pybibupdate functionality

examples/example.bib

@@ -13,21 +13,17 @@ @article{MR0026286
     DOI = "10.1002/j.1538-7305.1948.tb01338.x",
     URL = "https://doi.org/10.1002/j.1538-7305.1948.tb01338.x"
 }
+
 @article{10.1109/TIT.2006.885507,
-    AUTHOR = "Candes, Emmanuel J. and Tao, Terence",
-    TITLE = "Near-optimal signal recovery from random projections: universal encoding strategies?",
-    JOURNAL = "IEEE Trans. Inform. Theory",
-    FJOURNAL = "Institute of Electrical and Electronics Engineers. Transactions on Information Theory",
-    VOLUME = "52",
-    YEAR = "2006",
-    NUMBER = "12",
-    PAGES = "5406--5425",
-    ISSN = "0018-9448",
-    MRCLASS = "94A12 (41A25 94A13)",
-    MRNUMBER = "2300700",
-    MRREVIEWER = "L. L. Campbell",
-    DOI = "10.1109/TIT.2006.885507",
-    URL = "https://doi.org/10.1109/TIT.2006.885507"
+    title = "{Near}-optimal signal recovery from random projections: {Universal} encoding strategies?",
+    author = "Candes, Emmanuel J. and Tao, Terence",
+    year = "2006",
+    doi = "10.1109/TIT.2006.885507",
+    url = "https://doi.org/10.1109/TIT.2006.885507",
+    journal = "IEEE Trans. Inf. Theory",
+    volume = "52",
+    number = "12",
+    pages = "5406--5425"
 }
 
 @unpublished{math/0211159,
@@ -40,42 +36,36 @@ @unpublished{math/0211159
 }
 
 @article{PMID:271968,
+    title = "{DNA} sequencing with chain-terminating inhibitors.",
     author = "Sanger, F. and Nicklen, S. and Coulson, A. R.",
+    year = "1977",
     doi = "10.1073/pnas.74.12.5463",
     url = "https://doi.org/10.1073/pnas.74.12.5463",
-    year = "1977",
-    publisher = "Proceedings of the National Academy of Sciences",
+    journal = "Proc. Natl. Acad. Sci. U.S.A.",
     volume = "74",
     number = "12",
     pages = "5463--5467",
-    title = "{DNA} sequencing with chain-terminating inhibitors",
-    journal = "Proceedings of the National Academy of Sciences",
-    PMID = "271968"
+    pmid = "PMID:271968"
 }
 
 @inproceedings{10.1109/CVPR.2016.90,
+    title = "{Deep} residual learning for image recognition",
     author = "He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian",
-    doi = "10.1109/cvpr.2016.90",
-    url = "https://doi.org/10.1109/cvpr.2016.90",
     year = "2016",
-    publisher = "{IEEE}",
-    title = "{Deep} {Residual} {Learning} for {Image} {Recognition}",
-    booktitle = "2016 {IEEE} Conference on Computer Vision and Pattern Recognition ({CVPR})"
+    doi = "10.1109/CVPR.2016.90",
+    url = "https://doi.org/10.1109/CVPR.2016.90",
+    publisher = "IEEE Computer Society",
+    booktitle = "{Proc} {IEEE} {Comput} {Soc} {Conf} {Comput} {Vision} {Pattern} {Recognit}"
 }
 
 @article{10.4310/ATMP.1998.v2.n2.a1,
-    AUTHOR = "Maldacena, Juan",
-    TITLE = "The large {$N$} limit of superconformal field theories and supergravity",
-    JOURNAL = "Adv. Theor. Math. Phys.",
-    FJOURNAL = "Advances in Theoretical and Mathematical Physics",
-    VOLUME = "2",
-    YEAR = "1998",
-    NUMBER = "2",
-    PAGES = "231--252",
-    ISSN = "1095-0761",
-    MRCLASS = "81T30 (81T60 83E30)",
-    MRNUMBER = "1633016",
-    MRREVIEWER = "Douglas J. Smith",
-    DOI = "10.4310/ATMP.1998.v2.n2.a1",
-    URL = "https://doi.org/10.4310/ATMP.1998.v2.n2.a1"
+    title = "{The} large {N} {Limit} of superconformal field theories and supergravity",
+    author = "Maldacena, Juan",
+    year = "1998",
+    doi = "10.4310/ATMP.1998.v2.n2.a1",
+    url = "https://doi.org/10.4310/ATMP.1998.v2.n2.a1",
+    journal = "Adv. Theor. Math. Phys.",
+    volume = "2",
+    number = "2",
+    pages = "231--252"
 }

examples/example_biber.tex

@@ -4,7 +4,7 @@
 
 \usepackage[backend=biber, datamodel=pmid-mr-arxiv,url=false]{biblatex}
 \usepackage{hyperref} 
-\addbibresource{example.bib}
+\addbibresource{example.bib} 
 
