A simple Nodejs wrapper for PICT, this brings the goodness of pairwise (aka all pairs) testing into the Javascript world. All models can be defined using either a simple JSON structure, inline text, or file-based models are also supported. The generated set of test cases will be in an easy to consume JSON array.
For a full overview of capabilities of PICT, visit Microsoft's gitub repo or PICT's detailed documentation
npm install pict-pairwise-testing
const pict = require('pict-pairwise-testing').pict
// model for a fictitious mortgage calculator application
const model = {
parameters: [
{ property: 'Principal', values: [300000, 350000, 4000000] },
{ property: 'Number of years', values: [20, 21, 22, 23, 24] },
{ property: 'Repayment type', values: ['Interest only 1 year', 'Interest only 2 years'] },
{ property: 'Interest Rate', values: [1, 1, 99, 2.99, 3.5, 4] }
],
};
let result = pict(model);
console.log('Generated test cases', result.testCases)
Pict will then generate the following test cases:
[
{
principal: '350000',
number_of_years: '23',
repayment_type: 'Interest only 1 year',
interest_rate: '1'
},
{
principal: '300000',
number_of_years: '21',
repayment_type: 'Interest only 2 years',
interest_rate: '99'
},
{
principal: '4000000',
number_of_years: '21',
repayment_type: 'Interest only 1 year',
interest_rate: '1'
},
...
...
...
{
principal: '4000000',
number_of_years: '23',
repayment_type: 'Interest only 2 years',
interest_rate: '3.5'
},
{
principal: '300000',
number_of_years: '22',
repayment_type: 'Interest only 2 years',
interest_rate: '1'
}
]
Inline text models are also supported, this can be achieved by defining the model as follows:
const pict = require('pict-pairwise-testing').pict
const model = `
Principal: 300000,350000,4000000
Number of years: 20,21,22,23,24
Repayment type: Interest only 1 year,Interest only 2 years
Interest Rate: 1,1,99,2.99,3.5,4
`
let result = pict(model);
console.log('Generated test cases', result.testCases)
You can even use old school PICT model file by using the package like so:
my_model.pict
Principal: 300000,350000,4000000
Number of years: 20,21,22,23,24
Repayment type: Interest only 1 year,Interest only 2 years
Interest Rate: 1,1,99,2.99,3.5,4
const pict = require('pict-pairwise-testing').pict
const __dirname = path.resolve(path.dirname(''));
const modelFile = path.resolve(__dirname, 'fixtures/my_model.model');
let results = pict(modelFile);
console.log('Generated test cases', result.testCases)
The following set of command-line options are supported:
key | mapped to PICT option | description |
---|---|---|
order_of_combinations | /o:N | - Order of combinations (default: 2) |
randomize_generation | /r[:N] | - Randomize generation, N - seed |
case_sensitive_model_evaluation | /c | - Case-sensitive model evaluation |
show_model_statistics | /s | - Show model statistics |
CLI parameters can be invoked by passing an option object into PICT:
const pict = require('pict-pairwise-testing').pict
const model = {
parameters: [
{ property: 'Status', values: ['Open', 'Closed'] },
{ property: 'Threshold', values: [10, 100, 500] },
{ property: 'Approved', values: ['yes', 'no'] },
],
};
let result = pict(model, {
options: { randomize_generation: 10, order_of_combinations: 1 },
});
The show_model_statistics
option can be used to preview model statistics derived by PICT:
const pict = require('pict-pairwise-testing').pict
const model = {
parameters: [
{ property: 'Status', values: ['Open', 'Closed'] },
{ property: 'Threshold', values: [10, 100, 500] },
{ property: 'Approved', values: ['yes', 'no'] },
],
};
let result = pict(model, {
options: { show_model_statistics: true }
});
console.log(result.statistics)
This yields
{
combinations: 16,
generated_tests: 6,
}
- separator_for_values: /d:C - Separator for values (default: ,)
- separator_for_aliases: /a:C - Separator for aliases (default: |)
- negative_value_prefix: /n:C - Negative value prefix (default: ~)
Constraints can be defined by passing a constraints array into the model, for example:
const pict = require('pict-pairwise-testing').pict
const modelWithConstraints = {
parameters: [
{ property: 'Size', values: [10, 100, 500, 1000, 5000, 10000, 40000] },
{ property: 'Format method', values: ['quick', 'slow'] },
{ property: 'File system', values: ['FAT', 'FAT32', 'NTFS'] },
{
property: 'Cluster size',
values: [512, 1024, 2048, 4096, 8192, 16384, 32768, 65536],
},
{ property: 'Compression', values: ['on', 'off'] },
],
constraints: [
'IF [File system] = "FAT" THEN [Size] <= 4096;',
'IF [File system] = "FAT32" THEN [Size] <= 32000;',
],
};
let result = pict(modelWithConstraints);
Sub-models can be defined by passing a submodels array into the model, for example:
const pict = require('pict-pairwise-testing').pict
const modelWithSubmodel = {
parameters: [
{ property: 'PLATFORM', values: ['x86', 'x64', 'arm'] },
{ property: 'CPUS', values: [1, 2, 4] },
{ property: 'RAM', values: ['1GB', '4GB', '64GB'] },
{ property: 'HDD', values: ['SCSI', 'IDE'] },
{ property: 'OS', values: ['Win7', 'Win8', 'Win10'] },
{ property: 'Browser', values: ['Edge', 'Opera', 'Chrome', 'Firefox'] },
{ property: 'APP', values: ['Word', 'Excel', 'Powerpoint'] },
],
submodels: ['{ PLATFORM, CPUS, RAM, HDD } @ 3', '{ OS, Browser } @ 2'],
};
let result = pict(modelWithSubmodel);
There is nothing special about this package. All its doing is parsing the JSON pict model into text model which PICT can understand. PICT then performs the necessary heavy lifting to generate the test cases. Under the hood, this package fully relies on the PICT executable from Microsoft's gitub repo
[MIT][mit]