readme.markdown #1

  • //
  • guest/
  • perforce_software/
  • helix-web-services/
  • main/
  • source/
  • client-javascript/
  • node_modules/
  • falafel/
  • readme.markdown
  • Markdown
  • View
  • Commits
  • Open Download .zip Download (3 KB)

falafel

Transform the ast on a recursive walk.

browser support

build status

This modules uses acorn to create an AST from source code.

falafel döner

example

array.js

Put a function wrapper around all array literals.

var falafel = require('falafel');

var src = '(' + function () {
    var xs = [ 1, 2, [ 3, 4 ] ];
    var ys = [ 5, 6 ];
    console.dir([ xs, ys ]);
} + ')()';

var output = falafel(src, function (node) {
    if (node.type === 'ArrayExpression') {
        node.update('fn(' + node.source() + ')');
    }
});
console.log(output);

output:

(function () {
    var xs = fn([ 1, 2, fn([ 3, 4 ]) ]);
    var ys = fn([ 5, 6 ]);
    console.dir(fn([ xs, ys ]));
})()

methods

var falafel = require('falafel')

falafel(src, opts={}, fn)

Transform the string source src with the function fn, returning a string-like transformed output object.

For every node in the ast, fn(node) fires. The recursive walk is a pre-traversal, so children get called before their parents.

Performing a pre-traversal makes it easier to write nested transforms since transforming parents often requires transforming all its children first.

The return value is string-like (it defines .toString() and .inspect()) so that you can call node.update() asynchronously after the function has returned and still capture the output.

Instead of passing a src you can also use opts.source.

All of the opts will be passed directly to acorn.

custom parser

You may pass in an instance of acorn to the opts as opts.parser to use that version instead of the version of acorn packaged with this library.

var acorn = require('acorn-jsx');

falafel(src, {parser: acorn, plugins: { jsx: true }}, function(node) {
  // this will parse jsx
});

nodes

Aside from the regular esprima data, you can also call some inserted methods on nodes.

Aside from updating the current node, you can also reach into sub-nodes to call update functions on children from parent nodes.

node.source()

Return the source for the given node, including any modifications made to children nodes.

node.update(s)

Transform the source for the present node to the string s.

Note that in 'ForStatement' node types, there is an existing subnode called update. For those nodes all the properties are copied over onto the node.update() function.

node.parent

Reference to the parent element or null at the root element.

install

With npm do:

npm install falafel

license

MIT

# falafel

Transform the [ast](http://en.wikipedia.org/wiki/Abstract_syntax_tree) on a
recursive walk.

[![browser support](http://ci.testling.com/substack/node-falafel.png)](http://ci.testling.com/substack/node-falafel)

[![build status](https://secure.travis-ci.org/substack/node-falafel.png)](http://travis-ci.org/substack/node-falafel)

This modules uses [acorn](https://npmjs.org/package/acorn) to create an AST from
source code.

![falafel döner](http://substack.net/images/falafel.png)

# example

## array.js

Put a function wrapper around all array literals.

``` js
var falafel = require('falafel');

var src = '(' + function () {
    var xs = [ 1, 2, [ 3, 4 ] ];
    var ys = [ 5, 6 ];
    console.dir([ xs, ys ]);
} + ')()';

var output = falafel(src, function (node) {
    if (node.type === 'ArrayExpression') {
        node.update('fn(' + node.source() + ')');
    }
});
console.log(output);
```

output:

```
(function () {
    var xs = fn([ 1, 2, fn([ 3, 4 ]) ]);
    var ys = fn([ 5, 6 ]);
    console.dir(fn([ xs, ys ]));
})()
```

# methods

``` js
var falafel = require('falafel')
```

## falafel(src, opts={}, fn)

Transform the string source `src` with the function `fn`, returning a
string-like transformed output object.

For every node in the ast, `fn(node)` fires. The recursive walk is a
pre-traversal, so children get called before their parents.

Performing a pre-traversal makes it easier to write nested transforms since
transforming parents often requires transforming all its children first.

The return value is string-like (it defines `.toString()` and `.inspect()`) so
that you can call `node.update()` asynchronously after the function has
returned and still capture the output.

Instead of passing a `src` you can also use `opts.source`.

All of the `opts` will be passed directly to
[acorn](https://npmjs.org/package/acorn).

## custom parser

You may pass in an instance of acorn to the opts as `opts.parser` to use that
version instead of the version of acorn packaged with this library.

```js
var acorn = require('acorn-jsx');

falafel(src, {parser: acorn, plugins: { jsx: true }}, function(node) {
  // this will parse jsx
});
```

# nodes

Aside from the regular [esprima](http://esprima.org) data, you can also call
some inserted methods on nodes.

Aside from updating the current node, you can also reach into sub-nodes to call
update functions on children from parent nodes.

## node.source()

Return the source for the given node, including any modifications made to
children nodes.

## node.update(s)

Transform the source for the present node to the string `s`.

Note that in `'ForStatement'` node types, there is an existing subnode called
`update`. For those nodes all the properties are copied over onto the
`node.update()` function.

## node.parent

Reference to the parent element or `null` at the root element.

# install

With [npm](http://npmjs.org) do:

```
npm install falafel
```

# license

MIT
# Change User Description Committed
#1 19553 swellard Move and rename clients
//guest/perforce_software/helix-web-services/main/source/clients/2016.1.0/javascript/node_modules/falafel/readme.markdown
#1 19053 tjuricek Rebuild JavaScript Client SDK.

The JavaScript client now is a "typed" approach that tends to be similar in approach to the other clients, based on the swagger definition for the platform version.

Importantly, client SDK tests are individual scripts (that run under node) that are actually controlled via TestNG. This approach now lets us use a consistent test reporting format so we can at least collect reports from each of the jobs.

The documentation is still in progress, that I want to validate as the tests are generated.