Reduce HTML pages download size (was HTML Squeezer).
htmlPress reduces the download size of HTML pages without actually changing their looks. This is achieved by removing comments, extra whitespace characters, and anything else that your browser won't display on the screen. htmlPress is implemented as a MacPerl script.
What's new in htmlPress
Added compression support for style sheets (CSS).
CSS id names can now be replaced by short generic names. For that, all id names must be defined in the configuration file. No replacement is done if the generic id is already in use.
Better XML support (various), e.g. a new option --(no)lint to run xmllint (default on).
Added option --(no)link to check all hyperlinks (default off).
Introduced a &sign_off subroutine that the user can define in much the same way as &html_rules. Whereas &html_rules is called at the beginning of our processing, &sign_off is evoked at the very end. You can use this, for instance, to issue warnings on whatever you do not like about the result.
Retired $isdst variable as it was not used.
The file naming fall-back routine now checks whether a file rule has been used used before applying default settings.
Changed variable $isxml to $type (eq "xml").
Added support for € -> .
New <> macro to declare the document type. The inserted DOCTYPES adhere to W3C standards for HTML 4.01 XHTML 1.0 and 1.1 documents. Use e.g. <>.
While not intended as a complete syntax checker, a number of consistency checks have been added.
New option --(no)syntax to switch test off (default on).
Added a check for lowercase which is not allowed.
Added a check for undetected <>.
Simplified the code through stronger use of "unless".
Processing now stops at "__END__" (in a separate line).
Changed the format of the debug messages (to STDERR).
New subs "msg", "err" to streamline messages.