 \makeatletter
 \DeclareFieldFormat{eprint}{\href{https://arxiv.org/\abx@arxivpath/#1}{\nolinkurl{arXiv:#1}}}

examples/example_bibtex.tex

@@ -5,10 +5,10 @@
 \usepackage{hyperref}
 \bibliographystyle{plain}
 
-\begin{document}
-
- \nocite{MR0026286,10.1109/TIT.2006.885507,math/0211159,PMID:271968,10.1109/CVPR.2016.90,10.4310/ATMP.1998.v2.n2.a1} 
+\begin{document} 
+ 
+\nocite{MR0026286,10.1109/TIT.2006.885507,math/0211159,PMID:271968,10.1109/CVPR.2016.90,10.4310/ATMP.1998.v2.n2.a1}   
 
 \bibliography{example}
 
-\end{document}    
+\end{document}     

pybibget.egg-info/entry_points.txt

@@ -1,3 +1,4 @@
 [console_scripts]
 pybibget = pybibget:pybibget
 pybibparse = pybibget:pybibparse
+pybibupdate = pybibget:pybibupdate

pybibget/__init__.py

@@ -3,51 +3,54 @@
 import re
 import sys
 import logging as log
-from enum import Enum
-import pybtex.database
-import pybibget.bibentry as bibentry
+log.getLogger('asyncio').setLevel(log.WARNING)
+from pybibget.bibentry import Bibget
+from pybtex.database import parse_string
 
+def add_optional_args(parser):
+    parser.add_argument('-v', '--verbose', action='store_true', help='verbose output')
+    parser.add_argument('-d', '--debug', action='store_true', help='debug output')
+    parser.add_argument('--skip-doi-msc', action='store_true', help='skip MathSciNet lookup for DOIs')
+
 
-def pybibget() -> None:
+def pybibget():
     """
     Reads citation keys from command line and calls get_citations()
     """
     parser = argparse.ArgumentParser(prog='pybibget', description='Command line utility to automatically retrieve BibTeX citations from MathSciNet, arXiv and PubMed')
-
     parser.add_argument('keys', type=str, metavar='citekeys', nargs='*', help='MathSciNet (MRxxxxx), arXiv (xxxx.xxxxx), PubMed (PMID:xxxxxxxx) or DOI (10.xxx/xxxxx) citation keys (separated by spaces)')
-    parser.add_argument('-v', '--verbose', action='store_true', help='verbose output')
-    parser.add_argument('-d', '--debug', action='store_true', help='debug output')
-    parser.add_argument('-f', action='store', dest='file_name', help='Append output to file (default: write output to stdout)')
-
+    parser.add_argument('-w', action='store', dest='file_name', help='Append output to file (default: write output to stdout)')
+    add_optional_args(parser)
     args = parser.parse_args()
-    kwargs = {'file': args.file_name}
+    kwargs = {'file': args.file_name }
     if args.debug:
         kwargs['verbose'] = log.DEBUG
     elif args.verbose:
         kwargs['verbose'] = log.INFO
     if not args.keys:
         parser.print_help()
         exit(1)
-    asyncio.run(get_citations(args.keys, **kwargs))
+
+    get_citations(args.keys, **kwargs)
 
