Skip to content

Latest commit

 

History

History
205 lines (154 loc) · 5.62 KB

Backend-API.md

File metadata and controls

205 lines (154 loc) · 5.62 KB
Raw

SAVE Cloud Backend API

Overview

The process of execution submission is very simple, all that you have to do is just to provide your credentials and information about your project, which you would like to evaluate and do few requests.
In this document will be represented the examples, which use cURL as instrument for working with SAVE Cloud API.

To make the requests more universal, we will store major information, which is required by SAVE Cloud in requests, by into variables in command prompt.

Authorization configuration

Each request, performed to API, require credentials, which could be configured like below, where

  • SAVE_CLOUD_URL - SAVE Cloud url and port
  • SAVE_CLOUD_AUTH - basic authorization header in default format username:token.
    token is unique token, which could be created in personal settings in SAVE Cloud system.

Just open up a command prompt and enter the following commands:

SAVE_CLOUD_URL=https://saveourtool.com:443

SAVE_CLOUD_AUTH='Basic Z2l0aHViQHVzZXJuYW1lOnRva2Vu'

Configuration of information about evaluated tool

To specify information about project, that you would like to evaluate, it's necessary to create the fields below in command prompt. The organizationName, projectName are required. IDs of test suites should be provided for execution. Also need to provide a list of additional files:

# Required
organizationName=Huawei

# Required
projectName='"save"'

# Optional
sdkName='"openjdk"'
sdkVersion='"11"'

# Required
testSuites='[1, 2]'

# Required
additionalFiles='[fileName1:123;fileName2:321]'

Note: The organizationName here is intentionally don't contain quotes, since this value will be used in requests url, while other in requests bodies.

Optional configuration

Expand 
# Optional
execCmd=null

# Optional
# Batch size controls how many files will be processed at the same time.
batchSize=null

API for execution

Upload additional files into SAVE Cloud storage

If your tests require additional files for execution, they could be uploaded by following post request:

curl -X POST "${SAVE_CLOUD_URL}/api/v1/files/upload" \
-H "Authorization: ${SAVE_CLOUD_AUTH}" \
-F "file=@your-file-name"

It will return json with metadata about your file, which will be required
in the execution request later.

The format will have the following form:

{
  "name": "your-file-name",
  "uploadedMillis":1658227620000,
  "isExecutable":false
}

By default, the isExecutable argument, which indicates, whether the provided file
should be executable, or not, is false.
However, for the execution request from the next section, you can set this argument with value, that you need.

Execution submission

For execution request you will need full information about your project, which could be saved into variable project by following request:

project=$(curl -X GET "${SAVE_CLOUD_URL}/api/v1/projects/get/organization-name?name=save&organizationName=${organizationName}" \
-H "Authorization: ${SAVE_CLOUD_AUTH}")

Unifying all above, here the examples of how to submit execution via curl:

Request for execution submission 
curl -X POST "${SAVE_CLOUD_URL}/api/v1/run/trigger" \
-H "Authorization: ${SAVE_CLOUD_AUTH}" \
-H "Content-Type: application/json" \
-d "{
    \"projectCoordinates\": {
        \"organizationName\": ${organizationName},
        \"projectName\": ${projectName},
    },
    \"testSuiteIds\": [${testSuiteId1}, ${testSuiteId1}],
    \"files\": [
        {
            \"name\": ${fileKey1.name},
            \"uploadedMillis\": ${fileKey1.uploadedMillis}
        },
        {
            \"name\": ${fileKey2.name},
            \"uploadedMillis\": ${fileKey2.uploadedMillis}
        }
    ],
    \"sdk\": {
      \"name\": ${sdkName},
      \"version\": ${sdkVersion}
    },
    \"execCmd\": ${execCmd},
    \"batchSizeForAnalyzer\": ${batchSizeForAnalyzer}"

These requests will also return execution id, which could be used for getting the execution results or for rerun command.

For example:

Clone pending, execution id is 42

To get execution results, knowing the execution id, use the following request:

# Taken after submitExecutionRequest request
executionId=42

curl -X GET "${SAVE_CLOUD_URL}/api/v1/executionDto?executionId=${executionId}" \
-H "Authorization: ${SAVE_CLOUD_AUTH}"

You are also able to get results of latest execution by /latestExecution endpoind, with the following get request:

curl -X GET "${SAVE_CLOUD_URL}/api/v1/latestExecution?name=save&organizationName=${organizationName}" \
-H "Authorization: ${SAVE_CLOUD_AUTH}"

The response format will look like:

{
  "id": 42, # execution id
  "status": "FINISHED", # execution status, i.e. running, finished and so on
  "type":"PRIVATE_TESTS", # testing type
  "version": "264e5feb8f4c6410d70536d6fc4bdf090df62287", # commit hash
  "startTime": 1651856549, # start time of execution in Unix format
  "endTime": 1651856797, # end time of execution in Unix format
  "runningTests":0, # number of running tests at this moment
  "passedTests":20, # number of passed tests
  "failedTests":3, # number of failed tests
  "skippedTests":1, # number of skipped tests, i.e., because of configuration 
  "additionalFiles": [ # the list of additional files
    "/file-1",
    "/file-2",
    ]
}

If you would like to rerun some of your executions, you can use /run/re-trigger endpoint:

curl -X POST "${SAVE_CLOUD_URL}/api/v1/run/re-trigger?executionId=${executionId}" \
-H "Authorization: ${SAVE_CLOUD_AUTH}"