regluit/core/management/commands/recluster_singletons.py

53 lines
2.4 KiB
Python

"""
a variety of errors can cause works to not get clustered by add_editions, or a new isbn can take time to get incorporated into clustering services.
The signature of both problems is a work with only one related edition, a singleton.
This script goes through all singleton works and attempts to add_related. 'xx' works are excluded from being source works
"""
from itertools import islice
from django.core.management.base import BaseCommand
from django.db.models import Count
from regluit.core import models, bookloader
class Command(BaseCommand):
help = "add and merge editions for singleton works"
args = "<language> <max> <start>"
def handle(self, language, max=100, start=0, **options):
print "Number of singleton Works with language = %s: %s" % (language, models.Work.objects.annotate(num_editions=Count('editions')).filter(num_editions=1, language=language).count())
try:
max = int(max)
except:
max = 100
try:
start = int(start)
except:
start = 0
for (i, work) in enumerate(islice(models.Work.objects.annotate(num_editions=Count('editions')).filter(num_editions=1, language=language),start,start+max)):
#check that there's still only one edition
print "%d %s id:%s #editions:%d #isbn:%s -->" % (i, work.title.encode('ascii','ignore'), work.id, work.editions.count(), work.first_isbn_13()),
work_id=work.id
if work.editions.count() != 1:
print
continue
isbn=work.first_isbn_13()
if isbn:
new_work = bookloader.relate_isbn( isbn )
if new_work is None:
print "failed to get edition"
elif new_work.id != work_id:
print "added edition to work %s with %s editions" % (new_work.id, new_work.editions.count())
else:
if work.editions.count()>1:
print "singleton joined to new edition"
else:
print "singleton edition not moved"
else:
print "no ISBN for this work and therefore no new editions"
print "Updated Number of singleton Works with language = %s: %s" % (language,models.Work.objects.annotate(num_editions=Count('editions')).filter(num_editions=1, language=language).count() )