README

Known Issues

Perl formatting is really crappy. Partly because the author is lazy and partly because Perl is hard to parse.
On some browsers, <code> elements with newlines in the text which use CSS to specify white-space:pre will have the newlines improperly stripped if the element is not attached to the document at the time the stripping is done. Also, on IE 6, all newlines will be stripped from <code> elements because of the way IE6 produces innerHTML. Workaround: use <pre> for code with newlines.

Change Log

Added tests for PHP support to address issue 3.
Fixed bug: prettyPrintOne was not halting. This was not reachable through the normal entry point.
Fixed bug: recursing into a script block or PHP tag that was not properly closed would not silently drop the content. (test)
Fixed bug: was eating tabs (test)
Fixed entity handling so that the caveat

Caveats: please properly escape less-thans. x<y instead of x<y, and use " instead of " for string delimiters.

is no longer applicable.
Added noisefree's C# patch
Added a distribution that has comments and whitespace removed to reduce download size from 45.5kB to 12.8kB.

Added language specific formatters that are triggered by the presence of a lang-<language-file-extension>
Fixed bug: python handling of '''string'''
Fixed bug: / in regex [charsets] should not end regex

Language handlers for F#, OCAML, SQL
Support for nocode spans to allow embedding of line numbers and code annotations which should not be styled or otherwise affect the tokenization of prettified code. See the issue 22 testcase.

Language handlers for Visual Basic, Haskell, CSS, and WikiText
Added .mxml extension to the markup style handler for Flex MXML files. See issue 37.
Added .m extension to the C style handler so that Objective C source files properly highlight. See issue 58.
Changed HTML lexer to use the same embedded source mechanism as the wiki language handler, and changed to use the registered CSS handler for STYLE element content.

Rewrote to improve performance on large files. See benchmarks.
Fixed bugs with highlighting of Haskell line comments, Lisp number literals, Lua strings, C preprocessor directives, newlines in Wiki code on Windows, and newlines in IE6.

Added support for line numbers. Bug 22
Added YAML support. Bug 123
Added VHDL support courtesy Le Poussin.
IE performance improvements. Bug 102 courtesy jacobly.
A variety of markup formatting fixes courtesy smain and thezbyg.
Fixed copy and paste in IE[678].
Changed output to use   instead of   so that the output works when embedded in XML. Bug 108.

Added a themes gallery to showcase contributed styles.
Added support for XQuery courtesy Patrick Wied, Nemerle courtesy Zimin A.V., and Latex support courtesy Martin S.

Fixed IE newline issues, and copying/pasting of prettified source code from IE. This required significant internal changes but involves no API changes. Caveat: prettyPrintOne injects the HTML passed to it into a <pre> element. If the HTML comes from a trusted source, this may allow XSS. Do not do this. This should not be a problem for existing apps since the standard usage is to rewrite the HTML and then inject it, so anyone doing that with untrusted HTML already has an XSS vulnerability. If you sanitize and prettify HTML from an untrusted source, sanitize first.

Added a one script autoload&run mechanism and a way to embed hints in processing instructions/comments. See example.