def addFiles(self, files):
nbfiles = len(files)
# make sources struct
hashlist = pHash.DPptrArray(nbfiles)
if (hashlist is None):
self.log.error("mem alloc error")
raise MemoryError("mem alloc error")
# make a datapoint for each file
count = 0
tmphash = 0x00000000
for f in files:
tmpdp = self.makeDatapoint(f)
tmpdp.thisown = 0
hashlist[count] = tmpdp
self.log.debug("file[%d] = %s" % (count, f))
count += 1
#end for files
self.log.debug("add %d files to file %s" % (count, self.db))
nbsaved = 0
if (not self.__DbExists()):
# add all files to MVPTree
ret = pHash.ph_save_mvptree(self.mvpfile, hashlist.cast(), count)
(retcode, nbsaved) = ret, count
else:
# add all files to MVPTree
ret = pHash.ph_add_mvptree(self.mvpfile, hashlist.cast(), count)
if (type(ret) is int):
self.log.error("error on ph_add_mvptree")
raise PHashException("error on ph_add_mvptree")
(retcode, nbsaved) = ret
# common error handling
if (retcode != pHash.PH_SUCCESS and retcode != pHash.PH_ERRCAP):
self.log.warning("could not complete query, %d" % (retcode))
raise PHashException("could not complete query, %d" % (retcode))
self.log.debug("number saved %d out of %d, ret code %d" %
(nbsaved, count, retcode))
return
def addFiles(self,files):
nbfiles=len(files)
# make sources struct
hashlist=pHash.DPptrArray(nbfiles)
if ( hashlist is None):
self.log.error("mem alloc error")
raise MemoryError("mem alloc error")
# make a datapoint for each file
count=0
tmphash=0x00000000
for f in files:
tmpdp=self.makeDatapoint(f)
tmpdp.thisown=0
hashlist[count]=tmpdp
self.log.debug("file[%d] = %s"%( count, f ) )
count+=1
#end for files
self.log.debug("add %d files to file %s"%(count,self.db))
nbsaved=0
if (not self.__DbExists()):
# add all files to MVPTree
ret = pHash.ph_save_mvptree(self.mvpfile, hashlist.cast(), count)
(retcode,nbsaved)=ret,count
else:
# add all files to MVPTree
ret = pHash.ph_add_mvptree(self.mvpfile, hashlist.cast(), count)
if (type(ret) is int):
self.log.error("error on ph_add_mvptree")
raise PHashException("error on ph_add_mvptree")
(retcode,nbsaved)=ret
# common error handling
if (retcode != pHash.PH_SUCCESS and retcode != pHash.PH_ERRCAP):
self.log.warning("could not complete query, %d"%(retcode))
raise PHashException("could not complete query, %d"%(retcode))
self.log.debug("number saved %d out of %d, ret code %d"%( nbsaved,count,retcode))
return
def main(argv):
'''
'''
#locale.setlocale(locale.LC_ALL,'fr_FR')
#logger=logging.getLogger('root')
#logger.setLevel(logging.DEBUG)
#logging.basicConfig(level=logging.INFO)
logging.basicConfig(level=logging.DEBUG)
print pHash.ph_about()
if (len(argv) < 2):
print "not enough input args"
print "usage: %s dirname filename" % (sys.argv[0])
return -1
dir_name = argv[0] #;/* name of dir to retrieve image files */
filename = argv[1] #;/* name of file to save db */
mvpfile = pHash.MVPFile()
mvpfile.branchfactor = 2
mvpfile.pathlength = 5
mvpfile.leafcapacity = 23 #50
mvpfile.pgsize = 4096 #8192
mvpfile.filename = filename
#mvpfile.hashdist = distancefunc #save: ret code 17
pHash.my_set_callback(mvpfile, distancefunc)
mvpfile.hash_type = pHash.UINT64ARRAY
nbfiles = 0
print "dir name: %s" % (dir_name)
for root, dirs, files in os.walk(dir_name):
nbfiles = len(files)
print "nbfiles = %d" % nbfiles
#allocate a list of nbfiles elements # hashlist = (DP**)malloc(nbfiles*sizeof(DP*));
hashlist = pHash.DPptrArray(nbfiles)
count = 0
for i in range(0, nbfiles):
filename = os.path.normpath(os.path.join(root, files[i]))
# malloc DP or use pHash.DP() // pHash.ph_malloc_datapoint(mvpfile.hash_type)
#tmp=pHash.ph_malloc_datapoint(mvpfile.hash_type)
tmp = pHash.DP()
if (tmp is None):
print "mem alloc error"
