Tensoract SDK

Tensoract Studio provides python SDK which is thin wrapper on REST APIs https://api.tensoract.com/docs/

The SDK offers a number of useful methods to automate operational tasks in the workbench.

Generate API key

All SDK / API operations require the client to authenticate using an API key. An user with “API User” role will have privilege to generate the API key. Following are the instructions to generate API Key:

• Login as the user and click the User name in upper right corner as depicted below.

  

• Click Account Setting

• Click Create New API Key

  

• Take note of the generated API Key.

Installation

Tensoract Python SDK can be installed using pip:

$ pip install tensoract

Once installed, proceed to the following section to initialize the client using your API key.

Client

The client can be initialized as shown below:

Request Syntax

from pprint import pprint
from tensoract.client import Tensoract
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

type api_url:

String

param api_url:

The API URL endpoint. This is an optional argument for private Installations.

type api_key:

string

param api_key:

API Key to authenticate the client.

Modules

Tensoract Studio offers a vast set of functionalities ranging from user management to managing entire project lifecyle involving datasets and tasks. Following are available functions in each module.

Project

Below methods operate within the Project context

• create_project()

• get_project()

• update_project()

• get_projects()

• delete_project()

create_project(body)

Creates a new project programatically using create_project

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")


body = {
"project_name": "TestProject",
"project_type": "NER",
"enable_text_mode_option": True,
"disable_quality_audit": True,
...
}
# check API docs for the full JSON body structure https://api.tensoract.com/docs/#/projects/upload_project

pprint(client.create_project(body))

Parameters:

body (JSON Object) – JSON object with elements defined as here and further explained at ner

get_project(project_id)

Retrieves project details based on project_id. For more information see get_project.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke get_project method
response = client.get_project(project_id="449354de1168469a8229f605")
pprint(response)

Parameters:

project_id (String) – Internal project_id assigned to the Project within Tensoract Workbench.

Response

The requested project’s configuration is returned in JSON format.

{
"project_name": "TestProject",
"project_type": "NER|OCR|",
"enable_text_mode_option": True,
"disable_quality_audit": True,
...
}

update_project(project_id, body)

Updates configuration of an existing project in the workbench based on the provided project_id and body JSON parameter.

Sample Invocation

from pprint import pprint
from Tensoract import Tensoract


# Initialize Tensoract  client
client = tensoract.Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Set the JSON config in body
body = {
"project_name": "TestProject",
"project_type": "NER",
"enable_text_mode_option": True,
"disable_quality_audit": True,
...
}

# Invoke update_project method
response = client.update_project(project_id="449354de1168469a8229f605",body=body)
pprint(response)

Parameters:

• project_id (String) – Internal project_id assigned to the Project within Tensoract Workbench.

• body (JSON Object) –

  JSON object with elements defined as here and further explained at ner

Response

The updated project’s new configuration is returned in JSON format.

{
"project_name": "TestProject",
"project_type": "NER",
"enable_text_mode_option": True,
"disable_quality_audit": True,
...
}

get_projects(project_id, project_name, active)

Fetches a list of projects that meet the criteria defined by the parameters (project_id or project_name or active status)

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke get_projects method
response = client.get_projects(project_id=None,project_name="NER_Project",active=True)
pprint(response)

Parameters:

• project_id (String) – Internal project_id assigned to the Project within Tensoract Workbench.

• project_name (String) – Project Name defined for this project in Tensoract workbench.

• active (Boolean) – Project state is active or not.

All the above three parameters can be applied as filters with AND criteria. When all three parameters are set to None, then no filter is applied and all projects are returned

Response

JSON list of projects is returned as below

[
  {'company': '60f38299446de8dabe9207e5',
  'company_name': 'Objectways',
  'project_id': '8205f4dd2319fa0fc33b14a5',
  'project_name': 'NER_Project',
  'project_type': 'NER'}
]

delete_project(project_id)

Deletes an existing project in the workbench based on the provided project_id parameter.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke delete_project method
response = client.delete_project(project_id="8bc05ec26ab161e218a0e842",)
pprint(response)

Parameters:

project_id (String) – Internal project_id assigned to the Project within Tensoract Workbench.

