1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
|
#!/usr/bin/env python
# -*- coding: utf-8 -*-
import csv
import datetime
import hashlib
import sys
from django.core.management.base import BaseCommand
from ishtar_common.models import Document
BLOCKSIZE = 65536
def get_hexdigest(filename):
m = hashlib.sha256()
with open(filename, 'rb') as afile:
buf = afile.read(BLOCKSIZE)
while len(buf) > 0:
m.update(buf)
buf = afile.read(BLOCKSIZE)
return m.hexdigest()
class Command(BaseCommand):
help = 'Re-associate similar images in the database'
def add_arguments(self, parser):
parser.add_argument(
'--merge-title', type=str, default='', dest='merged-title',
help='If specified when title differs the given title will be '
'used.')
parser.add_argument(
'--output-path', type=str, default='', dest='output-path',
help='Output path for results CSV files. Default to current path.')
parser.add_argument(
'--ignore-reference', dest='ignore-reference', action='store_true',
help='Ignore the reference on diff between models.')
parser.add_argument(
'--quiet', dest='quiet', action='store_true',
help='Quiet output.')
def handle(self, *args, **options):
quiet = options['quiet']
ignore_ref = options['ignore-reference']
merged_title = options['merged-title']
output_path = options['output-path']
q = Document.objects.filter(image__isnull=False).exclude(
image='')
hashes = {}
missing_images = []
count = q.count()
out = sys.stdout
if not quiet:
out.write("* {} images\n".format(count))
for idx, doc in enumerate(q.all()):
if not quiet:
out.write("\r* hashes calculation: {} %".format(
int(float(idx + 1) / count * 100)))
out.flush()
path = doc.image.path
try:
hexdigest = get_hexdigest(path)
except IOError:
missing_images.append(doc.pk)
continue
if hexdigest not in hashes:
hashes[hexdigest] = []
hashes[hexdigest].append(doc.pk)
if not quiet:
out.write("\n* {} missing images\n".format(len(missing_images)))
attributes = ['title', 'associated_file', 'internal_reference',
'source_type', 'description']
if not ignore_ref:
attributes.append('reference')
nb_conflicted_items = 0
nb_merged_items = 0
distinct_image = 0
conflicts = []
merged = []
count = len(hashes)
for idx, hash in enumerate(hashes):
if not quiet:
out.write("\r* merge similar images: {} %".format(
int(float(idx + 1) / count * 100)))
out.flush()
if len(hashes[hash]) < 2:
distinct_image += 1
continue
items = [Document.objects.get(pk=pk) for pk in hashes[hash]]
ref_item = items[0]
other_items = items[1:]
for item in other_items:
ref_item = Document.objects.get(pk=ref_item.pk)
conflicted_values = []
for attr in attributes:
ref_value = getattr(ref_item, attr)
other_value = getattr(item, attr)
if ref_value:
if not other_value:
continue
if other_value != ref_value:
if attr == 'title' and merged_title:
setattr(ref_item, 'title', merged_title)
else:
conflicted_values.append(
(attr, ref_value, other_value)
)
else:
if not other_value:
continue
setattr(ref_item, attr, other_value)
if conflicted_values:
nb_conflicted_items += 1
for attr, ref_value, other_value in conflicted_values:
conflicts.append([
ref_item.pk, ref_item.reference.encode('utf-8'),
ref_item.cache_related_label.encode('utf-8'),
ref_item.image.name.encode('utf-8'),
item.pk, item.reference.encode('utf-8'),
item.cache_related_label.encode('utf-8'),
item.image.name.encode('utf-8'),
attr, unicode(ref_value).encode('utf-8'),
unicode(other_value).encode('utf-8')
])
continue
merged.append([
ref_item.pk, ref_item.reference.encode('utf-8'),
ref_item.cache_related_label.encode('utf-8'),
ref_item.image.name.encode('utf-8'),
item.pk, item.reference.encode('utf-8'),
item.cache_related_label.encode('utf-8'),
item.image.name.encode('utf-8'),
])
for author in ref_item.authors.all():
if author not in item.authors:
item.authors.add(author)
for author in item.authors.all():
if author not in ref_item.authors:
ref_item.authors.add(author)
for rel_attr in Document.RELATED_MODELS:
ref_rel_items = [
r.pk for r in getattr(ref_item, rel_attr).all()]
for rel_item in getattr(item, rel_attr).all():
if rel_item.pk not in ref_rel_items:
getattr(ref_item, rel_attr).add(rel_item)
ref_item.skip_history_when_saving = True
ref_item.save()
item.delete()
nb_merged_items += 1
if not quiet:
out.write(u"\n")
n = datetime.datetime.now().isoformat().split('.')[0].replace(':', '-')
if conflicts:
filename = output_path + u"{}-conflict.csv".format(n)
with open(filename, 'w') as csvfile:
writer = csv.writer(csvfile)
writer.writerow(
["Document 1 - pk", "Document 1 - Ref",
"Document 1 - related", "Document 1 - image path",
"Document 2 - pk", "Document 2 - Ref",
"Document 2 - related", "Document 2 - image path",
"Attribute", "Document 1 - value", "Document 2 - value"
]
)
for conflict in conflicts:
writer.writerow(conflict)
if not quiet:
out.write(u"* {} conflicted items ({})\n".format(
nb_conflicted_items, filename))
if merged:
filename = output_path + u"{}-merged.csv".format(n)
with open(filename, 'w') as csvfile:
writer = csv.writer(csvfile)
writer.writerow(
["Document 1 - pk", "Document 1 - Ref",
"Document 1 - related", "Document 1 - image path",
"Document 2 - pk", "Document 2 - Ref",
"Document 2 - related", "Document 2 - image path",
]
)
for merge in merged:
writer.writerow(merge)
if not quiet:
out.write(u"* {} merged items ({})\n".format(nb_merged_items,
filename))
if not quiet:
out.write("* {} distinct images\n".format(distinct_image))
|
