def get_info(self, xmlfile):
#返回字典
try:
xml = parse_xml(xmlfile, show_log=False)
#判断文件
xml.xml_exists()
xml.get_root()
all = {}
#dirc={}
childs = xml.get_element_children(xml.root)
for child in childs:
plat = child.get("platform")
cid = child.get("id")
ip = child.get("ip")
sshport = int(child.get("port"))
mark = plat + "_s" + cid
dirc = {mark: {}}
dirc[mark].update({
"ip": ip,
"id": cid,
"platform": plat,
"port": sshport
})
all.update(dirc)
if self.show_log:
for agent in all:
print agent
return all
except Exception as err:
print "获取失败.错误如下:", err
def read_label_from_xml(label_path):
"""Read label from xml file.
# Returns:
labe l_dic (dictionary): labels for one sequence.
size (list): Bounding Box Size. [l, w. h]?
"""
labels = pt.parse_xml(label_path)
label_dic = {}
for label in labels:
first_frame = label.first_frame
num_frames = label.num_frames
size = label.size
obj_type = label.object_type
for index, place, rot in zip(
range(first_frame, first_frame + num_frames), label.trans,
label.rots):
if index in label_dic.keys():
label_dic[index]["trans"] = np.vstack(
(label_dic[index]["trans"], place))
label_dic[index]["size"] = np.vstack(
(label_dic[index]["size"], np.array(size)))
label_dic[index]["rot"] = np.vstack(
(label_dic[index]["rot"], rot))
else:
label_dic[index] = {}
label_dic[index]["trans"] = place
label_dic[index]["rot"] = rot
label_dic[index]["size"] = np.array(size)
return label_dic, size
def get_info(self,xmlfile):
#返回字典
try:
xml=parse_xml(xmlfile,show_log=False)
#判断文件
xml.xml_exists()
xml.get_root()
all={}
#dirc={}
childs=xml.get_element_children(xml.root)
for child in childs:
plat=child.get("platform")
cid=child.get("id")
ip=child.get("ip")
sshport=int(child.get("port"))
mark=plat+"_s"+cid
dirc={mark:{}}
dirc[mark].update({"ip":ip,"id":cid,"platform":plat,"port":sshport})
all.update(dirc)
if self.show_log:
for agent in all:
print agent
return all
except Exception as err:
print "获取失败.错误如下:",err
def get_repeat_ip(self, xmlfile):
try:
xml = parse_xml(xmlfile, show_log=False)
#判断文件
xml.xml_exists()
if self.writefile:
name = xmlfile.split(".")[0]
namefile = "%s_ip.txt" % name
if os.path.exists(namefile):
os.remove(namefile)
file = open(namefile, 'w+')
xml.get_root()
childs = xml.get_element_children(xml.root)
ips = []
for child in childs:
ips.append(child.get("ip"))
ips = sorted(list(set(ips)))
if self.show_log:
for ip in ips:
print ip
if self.writefile:
file.write(ip)
file.write("\n")
if self.writefile:
file.close()
return ips
except Exception as err:
print err
def get_repeat_ip(self,xmlfile):
try:
xml=parse_xml(xmlfile,show_log=False)
#判断文件
xml.xml_exists()
if self.writefile:
name=xmlfile.split(".")[0]
namefile="%s_ip.txt"%name
if os.path.exists(namefile):
os.remove(namefile)
file=open(namefile,'w+')
xml.get_root()
childs=xml.get_element_children(xml.root)
ips=[]
for child in childs:
ips.append(child.get("ip"))
ips=sorted(list(set(ips)))
if self.show_log:
for ip in ips:
print ip
if self.writefile:
file.write(ip)
file.write("\n")
if self.writefile:
file.close()
return ips
except Exception as err:
print err
def main():
# List organization names found in document
orgs = []
file = input('Enter the name of a file you wish to enhance: ')
namespaces = {'gmi': "http://www.isotc211.org/2005/gmi", 'gmd': "http://www.isotc211.org/2005/gmd", 'gco':
"http://www.isotc211.org/2005/gco", 'gml': "http://www.opengis.net/gml/3.2", 'gmx':
"http://www.isotc211.org/2005/gmx", 'gsr': "http://www.isotc211.org/2005/gsr",
'gss': "http://www.isotc211.org/2005/gss", 'gts': "http://www.isotc211.org/2005/gts",
'xlink': "http://www.w3.org/1999/xlink", 'xsi': "http://www.w3.org/2001/XMLSchema-instance",
'saxon': "http://saxon.sf.net/", 'srv': "http://www.isotc211.org/2005/srv",
'schemaLocation': "http://www.isotc211.org/2005/gmi http://www.ngdc.noaa.gov/metadata/"
"published/xsd/schema.xsd"}
#tree = ET.parse(urlopen(file))
tree = ET.parse(file)
root = tree.getroot()
for prefix in namespaces:
ET.register_namespace(prefix, namespaces[prefix])
# Read document and populate orgs list with raw organization names found
if parse_xml(root, orgs):
if len(orgs) == 0:
print('No organization names could be found in the document. Exiting...')