 
 def pybibparse():
     """
     Reads latex file name from the command line, parses the .blg file and calls get_citations()
     """
-    parser = argparse.ArgumentParser(prog='pybibget', description='Command line utility to automatically retrieve missing BibTeX citations from MathSciNet, arXiv and PubMed')
-
-    parser.add_argument('file_name', type=str, metavar='file_name', nargs=1, help='base file name of main tex file (without .tex extension)')
-    parser.add_argument('-v', '--verbose', action='store_true', help='verbose output')
-    parser.add_argument('-d', '--debug', action='store_true', help='debug output')
-    parser.add_argument('-w', '--write', metavar="file_name", action='store', nargs='?', const=" ", type=str, help='Append output to .bib file (default: write output to stdout). The .bib file is automatically detected from the .blg file. If no .bib file is found, the .bib file has to be specified explicitly via "-w file_name.bib".')
-
+    parser = argparse.ArgumentParser(prog='pybibget', description='Command line utility to automatically retrieve BibTeX citations from MathSciNet, arXiv and PubMed')
+    parser.add_argument('file_name', type=str, metavar='tex_file(.tex)', nargs=1, help='LaTeX file to be parsed for missing citations')
+    parser.add_argument('-w', action='store', dest='write', metavar="output.bib", nargs='?', const=" ", help='Append output to file (default: write output to stdout). A bib file name can be specified via "-w file_name.bib" but usually the .bib file is found automatically.')
+    add_optional_args(parser)
     args = parser.parse_args()
     if not args.file_name:
         parser.print_help()
         sys.exit()
+    if args.file_name[0].endswith(".tex"):
+        base_file_name = args.file_name[0][:-4]
+    else:
+        base_file_name = args.file_name[0]
 
-    base_file_name = args.file_name[0]
     with open(base_file_name+".blg") as file:
         blg_file = file.read()
     missing_cites = re.findall(r"I didn't find a database entry for '([A-Za-z0-9\.\-_ :\/]*)'", blg_file) \
@@ -68,26 +71,39 @@ def pybibparse():
                 print("No .bib file found. Please specify the .bib file via '-w file_name.bib'")
                 sys.exit()
             kwargs['file'] = bib_file_names[0] if args.write == " " else args.write
-        asyncio.run(get_citations(missing_cites, **kwargs))        
+        get_citations(missing_cites, **kwargs)        
     else:
         print("No missing citations found. Make sure that biber/bibtex is run successfully before running pybibget.")
 
+def pybibupdate():
+    parser = argparse.ArgumentParser(prog='pybibget', description='Command line utility to update BibTeX citations from MathSciNet and Scopus')
+    parser.add_argument('file_name', type=str, metavar='bib_file(.bib)', help='bib file to be parsed for citations')
+    args = parser.parse_args()
+    if not args.file_name:
+        parser.print_help()
+        sys.exit()
+    if not args.file_name.endswith(".bib"):
+        args.file_name += ".bib"
+    with open(args.file_name) as file:
+        bib_file = file.read()
+    bibliography = parse_string(bib_file, 'bibtex').entries
+
+    log.basicConfig(format="%(levelname)s: %(message)s", level=log.WARNING)
+
+    bibget = Bibget(mathscinet=True)
+    updated_bibliography = asyncio.run(bibget.update_all(bibliography))
+    with open(args.file_name, 'w') as file:
+        file.write(updated_bibliography.to_string('bibtex'))
+        print(f"Wrote the updated bibliography to {args.file_name}.")
 
-async def get_citations(keys, verbose=log.WARN, file=None):
+def get_citations(keys, verbose=log.WARNING, file=None):
     """
     Retrieves BibTeX entries for given citation keys and writes them to file or stdout
     """
     log.basicConfig(format="%(levelname)s: %(message)s", level=verbose)
 
-    bibentries = await asyncio.gather(*[bibentry.getbibentry(key) for key in keys],return_exceptions=True)
-    bib_data = pybtex.database.BibliographyData()
-    for entry_key in bibentries:
-        if type(entry_key) in [ValueError, TypeError]:
-            log.error(entry_key)
-        else:
-            entry,key = entry_key
-            bib_data.entries[key] = entry
-
+    bibget = Bibget(mathscinet=True)
+    bib_data = asyncio.run(bibget.citations(keys))
     number_of_entries = len(bib_data.entries)
     bib_data = bib_data.to_string('bibtex')
     if file:
@@ -96,6 +112,7 @@ async def get_citations(keys, verbose=log.WARN, file=None):
             print(f"Successfully appended {number_of_entries} BibTeX entries to {file}.")
     else:
         print("\n"+bib_data)
+    return number_of_entries
 
 
 if __name__ == '__main__':