def delete_from_db(condition=None):
dbm = DBManager(expanduser(settings.IMAGES_DB))
dbm.connect()
count = dbm.delete('images', condition)
dbm.commit()
dbm.disconnect()
return count
def delete_from_db(condition=None):
dbm = DBManager(expanduser(settings.IMAGES_DB))
dbm.connect()
count = dbm.delete('images', condition)
dbm.commit()
dbm.disconnect()
return count
def clear_duplicate_images(arg):
dbm = DBManager(expanduser(settings.IMAGES_DB))
dbm.connect()
jobname = arg
result = dbm.query('SELECT path FROM images WHERE job = \'%s\' LIMIT 1'%jobname)
if len(result) == 0:
print('no such job')
imagepath = result[0]['path']
jobpath = imagepath[0:imagepath.rfind(jobname)+len(jobname)]
print jobpath
filelist = []
for root, dirs, files in os.walk(jobpath):
for filename in files:
md5hash = md5()
filepath = joinpath(root, filename)
with open(filepath, 'rb') as f:
md5hash.update(f.read())
filehash = md5hash.hexdigest()
filelist.append(FileItem(filepath, filehash, False))
dups_total = 0
for i in range(0, len(filelist)):
if filelist[i].dup:
continue
hash = filelist[i].hash
same_files = [(filelist[i].path, os.stat(filelist[i].path).st_mtime)]
for j in range(i + 1, len(filelist)):
if filelist[j].hash == hash:
same_files.append((filelist[j].path, os.stat(filelist[j].path).st_mtime))
filelist[j] = FileItem(None, None, True)
if len(same_files) > 1:
min_mtime = sys.float_info.max
keep = -1
for i in range(0, len(same_files)):
if same_files[i][1] < min_mtime:
min_mtime = same_files[i][1]
keep = i
for i in range(0, len(same_files)):
if i != keep:
dups_total += 1
print('deleting %s'%same_files[i][0])
try:
os.remove(same_files[i][0])
except OSError as e:
print(e.message)
dbm.query('UPDATE images SET path = \'#duplicate\' WHERE path = \'%s\''%same_files[i][0])
dbm.commit()
dbm.disconnect()
print('%d duplicate images deleted.'%dups_total)
def clear_duplicate_images(arg):
dbm = DBManager(expanduser(settings.IMAGES_DB))
dbm.connect()
jobname = arg
result = dbm.query('SELECT path FROM images WHERE job = \'%s\' LIMIT 1' %
jobname)
if len(result) == 0:
print('no such job')
imagepath = result[0]['path']
jobpath = imagepath[0:imagepath.rfind(jobname) + len(jobname)]
print jobpath
filelist = []
for root, dirs, files in os.walk(jobpath):
for filename in files:
md5hash = md5()
filepath = joinpath(root, filename)
with open(filepath, 'rb') as f:
md5hash.update(f.read())
filehash = md5hash.hexdigest()
filelist.append(FileItem(filepath, filehash, False))
dups_total = 0
for i in range(0, len(filelist)):
if filelist[i].dup:
continue
hash = filelist[i].hash
same_files = [(filelist[i].path, os.stat(filelist[i].path).st_mtime)]
for j in range(i + 1, len(filelist)):
if filelist[j].hash == hash:
same_files.append(
(filelist[j].path, os.stat(filelist[j].path).st_mtime))
filelist[j] = FileItem(None, None, True)
if len(same_files) > 1:
min_mtime = sys.float_info.max
keep = -1
for i in range(0, len(same_files)):
if same_files[i][1] < min_mtime:
min_mtime = same_files[i][1]
keep = i
for i in range(0, len(same_files)):
if i != keep:
dups_total += 1
print('deleting %s' % same_files[i][0])
try:
os.remove(same_files[i][0])
except OSError as e:
print(e.message)
dbm.query(
'UPDATE images SET path = \'#duplicate\' WHERE path = \'%s\''
% same_files[i][0])
dbm.commit()
dbm.disconnect()
print('%d duplicate images deleted.' % dups_total)
class ImageStorePipeline(object):
def __init__(self):
if exists(settings.IMAGES_DB):
self._dbm = DBManager(settings.IMAGES_DB)
self._dbm.connect()
self._nodb = False
log.debug('opened db: %s' % settings.IMAGES_DB)
else:
self._nodb = True
log.debug('could not open db: %s' % settings.IMAGES_DB)
def process_item(self, item, spider):
if isinstance(item, ImageItem):
images = item.get('images', None)
final_storepath = joinpath(settings.IMAGES_STORE_FINAL,
spider.jobname)
if images:
for d in item['images']:
ext = d['path'][d['path'].rfind('.') + 1:]
filebasename, ext = self.get_filename(d['url'])
final_path = joinpath(final_storepath,
filebasename + '.' + ext)
i = 0
while exists(final_path):
log.debug(final_path + ' exists')
final_path = joinpath(
final_storepath,
filebasename + '_%02d' % i + '.' + ext)
i += 1
try:
os.rename(joinpath(settings.IMAGES_STORE, d['path']),
final_path)
log.debug('moved to: ' + final_path)
spider.update_monitor(final_path)
if not self._nodb:
self._dbm.insert('images',
(d['url'], final_path,
spider.jobname, int(time())))
except OSError as e:
log.error(e)
if not self._nodb:
self._dbm.commit()
return item
def get_filename(self, url):
url_parts = url.split('/')
del url_parts[0:2]
filename = url_parts.pop()
ext = filename[filename.rfind('.') + 1:]
filename = filename[0:filename.rfind('.')]
url_parts.append(filename)
part = filename
words = []
while not any([len(w) > 2 for w in words]) and len(url_parts) > 0:
part = url_parts.pop()
part = part.replace('-', '_').replace('.', '_').replace('+', '_')
pwords = part.split('_')
pwords = [
w for w in pwords
if (len(w) > 0 and len(w) <= 2) or (len(w) > 2 and (
len([c for c in w if
(ord(c) >= 48 and ord(c) <= 57)]) < len(w) / 2))
]
words = pwords + words
final = ''
if len(words) > 0:
for w in words:
final += w + '_'
final = final.rstrip('_')
else:
final = 'image.'
return final, ext
def __del__(self):
self._dbm.disconnect()