Response

Status of the delete operation is returned as a JSON

{
'message': 'project deleted',
'status': 1
}

Tasks

Below methods operate within the Tasks context

• add_task()

• add_bulk_tasks()

• add_labels_to_task()

• add_task_file()

• get_task()

• get_tasks()

• delete_task()

• export_tasks()

• get_word_boxes()

add_task(body)

Add a single task item to a given project

Sample Invocation

from pprint import pprint
    from tensoract.client import Tensoract
    # Initialize Tensoract client
    client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

        body= {

            "project_id": "179cd15e334f9a63e2a9632a",
            "item_id": "5835e22985bb0ca47a9fa96e"

        }

# Invoke add_task method
pprint(client.add_task(body))

Parameters:

body (JSON Object) –

JSON Object with following elements

• status - “success” or an error status

• value - Count of task added to project

• task_ids - An array of task_ids

Response

Task details returned as a JSON

{
    "status": "success",
    "value": "1 item(s) added to project",
    "task_ids": [
        "c1b160d50814fdf190eb12fc"
    ]
}

add_bulk_tasks(body)

Add a list of task items to a given project

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Define bulk task JSON payload
body = {
"project_id": "cd1a965e334a9a63e2f17932",
"task_list": [
  {
    "source": "s3://demo-examples/pdfs/document.pdf",
    "annotations": "s3://demo-examples/pdfs/annotation.json"
  },
  {
    "source": "s3://demo-examples/pdfs/document2.pdf"
  },
  {
    "source": "s3://demo-examples/pdfs/document3.pdf",
    "annotations": {
      "tags": [
        {
          "page": 1,
          "range": [
            192,
            198
          ],
          "text": "Oxford",
          "id": 1,
          "type": "NAME"
        }
      ]
    }
  }
]
}

# Invoke method to add bulk tasks
pprint(client.add_bulk_tasks(body))

Parameters:

body (JSON Object) –

JSON Object with following elements

• status - “success” or an error status

• value - Count of task added to project

• task_ids - An array of task_ids

Response

Status of the operation is returned as a JSON

{
       "status": "success",
       "value": "2 item(s) added to project",
       "task_ids": [
           "aa7b5455f31a2910e9e41e12",
           "6dae0d4817c9b9b1985e2a0a"
       ]
  }

add_labels_to_task(task_id, body))

Add labels to a given task

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")
task_id = "4d572fef4ec4bc9d86bd7b2f"

body = {
"annotations": "s3://demo-examples/pdfs/annotation.json"
}

pprint(client.add_labels_to_task(task_id=task_id,body=body))

Parameters:

• task_id (String) – Internal task_id for the given task in Tensoract workbench

• body (JSON Object) –

  JSON Object with following elements

  • annotations - s3 reference to the json payload containing annotations for the given task.

Response

Status of the operation is returned as a JSON

{'task_id': '4d572fef4ec4bc9d86bd7b2f'}

add_task_file(project_id, file_path, mime_type, annotations)

Add a local file as a task item to given project

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

data = client.add_task_file(
project_id="449354de1168469a8229f605",
file_path="examples/pdfs/document.pdf",
mime_type="application/pdf",
annotations=None
)

pprint(data)

Parameters:

• project_id (String) – Internal project_id for the given project in Tensoract workbench

• file_path (String) – local file path to the document/image to be added to the task

• mime_type (String) – mime_type of the local file to be added

• annotations (String) – annotations to be added to the task item

Response

Status of the operation is returned as a JSON

{'_id': '6311e17a412bf7b26e6f6649',
'active': True,
'company': '60f38299446de8dabe9207e5',
'created': 1662116218,
'deep_link': None,
'fileId': 'fc879a309df1b8c1efd7f380',
'modified': 1662116218,
'name': 'document.pdf',
'preview': '',
'project': '449354de1168469a8229f605',
'qcEmail': [],
'source': 'file:449354de1168469a8229f605/fc879a309df1b8c1efd7f380',
'state': 16,
'type': 'application/pdf',
 'userEmail': []}

get_task(task_id)

