Python API

Azure

class RPA.Cloud.Azure.Azure(region: str = 'northeurope', robocloud_vault_name: str = None)

Bases: RPA.Cloud.Azure.ServiceTextAnalytics, RPA.Cloud.Azure.ServiceFace, RPA.Cloud.Azure.ServiceComputerVision, RPA.Cloud.Azure.ServiceSpeech

Azure is a library for operating with Microsoft Azure API endpoints.

List of supported service names:

Azure authentication

Authentication for Azure is set with service subscription key which can be given to the library in two different ways.

  • Method 1 as environment variables, either service specific environment variable for example AZURE_TEXTANALYTICS_KEY or with common key AZURE_SUBSCRIPTION_KEY which will be used for all the services.

  • Method 2 as Robocorp Vault secret. The vault name needs to be given in library init or with keyword Set Robocloud Vault. Secret keys are expected to match environment variable names.

Method 1. subscription key using environment variable

*** Settings ***
Library   RPA.Cloud.Azure

*** Tasks ***
Init Azure services
    # NO parameters for client, expecting to get subscription key
    # with AZURE_TEXTANALYTICS_KEY or AZURE_SUBSCRIPTION_KEY environment variable
    Init Text Analytics Service

Method 2. setting Robocloud Vault in the library init

*** Settings ***
Library   RPA.Cloud.Azure  robocloud_vault_name=azure

*** Tasks ***
Init Azure services
    Init Text Analytics Service  use_robocloud_vault=${TRUE}

Method 2. setting Robocloud Vault with keyword

*** Settings ***
Library   RPA.Cloud.Azure

*** Tasks ***
Init Azure services
    Set Robocloud Vault          vault_name=googlecloud
    Init Text Analytics Service  use_robocloud_vault=${TRUE}

References

List of supported language locales - Azure locale list

List of supported region identifiers - Azure region list

Examples

Robot Framework

This is a section which describes how to use the library in your Robot Framework tasks.

*** Settings ***
Library  RPA.Cloud.Azure

*** Variables ***
${IMAGE_URL}   IMAGE_URL
${FEATURES}    Faces,ImageType

*** Tasks ***
Visioning image information
   Init Computer Vision Service
   &{result}   Vision Analyze  image_url=${IMAGE_URL}  visual_features=${FEATURES}
   @{faces}    Set Variable  ${result}[faces]
   FOR  ${face}  IN   @{faces}
      Log  Age: ${face}[age], Gender: ${face}[gender], Rectangle: ${face}[faceRectangle]
   END

Python

This is a section which describes how to use the library in your own Python modules.

library = Azure()
library.init_text_analytics_service()
library.init_face_service()
library.init_computer_vision_service()
library.init_speech_service("westeurope")

response = library.sentiment_analyze(
   text="The rooms were wonderful and the staff was helpful."
)
response = library.detect_face(
   image_file=PATH_TO_FILE,
   face_attributes="age,gender,smile,hair,facialHair,emotion",
)
for item in response:
   gender = item["faceAttributes"]["gender"]
   age = item["faceAttributes"]["age"]
   print(f"Detected a face, gender:{gender}, age: {age}")

response = library.vision_analyze(
   image_url=URL_TO_IMAGE,
   visual_features="Faces,ImageType",
)
meta = response['metadata']
print(
   f"Image dimensions meta['width']}x{meta['height']} pixels"
)

for face in response["faces"]:
   left = face["faceRectangle"]["left"]
   top = face["faceRectangle"]["top"]
   width = face["faceRectangle"]["width"]
   height = face["faceRectangle"]["height"]
   print(f"Detected a face, gender:{face['gender']}, age: {face['age']}")
   print(f"      Face rectangle: (left={left}, top={top})")
   print(f"      Face rectangle: (width={width}, height={height})")

library.text_to_speech(
    text="Developer tools for open-source RPA leveraging the Robot Framework ecosystem",
    neural_voice_style="cheerful",
    target_file='output.mp3'
)
ROBOT_LIBRARY_DOC_FORMAT = 'REST'
ROBOT_LIBRARY_SCOPE = 'GLOBAL'
class RPA.Cloud.Azure.AzureBase

Bases: object

Base class for all Azure servives.

TOKEN_LIFESPAN is in seconds, token is valid for 10 minutes so max lifetime is set to 9.5 minutes = 570.0 seconds

COGNITIVE_API = 'api.cognitive.microsoft.com'
TOKEN_LIFESPAN = 570.0
logger = None
region = None
robocloud_vault_name: str = None
services: dict = {}
set_robocloud_vault(vault_name)

Set Robocorp Vault name

Parameters

vault_name – Robocorp Vault name

token = None
token_time = None
class RPA.Cloud.Azure.ServiceComputerVision

Bases: RPA.Cloud.Azure.AzureBase

Class for Azure Computer Vision service

init_computer_vision_service(region: str = None, use_robocloud_vault: bool = False) → None

Initialize Azure Computer Vision

Parameters

  • region – identifier for service region

  • use_robocloud_vault – use secret stored into Robocorp Vault

vision_analyze(image_file: str = None, image_url: str = None, visual_features: str = None, json_file: str = None) → dict

Identify features in the image

Parameters

  • image_file – filepath of image file

  • image_url – URI to image, if given will be used instead of image_file

  • visual_features – comma separated list of features, for example. “Categories,Description,Color”

