microsofttranslator / text-translation-api-v3-python

page_type	name	description	urlFragment	languages	products
sample	Microsoft Translator Python samples (v3)	This repository includes Python code samples for Microsoft Translator.	translator-python-v3	python	azure azure-cognitive-services azure-translator

Translator API V3 - Python Samples

This repository includes Python code samples for Microsoft Translator. The samples are designed to run on Python 2.7.x and Python 3.x. Each sample corresponds to a Quickstart activity on doc.microsoft.com, including:

• Translating text
• Transliterating text
• Identifying the language of source text
• Getting alternate translations
• Getting a complete list of supported languages
• Determining sentence length

Get started with the Translator quickstart.

Prerequisites

Here's what you'll need before you use these samples:

• Your favorite IDE or text editor
• Python 2.7.x or 3.x
• An Azure subscription - Sign-up for a free account!
• A Translator resource - Create a Translator resource

Code samples

This repository includes a sample for each of the methods made available by the Microsoft Translator API v3. To use each of the samples, follow these instructions:

• Create a new project in your favorite IDE or editor.

• Copy the code from one of the samples into your project.

• Set your subscription key, the Azure region of your resource and the Text Translation endpoint as environment variables TRANSLATOR_TEXT_SUBSCRIPTION_KEY TRANSLATOR_TEXT_REGION TRANSLATOR_TEXT_ENDPOINT

• Run the program. For example: python Translate.py.

Resources

• What is Translator?
• v3 Translator API Reference
• Supported languages

text-translation-api-v3-python's Issues

Improvement in code sample - rename variable

Hello,

In some of the examples there is:

request = requests.post(constructed_url, headers=headers, json=body)
response = request.json()

Would it be possible to rename request for response (for example):

response = requests.post(constructed_url, headers=headers, json=body)
response = response.json()

I'm just mentioning it because post() returns a requests.models.Response object.

Regards

Http connection pool error

Hi When i am trying to make a post request i am getting the below error

requests.exceptions.SSLError: HTTPSConnectionPool(host='api.cognitive.microsofttranslator.com', port=443): Max retries exceeded with url: /translate?api-version=3.0&to=de (Caused by SSLError(SSLEOFError(8, 'EOF occurred in violation of protocol (_ssl.c:748)'),))

I have tried adding proxy and set the verify = false but it didnot work .please help me with resolution

Adds spaces around markdown variables like ** while translating , thus breaking the markdown syntax

Faced this issue while translating from English to Indian languages like Hindi, Kannada and Tamil

Result output broken in translate.py

Text-Translation-API-V3-Python/Translate.py

Line 41 in dc2abc1

print(json.dumps(response, sort_keys=True, indent=4, separators=(',', ': ')))

If the output is not English, ensure_ascii=False should be provided in json.dumps as below,
otherwise, the result could be broken.

print(json.dumps(response, sort_keys=True, indent=4, ensure_ascii=False, separators=(',', ': ')))

JSONDecodeError("Expecting value", s, err.value) from None

I used the sample code, used the latest endpoint, but I keep getting this error:

raise JSONDecodeError("Expecting value", s, err.value) from None

JSONDecodeError: Expecting value

please help!

Error : the JSON object must be str, not 'bytes'

When I run the sample code with my subscription keys, I get the following error.

Error :

File "T.py", line 42, in
output = json.dumps(json.loads(result), indent=4, ensure_ascii=False)
File "/usr/lib/python3.5/json/init.py", line 312, in loads
s.class.name))
TypeError: the JSON object must be str, not 'bytes'

Requests module installed but not recognized

Maybe not a bug in this code, but is anyone having trouble with the requests module not being recognized in your IDE? I tried to create this sample in just a .py file too (so I know it's not my IDE), but same issue. When I check my python libraries with pip list, it shows requests 2.20.0 is indeed installed.

Translate.py contains two "to" parameters in the params object

I apologize in case I'm wrong, I'm very new to the Azure translator Python API. My sense is that in this part:

params = '&to=de&to=it'

one of the to parameters should actually be a language to translate from

Simplified translate.py example without environment variables and JSON output, but as text

I can't understand the rationale behind the use of environment variables in the Translate.py file: https://github.com/MicrosoftTranslator/Text-Translation-API-V3-Python/blob/master/Translate.py

Why set the variables to values, and then check if they are available with their name (not value) in the os environment, only to be assigned to other variables? Why not do it directly? So instead of:

key_var_name = 'TRANSLATOR_TEXT_SUBSCRIPTION_KEY'
if not key_var_name in os.environ:
    raise Exception('Please set/export the environment variable: {}'.format(key_var_name))
subscription_key = os.environ[key_var_name]

just:

subscription_key = 'TRANSLATOR_TEXT_SUBSCRIPTION_KEY'

As a simple example? If you want to use environment variables, then have a associated .env file and import it, describe how to use it. To make it more beginner friendly. And have an option to just export it as a text, not JSON.