Retrieve a task

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

data = client.get_task(task_id="449354de1168469a8229f605")
pprint(data)

Parameters:

task_id (String) – Internal task_id for the given task in Tensoract workbench

Response

Task details returned as a JSON

{
'company': '60f38299446de8dabe9207e5',
'company_name': 'Objectways',
'file_type': 'application/pdf',
'plain_text': {'1': '2/27/2021 Amazon.com: Customer reviews: OPI Nail '
                 'Lacquer, A Great Opera Tunity '
                 'https://www.amazon.com/gp/customer -reviews/B00014353Y/ '
                 '1/5 Skip to main content Top reviews All reviewers All '
                 '.........................................................'
                 'Select your address Returns & Orders All Account Hello, '
                 'Sign in 0'}
'project_id': '557b89e25a0588898fae91f0',
'project_name': 'NER_PROJECT_AUTO',
'project_type': 'NER',
'source': 's3://Tensoract-demo-examples/sample.pdf',
'state': 0,
'state_description': 'Unassigned',
'task_id': '449354de1168469a8229f605'

}

get_tasks(project_id, task_id, file_name, file_type, trail)

Retrieve a list of tasks matching the filter criteria

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

tasks = client.get_tasks(
project_id="557b89e25a0588898fae91f0",
task_id="16e2a709994421da996cef4f",
file_name=None,
file_type=None,
trail = False
)
pprint(tasks)

type project_id:

String

param project_id:

Internal project_id for the given project in Tensoract workbench

type task_id:

String

param task_id:

Internal task_id for the given task in Tensoract workbench

type file_name:

String

param file_name:

file_name metadata assigned to the task

type file_type:

String

param file_type:

file_type metadata assigned to the task

type trail:

Boolean

param trail:

trail

Response

JSON List of tasks matching the filter criteria.

[
  {
  'company': '60f38299446de8dabe9207e5',
  'file_type': 'application/pdf',
  'project_id': '557b89e25a0588898fae91f0',
  'source': 's3://Tensoract-demo-examples/sample.pdf',
  'state': 0,
  'state_description': 'Unassigned',
  'task_id': '16e2a709994421da996cef4f'
  }
]

delete_task(task_id)

Delete the given task item

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

data = client.delete_task(task_id="449354de1168469a8229f605")
pprint(data)

Parameters:

task_id (String) – Internal task_id for the given task in Tensoract workbench

Response

Status of the operation is returned as a JSON

{'message': 'Task deleted sucessfuly', 'status': 1}

export_tasks(project_id)

Export all tasks from a given project as JSON list

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.export_tasks(project_id="449354de1168469a8229f605")
pprint(response)

Parameters:

project_id (String) – Internal project_id for the given project in Tensoract workbench

Response

Export status with url

{'message': 'The export is being generated, please try again in a few seconds '
         'to check the export status',
  'progress': '0%',
   'status': 'IN PROGRESS'}


 {'export_url': 'https://devdocumentsdemo.objectways.com/export_ner_project_auto_2022-09-02T09:23:34.9944468.manifest',
  'message': 'Use the export_url to download the project export file',
   'progress': '100%',
   'status': 'COMPLETED'}

get_word_boxes(body)

Get Bounding Box coordinates of all pages in a document. Optionally pass “page” number to restrict by the page

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.get_word_boxes(task_id="449354de1168469a8229f605", page = 1)
pprint(response)

Parameters:

• task_id (String) – Internal task_id for the given task in Tensoract workbench

• page (Integer) – Page number to which this operation should be restricted to.

Response

JSON list of Bounding box co-ordinates with related details

[
  [
    {
    'box': [0.04322405692755147,
       0.020058986644677534,
       0.11036431902612907,
       0.030159995920771787],
       'range': [0, 9],
       'text': '2/27/2021'
    },
    {
    'box': [0.3196231438320821,
       0.020058986644677534,
       0.4089945621543309,
       0.030159995920771787],
       'range': [10, 21],
       'text': 'Amazon.com:'},
    {
    'box': [0.2981259968948721,
       0.9708164847570528,
       0.4446869530303018,
       0.9809174940331471],
       'range': [114, 134],
       'text': '-reviews/B00014353Y/'
    }
  ]
]

Teams

Below methods operate within the Teams context and can be used to manage team members in a project.

