def fail(self, message = None):
if self._colorize:
self._spinner.fail(color(self._current_message, 'error', self._colorize))
self._spinner.stop()
self.start(message)
self._spinner.fail(color(message, 'error', self._colorize))
self._spinner.stop()
else:
msg('ERROR: ' + message)
def update(self, message = None):
if self._quiet:
return
if self._colorize:
self._spinner.succeed(color(self._current_message, 'info', self._colorize))
self._spinner.stop()
self.start(message)
else:
msg(message)
def warn(self, message=None):
if self._quiet:
return
if self._colorize:
self._spinner.succeed(
color(self._current_message, 'info', self._colorize))
self._spinner.stop()
self._spinner.warn(color(message, 'warn', self._colorize))
self._current_message = ''
else:
msg('WARNING: ' + message)
def start(self, message = None):
if self._quiet:
return
if message is None:
message = ''
if self._colorize:
text = color(message, 'info')
self._spinner = Halo(spinner='bouncingBall', text = text)
self._spinner.start()
self._current_message = message
else:
msg(message)
def all_results(self, path):
'''Returns all the results from the service as a Python dict.'''
# Check if we already processed it.
if path in self._results:
return self._results[path]
if __debug__: log('Reading {}', path)
with io.open(path, 'rb') as image_file:
image_data = image_file.read()
# Google Cloud Vision API docs state that images cannot exceed 20 MB:
# https://cloud.google.com/vision/docs/supported-files
if len(image_data) > 20 * 1024 * 1024:
text = 'Error: file "{}" is too large for Google service'.format(
path)
msg(text, 'warn')
return text
try:
if __debug__: log('Building Google vision API object')
client = gv.ImageAnnotatorClient()
image = gv.types.Image(content=image_data)
context = gv.types.ImageContext(
language_hints=['en-t-i0-handwrit'])
# Iterate over the known API calls and store each result.
results = {}
for feature in self._known_features:
if __debug__:
log('Sending image to Google for {} ...', feature)
response = getattr(client, feature)(image=image,
image_context=context)
if __debug__: log('Received result.')
results[feature] = MessageToDict(response)
self._results[path] = results
return results
except google.api_core.exceptions.PermissionDenied as err:
text = 'Authentication failure for Google service -- {}'.format(
err)
raise ServiceFailure(text)
except Exception as err:
text = 'Error: failed to convert "{}": {}'.format(path, err)
return text
def all_results(self, path):
'''Returns all the results from the service as a Python dict.'''
# Check if we already processed it.
if path in self._results:
return self._results[path]
vision_base_url = "https://westus.api.cognitive.microsoft.com/vision/v2.0/"
text_recognition_url = vision_base_url + "recognizeText"
headers = {
'Ocp-Apim-Subscription-Key': self.credentials,
'Content-Type': 'application/octet-stream'
}
params = {'mode': 'Handwritten'}
image_data = open(path, 'rb').read()
# https://docs.microsoft.com/en-us/azure/cognitive-services/computer-vision/home
# states "The file size of the image must be less than 4 megabytes (MB)"
if len(image_data) > 4 * 1024 * 1024:
text = 'File "{}" is too large for Microsoft service'.format(path)
msg(text, 'warn')
return text
# Post it to the Microsoft cloud service.
if __debug__: log('Sending file to MS cloud service')
response = requests.post(text_recognition_url,
headers=headers,
params=params,
data=image_data)
try:
response.raise_for_status()
except HTTPError as err:
# FIXME this might be a good place to suggest to the user that they
# visit https://blogs.msdn.microsoft.com/kwill/2017/05/17/http-401-access-denied-when-calling-azure-cognitive-services-apis/
if response.status_code in [401, 402, 403, 407, 451, 511]:
text = 'Authentication failure for MS service -- {}'.format(
err)
raise ServiceFailure(text)
elif response.status_code == 429:
text = 'Server blocking further requests due to rate limits'
raise ServiceFailure(text)
elif response.status_code == 503:
text = 'Server is unavailable -- try again later'
raise ServiceFailure(text)
else:
text = 'Encountered network communications problem -- {}'.format(
err)
raise ServiceFailure(text)
except Exception as err:
import pdb
pdb.set_trace()
msg('MS rejected "{}"'.format(path), 'warn')
return ''
# The Microsoft API for extracting handwritten text requires two API
# calls: one call to submit the image for processing, the other to
# retrieve the text found in the image. We have to poll and wait
# until a result is available.
analysis = {}
poll = True
if __debug__: log('Polling MS for results ...')
while (poll):
response_final = requests.get(
response.headers["Operation-Location"], headers=headers)
analysis = response_final.json()
time.sleep(1)
if ("recognitionResult" in analysis):
poll = False
if ("status" in analysis and analysis['status'] == 'Failed'):
poll = False
if __debug__: log('Results received.')
self._results[path] = analysis
return analysis