exit()
option = input(choice)
if option == 'I':
for each in orgs:
print(each.name)
next_thing = input('(E)nhance organization or press any other key to continue '
'iterating through organizations found.\n')
if next_thing == 'E':
each.validate_in_viaf()
print(each.enhancement_info())
if next_thing == 'C':
continue
enhanced_orgs = []
enhance_doc = input('Enhance document? If no, program will exit [Y/N]: ')
if enhance_doc == 'N':
exit()
if enhance_doc == 'Y':
for each in orgs:
if each.validated:
enhanced_orgs.append(each)
if len(enhanced_orgs) == 0:
print('No enhancements could be made. Exiting.')
exit()
# Enhance xml document
enhance_xml(root, enhanced_orgs)
rough_string = ET.tostring(root, 'utf-8')
parsed = xml.dom.minidom.parseString(rough_string)
#print(parsed.toprettyxml('\t'))
f = open('enhanced.xml', 'wb+')
f.write(parsed.toxml('utf-8'))
f.close()
print('Enhanced document created. See enhanced.xml')
else:
print('Problem reading document')
def snip_illustrations(zp, filename, id, vol, page):
try:
img_color = cv2.imread(extract_jp2(zp, filename))
delete_jp2(filename)
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(id)
if not os.path.exists(os.path.join(cover_path, guesseddate)):
os.mkdir(os.path.join(cover_path, guesseddate))
cv2.imwrite(os.path.join(cover_path, guesseddate, "{0}_{1}_{2}_{3}.jpg".format(id, vol, make_safe(title), guesseddate)), img_color, [cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93] )
print("{0}_{1}_{2}_{3}.jpg".format(id, vol, make_safe(title), guesseddate))
except KeyError, e:
r.rpoplpush(workerid, "coverproblems")
def read_objects(tracklet_file, num_frames):
objects = [] # grouped by frames
for n in range(num_frames):
objects.append([])
# read tracklets from file
tracklets = parse_xml(tracklet_file)
num = len(tracklets)
for n in range(num):
tracklet = tracklets[n]
# this part is inspired by kitti object development kit matlab code: computeBox3D
h, w, l = tracklet.size
trackletBox = np.array([ # in velodyne coordinates around zero point and without orientation yet\
[-l / 2, -l / 2, l / 2, l / 2, -l / 2, -l / 2, l / 2, l / 2],
[w / 2, -w / 2, -w / 2, w / 2, w / 2, -w / 2, -w / 2, w / 2],
[0.0, 0.0, 0.0, 0.0, h, h, h, h]])
# loop over all data in tracklet
t = tracklet.firstFrame
for translation, rotation, state, occlusion, truncation, amtOcclusion, amtBorders, absoluteFrameNumber in tracklet:
# determine if object is in the image; otherwise continue
if truncation not in (TRUNC_IN_IMAGE, TRUNC_TRUNCATED):
continue
# re-create 3D bounding box in velodyne coordinate system
yaw = rotation[2] # other rotations are 0 in all xml files I checked
assert np.abs(rotation[:2]).sum() == 0, 'object rotations other than yaw given!'