• add_team_member()

• get_project_team_members()

• remove_project_team_member()

add_team_member(project_id, email, role)

Add a team member to collaborate on an existing project.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

member = client.add_team_member(project_id="449354de1168469a8229f605", email="someone@email.com", role="annotator|reviewer|supervisor")

Parameters:

• project_id (String) – Internal project_id for the given project in Tensoract workbench

• email (String) – Email id of an existing user to be added to the project as a collaborator

• role (String) – role assigned to the team member. This should be one of annotator or reviewer or supervisor

Response

Status of the operation is returned as a JSON

{'message': 'user added to the project team', 'status': 1}

get_project_team_members(project_id)

Return list of team members assigned to a project.

Sample Invocation

  from pprint import pprint
  from tensoract.client import Tensoract

  # Initialize Tensoract  client
  client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

  team = client.get_project_team_members(project_id="449354de1168469a8229f605")


:type project_id: String
:param project_id:
      Internal project_id for the given project in Tensoract  workbench

**Response**

JSON list of team members in the given project

::

  [{'email': '1@an.com', 'role': 'annotator'},
   'email': 'bobmorgan@me.com', 'role': 'annotator'},
   'email': 'q1@qc.com', 'role': 'reviewer'}]

remove_project_team_member(project_id, email)

Remove the given team member from an existing project.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

team = client.remove_project_team_member(project_id="449354de1168469a8229f605", email="someone@email.com" )

Parameters:

• project_id (String) – Internal project_id for the given project in Tensoract workbench

• email (String) – Email id of an existing user to be removed from the project as a collaborator

Response

Status of the operation is returned as a JSON

{'message': 'user deleted from project team', 'status': 1}

Dataset

Below methods operate within the Dataset context

• create_dataset()

• get_dataset()

• update_dataset()

• list_datasets()

• delete_dataset()

create_dataset(body)

Creates a new dataset based on configuration settings provided in the JSON body.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")
body = {
'dataset_name': 'Test Dataset Demo',
'dataset_description': 'This is a test Dataset',
'dataset_items': 0,
'dataset_type': 'Image',
....
}
# check API docs for the full JSON body structure https://api.tensoract.com/docs/#/datasets/upload_dataset
pprint(client.create_dataset(body))

Parameters:

body (JSON Object) –

JSON object with elements defined as here

get_dataset(dataset_id)

Fetches configuration details of an existing dataset in the workbench based on the provided dataset_id parameter.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke get_dataset method
response = client.get_dataset(dataset_id="449354de1168469a8229f605")
pprint(response)

Parameters:

dataset_id – Internal dataset_id assigned to the Dataset within Tensoract Workbench.

Response

The requested dataset’s configuration is returned in JSON format.

{
'dataset_name': 'Test Dataset Demo',
'dataset_id': '5c1d6892f4712fde11d8e25a',
'dataset_description': 'This is a test Dataset',
'dataset_items': 0,
'dataset_type': 'Image'
.....
}

update_dataset(dataset_id, body)

Updates configuration of an existing dataset in the workbench based on the provided dataset_id and body JSON parameter.

Sample Invocation

   from pprint import pprint
   from tensoract.client import Tensoract

   # Initialize Tensoract  client
   client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

   # Set the JSON config in body
   body = {
           'dataset_name': 'Test Dataset Demo',
           'dataset_description': 'This is a test Dataset',
           'dataset_items': 0,
           'dataset_type': 'Image'
           ......
          }

# Invoke update_dataset method
 response = client.update_dataset(dataset_id="449354de1168469a8229f605",body=body)
 pprint(response)

Parameters:

• dataset_id – Internal dataset_id assigned to the Dataset within Tensoract Workbench.

• body (JSON Object) –

  JSON object with elements defined as here

Response

The updated dataset’s new configuration is returned in JSON format.

