CLI tools for SourceCode
npm install @sourcecodeoer/cliThe purpose of this cli tool is to easily bulky export / import exercises for the API.
- Pandoc (for the description conversion)
``sh`
npm install --save-dev @sourcecodeoer/cli
This tool was build on top of yargs.
For the full list of commands and options, everything is explained with the help command :
`sh`
npx sourcecode-cli --help
To be platform independent, we define the following rule :
Each result file generated by one of the commands must following our specification.
Of course, nothing blocks you to extend our specification for your own purpose.
An small example before explanation :
`json`
{
"exercises":[
{
"title":"An exercise",
"description": "Some description here",
"url": "https://inginious.info.ucl.ac.be/courselist",
"tags":[
{
"autoGenerated":true,
"category_id":"_PROGRAMMING-LANGUAGE_",
"text":"c"
},
{
"text": "S3",
"category": 2
}
]
}
],
"own_categories":{
"0":"thématique",
"1":"Misconception",
"2":"autres"
},
"extraction_date": "2019-11-26T14:04:34.107Z",
"url": "https://github.com/UCL-INGI/LSINF1252"
}
Only two properties are required inside : and "own_categories". "extraction_date"
Extra properties can be defined if you want (like and "url" to keep track of the source / extraction date).
Inside the key "exercises" , we have an array of exercises metadata. "title"
As described in the endpoint /api/bulk_create_exercises,
some attributes for an exercise are required
(like , tags, description) whereas some are optional (like "url" or "file").
As you can see, whatever the platform, there is some common tags categories that should be used whenever it is possible.
An possible example might be:
`json`
{
"_PLATFORM_":"plateforme",
"_SOURCE_":"source",
"_COURSE_":"cours",
"_EXERCISE-TYPE_":"type d'exercice",
"_PROGRAMMING-LANGUAGE_":"langage",
"_AUTHOR_":"auteur"
}
To distinguish them from your own tag categories, when you used one of them for a tag, we must specify the property
"autoGenerated" to true and category_id to the right key of previously explained object :
`json`
{
"autoGenerated":true,
"category_id":"_PROGRAMMING-LANGUAGE_",
"text":"c"
}
For tags categories of your own platform, You must describe them in "own_categories" property as an mapping object :{}
( if you have none)
`json`
{
"0":"thématique",
"1":"Misconception",
"2":"autres"
}
Then use them in tags as shown in the example ("category" with your own key) :`
{
"text":"S3",
"category":2
}
If you wish to add the sources of an exercise, it should be a zip file.
Two choices are possible :
1. If you already have the file, you should use the key "file" with value a absolute path of this file :
`json`
{
"file": "/home/jy95/folder/mysources.zip"
}
2. If you don't have the file, you can delegate the creation of zip file to archiver command thanks to the key "archive_properties" structured like that :
`json`
{
"archive_properties": {
"folders": [],
"files": []
}
}
These two sub properties must be present : use an empty array if you don't have any folder or file.
Each item of these arrays should be relative path so can build the absolute path with the option baseFolder (in case you change location of an exercise folder afterward)"file"
and automatically add the property in each exercise.
Using the property "file" in each exercise, the uploader command will do for you the mapping between files and exercises when you want to upload exercises into the API. auto_generated_tags_categories
(don't forget to use option in uploader if you used other common tag categories)
- inginious-git : for INGInious tasks / course(s) stored on a git repository
Simply create a module file that looks like this :
`node`
module.exports = async function (options) {
// TODO
}
This module should return an valid object to the specification.
The given options` parameter is simply the argv object created by yargs.
Check out inginious-git if you want to see an example about how to use it.