rotMat = np.array([
[np.cos(yaw), -np.sin(yaw), 0.0],
[np.sin(yaw), np.cos(yaw), 0.0],
[0.0, 0.0, 1.0]
])
cornerPosInVelo = np.dot(rotMat, trackletBox) + np.tile(translation, (8, 1)).T
# calc yaw as seen from the camera (i.e. 0 degree = facing away from cam), as opposed to
# car-centered yaw (i.e. 0 degree = same orientation as car).
# makes quite a difference for objects in periphery!
# Result is in [0, 2pi]
x, y, z = translation
yawVisual = (yaw - np.arctan2(y, x)) % (2 * math.pi)
o = type('', (), {})()
o.box = cornerPosInVelo.transpose()
o.type = tracklet.objectType
o.tracklet_id = n
objects[t].append(o)
t = t + 1
return objects
def gather_signals(filepath):
print("Gathering signals for '{0}'".format(filepath))
filename = filepath.split("/")[-1]
id, vol, page, imgno = breakdown_imagename(filename.split("/")[-1])
metadata = parse_xml(id)
cv_signals = []
for cv_measure in CV_SIGNALS:
signal = cv_measure(filepath)
if isinstance(signal, tuple):
cv_signals.extend(signal)
else:
cv_signals.append(signal)
return {'metadata': metadata,
'identifier': (id, vol, page, imgno),
'cv_signals': cv_signals}
def convert_component(path: Union[str, PathLike], mc=False) -> Path:
root_file = Path(path)
if not root_file.exists():
raise FileNotFoundError(f'File does not exists: {root_file}')
elif root_file.suffix != '.xml':
raise ValueError(f'Expected xml file, got {root_file.suffix}')
rename_to = root_file.with_suffix('.dot')
if rename_to.exists():
pass
#return rename_to
graph_repr = parse_xml(root_file)
modal_to_dot(str(rename_to.absolute()), graph_repr, mc=mc)
return rename_to
def get_all_info(self, xmlfile):
xml = parse_xml(xmlfile, show_log=False)
#判断文件
xml.xml_exists()
xml.get_root()
all = []
childs = xml.get_element_children(xml.root)
for child in childs:
l = []
cid = child.get("id")
ip = child.get("ip")
plat = child.get("platform")
l.append(cid)
l.append(ip)
l.append(plat)
all.append(l)
return all
def get_all_info(self,xmlfile):
xml=parse_xml(xmlfile,show_log=False)
#判断文件
xml.xml_exists()
xml.get_root()
all=[]
childs=xml.get_element_children(xml.root)
for child in childs:
l=[]
cid=child.get("id")
ip=child.get("ip")
plat=child.get("platform")
l.append(cid)
l.append(ip)
l.append(plat)
all.append(l)
return all
def convert(args):
default_options = {
'input': '',
'output': '',
'mc': False,
'auto_group': False
}
default_options.update(vars(args))
ipath = Path(default_options['input'])
opath = Path(default_options['output'])
pprint(default_options)
if not ipath.exists():
log.error(f'Given path does not exists, can not read: {ipath}')
exit(-1)
if ipath.suffix not in ['.dot', '.xml']:
log.error(f'Unknown suffix: {ipath.suffix}')
exit(-1)
if opath.suffix not in ['.dot', '.xml']:
log.error(f'Unknown suffix: {opath.suffix}')
exit(-1)
mts_repr = None
if ipath.suffix == '.dot':
log.debug(f'Read input (dot): {ipath}')
mts_repr = parse_dot(ipath)
elif ipath.suffix == '.xml':
log.debug(f'Read input (xml): {ipath}')
mts_repr = parse_xml(ipath)
if opath.suffix == '.dot':
log.info(f'Writing to {opath} ...')
modal_to_dot(path=opath,
mts=mts_repr,
mc=default_options['mc'],
derive_groups=default_options['auto_group'])
elif opath.suffix == '.xml':
log.info(f'Writing to {opath} ...')
