            if original_title.endswith('.'):
                original_title = original_title[:-1]

        # TODO: happening in alpha order, not handling multi-language well.
        # also need to convert lang codes: https://www.nlm.nih.gov/bsd/language_table.html
        language = medline.Article.Language
        if language:
            language = language.string
            # TODO: map to two-letter
            if language in ("und", "un"):
                # "undetermined"
                language = None

        ### Journal/Issue Metadata
        # MedlineJournalInfo is always present
        container = dict()
        container_extra = dict()
        mji = medline.MedlineJournalInfo
        if mji.find("Country"):
            container_extra['country_name'] = mji.Country.string
        if mji.find("ISSNLinking"):
            container['issnl'] = mji.ISSNLinking.string

        journal = medline.Article.Journal
        issnp = journal.find("ISSN", IssnType="Print")
        if issnp:
            container_extra['issnp'] = issnp.string

        pub_date = journal.PubDate
        release_date = None
        if pub_date.find("MedlineDate"):
            release_year = int(pub_date.MedlineDate.string.split()[0][:4])
        else:
            release_year = int(pub_date.Year.string)
            if pub_date.find("Day") and pub_date.find("Month"):
                release_date = datetime.date(
                    release_year,
                    MONTH_ABBR_MAP[pub_date.Month.string],
                    int(pub_date.Day.string))
                release_date = release_date.isoformat()
       
        ji = journal.JournalIssue
        volume = None
        if ji.find("Volume"):
            volume = ji.Volume.string
        issue = None
        if ji.find("Issue"):
            issue = ji.Issue.string
        if journal.find("Title"):
            container['name'] = journal.Title.string

        if extra_pubmed:
            extra['pubmed'] = extra_pubmed
        if not extra:
            extra = None

        ### Abstracts
        # "All abstracts are in English"
        abstracts = []
        first_abstract = medline.find("AbstractText")
        if first_abstract and first_abstract.get('NlmCategory'):
            joined = "\n".join([m.get_text() for m in medline.find_all("AbstractText")])
            abstracts.append(dict(
                content=joined,
                mimetype="text/plain",
                lang="en",
            ))
        else:
            for abstract in medline.find_all("AbstractText"):
                abstracts.append(dict(
                    content=abstract.get_text().strip(),
                    mimetype="text/plain",
                    lang="en",
                ))
                if abstract.find('math'):
                    abstracts.append(dict(
                        # strip the <AbstractText> tags
                        content=str(abstract)[14:-15],
                        mimetype="application/mathml+xml",
                        lang="en",
                    ))
        if not abstracts:
            abstracts = None

        ### Contribs
        contribs = []
        if medline.AuthorList:
            for author in medline.AuthorList.find_all("Author"):
                contrib = dict(
                    role="author",
                )
                if author.ForeName:
                    contrib['raw_name'] = "{} {}".format(author.ForeName.string, author.LastName.string)
                elif author.LastName:
                    contrib['raw_name'] = author.LastName.string
                contrib_extra = dict()
                orcid = author.find("Identifier", Source="ORCID")
                if orcid:
                    # needs re-formatting from, eg, "0000000179841889"
                    orcid = orcid.string
                    if orcid.startswith("http://orcid.org/"):
                        orcid = orcid.replace("http://orcid.org/", "")
                    elif orcid.startswith("https://orcid.org/"):
                        orcid = orcid.replace("https://orcid.org/", "")
                    elif not '-' in orcid:
                        orcid = "{}-{}-{}-{}".format(
                            orcid[0:4],
                            orcid[4:8],
                            orcid[8:12],
                            orcid[12:16],
                        )
                    contrib_extra['orcid'] = orcid
                affiliation = author.find("Affiliation")
                if affiliation:
                    contrib['raw_affiliation'] = affiliation.string
                if author.find("EqualContrib"):
                    # TODO: schema for this?
                    contrib_extra['equal_contrib'] = True
                if contrib_extra:
                    contrib['extra'] = contrib_extra
                contribs.append(contrib)

            if medline.AuthorList['CompleteYN'] == 'N':
                contribs.append(dict(raw_name="et al."))
        if not contribs:
            contribs = None

        ### References
        refs = []
        if pubmed.ReferenceList:
            for ref in pubmed.ReferenceList.find_all('Reference'):
                ref_obj = dict()
                ref_extra = dict()
                ref_pmid = ref.find("ArticleId", IdType="pubmed")
                if ref_pmid:
                    ref_extra['pmid'] = ref_pmid.string
                ref_raw = ref.Citation
                if ref_raw:
                    ref_extra['raw'] = ref_raw.string
                if ref_extra:
                    ref_obj['extra'] = ref_extra
                refs.append(ref_obj)
        if not refs:
            refs = None

        re = dict(
            work_id=None,
            title=title,
            original_title=original_title,
            release_type=release_type,
            release_status=release_status,
            release_date=release_date,
            release_year=release_year,
            doi=doi,
            pmid=int(medline.PMID.string), # always present
            pmcid=pmcid,
            #isbn13     # never in Article
            volume=volume,
            issue=issue,
            pages=pages,
            #publisher  # not included?
            language=language,
            #license_slug   # not in MEDLINE

            # content, mimetype, lang
            abstracts=abstracts,

            # raw_name, role, raw_affiliation, extra
            contribs=contribs,

            # key, year, container_name, title, locator
            # extra: volume, authors, issue, publisher, identifiers
            refs=refs,

            #   name, type, publisher, issnl
            #   extra: issnp, issne, original_name, languages, country
            container=container,

            # extra:
            #   withdrawn_date
            #   translation_of
            #   subtitle
            #   aliases
            #   container_name
            #   group-title
            #   pubmed: retraction refs
            extra=extra,
        )

        return re

if __name__=='__main__':
    parser = PubMedParser()
    parser.parse_file(open(sys.argv[1]))