Add BibTeX export for ScopusSearch #302
Conversation
First implementation, might need some cleanup.
Michael-E-Rose
force-pushed
the
master
branch
4 times, most recently
from
403e8a5 to
12b1e6e
Compare
| type_conference_paper = "Conference Paper" | ||
| type_conference_review = "Conference Review" | ||
| type_article = "Article" | ||
| type_review = "Review" | ||
| type_short_survey = "Short Survey" | ||
| type_editorial = "Editorial" | ||
| type_note = "Note" | ||
| type_letter = "Letter" | ||
| type_data_paper = "Data Paper" | ||
| type_erratum = "Erratum" | ||
| type_book_chapter = "Book Chapter" | ||
| type_book = "Book" | ||
| type_report = "Report" | ||
| type_retracted = "Retracted" | ||
| type_none = None | ||
|
|
||
| aggregation_type_conference_proceedings = "Conference Proceeding" | ||
| aggregation_type_journal = "Journal" | ||
| aggregation_type_trade_journal = "Trade Journal" | ||
| aggregation_type_book_series = "Book Series" | ||
| aggregation_type_book = "Book" | ||
| aggregation_type_report = "Report" | ||
| aggregation_type_none = None | ||
|
|
||
| bib_tex_type_article = "Article" | ||
| bib_tex_type_in_proceedings = "InProceedings" | ||
| bib_tex_type_in_collection = "InCollection" | ||
| bib_tex_type_book = "Book" | ||
| bib_tex_type_techreport = "TechReport" |
There was a problem hiding this comment.
All these variables just contain one string. Isn't it easier to just use the strings instead of the variables?
There was a problem hiding this comment.
Also possible. I chose this way to allow reuse, a better structure, making similar strings distinguishable (bib_tex vs. Scopus), and possibly in the future allowing modularity (for example by storing such strings in a different file).
But I'll adhere to the project's style for that!
| if (document_type in [type_article, type_review, type_short_survey, type_editorial, type_note, type_letter, type_data_paper, type_erratum, type_conference_review, type_conference_paper, type_retracted, type_none] and aggregation_type == aggregation_type_journal) or (document_type in [type_article, type_review, type_short_survey, type_note] and aggregation_type == aggregation_type_trade_journal) or (document_type == type_article and aggregation_type == aggregation_type_none): | ||
| bib_tex_type = bib_tex_type_article | ||
| if aggregation_type == aggregation_type_conference_proceedings or (document_type == type_conference_paper and aggregation_type in [aggregation_type_book, aggregation_type_none]): | ||
| bib_tex_type = bib_tex_type_in_proceedings | ||
| elif aggregation_type == aggregation_type_book_series or (document_type in [type_book_chapter, type_article, type_editorial] and aggregation_type == aggregation_type_book): | ||
| bib_tex_type = bib_tex_type_in_collection |
There was a problem hiding this comment.
It would be more readable to define suitable doctypes as a set beforehand:
in_collection = {"ed", "ar", "ch"}
etc.
There was a problem hiding this comment.
Mh, you mean only for the subcheck document_type in [...]?
| fields = self.add_bibtex_field(fields, "author", authors) | ||
| fields = self.add_bibtex_field(fields, "title", result.title) | ||
| fields = self.add_bibtex_field(fields, "date", result.coverDate) |
There was a problem hiding this comment.
Can you turn this and the following statements into a loop of tuples?
[("author", authors), ("title", result.title), ...]
There was a problem hiding this comment.
Not sure if I recall this correctly, but I might have followed the style of the other existing bibtex export function in pybliometrics.
The problem with putting this in a loop is that there are some fields depending on checks. So the list would then need to be constructed in a similar manner, not making the code much shorter. Still probable makes sense to eliminate a bit of duplication!
| bib_library.add(entry) | ||
|
|
||
| # Check whether the addition was successful or resulted in a duplicate block that needs fixing. | ||
| for i in range(26): |
There was a problem hiding this comment.
Where does the 26 come from?
There was a problem hiding this comment.
i is later used for shifting letters from a to z.
| # Check whether the addition was successful or resulted in a duplicate block that needs fixing. | ||
| for i in range(26): | ||
| failed_blocks = bib_library.failed_blocks | ||
| if failed_blocks: |
There was a problem hiding this comment.
What exactly is happening here?
There was a problem hiding this comment.
The used python library has problems adding a new entry if the key already exists in the library. So if there is already "@author" in the library, no other entry with @Book{Author, ...} can be added. Then, the addition will result in a failed block. This checks whether the addition failed. If so, we try to add the entry as @Book{Authora, ...} instead. If that fails, it will just move through the alphabet until it reaches @Book{Authorz, ...}. Not ideal, but this is mainly dealing with shortcomings of the used library...
Michael-E-Rose
force-pushed
the
master
branch
3 times, most recently
from
9cfdac3 to
9d182fa
Compare
|
Dear @claell , do we want to continue here? I very much appreciate your effort and willingness to improve pybliometrics, but I do not understand all the code. And here and there we are able to improve the code, which helps other users because of clarity and transparency. |
|
Ah, I guess your review wasn't seen by me, before. I'll add some comments! |
First implementation, might need some cleanup.
I thought, I will already start the PR, so feedback can be given.
Fixes #295