merging_wip.py

from catharbot import catharbot
bot = catharbot.CatharBot()
# Work In Progress
# uses the 'merging' branch of Catharbot, https://github.com/hornc/catharbot/tree/merging

def extract_olid(olid):
    """Convert a string like '/authors/OL1412764A' to just 'OL1412764A'"""
    return olid.split('/')[-1]

def remove_editions(duplicates, docs):
    """ Removes editions by [OLID str] from a merge changeset docs ([JSON dicts])
        reason: to prevent duplicated edition documents that merge_works() reassigns
                and merge_editions() makes into redirects.
        TODO: refactor to make this process clearer / cleaner
    """
    return [ d for d in docs if extract_olid(d['key']) not in duplicates ]


def test_remove_editions():
    dupes = ['A', 'C']
    docs  = [{'key': '/books/A'}, {'key': '/books/B'}, {'key': '/books/C'}]
    assert(remove_editions(dupes, docs) == [{'key': '/books/B'}])


# TODO: change merge_works() and merge_editions() to use **kwargs
# e.g. merge_works(master="OL1234W", duplicates=["OL345W", "OL678W"])

# example changeset: https://openlibrary.org/recentchanges/2017/09/22/bulk_update/53034676
#    full_merge(master='OL24869802M', duplicates=['OL25426847M', 'OL24928157M'])

def full_merge(**kwargs):
    """ Merge identical editions and their works
        kwargs:
            master: Master edition OLID (str - required)
          and one of
            duplicate: Duplicate edtion OLID (str)
            duplicates: list of edition OLIDs ([str])
          simple: (bool) Do not merge data, just perform redirects, defaults to False
            Simple merge is faster and can be used when merging 'bad' data into a good record.
    """
    # TODO: this works, but refactor for clarity!
    master = kwargs['master']
    simple = kwargs.setdefault('simple', False)
    duplicates = kwargs.setdefault('duplicates', [])
    if 'duplicate' in kwargs:
        duplicates.append(kwargs['duplicate'])
    print "Merge %s into %s" % (duplicates, master)
    master_edition = bot.load_doc(master)
    dupe_editions = [ bot.load_doc(e) for e in duplicates ]
    changeset = []
    if not simple:
        merged_edition = bot.merge_into_work(master_edition, dupe_editions)
        changeset.append(merged_edition)
    # are there extra works to merge?
    master_w_olid = extract_olid(master_edition['works'][0]['key'])
    dupe_w_olids  = [ extract_olid(e['works'][0]['key']) for e in dupe_editions if extract_olid(e['works'][0]['key']) != master_w_olid ]
    if len(dupe_w_olids) > 1 or master_w_olid not in dupe_w_olids:
        master_work = bot.load_doc(master_w_olid)
        dupe_works  = [ bot.load_doc(w) for w in dupe_w_olids ]
        if not simple:
            merged_work = bot.merge_into_work(master_work, dupe_works)
            changeset.append(merged_work)
        changeset += bot.merge_works(dupe_w_olids, master_w_olid)
        # remove reassigned duplicate editions from changeset that will be made into redirects
        changeset = remove_editions(duplicates, changeset)
    changeset += bot.merge_editions(duplicates, master)
    return changeset

move_editions.py

# move_editions(['OL559079M'], 'OL2420021W')
from olclient.openlibrary import OpenLibrary
ol = OpenLibrary()

def move_editions(edition_list, master_work):
  changeset = []
  for olid in edition_list:
    edition = ol.Edition.get(olid)
    edition.work_olid = master_work
    changeset.append(edition)
  ol.save_many(changeset, "move to work %s" % master_work)