HTML Resource Embedder

HTML Resource Embedder (htmlresemb) tries to embed all external resources like CSS, JavaScript and image files into an HTML document. It reads the content of those files and modifies the corresponding tags (link, style and img) accordingly. CSS and JavaScript files are embedded into a CDATA section, so in most cases there shouldn't be a problem with characters like > in the resource files. Image files will be converted to a Base64 encoded data URI.

The resulting document is in most cases "portable", that is, it can be copied/moved to other locations without breaking the layout. Please note that currently embedding of font files isn't supported. This may be added in the future but can be a license problem depending on the font.

Usage

htmlresemb -i <infile> -o <outfile> [options]

• <infile>
  Use for input. If -i <infile> is ommitted, input is read from stdin. In this case a document root path must be supplied via -docroot, otherwise loading of resources referenced in the document is impossible.

• <outfile>
  Use <outfile> for output. If -o <outfile> is ommitted, output goes to stdout.

Available options (others are silently ignored):

• -h, --help
  Print help message and exit.

• -docroot <path>
  A path which is used to resolve relative filenames. Relative paths and environment variables can be used in path.

• -css
  Embed CSS files.

• -js
  Embed JavaScript files.

• -img
  Embed image files.

If none of the switches -css, -js and -img is used, all resources will be embedded.

Return values:

• 0  Success.
• 1  Invalid/incomplete options or -h / --help was used.
• 2  An error occured.

Warning or error messages during processing are written to stderr.

Status

htmlresemb is by no means a perfect solution, it's a very quick spin off from one of our other internal projects. In general it does it's best and should just work, but it hasn't been tested with many different documents. For example encodings other than UTF-8 will fail (mostly). Reading of the source document is done with an XML parser, so this also may fail completely (although htmlresemb tries to fix broken HTML to some extent). Also the output document might contain slightly different whitespace.

Development

htmlresemb is written in Free Pascal, the Mac and Windows binaries (see Releases) were created with compiler version 2.6.4 and Lazarus 1.4.4. You also need the LazUtils package (part of FPC) and Internet Tools. The Internet Tools version used for htmlresemb is available as a Zip file in the Releases section.

The source code contains a slightly modified version of xmlwriter.pas to overcome indenting and whitespace problems when creating the result document. If you want to use a newer compiler version this maybe no longer necessary, but it's very likely that you have to change larger parts of the source code due to the new Unicode string handling in FPC 3.x. The Lazarus project file depends on strip as a post compile command, on the Mac this should be available if the Apple developer tools are installed, on Windows using Cygwin/MinGW is one possible solution.

Credits

• Free Pascal
• Lazarus
• Benito van der Zander (Internet Tools)

License

MIT © idesis GmbH, Rellinghauser Straße 334F, D-45136 Essen

See the LICENSE file for details.