modal_to_xml(path=opath,
mts=mts_repr,
mc=default_options['mc'],
derive_groups=default_options['auto_group'])
def snip_illustrations(zp, filename, id, vol, page):
try:
img_color = cv2.imread(extract_jp2(zp, filename))
delete_jp2(filename)
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(
id)
if not os.path.exists(os.path.join(cover_path, guesseddate)):
os.mkdir(os.path.join(cover_path, guesseddate))
cv2.imwrite(
os.path.join(
cover_path, guesseddate,
"{0}_{1}_{2}_{3}.jpg".format(id, vol, make_safe(title),
guesseddate)), img_color,
[cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
print("{0}_{1}_{2}_{3}.jpg".format(id, vol, make_safe(title),
guesseddate))
except KeyError, e:
r.rpoplpush(workerid, "coverproblems")
def get_md(item):
job, flickr_id = item.strip().split("\t")
img = job.strip()[24:]
location = os.path.join(embellishments_path, img)
id, vol, page, _ = img.split("/", 1)[1].split("_", 3)
metadata = parse_xml(id)
identifier, title, author, pubplace, publisher, guesseddate = metadata
decoded = map(lambda x: x.encode("utf-8"), [
title, author, guesseddate, pubplace, publisher, vol,
str(int(page)), identifier
])
uploaded_title = u"Image taken from page {1} of '{0}'".format(
decoded[0].decode("utf-8"), page.decode("utf-8"))
try:
uploaded_title = u"Image taken from page {1} of '{0}'".format(
decoded[0].decode("utf-8"),
str(int(page)).decode("utf-8"))
except:
pass
if os.path.exists(location):
try:
rosetta = r.get("s:" + id)
if rosetta:
d, a, u = rosetta.strip().split("\t")
adjusted_ark_number = "{0:012x}".format(
int(a.split("_")[1], 16) - 1)
adj_ark = "{0}_{1}".format(
a.split("_")[0], adjusted_ark_number.upper())
hexpage = "0x000001"
try:
hexpage = "{0:06x}".format(int(page)).upper()
except:
pass
additional = """
<ul><li>Open the page in the <a href="http://itemviewer.bl.uk/?itemid={2}#{1}.0x{0}">British Library's itemViewer (page: {3})</a></li>
<li><a href="http://access.dl.bl.uk/{2}">Download the PDF for this book</a>
""".format(hexpage, adj_ark, u, page)
decoded += [additional]
else:
decoded += [u""]
except Exception, e:
print e
def get_md(item):
job, flickr_id = item.strip().split("\t")
img = job.strip()[24:]
location = os.path.join(embellishments_path, img)
id, vol, page, _ = img.split("/",1)[1].split("_",3)
metadata = parse_xml(id)
identifier, title, author, pubplace, publisher, guesseddate = metadata
decoded = map(lambda x: x.encode("utf-8"), [title, author, guesseddate, pubplace, publisher, vol, str(int(page)), identifier])
uploaded_title=u"Image taken from page {1} of '{0}'".format(decoded[0].decode("utf-8"), page.decode("utf-8"))
try:
uploaded_title = u"Image taken from page {1} of '{0}'".format(decoded[0].decode("utf-8"), str(int(page)).decode("utf-8"))
except:
pass
if os.path.exists(location):
try:
rosetta = r.get("s:"+id)
if rosetta:
d, a, u = rosetta.strip().split("\t")
adjusted_ark_number = "{0:012x}".format(int(a.split("_")[1], 16) - 1)
adj_ark = "{0}_{1}".format(a.split("_")[0], adjusted_ark_number.upper())
hexpage = "0x000001"
try:
hexpage = "{0:06x}".format(int(page)).upper()
except:
pass
additional = """
<ul><li>Open the page in the <a href="http://itemviewer.bl.uk/?itemid={2}#{1}.0x{0}">British Library's itemViewer (page: {3})</a></li>
<li><a href="http://access.dl.bl.uk/{2}">Download the PDF for this book</a>
""".format(hexpage, adj_ark, u, page)
decoded += [additional]
else:
decoded += [u""]
except Exception, e:
print e
def snip_illustrations(zp,
filename,
altoxmloldpath,
scale=1.1,
threshold=1200 * 600):
id, vol, page = altopath_to_idvol(altoxmloldpath)
altofilepath = old_winpath_to_vmpath(altoxmloldpath)
doc = []
with open(altofilepath, "r") as xmlfile:
doc = ET.fromstring(xmlfile.read())
page_shape, images = get_illustration_coords(doc)
if images and [x for x in images if x[2] * x[3] < threshold]:
counter = 0
skippable = True
for rect in images:
if rect[2] * rect[3] < threshold:
skippable = False
if skippable:
print("No suitable images found within '{0}' - skipping".format(
filename))
return
else:
print("Page contains a relevant illustration, not skipping.")
