def handle(self, *args, **options):
verbosity = int(options['verbosity']) # 1 = normal, 0 = minimal, 2 = all
v_normal = 1
v_all = 2
if options['pdf_only'] and options['skip_pdf_reload']:
raise CommandError("Options -s and -p are not compatible")
# check for required settings
if not hasattr(settings, 'EXISTDB_ROOT_COLLECTION') or not settings.EXISTDB_ROOT_COLLECTION:
raise CommandError("EXISTDB_ROOT_COLLECTION setting is missing")
return
if len(args):
files = args
else:
# Note: copied from prep_ead manage command; move somewhere common?
files = set()
svn = svn_client()
for archive in Archive.objects.all():
# update to make sure we have latest version of everything
svn.update(str(archive.svn_local_path)) # apparently can't handle unicode
files.update(set(glob.iglob(os.path.join(archive.svn_local_path, '*.xml'))))
if verbosity == v_all:
print 'Documents will be loaded to configured eXist collection: %s' \
% settings.EXISTDB_ROOT_COLLECTION
if options['skip_pdf_reload']:
print "** Skipping PDFs cache reload"
db = ExistDB()
loaded = 0
errored = 0
pdf_tasks = {}
start_time = datetime.now()
if not options['pdf_only']:
# unless PDF reload only has been specified, load files
for file in files:
try:
# full path location where file will be loaded in exist db collection
dbpath = settings.EXISTDB_ROOT_COLLECTION + "/" + os.path.basename(file)
errors = check_ead(file, dbpath)
if errors:
# report errors, don't load
errored += 1
print "Error: %s does not pass publication checks; not loading to eXist." % file
if verbosity >= v_normal:
print " Errors found:"
for err in errors:
print " %s" % err
else:
with open(file, 'r') as eadfile:
success = db.load(eadfile, dbpath, overwrite=True)
if success:
loaded += 1
if verbosity >= v_normal:
print "Loaded %s" % file
# load the file as a FindingAid object to get the eadid for PDF reload
ead = load_xmlobject_from_file(file, FindingAid)
# trigger PDF regeneration in the cache and store task result
# - unless user has requested PDF reload be skipped
if not options['skip_pdf_reload']:
pdf_tasks[ead.eadid.value] = reload_cached_pdf.delay(ead.eadid.value)
# NOTE: unlike the web admin publish, this does not
# generate TaskResult db records; task outcomes will be
# checked & reported before the script finishes
else:
errored += 1
print "Error: failed to load %s to eXist" % file
except ExistDBException, e:
print "Error: failed to load %s to eXist" % file
print e.message()
errored += 1
# output a summary of what was done
print "%d document%s loaded" % (loaded, 's' if loaded != 1 else '')
print "%d document%s with errors" % (errored, 's' if errored != 1 else '')
def publish(request):
"""
Admin publication form. Allows publishing an EAD file by updating or adding
it to the configured eXist database so it will be immediately visible on
the public site. Files can only be published if they pass an EAD sanity check,
implemented in :meth:`~findingaids.fa_admin.utils.check_ead`.
On POST, sanity-check the EAD file specified in request from the configured
and (if it passes all checks), publish it to make it immediately visible on
the site. If publish is successful, redirects the user to main admin page
with a success message that links to the published document on the site.
If the sanity-check fails, displays a page with any problems found.
"""
# formerly supported publish from filename, but now only supports
# publish from preview
if 'preview_id' not in request.POST:
messages.error(request, "No preview document specified for publication")
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
id = request.POST['preview_id']
# retrieve info about the document from preview collection
try:
# because of the way existdb.query.queryset constructs returns with 'also' fields,
# it is simpler and better to retrieve document name separately
ead = get_findingaid(id, preview=True)
ead_docname = get_findingaid(id, preview=True, only=['document_name'])
filename = ead_docname.document_name
except (ExistDBException, Http404): # not found in exist OR permission denied
messages.error(request,
'''Publish failed. Could not retrieve <b>%s</b> from preview collection.