{
  'dataset_name': 'Test Dataset Demo',
  'dataset_description': 'This is a test Dataset',
  'dataset_items': 0,
  'dataset_type': 'Image'
......
}

list_datasets(dataset_id, dataset_name, active)

Fetches a list of datasets that meet the criteria defined by the parameters (dataset_id or dataset_name or active status)

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke list_datasets method
response = client.list_datasets(dataset_id=None,dataset_name="Test Dataset Demo",active=True)
pprint(response)

Parameters:

• dataset_id – Internal dataset_id assigned to the Dataset within Tensoract Workbench.

• dataset_name (String) – Dataset Name defined for this dataset in Tensoract workbench.

• active (Boolean) – Dataset state is active or not.

All the above three parameters can be applied as filters with AND criteria. When all three parameters are set to None, then no filter is applied and all datasets are returned

Response

JSON list of datasets is returned as below

[
    {'company': '60f38299446de8dabe9207e5',
    'company_name': 'Objectways',
    'dataset_id': '2dfc93fb1723921e8bad8027',
    'dataset_name': 'Test Dataset Demo',
    'dataset_type': 'Image',}

]

delete_dataset(dataset_id)

Deletes an existing dataset in the workbench based on the provided dataset_id parameter.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

# Invoke delete_dataset method
response = client.delete_dataset(dataset_id="8bc05ec26ab161e218a0e842")
pprint(response)

Parameters:

dataset_id (String) – Internal dataset_id assigned to the Dataset within Tensoract Workbench.

Response

Status of the delete operation is returned as a JSON

{
'message': 'dataset deleted',
'status': 1
  }

Dataset Items

Below methods operate within the Dataset Items context

• create_dataset_item()

• get_dataset_item()

• delete_dataset_item()

• list_dataset_items()

create_dataset_item(body)

Add dataset items to a given dataset

Admin can post the metadata with the items as below

Sample Invocation

 from pprint import pprint
 from tensoract.client import Tensoract

 # Initialize Tensoract  client

client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

 body = {
 "dataset_id": "208eb554d3fa1e161b2c6a44",
 "dataset_items": [
 {
 "meta_data": {
   "batch": "10",
   "test": "cell"
 },
 "source": "s3://mybucket/test-dataset/file1.tiff"
 },
 {
 "meta_data": {
   "batch": "1",
   "test": "cell"
 },
 "source": "s3://mybucket/test-dataset/file2.tiff"
 }
 ....
 ....
 ]
 }
 # Invoke create_dataset_item method
 pprint(client.create_dataset_item(body))

Parameters:

body (JSON Object) –

JSON Object with following elements

• dataset_id - datatset to which the items needs to be added

• Metadata - Metadata can de added with dataset item

• source - link to S3 source of the document/image.

Response

Status of operation is returned as a JSON

{'inserted_items': 11, 'status': 1}

get_dataset_item(item_id)

Retrieve a dataset item

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

data = client.get_dataset_item(item_id="449354de1168469a8229f605")
pprint(data)

Parameters:

item_id (String) –

Internal item_id for the given dataset item in Tensoract workbench

Response

Dataset Item details returned as a JSON

[{'company': '60f38299446de8dabe9207e5',
'company_name': 'Objectways',
'dataset_id': '2ed8316fe3691cea93b8d70c',
'file_name': '10_2a68e073-b0b2-eb11-80cb-484d7e9f90c1.tiff',
'file_type': 'image/tiff',
'intermediate_source': 's3://Tensoract-demo-examples/sample.png',
'item_id': '0d1f56964171e372afa90323',
'source': 's3://Tensoract-demo-examples/sample.tiff',
'state': 0,
'version': [0]}]

delete_dataset_item(item_id)

Delete the given dataset item

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

data = client.delete_dataset_item(item_id="6bae961fcebe81cbacbedec6")
pprint(data)

Parameters:

item_id (String) – Internal item_id for the given dataset item in Tensoract workbench

Response

Status of the operation is returned as a JSON

{
  "message": "dataset item deleted",
"status": 1
}

list_dataset_items(dataset_id, dataset_version)

Retrieve a list of dataset items matching the filter criteria

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

dataset_items = client.list_dataset_items(dataset_id="2ed8316fe3691cea93b8d70c",dataset_version="0")

Parameters:

• dataset_id – Internal dataset_id for the given dataset in Tensoract workbench

• dataset_version (String) – Internal dataset_version for the given datasetk in Tensoract workbench

Response

JSON List of tasks matching the filter criteria.

