Data Thing, by The APS Group

Available, for your convenience, at http://theapsgroup.github.io/Data-Thing/

General purpose json fiddle tool. I found myself writing many small nodejs scripts just to transform data in another format. The overhead of reading and saving files is removed, furthermore a real-time feedback mechanism (better than console and nodemon) is added.

Load a json file by opening or dragging in.
Create a script by opening or creating from scratch. jsonld, Q and TreeLib are available as globals, more (or a more generic way to inject dependencies) may follow.
Use the "magic" variable input and return a new output. Most simple script is return input. For XQuery, the "magic" variable is $input.
Save the result.

General way of working

This thing needs a bit of getting used to. Some guidelines:

match the input mode to what you are working with
set modes to xml and xsl when working with those
set output mode to 'JSON' first when working with js transformations
make sure you output what you want, taking errors into account
switch mode to text or tsv when you have a data structure that supports it (see next)

Working with text or xml

By setting the input mode to text, the input is interpreted as an array of strings, which can be used by JavaScript transformation scripts.

By setting the input mode to xml, the input is interpreted as jsonml, which is a JSON serialization of xml. This can be used by JavaScript or XSL transformation scripts.

By setting the input mode to tsv, the input is interpreted as tab separated values (with or without header). You'll get this when you paste from excel. This can be used by JavaScript transformation scripts.

By setting the output mode to text, the output MUST be an array of strings.

By setting the output mode to xml, the output MUST be valid jsonml, which is automatically produced by a correct xsl transformation

By setting the output mode to TSV, the output MUST be an array of objects.

Working with HTML

Getting structured data out of HTML is easy using scrapio. Scrapio provides an easy way to extract some json structure from HTML, using a template:

scrapio.load(input).tmpl({
    title: scrapio('h1').text(),
    content: scrapio('section').text()
});

Output SQL

The following helper functions are available to output various text based file formats:

makeSQLInsert(arr,tableName);
makeSQLUpdate(arr,tableName,key);

These all expect an array of the same objects.

Furthermore, the following helpers are available for your convenience.

keys(obj);
values(obj);

Linking gists

By creating a gist with at least one of the following files:

input.js
input.txt
input.xml
input.tsv
script.js
script.xsl
script.xq

And calling the Data Thing with a gist id, you can use the stored scripts.

Examples:

JSON: http://theapsgroup.github.io/Data-Thing/#9ac1897316509d397d15
XML + XSL: http://theapsgroup.github.io/Data-Thing/#afb53e880965541454d7
XML + XQuery: http://theapsgroup.github.io/Data-Thing/#7db6cc513ceab177ebe8b6e196a29378
YAML: http://theapsgroup.github.io/Data-Thing/#68e5961db3901ba161bf
HTML + srapio: http://theapsgroup.github.io/Data-Thing/#2442a8f7a5873244feec9b3df331aa3d
JSON-LD: http://theapsgroup.github.io/Data-Thing/#9848b7ef1f48cc77a445f1cf58868a20 (try NQuads as output method)

Saving gists

By clicking the "save as gist" button, an anonymous gist with the input and script files will be created. Note that this gist is secret, but public nonetheless. Anyone with access to the url can view your data. Also, it cannot be easily deleted

Example

Suppose you have the following JSON (randomly generated data): http://beta.json-generator.com/PSKfAr_

Getting specific data

Return the email adress of the zeroth record

return input[0].email;

Filtering

Return only records with blue eyes:

return input.filter(function(r) {
    return r.eyeColor==='blue';
})

Mapping (creating a new / modified record for every record)

Return only eye color and name:

return input.map(function(r) {
    return {
        eyeColor: r.eyeColor,
        name: r.name
    }
});

A combination of functions:

To get the names of blue-eyed records:

return input.filter(function(r) {
    return r.eyeColor==='blue';
}).map(function(r) {
    return {
        eyeColor: r.eyeColor,
        name: r.name
    }
});

Or to create groups by eye color:

return input.reduce(function(groups,r) {
    if (!groups[r.eyeColor]) {
        groups[r.eyeColor] = [];
    }
    groups[r.eyeColor].push({
        name: r.name
    });
    return groups;
},{});

theapsgroup / Data-Thing