Keeping child nodes that in turn gets parsed recursively.

Update README with advanced usage of `replace` method

- Try to follow the 80 column rule for readability.
- Reword sentences to make them more understandable.
- Gives thanks to `htmlparser2` and contributors.
- Add lint script to testing documentation.

Tidy and reword README for readability and add contributors

Follow example: https://opensource.org/licenses/MIT

Follow standard config:
https://webpack.github.io/docs/configuration.html

Export to `./dist/html-react-parser.min.js`

Specify `react` in "externals":
https://github.com/webpack/webpack/tree/master/examples/externals

npm scripts:
- build (unminified UMD bundle)
- build-min (minified UMD bundle)

`webpack.config.js`:
- Remove "entry" and "output.filename" now that paths are passed
  via CLI

Remove branches where `process.env.NODE_ENV !== "production"`
especially for React.

Build the unminified and minified client-side bundles before
publish.

Use webpack to build UMD bundle

For the client parser, when the first element of the HTML string
has attributes, the regex will improperly capture everything and
make the tag name invalid.

The fix is to ensure only the tag name is matched and whitespace
is trimmed.

Fixes #23

Confirm that for the client parser, if the first HTML element has
attributes, it should properly match only the tag name and no
longer throw an error.

Issue #23

Fix regex bug on client parser

- Update README (#17 and #21):
  - Improve documentation of `replace` option with the help of
    @poacher2k
  - Tidy and reword some parts
- Use webpack to build UMD bundle (#22)
  - Build to `./dist/` directory before publish
- Fix regex bug on client parser (#24)
  - Add test for `window.DOMParser`