img_color = ""
#img_color = cv2.imread(extract_jp2(zp, filename))
#delete_jp2(filename)
#h,w,_ = img_color.shape
#current_page_shape = (w,h)
#dh = float(h)/float(page_shape[1])
#dw = float(w)/float(page_shape[0])
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(
id)
for rect in images:
counter += 1
if rect[2] * rect[3] < threshold:
print("Attempting to slice '{3}' from {0}_{1}_{2}".format(
id, vol, page, rect))
#scaled = increase_size(rect, scale, (dw,dh), current_page_shape)
#boundary = get_rect(scaled)
if len(guesseddate) != 4:
guesseddate = "Unknown"
if not os.path.exists(os.path.join(embellish_path,
guesseddate)):
os.mkdir(os.path.join(embellish_path, guesseddate))
img_filename = os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(
id, vol, page, counter, make_safe(title), guesseddate))
short_filename = os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(
id, vol, page, counter,
make_safe(title)[:80], guesseddate))
# if not already captured this, switch to new short title and carry on
if not os.path.exists(img_filename):
img_filename = short_filename
# Has this been captured before?
try:
if not os.path.exists(img_filename):
if img_color == "":
img_color = cv2.imread(
extract_jp2(zp, filename))
delete_jp2(filename)
h, w, _ = img_color.shape
current_page_shape = (w, h)
dh = float(h) / float(page_shape[1])
dw = float(w) / float(page_shape[0])
scaled = increase_size(rect, scale, (dw, dh),
current_page_shape)
boundary = get_rect(scaled)
cv2.imwrite(
img_filename,
img_color[boundary[0][1]:boundary[1][1],
boundary[0][0]:boundary[1][0]],
[cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
print("{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(
id, vol, page, counter, make_safe(title),
guesseddate))
if os.path.isfile(img_filename):
r.lpush("secondpassembellish", img_filename)
noplates = r.llen("secondpassembellish")
print("{1} - {0}".format(
img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("embellisherror", img_filename)
else:
print("Already captured, skipping")
except:
cv2.imwrite(
os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}.jpg".format(
id, vol, page, counter, guesseddate)),
img_color[boundary[0][1]:boundary[1][1],
boundary[0][0]:boundary[1][0]],
[cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
print("{0}_{1}_{2}_{3}_{4}.jpg".format(
id, vol, page, counter, guesseddate))
else:
print("Already captured, skipping")
else:
print(
"All illustration areas in {1}(vol:{2}, pg: {3} are above threshold area size (currently: {0})"
.format(threshold, id, vol, page))
import flickr_api
a = flickr_api.auth.AuthHandler.load("BLLibraryAuth")
flickr_api.set_auth_handler(a)
if __name__ == "__main__":
workerid = "uploadwrk"+sys.argv[1]
while(True):
job = get_job(workerid)
while(job):
if job != "":
img = job.strip()
location = os.path.join(embellishments_path, img)
# get id
id, vol, page, _ = img.split("/",1)[1].split("_",3)
metadata = parse_xml(id)
"""
Arguments:
photo_file
The file to upload.
title (optional)
The title of the photo.
description (optional)
A description of the photo. May contain some limited HTML.
tags (optional)
A space-seperated list of tags to apply to the photo.
is_public, is_friend, is_family (optional)
Set to 0 for no, 1 for yes. Specifies who can view the photo.
safety_level (optional)
Set to 1 for Safe, 2 for Moderate, or 3 for Restricted.
content_type (optional)
def snip_illustrations(zp, filename, altoxmloldpath, scale = 1.1, threshold = 1200*600, zipfilename=""):
id, vol, page = altopath_to_idvol(altoxmloldpath)
start, end = get_extent(id, zipfilename)
if int(page) == start or int(page) == end:
print("This looks like the book cover - page '{0}' - extent ({1}, {2})".format(int(page), start, end))
return
altofilepath = old_winpath_to_vmpath(altoxmloldpath)
doc = []
with open(altofilepath, "r") as xmlfile:
doc = ET.fromstring(xmlfile.read())
page_shape, images = get_illustration_coords(doc)
if images and [x for x in images if x[2]*x[3] > threshold]:
counter = 0
img_color = cv2.imread(extract_jp2(zp, filename))
delete_jp2(filename)
h,w,_ = img_color.shape
current_page_shape = (w,h)
dh = float(h)/float(page_shape[1])
dw = float(w)/float(page_shape[0])
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(id)
for rect in images:
counter += 1
if rect[2]*rect[3] > threshold:
print("Attempting to slice '{1}' from {0}".format(filename, rect))
scaled = increase_size(rect, scale, (dw,dh), current_page_shape)
boundary = get_rect(scaled)
if len(guesseddate) != 4:
guesseddate = "Unknown"
if not os.path.exists(os.path.join(embellish_path, guesseddate)):
os.mkdir(os.path.join(embellish_path, guesseddate))
img_filename = os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(id, vol, page, counter, make_safe(title), guesseddate))
short_filename = os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(id, vol, page, counter, make_safe(title)[:80], guesseddate))
# if not already captured this, switch to new short title and carry on
if not os.path.exists(img_filename):