Problem with alignment information.

I have next code:

def translate(
        from_lang: str,
        to_lang: str,
        texts: List[str],
        include_alignment: bool = True
    ) -> List[JSONType]:
        constructed_url = "https://api.cognitive.microsofttranslator.com/translate"
        params = {
            "api-version": "3.0",
            "from": from_lang,
            "to": to_lang,
            "includeAlignment": include_alignment,
            "includeSentenceLength": True,
        }
        headers = {
            "Ocp-Apim-Subscription-Key": "*******",
            "Ocp-Apim-Subscription-Region": "GLOBAL",
            "Content-type": "application/json",
            "X-ClientTraceId": str(uuid.uuid4()),
        }
        body = [{"text": text} for text in texts]
        request = requests.post(
            constructed_url, params=params, headers=headers, json=body
        )
        resp = request.json()
        return resp
result = translate("de", "en", ["Artikelnummer des Kunden"])

and i have a result
{'translations': [{'text': "Customer's item number", 'to': 'en', 'alignment': {'proj': '0:12-11:14 0:12-16:21 14:16-0:9'}, 'sentLen': {'srcSentLen': [24], 'transSentLen': [22]}}]}
so if we look at result:
Original: Artikelnummer des Kunden
Translation: Customer's item number
we see that we have no alignment for the original word "Kunden".

where can I find "endpoint" and "subscription_key" from my azure dashboard?

I can find "key1" "key2" and "endpoint" from my dashboard.
and endpoint looks like this: https://eng2jap.cognitiveservices.azure.com/sts/v1.0/issuetoken
I got 404 no found when using key1 and this endpoint by the python code.
please give me some advice. Thanks!

"message": "Resource not found"

Examples not working for me. Can anyone confirm that?

I got always this error?

{
    "error": {
        "code": "404",
        "message": "Resource not found"
    }
}

translation API doesn't work in python2.7?

I am developing an web App that calls the Text Translation API. My whole project is developed in Python2.7 so I try to call the APIs in Python2.7. But I always get the failure result:

send: 'POST /translate?api-version=3.0&to=zh HTTP/1.1\r\nHost: api.cognitive.microsofttranslator.com\r\nAccept-Encoding: identity\r\nContent-Length: 19\r\nOcp-Apim-Subscription-Key: [my-sub-key]\r\nContent-type: application/json\r\nX-ClientTraceId: 7ade7c69-7100-4afb-8bb7-676d594637c4\r\n\r\n'
send: '[{"Text": "Hello"}]'
DEBUG:root:
reply: 'HTTP/1.1 404 Not Found\r\n'

My code is in https://github.com/yangpancn/alchemy/blob/master/ms_translator.py

could someone provide a demo code how to use the API in python2.7?
Thanks in advance!

Count of translated characters

There is a limit of 12M characters per month to the F0 pricing tier. I wanted to clarify what if I have given let's say 10M characters to the API, out of which only 1M characters need translation.
For eg. I want to have all characters in English, but only 1M characters are non-English, which means effectively the API is working on 1M characters only.
So will I be charged for complete 10M characters, or only 1M characters?

line 23，params = '&to=de&to=it' the first to should be from

Translate.py file: line 23，params = '&to=de&to=it' the first to should be from, '&from=de&to=it'

does not give output when translating between languages of 2 scripts

when i have deployed the python code via spawn with my node js to build my api i am not getting the outpt if i translate a text from english to hindi

About the Translate.py sample

Hello ,
I couldn't run the script Translate.py until I forced the parameter 'Ocp-Apim-Subscription-Region' in the header to the region of my subscription , to me 'francecentral' .
Otherwise great sample.
Thank you .

Ocp-Apim-Subscription-Region is missing from the headers

I was not able to make the Translate.py work, because all I got back from the API was:

{
    "error": {
        "code": 401000,
        "message": "The request is not authorized because credentials are missing or invalid."
    }
}

I was investigating the documentation and it seems, that the Ocp-Apim-Subscription-Region header is required for the authentication.