RoboTA config¶

RoboTA reads in various types of data from different sources. This could be data about software engineering objects such as git commits or it could be about the human elements of software engineering such as interactions with an issue board or bug tracker.

These data types and their sources are specified in the robota-config file.

Data types¶

Each data type provides a different type of information to RoboTA that is used somewhere in the code. Depending on what parts of RoboTA are being used, not all the data types may be required.

Every data type requires a ‘data_source’ key which says where the data comes from. Some data types have additional keys which are required to specify details of the data type.

marking_config¶

The location of the set up files for RoboTA marking.

Valid data sources:

local_path
gitlab

Required keys:

course_config_file: The path of the Course config file
mark_scheme_file: The path of the file that specifies the mark scheme
build_config_file: The path to the file that has information about the build config

All of these keys may specify sub-folder(s) in the git repository, e.g.: course_config_file: config_files/course/course_config.yaml

issues¶

The location of issues. Issues are used in marking students planning and teamwork.

Valid sources:

gitlab
github

ci¶

A CI server hosting tests assessing student code.

Valid sources:

jenkins

repository¶

A source of information about commits, tags, events, branches and files in the student work repository.

Valid sources:

gitlab
github
local_repository

remote_provider¶

A cloud provider that hosts git repositories. Provides info about pull/merge requests and team members

Valid sources:

gitlab
github

attendance¶

Information about student attendance at a class, workshop or event

Valid sources:

benchmark

student_details¶

Information about student names and usernames

Valid sources:

gitlab
local_path

student_emails¶

The relationship between student usernames and email addresses

Valid sources:

gitlab
local_path

required keys:

name_list_file - The path of the file containing the mapping between names and email addresses

ta_marks¶

Manually assigned marks that can be used to override RoboTA marks in the marking report

Valid sources:

gitlab
local_path

required keys:

ta_marks_file - The path of the file containing the TA marks.

Data Sources¶

Each data source specified in a data type should correspond to a data source in the data sources section. For flexibility the data sources are specified by name which means that each data source could be used for multiple data types.

Each data source has a type. Each data source type is treated differently by the code.

The different data sources are:

local_path
gitlab
github
local_repository
jenkins
benchmark

Below are specified the keys which are required for each data source type.

local_path¶

Download files from a local directory on the machine on which RoboTA is running.

required sub-values:

path - The path to look in for the file(s).

gitlab¶

Connect to a remote gitlab instance to retrieve information or files.

required sub-values:

url: The url of the gitlab instance
project: The name of the gitlab project to load
token: An authentication token to connect to the gitlab instance

optional sub-values:

branch: Which git branch to assess - defaults to ‘master’

github¶

Connect to a remote GitHub instance to retrieve information or files.

required sub-values:

url: The url of the GitHub instance
project: The name of the GitHub project to load
token: An authentication token to connect to the GitHub instance

optional sub-values:

branch: Which git branch to assess - defaults to ‘master’

local_repository¶

Connect to a repository on the local machine.

required sub-values:

path: The path of the git repository.

optional sub-values:

branch: Which branch to consider - defaults to ‘master’

jenkins¶

Connect to a remote Jenkins instance to retrieve job information

required sub-values:

url: The url of the Jenkins instance
username: Username used for authentication
token: Token used for authentication
project_name: The name of the project containing the tests
folder_name: The name of the folder in the project containing the tests

benchmark¶

A University of Manchester service that has information about students

required sub-values:

url: The URL of the benchmark instance
token: Token used for authentication

Note that a connection to benchmark requires either being on campus or use of the UoM VPN.

Example Config¶

# This is an example robota config file.
# This file is used it to store RoboTA config variables and credentials locally.
# You should NOT commit any credentials to git.

# The 'data_types' section specifies where the data to run RoboTA comes from. The keys are
the data type and are fixed as they are specified in the code.
# The data source key is mandatory for each data type. Other key: value pairs are passed
# into the code to be used for configuration of that data type.

data_types:
    issues:
        data_source: github_repo
    repository:
        data_source: github_repo
    remote_provider:
        data_source: github_repo

# Details of data sources. The name of each data source corresponds to those specified in the data_types section above.
# Keys and values are specific to the data source.

data_sources:
    github_repo:
        type: github
        url: www.github.com
        project: merrygoat/chi4
        token: xxx-xxx-xxx

    local_repository:
        type: local_path
        directory: C:/robota/chi_4

In this case repository could probably be set to the data source: github_repo, but it might be useful to set it to local_repository if the repository was already synced locally and was large. Operating on large repositories locally is likely to be more efficient in most cases than querying them through the API.

Variable Substitution¶

To improve automation, named keys in the config file can be specified which are replaced by values at run time. Strings to be substituted should be enclosed in curly brackets.

The second argument of the robota_core.config_readers.get_robota_config() method is a dictionary of substitutions. The keys are the variable to replace and the values are the values to substitute in.

An example robota config might look like:

data_types:
    repository:
        data_source: student_repo

data_sources:
    student_repo:
        type: github
        url: www.gitlab.com
        project: UoMProgramming/first_year/Team{team_number}
        token: xxx-xxx-xxx

To loop assessment over many teams you could read in the config in a loop using the robota_core.config_readers.get_robota_config() method. On the first loop the substitution_vars parameter would be {“team_number”: 01}, on the second loop, {“team_number”: 02} etc.