return -4
tmp.hash_type = mvpfile.hash_type
# say to python, to NOT garbage collect DP reference.
tmp.thisown = 0
# store the ref.
hashlist[count] = tmp
# calculate image hash
ret, tmphash = pHash.ph_dct_imagehash(filename)
if (ret < 0):
print "unable to get hash"
continue
# we can't assign .hash to hashlist[count].hash = tmphash because .hash is a pointer
# we use ulong64Ptr instead of voidPtr because .. it's a ulong64 ? casting is dynamic
hashlist[count].hash = pHash.copy_ulong64Ptr(tmphash)
#
print "files[%d]: %s hash = %x" % (i, filename, tmphash)
hashlist[count].id = filename
hashlist[count].hash_length = 1
count += 1
# method with DPptrArray being a DP **
hashlistf = hashlist.cast()
ret = pHash.ph_save_mvptree(mvpfile, hashlistf, count)
# ret 11 is null hash distance function
# save: ret code 17 has callback func. -> not enought hashdist ?
print "save: ret code %d" % (ret)
#free is done by GC .. ?
#for i in range(0,nbfiles):
# pHash.free(hashlist[i].hash)
#free is NOT done by GC
for i in range(0, nbfiles):
pHash.ph_free_datapoint(hashlist[i])
def main(argv):
'''
'''
#locale.setlocale(locale.LC_ALL,'fr_FR')
#logger=logging.getLogger('root')
#logger.setLevel(logging.DEBUG)
#logging.basicConfig(level=logging.INFO)
logging.basicConfig(level=logging.DEBUG)
print pHash.ph_about()
if (len(argv) < 2):
print "not enough input args"
print "usage: %s dirname filename"% (sys.argv[0])
return -1
dir_name = argv[0] #;/* name of dir to retrieve image files */
filename = argv[1] #;/* name of file to save db */
mvpfile= pHash.MVPFile()
mvpfile.branchfactor = 2
mvpfile.pathlength = 5
mvpfile.leafcapacity = 23 #50
mvpfile.pgsize = 4096 #8192
mvpfile.filename = filename
#mvpfile.hashdist = distancefunc #save: ret code 17
pHash.my_set_callback(mvpfile,distancefunc)
mvpfile.hash_type = pHash.UINT64ARRAY
nbfiles = 0
print "dir name: %s"%( dir_name)
for root, dirs, files in os.walk(dir_name):
nbfiles=len(files)
print "nbfiles = %d"% nbfiles
#allocate a list of nbfiles elements # hashlist = (DP**)malloc(nbfiles*sizeof(DP*));
hashlist=pHash.DPptrArray(nbfiles)
count = 0
for i in range(0,nbfiles):
filename=os.path.normpath(os.path.join(root,files[i]) )
# malloc DP or use pHash.DP() // pHash.ph_malloc_datapoint(mvpfile.hash_type)
#tmp=pHash.ph_malloc_datapoint(mvpfile.hash_type)
tmp=pHash.DP()
if (tmp is None):
print "mem alloc error"
return -4
tmp.hash_type = mvpfile.hash_type
# say to python, to NOT garbage collect DP reference.
tmp.thisown = 0
# store the ref.
hashlist[count]=tmp
# calculate image hash
ret,tmphash=pHash.ph_dct_imagehash(filename)
if ( ret < 0):
print "unable to get hash"
continue
# we can't assign .hash to hashlist[count].hash = tmphash because .hash is a pointer
# we use ulong64Ptr instead of voidPtr because .. it's a ulong64 ? casting is dynamic
hashlist[count].hash = pHash.copy_ulong64Ptr(tmphash)
#
print "files[%d]: %s hash = %x"%( i, filename, tmphash )
hashlist[count].id = filename
hashlist[count].hash_length = 1
count+=1
# method with DPptrArray being a DP **
hashlistf=hashlist.cast()
ret = pHash.ph_save_mvptree(mvpfile, hashlistf, count)
# ret 11 is null hash distance function
# save: ret code 17 has callback func. -> not enought hashdist ?
print "save: ret code %d"%(ret)
#free is done by GC .. ?
#for i in range(0,nbfiles):
# pHash.free(hashlist[i].hash)
#free is NOT done by GC
for i in range(0,nbfiles):
pHash.ph_free_datapoint(hashlist[i])