img_filename = short_filename
# Has this been captured before?
try:
if not os.path.exists(img_filename):
cv2.imwrite(img_filename, img_color[boundary[0][1]:boundary[1][1], boundary[0][0]:boundary[1][0]], [cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93] )
if os.path.isfile(img_filename):
r.lpush(plateimgs, img_filename)
noplates = r.llen(plateimgs)
print("Plate no: {1} - {0}".format(img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("plateerror", altoxmloldpath)
else:
print("Already captured, skipping")
except:
img_filename = os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}.jpg".format(id, vol, page, counter, guesseddate))
cv2.imwrite(img_filename, img_color[boundary[0][1]:boundary[1][1], boundary[0][0]:boundary[1][0]], [cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
if os.path.isfile(img_filename):
r.lpush(plateimgs, img_filename)
noplates = r.llen(plateimgs)
print("Plate no: {1} - {0}".format(img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("plateerror", altoxmloldpath)
else:
print("All illustration areas in {1}(vol:{2}, pg: {3} are below threshold area size (currently: {0})".format(threshold, id, vol, page))
def snip_illustrations(zp, filename, altoxmloldpath, scale = 1.1, threshold = 1200*600):
id, vol, page = altopath_to_idvol(altoxmloldpath)
altofilepath = old_winpath_to_vmpath(altoxmloldpath)
doc = []
with open(altofilepath, "r") as xmlfile:
doc = ET.fromstring(xmlfile.read())
page_shape, images = get_illustration_coords(doc)
if images and [x for x in images if x[2]*x[3] < threshold]:
counter = 0
skippable = True
for rect in images:
if rect[2]*rect[3] < threshold:
skippable = False
if skippable:
print("No suitable images found within '{0}' - skipping".format(filename))
return
else:
print("Page contains a relevant illustration, not skipping.")
img_color = ""
#img_color = cv2.imread(extract_jp2(zp, filename))
#delete_jp2(filename)
#h,w,_ = img_color.shape
#current_page_shape = (w,h)
#dh = float(h)/float(page_shape[1])
#dw = float(w)/float(page_shape[0])
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(id)
for rect in images:
counter += 1
if rect[2]*rect[3] < threshold:
print("Attempting to slice '{3}' from {0}_{1}_{2}".format(id, vol, page, rect))
#scaled = increase_size(rect, scale, (dw,dh), current_page_shape)
#boundary = get_rect(scaled)
if len(guesseddate) != 4:
guesseddate = "Unknown"
if not os.path.exists(os.path.join(embellish_path, guesseddate)):
os.mkdir(os.path.join(embellish_path, guesseddate))
img_filename = os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(id, vol, page, counter, make_safe(title), guesseddate))
short_filename = os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(id, vol, page, counter, make_safe(title)[:80], guesseddate))
# if not already captured this, switch to new short title and carry on
if not os.path.exists(img_filename):
img_filename = short_filename
# Has this been captured before?
try:
if not os.path.exists(img_filename):
if img_color == "":
img_color = cv2.imread(extract_jp2(zp, filename))
delete_jp2(filename)
h,w,_ = img_color.shape
current_page_shape = (w,h)
dh = float(h)/float(page_shape[1])
dw = float(w)/float(page_shape[0])
scaled = increase_size(rect, scale, (dw,dh), current_page_shape)
boundary = get_rect(scaled)
cv2.imwrite(img_filename, img_color[boundary[0][1]:boundary[1][1], boundary[0][0]:boundary[1][0]], [cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93] )
print("{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(id, vol, page, counter, make_safe(title), guesseddate))
if os.path.isfile(img_filename):
r.lpush("secondpassembellish", img_filename)
noplates = r.llen("secondpassembellish")
print("{1} - {0}".format(img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("embellisherror", img_filename)
else:
print("Already captured, skipping")
except:
cv2.imwrite(os.path.join(embellish_path, guesseddate, "{0}_{1}_{2}_{3}_{4}.jpg".format(id, vol, page, counter, guesseddate)),
img_color[boundary[0][1]:boundary[1][1], boundary[0][0]:boundary[1][0]], [cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
print("{0}_{1}_{2}_{3}_{4}.jpg".format(id, vol, page, counter, guesseddate))
else:
print("Already captured, skipping")
else:
print("All illustration areas in {1}(vol:{2}, pg: {3} are above threshold area size (currently: {0})".format(threshold, id, vol, page))
def snip_illustrations(zp,
filename,
altoxmloldpath,
scale=1.1,
threshold=1200 * 600,
zipfilename=""):
id, vol, page = altopath_to_idvol(altoxmloldpath)
start, end = get_extent(id, zipfilename)
if int(page) == start or int(page) == end:
print(
"This looks like the book cover - page '{0}' - extent ({1}, {2})".