Please reload and try again.''' % id)
# if ead could not be retrieved from preview mode, skip processing
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
# determine archive this ead is associated with
archive = None
if not ead.repository:
messages.error(request,
'''Publish failed. Could not determine which archive <b>%s</b> belongs to.
Please update subarea, reload, and try again.''' % id)
else:
archive_name = ead.repository[0]
# NOTE: EAD supports multiple subarea tags, but in practice we only
# use one, so it should be safe to assume the first should be used for permissions
try:
archive = Archive.objects.get(name=archive_name)
except ObjectDoesNotExist:
messages.error(request,
'''Publish failed. Could not find archive <b>%s</b>.''' % archive_name)
# bail out if archive could not be identified
if archive is None:
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
# check that user is allowed to publish this document
if not archive_access(request.user, archive.slug):
messages.error(request,
'''You do not have permission to publish <b>%s</b> materials.''' \
% archive.label)
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
errors = []
try:
# NOTE: *not* using serialized xml here, because it may introduce
# whitespace errors not present in the original file.
ok, response, dbpath, fullpath = _prepublication_check(request, filename, archive)
if ok is not True:
# publication check failed - do not publish
return response
# only load to exist if there are no errors found
db = ExistDB()
# get information to determine if an existing file is being replaced
replaced = db.describeDocument(dbpath)
try:
# move the document from preview collection to configured public collection
success = db.moveDocument(settings.EXISTDB_PREVIEW_COLLECTION,
settings.EXISTDB_ROOT_COLLECTION, filename)
# FindingAid instance ead already set above
except ExistDBException, e:
# special-case error message
errors.append("Failed to move document %s from preview collection to main collection." \
% filename)
# re-raise and let outer exception handling take care of it
raise e
except ExistDBException as err:
errors.append(err.message())
success = False
if success:
# request the cache to reload the PDF - queue asynchronous task
result = reload_cached_pdf.delay(ead.eadid.value)
task = TaskResult(label='PDF reload', object_id=ead.eadid.value,
url=reverse('fa:findingaid', kwargs={'id': ead.eadid.value}),
task_id=result.task_id)
task.save()
ead_url = reverse('fa:findingaid', kwargs={'id': ead.eadid.value})
change = "updated" if replaced else "added"
messages.success(request, 'Successfully %s <b>%s</b>. View <a href="%s">%s</a>.'
% (change, filename, ead_url, unicode(ead.unittitle)))
# redirect to main admin page and display messages
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
else:
return render(request, 'fa_admin/publish-errors.html',
{'errors': errors, 'filename': filename, 'mode': 'publish', 'exception': err})
settings.EXISTDB_ROOT_COLLECTION, filename)
# FindingAid instance ead already set above
except ExistDBException, e:
# special-case error message
errors.append("Failed to move document %s from preview collection to main collection." \
% filename)
# re-raise and let outer exception handling take care of it
raise e
except ExistDBException, e:
errors.append(e.message())
success = False
if success:
# request the cache to reload the PDF - queue asynchronous task
result = reload_cached_pdf.delay(ead.eadid.value)
task = TaskResult(label='PDF reload', object_id=ead.eadid.value,
url=reverse('fa:findingaid', kwargs={'id': ead.eadid.value}),
task_id=result.task_id)
task.save()
ead_url = reverse('fa:findingaid', kwargs={'id': ead.eadid.value})
change = "updated" if replaced else "added"
messages.success(request, 'Successfully %s <b>%s</b>. View <a href="%s">%s</a>.'
% (change, filename, ead_url, unicode(ead.unittitle)))
# redirect to main admin page and display messages
return HttpResponseSeeOtherRedirect(reverse('fa-admin:index'))
else:
return render(request, 'fa_admin/publish-errors.html',
{'errors': errors, 'filename': filename, 'mode': 'publish', 'exception': e})
