FeatureIDE-Service

A microservice proving an RESTful API for the feature model conversion using the FeatureIDE library.

Installation

How to Use CLI

Start in Intellij

Use the CLI Main Method to start the Application and use the configurations option in Intellij to configure Commandline Arguments.

Start in Terminal

Use "./gradlew run -Pcli="typeOfCLI" --args="fill in arguments"" in a Terminal

The different types are:

• conv - Converting files and checking
• slice - Slicing FMs
• conf - Configuration Samples
• prop - Decision Propagation

For Converting and Check

Options:
--path, -p -> Input path for file or directory. (always required) { String }
--check, -c -> Input path for the second file that should be checked with the first one. { String }
--all, -a [false] -> Parsers all files from path into all formats.
--dimacs, -d [false] -> Parses all files from path into dimacs files.
--uvl, -u [false] -> Parses all files from path into uvl files.
--sxfm, -sf [false] -> Parses all files from path into sxfm(xml) files.
--featureIde, -fi [false] -> Parses all files from path into featureIde(xml) files.

For Example:  <br />
-p "FeatureIDE.xml" -c "FeatureIDE2.dimacs"  --> Checks if the Hashes of both files are the same after Converting it to the same format <br />
-p "FeatureIDE.xml" -u  --> This command takes the FeatureIDE.xml File and converts it to an uvl file <br />

For Slicing

Options:
--path, -p -> Input path for file. (always required) { String }
--selection, -s -> The names of the features that should be sliced separated by ','. For example: Antenna,AHEAD. { String }

For Example:  <br />
-p "FeatureIDE.xml" -s "Antenna"  --> This command takes the FeatureIDE.xml File and slices out the Feature Antenna <br />

For Configuration Samples

Options:
--path, -p -> Input path for file. (always required) { String }
--algorithm, -alg -> The algorithm to generate a configuration sample as csv file { String }
--t, -t [0] -> The t wise pairing that should be covered by the configuration sample { Int }
--limit, -l [2147483647] -> The maximum amount of configurations for the configuration sample { Int }
--time, -ti [-1] -> The maximum time that an algorithm can use for the execution only with yasa { Int }

For Example:  <br />
-p "FeatureIDE.xml" -alg "yasa_50" -t 2 -l 10  --> This command takes the FeatureIDE.xml File and gives an sampling for the model with the parameters iterations 50, 2-wise and a limit of 10 <br />

For DecisionPropagation

Options:
--path, -p -> Input path for file. (always required) { String }
--selection, -s -> The names of the features that are already selected separated by ','. For example: Antenna,AHEAD. { String }

For Example:  <br />
-p "FeatureIDE.xml" -s "Antenna"  --> This command takes the FeatureIDE.xml File and returns a Array of Features that need to be selected to provide a valid configuration <br />

All of the Files are saved after the operation in the "files/output" directory which is completely cleared before saving any files.
--help, -h -> Usage info

How to Start Server API

Build and Run as Docker Container

1. Change dir to ktor-api directory
2. Run docker compose up

Building the jar takes some time (1-2 minutes).

Use it in Intellij

1. Run Main Method of Application

Use it in Terminal

1. Run "./gradlew run" in Commandline

How to Use API

CONVERT

Send the file you want to convert with an HTTP POST to the path /convert. The content of the HTTP POST need to be the following.

Input

• name: String - The name of the featuremodel
• typeOutput: Array<String> - The list of types to convert the file to (currently supported: dimacs, uvl, sxfm, featureIde)
• content: Array<Byte> - The content of the featuremodel

After this you receive a HTTP Message with the content "Request Accepted" and a Location header to now receive your converted file/s make a HTTP GET Request until you receive the HTTP STATUS OK. The content of the message will be following:

Output

• name: Array<String> - The name of the different files
• typeOutput: Array<String> - The list of types to convert the file to (currently supported: dimacs, uvl, sxfm, featureIde)
• content: Array<Byte> - The contents of the different files

SLICE

Send the file you want to slice with an HTTP POST to the path /slice. The content of the HTTP POST need to be the following.

Input

• name: String - The name of the featuremodel
• selection: Array<String> - The features that should be sliced
• content: Array<Byte> - The content of the featuremodel

After this you receive a HTTP Message with the content "Request Accepted" and a Location header. To receive your sliced file make a HTTP GET Request on the Location until you receive the HTTP STATUS OK. The content of the message will be following:

Output

• name: String - The name of the feature model
• selection: Array<String> - The features that were sliced
• content: Array<Byte> - The contents of the sliced feature model

CONFIGURATION

Send the file you want to generate a sample with an HTTP POST to the path /configuration. The content of the HTTP POST need to be the following.

Input

• name: String - The name of the featuremodel file
• algorithm: String - The algorithm that should be used to generate the sample. Special case for "yasa" with "yasa_Number" the iterations can be changed
• t: Int - The T-Wise configuration for the algorithm
• limit: Int - The maximum amount of samples that the algorithm should use
• content: Array<Byte> - The content of the featuremodel file in xml

After this you receive a HTTP Message with the content "Request Accepted" and a Location header to now receive your converted file/s make a HTTP GET Request until you receive the HTTP STATUS OK. The content of the message will be following:

Output

• name: String - The name of the featuremodel file
• algorithm: String - The algorithm that should be used to generate the sample. Special case for "yasa" with "yasa_Number" the iterations can be changed
• t: Int - The T-Wise configuration for the algorithm
• limit: Int - The maximum amount of samples that the algorithm should use
• content: Array<Byte> - The content of the CSV file that was generated

PROPAGATION

Send the file you want to get an decision propagation with an HTTP POST to the path /propagation. The content of the HTTP POST need to be the following.

Input

• name: String - The name of the featuremodel
• selection: Array<String> - The features that are selected
• deselection: Array<String> - The features that are deselected
• content: Array<Byte> - The content of the featuremodel

After this you receive a HTTP Message with the content "Request Accepted" and a Location header to now receive your converted file/s make a HTTP GET Request until you receive the HTTP STATUS OK. The content of the message will be following:

Output

• name: String - The name of the featuremodel
• selection: Array<String> - The features that are selected
• deselection: Array<String> - The features that are deselected
• impliedSelection: Array<String> - The features that need to be selected because of the first selection/deselection
• impliedDeselection: Array<String> - The features that need to be deselected because of the first selection/deselection
• openParents: Array<String> - The features where the selection/deselection led to open children
• openChildren: Array<String> - The features where one of the features needs to be selected to have a valid configuration
• content: Array<Byte> - The content of the featuremodel

Testing

To run all tests: gradle test