Import in current format only finds entries with ISBN numbers

[furicle]

According to https://wiki.librarything.com/index.php/Adding_and_importing_books this is expected behaviour

Universal Import does not read any data from your file other than valid ISBNs, if they're included. If you have books without ISBNs, this will not add them. If you have keyed reviews or tags into your desktop program, this will not add them.

Looks like the correct way to do this is to include exactly (? unclear, all fields, same order perhaps?) the entries as in the example file

Alternately, if you format your file with a precise set of columns, it will import everything in those columns. A sample spreadsheet is included on the Import page indicating the format needed. It includes the following fields: title, author, date, ISBN, publication information, tags, rating, review, entry date.

That file https://www.librarything.com/LibraryThingSample.csv has

'TITLE',"'AUTHOR (last, first)'",'DATE','ISBN',"'PUBLICATION INFO'",'TAGS','RATING','REVIEW',"'DATE READ'","'PAGE COUNT'","'CALL NUMBER'"

It's unclear

• what 'publication info' is
• what date 'date' is referring to
• what 'call number' is
• how multiple authours are handled
• how titles are handled (md, phd etc)

Need to line these up with the default fields available in Calibre and update script to match

[furicle]

Further research seems to say multiple authours in the csv import is not supported? - https://www.librarything.com/topic/347569

That thread suggest MARC format which is apparently some sort of standard... but would drastically change this script... https://wiki.librarything.com/index.php/LibraryThing_and_MARC

Do more testing with multiple authours as it may just use the first authour as a 'search' item to match against when not using ISBN...

[furicle]

Note that MARC is a whole series of formal standards - https://www.loc.gov/marc/marctools.html Digging into that is probably a huge scope increase I'm not interested in at this time.