@statebox/pnpro.js

PNPRO.js

Simple JS library to work with .pnpro files.

• Parse the file
• Convert to NLL
• Convert to NBPT

These files contain multiple "pages", to select a single net and convert it, run:

let pnproNet = parse(xml).nets["page title"]
let nll = to_nll(pnproNet)
let nbpt = to_nbpt(pnproNet)

API Details

Assuming you have the contents of a PNPRO file as a string called pnproProject, here is an overview of the functions exported by this library.

Peek into the project file and pick "pages" (nets) out of it:

• parse(pnproProject)

Convert a net into some target:

• to_nll(pnproNet) : NLL
• to_nbpt(pnproNet) : NBPT

More complete example

For Node:

let PNPRO = require('@statebox/pnpro.js')
let R = require('ramda')

// load a file as a string
let xmlStr = fs.readFileSync('file.pnpro', {encoding:'utf8'})

let project = PNPRO.parse(xmlStr)

let nlls = R.map(PNPRO.to_nll, project.nets) // {name => nll}
let nbpts = R.map(PNPRO.to_nll, project.nets) // {name => nbpt}

Parsed JSON Format

The two main keys are:

• .title ~ project title
• .nets ~ dictionary of {title => net}

Here is the full output of parse(..):

{
    "title": "Project Title",
    "nets": {
        "page title": {
            "title": "page title",
            "net": {
                "marking": {
                    "placeLabel": 1
                },
                "transitions": [
                    {
                        "label": "buy",
                        "x": 32.55,
                        "y": 11,
                        "pre": {
                            "Fresh": 1
                        },
                        "post": {
                            "Sold": 1
                        }
                    }
                ],
                "places": [
                    {
                        "label": "Fresh",
                        "x": 42,
                        "y": 22
                    }
                ]
            }
         }
    }
}

Background

We have various repo's that deal with PNRO

• https://github.com/statebox/gspn-parser
• https://github.com/statebox/purescript-pnpro-parser
• https://github.com/statebox/js-pnpro-parser

And now also https://github.com/statebox/pnpro.js

But Why?

There are some issues with the current approaches, mainly related to the library used to parse the XML.

• js-pnpro-parser uses the browser provided DOMParser()
• if js-pnro-parser is used from Node, xmldom is used
• gspn-parser uses cheerio which cannot be bundled well for the browser (due to cyclic dependencies)

Instead I propose we use this 100% JS + regex based one:

• https://github.com/NaturalIntelligence/fast-xml-parser

We need a simple library that works from node or the browser and doesn't have any complicated (native) dependencies and can be bundled properly.

It's just an XML parser anyway!