GraphQL

Queries and Mutations

• Endpoint: (filenet-root-link)/content-services-graphql/graphql
  • where home link comes before /acce/ and /navigator/ for the acce console and navigator, respectively
• XSRF Token: can be found in the page source of the graphiql page
  • XSRF/CSRF stands for Cross-Site Request Forgery, the token provides protection from this

Prerequisites

• xsrf authentication is enabled in the CR
• if using GraphiQL, this also must be enabled in the CR

Lambda Functions

Prerequisites

• Endpoint
• Authorization Header (Bearer Token)
• aws cli and credentials

Development Flow

1. If possible, test your query in the graphiql interface (Note: graphiql does not support file uploads as it cannot access local files)
2. Test your query as a curl command
3. Place the curl command into a Python function, making use of the json and requests libraries
4. Test running the Python function locally
5. Modify the Python function to run as a lambda function
6. Install the required python libraries into the lambda python environment
7. Invoke the function (ensure your role is allowed to execute the lambda function and if required, access s3 buckets)

Example: Uploading a file from an S3 bucket to Filenet

1. curl command:

Windows

curl -H 'Authorization: Basic <insert xsrf token here>' -H 'content-type: multipart/form-data' --verbose '(filenet-root-link)/content-services-graphql/graphql' -F graphql='{"query":"mutation ($contvar:String) { createDocument (repositoryIdentifier: \"<insert repo name here: ex. OS01>\" fileInFolderIdentifier: \"/<insert filenet folder path here ex. CUSTOMERS/TEST-LOAN>\" documentProperties: {name: \"<insert document name here>\" contentElements:{replace: [{type: CONTENT_TRANSFER contentType: \"<insert content type here: ex. text/plain>\" subContentTransfer: {content:$contvar} }]}} checkinAction: {} ) { id name }}", "variables":{"contvar":null}}' -F 'contvar=@\"C:/path/to/file/file.extension\"'

macOS/Linux

curl -H 'Authorization: Basic <insert xsrf token here>' -H 'content-type: multipart/form-data' --verbose '(filenet-root-link)/content-services-graphql/graphql' -F graphql='{"query":"mutation ($contvar:String) { createDocument (repositoryIdentifier: \"<insert repo name here: ex. OS01>\" fileInFolderIdentifier: \"/<insert filenet folder path here: ex. CUSTOMERS/TEST-LOAN>\" documentProperties: {name: \"<insert document name here>\" contentElements:{replace: [{type: CONTENT_TRANSFER contentType: \"<insert content type here: ex. text/plain>\" subContentTransfer: {content:$contvar} }]}} checkinAction: {} ) { id name }}", "variables":{"contvar":null}}' -F 'contvar=@/path/to/file/file.extension'

2. Python function:

file_upload_local_curl_query.py

import requests
import json

file_mac_linux = '/path/to/file/filename.extension'
file_windows = 'C:/path/to/file/filename.extension'
file = file_<platform>                                  # pick your platform
repo_ID = "<insert repo name here>"                     # ex. OS01
folder_path = "<insert folder path here>"                # ex. /CUSTOMERS/TEST-LOAN
document_name = "<insert document name here>"           # ex. curl upload test
content_type = "<insert content type here>"             # ex. text/plain
filenet_home_link = "<insert filenet home link>"
xsrf_token = "<insert xsrf token here>"

# Define the GraphQL query and variables
graphql_query = """
mutation ($contvar: String) {
    createDocument (
        repositoryIdentifier: repo_ID
        fileInFolderIdentifier: folder_path
        documentProperties: {
            name: document_name
            contentElements: {
                replace: [{
                    type: CONTENT_TRANSFER
                    contentType: content_type
                    subContentTransfer: {
                        content: $contvar
                    }
                }]
            }
        }
        checkinAction: {}
    ) {
        id
        name
    }
}
"""

variables = {"contvar": None}

# Define the GraphQL endpoint URL and headers
url = filenet_home_link + '/content-services-graphql/graphql'
headers = {
    'Authorization': 'Basic ' + xsrf_token,
}

operations = json.dumps({
            "query": graphql_query,
            "variables": variables
        })

data = {
    "operations": operations
}

# Prepare the multipart form data
files = {
    'contvar': (file, open(file, 'rb'), content_type)
}

# Send the request
response = requests.post(url, headers=headers, files=files, data=data)

# Handle the response
if response.status_code == 200:
    result = response.json()
    print(response)
    print(response.text)
    # Process the GraphQL response here