format(int(page), start, end))
return
altofilepath = old_winpath_to_vmpath(altoxmloldpath)
doc = []
with open(altofilepath, "r") as xmlfile:
doc = ET.fromstring(xmlfile.read())
page_shape, images = get_illustration_coords(doc)
if images and [x for x in images if x[2] * x[3] > threshold]:
counter = 0
img_color = cv2.imread(extract_jp2(zp, filename))
delete_jp2(filename)
h, w, _ = img_color.shape
current_page_shape = (w, h)
dh = float(h) / float(page_shape[1])
dw = float(w) / float(page_shape[0])
identifier, title, author, pubplace, publisher, guesseddate = parse_xml(
id)
for rect in images:
counter += 1
if rect[2] * rect[3] > threshold:
print("Attempting to slice '{1}' from {0}".format(
filename, rect))
scaled = increase_size(rect, scale, (dw, dh),
current_page_shape)
boundary = get_rect(scaled)
if len(guesseddate) != 4:
guesseddate = "Unknown"
if not os.path.exists(os.path.join(embellish_path,
guesseddate)):
os.mkdir(os.path.join(embellish_path, guesseddate))
img_filename = os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(
id, vol, page, counter, make_safe(title), guesseddate))
short_filename = os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}_{5}.jpg".format(
id, vol, page, counter,
make_safe(title)[:80], guesseddate))
# if not already captured this, switch to new short title and carry on
if not os.path.exists(img_filename):
img_filename = short_filename
# Has this been captured before?
try:
if not os.path.exists(img_filename):
cv2.imwrite(
img_filename,
img_color[boundary[0][1]:boundary[1][1],
boundary[0][0]:boundary[1][0]],
[cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
if os.path.isfile(img_filename):
r.lpush(plateimgs, img_filename)
noplates = r.llen(plateimgs)
print("Plate no: {1} - {0}".format(
img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("plateerror", altoxmloldpath)
else:
print("Already captured, skipping")
except:
img_filename = os.path.join(
embellish_path, guesseddate,
"{0}_{1}_{2}_{3}_{4}.jpg".format(
id, vol, page, counter, guesseddate))
cv2.imwrite(
img_filename, img_color[boundary[0][1]:boundary[1][1],
boundary[0][0]:boundary[1][0]],
[cv2.cv.CV_IMWRITE_JPEG_QUALITY, 93])
if os.path.isfile(img_filename):
r.lpush(plateimgs, img_filename)
noplates = r.llen(plateimgs)
print("Plate no: {1} - {0}".format(
img_filename, noplates))
else:
print("CV2 SECRETLY FAILED...")
r.lpush("plateerror", altoxmloldpath)
else:
print(
"All illustration areas in {1}(vol:{2}, pg: {3} are below threshold area size (currently: {0})"
.format(threshold, id, vol, page))
r = Redis()
wq = "q"
output = "output"
def get_job(workerid):
if r.llen(workerid) == 0:
status = r.rpoplpush(wq, workerid)
return status
else:
return r.lrange(workerid, 0, 0)[0]
def clear_job(workerid, job):
r.lrem(workerid, job, 1)
if __name__ == "__main__":
workerid = "wrk"+sys.argv[1]
while(True):
job = get_job(workerid)
while(job):
try:
row = parse_xml(job)
r.lpush(output, u"\t".join(row).encode("utf-8"))
clear_job(workerid, job)
except OSError, e:
clear_job(workerid, job)
job = get_job(workerid)
print("%s ran out of jobs - waiting for 10 seconds before checking again" % workerid)
time.sleep(10)