2.1 KiB
2.1 KiB
parse-bmfont-xml
Parses XML BMFont files.
Takes a string or Buffer:
var fs = require('fs')
var parse = require('parse-bmfont-xml')
fs.readFileSync(__dirname+'/Arial.fnt', function(err, data) {
var result = parse(data)
console.log(result.info.face) // "Arial"
console.log(result.pages) // [ 'sheet0.png' ]
console.log(result.chars) // [ ... char data ... ]
console.log(result.kernings) // [ ... kernings data ... ]
})
Also works in the browser, for example using XHR:
var parse = require('parse-bmfont-xml')
var xhr = require('xhr')
xhr({ uri: 'fonts/NexaLight32.xml' }, function(err, res, body) {
if (err)
throw err
var result = parse(body)
console.log(result.info.face)
})
The spec for the returned JSON object is here. The input XML should match the spec with a <font>
root element, see [test/Nexa Light-32.fnt](test/Nexa Light-32.fnt) for an example.
See Also
See text-modules for related modules.
Usage
result = parse(data)
Parses data
, a string or Buffer that represents XML data of an AngelCode BMFont file. The returned result
object looks like this:
{
pages: [
"sheet_0.png",
"sheet_1.png"
],
chars: [
{ chnl, height, id, page, width, x, y, xoffset, yoffset, xadvance },
...
],
info: { ... },
common: { ... },
kernings: [
{ first, second, amount }
]
}
If the data is malformed, an error will be thrown.
The browser implementation relies on xml-parse-from-string, which may not work in environments without valid DOM APIs (like CocoonJS).
License
MIT, see LICENSE.md for details.