{'company': '60f38299446de8dabe9207e5',
'company_name': 'Objectways',
'dataset_id': '3aa168a963f1d4dbe0b1f549',
'file_name': 'violet.tiff',
'file_type': 'image/tiff',
'item_id': '6bae961fcebe81cbacbedec6',
'meta_data': {'color': 'violet'},
'source': 'file:3aa168a963f1d4dbe0b1f549/6bae961fcebe81cbacbedec6',
'state': 0,
'tags': {'a dataset tag': 1},
'version': [0]}

Project Export

Below methods operate within the Project Export context

• export_project_tasks()

• list_project_exports()

• download_project_export()

export_project_tasks(project_id)

Export all tasks from a given project as JSON list

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract (api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

projects = client.export_project_tasks(project_id="1e80e389c318d5746169c5b6",
export_note="This is test export via API")
pprint(projects)

Parameters:

project_id (String) – Internal project_id for the given project in Tensoract workbench

Response

Export status

{'message': 'The export is being generated, please try again in a few seconds '
            'to check the export status',
            'progress': '0%'}


{'progress': '100%'}

list_project_exports(project_id)

This method enlists the exports exported in the project.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")
response = client.list_project_exports( project_id="449354de1168469a8229f605")

Parameters:

project_id (String) – Internal project_id for the given project in Tensoract workbench

Response

[{ 'created': 1667897879,
  'createdBy': 'abc@example.com',
  'exportId': '**************',
  'filter': '',
  'items': 41,
  'note': 'This is test export via API',
  'project': 'd2ff8489f14f0b1b59ec1df7',
  'projectName': 'test10',
  'selection': False}]

download_project_export(export_id)

It downloads the exports with given export id.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.download_project_export(export_id="449354de1168469a8229f605")

Response

{'export_url': 'https://devdocuments.objectways.com/export_2022-10-20T09:54:14.7355669_project1.manifest',
'message': 'Use the export_url to download the project export file',
'progress': '100%',
'status': 'COMPLETED'}

Dataset Export

Below methods operate within the Dataset Export context

• export_dataset()

• list_dataset_exports()

• download_dataset_export()

export_dataset(dataset_id)

Export all dataset items from a given dataset as JSON list.

Dataset can be exported with GroudTruth filter

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.export_dataset(
dataset_id="3aa168a963f1d4dbe0b1f549",
export_note="This is test export via API",
project_id="449354de1168469a8229f606")

Parameters:

dataset_id (String) – Internal dataset_id for the given dataset in Tensoract workbench

Response

Export status

{'progress': '100%'}

list_dataset_exports(dataset_id)

This method enlisits the exports exported in the dataset.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.list_dataset_exports(dataset_id="449354de1168469a8229f605")

Parameters:

dataset_id – Internal dataset_id for the given Dataset in Tensoract workbench

Response

[
    {'created': 1667901440,
    'createdBy': 'abc@example.com',
    'datasetId': '3aa168a963f1d4dbe0b1f549',
    'exportId': '***************',
    'filter': '',
    'items': 44,
    'note': 'This is test export via API',
    'project': '',
    'projects': [],
    'selection': False}

]

download_dataset_export(export_id)

It downloads the exports with given export id.

Sample Invocation

from pprint import pprint
from tensoract.client import Tensoract

# Initialize Tensoract  client
client = Tensoract(api_url="YOUR_API_URL",api_key="YOUR_API_KEY")

response = client.download_dataset_export(export_id="449354de1168469a8229f605")

Response

{'export_url': 'https://devdocuments.objectways.com/export_2022-08-20T14:55:09.5593106_goldenset1.manifest',
'message': 'Use the export_url to download the project export file',
'progress': '100%',
'status': 'COMPLETED'}