  • json_file – filepath to write results into

Returns

analysis in json format

See Computer Vision API for valid feature names and their explanations:

  • Adult

  • Brands

  • Categories

  • Color

  • Description

  • Faces

  • ImageType

  • Objects

  • Tags

vision_describe(image_file: str = None, image_url: str = None, json_file: str = None) → dict

Describe image with tags and captions

Parameters

  • image_file – filepath of image file

  • image_url – URI to image, if given will be used instead of image_file

  • json_file – filepath to write results into

Returns

analysis in json format

vision_detect_objects(image_file: str = None, image_url: str = None, json_file: str = None) → dict

Detect objects in the image

Parameters

  • image_file – filepath of image file

  • image_url – URI to image, if given will be used instead of image_file

  • json_file – filepath to write results into

Returns

analysis in json format

vision_ocr(image_file: str = None, image_url: str = None, json_file: str = None) → dict

Optical Character Recognition (OCR) detects text in an image

Parameters

  • image_file – filepath of image file

  • image_url – URI to image, if given will be used instead of image_file

  • json_file – filepath to write results into

Returns

analysis in json format

class RPA.Cloud.Azure.ServiceFace

Bases: RPA.Cloud.Azure.AzureBase

Class for Azure Face service

detect_face(image_file: str = None, image_url: str = None, face_attributes: str = None, face_landmarks: bool = False, recognition_model: str = 'recognition_02', json_file: str = None) → dict

Detect facial attributes in the image

Parameters

  • image_file – filepath of image file

  • image_url – URI to image, if given will be used instead of image_file

  • face_attributes – comma separated list of attributes, for example. “age,gender,smile”

  • face_landmarks – return face landmarks of the detected faces or not. The default value is False

  • recognition_model – model used by Azure to detech faces, options are “recognition_01” or “recognition_02”, default is “recognition_02”

  • json_file – filepath to write results into

Returns

analysis in json format

Read more about face_attributes at Face detection explained:

  • age

  • gender

  • smile

  • facialHair

  • headPose

  • glasses

  • emotion

  • hair

  • makeup

  • accessories

  • blur

  • exposure

  • nouse

init_face_service(region: str = None, use_robocloud_vault: bool = False) → None

Initialize Azure Face

Parameters

  • region – identifier for service region

  • use_robocloud_vault – use secret stored into Robocorp Vault

class RPA.Cloud.Azure.ServiceSpeech

Bases: RPA.Cloud.Azure.AzureBase

Class for Azure Speech service

audio_formats = {'MP3': 'audio-24khz-96kbitrate-mono-mp3', 'WAV': 'riff-24khz-16bit-mono-pcm'}
init_speech_service(region: str = None, use_robocloud_vault: bool = False) → None

Initialize Azure Speech

Parameters

  • region – identifier for service region

  • use_robocloud_vault – use secret stored into Robocorp Vault

list_supported_voices(locale: str = None, neural_only: bool = False, json_file: str = None)

List supported voices for Azure API Speech Services.

Parameters

  • locale – list only voices specific to locale, by default return all voices

  • neural_only – True if only neural voices should be returned, False by default

  • json_file – filepath to write results into

Returns

voices in json

Available voice selection might differ between regions.

text_to_speech(text: str, language: str = 'en-US', name: str = 'en-US-AriaRUS', gender: str = 'FEMALE', encoding: str = 'MP3', neural_voice_style: Any = None, target_file: str = 'synthesized.mp3')

Synthesize speech synchronously

Parameters

  • text – input text to synthesize

  • language – voice language, defaults to “en-US”

  • name – voice name, defaults to “en-US-AriaRUS”

  • gender – voice gender, defaults to “FEMALE”

  • encoding – result encoding type, defaults to “MP3”

  • neural_voice_style – if given then neural voice is used, example style. “cheerful”

  • target_file – save synthesized output to file, defaults to “synthesized.mp3”

Returns

synthesized output in bytes

Neural voices are only supported for Speech resources created in East US, South East Asia, and West Europe regions.

class RPA.Cloud.Azure.ServiceTextAnalytics

Bases: RPA.Cloud.Azure.AzureBase

Class for Azure TextAnalytics service

detect_language(text: str, json_file: str = None) → dict

Detect languages in the given text

Parameters

  • text – A UTF-8 text string

  • json_file – filepath to write results into

Returns

analysis in json format

find_entities(text: str, language: str = None, json_file=None) → dict

Detect entities in the given text

Parameters

  • text – A UTF-8 text string

  • language – if input language is known

  • json_file – filepath to write results into

Returns

analysis in json format

init_text_analytics_service(region: str = None, use_robocloud_vault: bool = False)

Initialize Azure Text Analyticts

Parameters

  • region – identifier for service region

  • use_robocloud_vault – use secret stored into Robocorp Vault

key_phrases(text: str, language: str = None, json_file: str = None) → dict

Detect key phrases in the given text

Parameters

  • text – A UTF-8 text string

  • language – if input language is known

  • json_file – filepath to write results into

Returns

analysis in json format

sentiment_analyze(text: str, language: str = None, json_file: str = None) → dict

Analyze sentiments in the given text

Parameters

  • text – A UTF-8 text string

  • language – if input language is known

  • json_file – filepath to write results into

Returns

analysis in json format