# Handle the error
else:
    print(f"Request failed with status code {response.status_code}")
    print(response.text)

3. Test this function by running python3 local_curl_query.py
4. Python function as a lambda handler:

file_upload_lambda_handler.py

import json
import requests
import boto3
import os

repo_ID = "<insert repo name here>"                                 # ex. OS01
folder_path = "<insert folder path here>"                           # ex. /CUSTOMERS/TEST-LOAN
document_name = "<insert document name here>"                       # ex. curl upload test
content_type = "<insert content type here>"                         # ex. text/plain
filenet_home_link = "<insert filenet home link>"
xsrf_token = "<insert xsrf token here>"
s3_bucket_name = '<insert filenet s3 bucket name here>'             # ex. filenet-test-bucket
file_from_s3_bucket = '<insert file to download from s3 bucket>'    # ex. IBMContentManagerDevice.txt
target_file = '/tmp/<insert filename here with .extension>'         # ex. test.txt

def lambda_handler(event = None, context = None):
    # Define your GraphQL query and variables
    graphql_query = """
    mutation ($contvar: String) {
        createDocument (
            repositoryIdentifier: repo_ID
            fileInFolderIdentifier: folder_path
            documentProperties: {
                name: document_name
                contentElements: {
                    replace: [{
                        type: CONTENT_TRANSFER
                        contentType: content_type
                        subContentTransfer: {
                            content: $contvar
                        }
                    }]
                }
            }
            checkinAction: {}
        ) {
            id
            name
        }
    }
    """
    
    variables = {"contvar": None}

    # Prepare the multipart form request
    url = filenet_home_link + '/content-services-graphql/graphql'
    headers = {
        'Authorization': 'Basic ' + xsrf_token,
    }

    operations = json.dumps({
                "query": graphql_query,
                "variables": variables
            })

    data = {
        "operations": operations
    }

    # Upload file to Lambda's /tmp directory
    s3 = boto3.client('s3')
    s3.download_file(s3_bucket_name, file_from_s3_bucket, target_file)

    files = {
        'contvar': (target_file, open(target_file, 'rb'), content_type)
    }

    # Send the request
    response = requests.post(url, headers=headers, files=files, data=data)

    # Handle the response
    if response.status_code == 200:
        result = response.json()
        # Process the GraphQL response here
    else:
        print(f"Request failed with status code {response.status_code}")
        # Handle the error

    print(response)
    print(response.text)
    return result

5. Install required python libraries into the lambda python environment

• Follow these aws cli setup steps if you haven't already
• Create a role that can execute lambda functions:

aws iam create-role --role-name filenet-lambda-s3 --assume-role-policy-document file://permissions.json

permissions.json

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "sts:AssumeRole"
            ],
            "Principal": {
                "Service": [
                    "lambda.amazonaws.com"
                ]
            }
        }
    ]
}

• give that role permission to access s3 buckets (in this example we grant permission to perform all s3 bucket actions):

aws iam put-role-policy --role-name filenet-lambda-s3 --policy-name s3_permissions --policy-document file://s3permissions.json

s3permissions.json

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "s3:*"
            ],
            "Resource": "*"
        }
    ]
}

• Continue to the environment setup to install the boto3, json and requests libraries
• When creating the package, replace the filename with file_upload_lambda_handler.py
• Then create the aws lambda function, replce the function name, as well as the role and zip file path parameters

6. Invoke the function, replacing the function name you chose in the previous step:

aws lambda invoke --function-name <insert function name here> out --log-type Tail --query 'LogResult' --output test | base64 -d

Your file should now appear in filenet under the chosen directory with the given name like so: