CanonicaliseISBN.py

#!/usr/bin/env python3
"""Iterate through a Zotero library and convert all the ISBNs to the
   standard form with hypens.
"""

from isbnlib import get_canonical_isbn, mask
from pyzotero import zotero
from tqdm import tqdm
from LibraryConfig import *

zot = zotero.Zotero(ZOT_LIBRARY_ID, 'user', ZOT_API_KEY)
print("Getting library data from Zotero")
nitems =  zot.num_items()
itemsgen =  zot.makeiter(zot.top(limit=10))
for batch in tqdm(itemsgen, total=int(nitems/10 + int(bool(nitems % 10)))):
    for item in batch:
        try:
            isbn = item['data']['ISBN']
        except KeyError:
            # Not a book-ish item
            continue
            
        if not isbn:
            # No ISBN listed
            continue

        # Transform to canonical (bare) form, then return to standardised
        # form with hyphens.
        canisbn = get_canonical_isbn(isbn.replace(' ', ''))

        if not canisbn:
            # This most likely means that the ISBN given is bogus
            # (e.g. has a faulty checksum). Some books have bogus
            # ISBNs printed on them, so they are used for cataloguing
            # by some libraries despite being formally invalid.
            print("Error extracting ISBN from "+str(isbn))
            continue

        newisbn = mask(canisbn)
        if newisbn != isbn:
            assert newisbn
            print("Updating "+str(isbn)+" to "+str(newisbn))
            item['data']['ISBN'] = newisbn
            if not zot.update_item(item):
                raise Exception("Zotero write failed for "+str(isbn))