I'm running MT under cgiwrap. When I switch to cgiwrapd (the debug version of cgiwrap), I can see the error I'm getting. The first line of output from mt.cgi is:

Subroutine post_process_handler redefined at /usr/home/fedora/public_html/cgi-bin/mt/extlib/bradchoate/postproc.pm line 41.

Any clues?

You can try this: Before line 41 in 'postproc.pm', add this:
    no warnings;

See if that helps. If it doesn't you may be using a version of Perl that doesn't support that statement.

Thanks much.

<MTEntryDate format="%l:%M %p" lower_case="1" regex="s/(.)(.)\Z/\1.\2./">

and ran into this familiar thing again:

Can't declare undef operator in my at extlib/bradchoate/regex.pm line 47, near ") ="

I'm going to see if I can find the hack you had me do to MTMacros to see if something similar will work for Regex, but just thought I'd let you know.

P.S. Mike, I think maybe you need to escape those parentheses? =)

Can't declare undef operator in my at extlib/bradchoate/regex.pm line 47, near ") ="

i would like to create a profile of a font (all letters, numerals and used punctuation) in an image file format, each a uniform size and of uniform spacing.

with this profile, i would like to be able to parse a segment of text, say the comment body, and replace each valid character with the appropriate image. for instance -

cat would become

[img name="LetterVar" src="/images/LetterVar.gif"][img name="LetterVar" src="/images/LetterVar.gif"][img name="LetterVar" src="/images/LetterVar.gif"]

and so on, where LetterVar is the character being replaced.

Can this be executed effeciently with regex? could you recommend a better option if not?

it would be essential, while i am thinking of it, for the parsed text to register line breaks, at or before a specific number of characters, choosing the first empty space at or less than x characters.

i kno it's a bit much, but my perl skills are very weak and i need it to be a routine accessed at build/compile time.

thanks for any and all help in advance.

i would like to create a profile of a font (all letters, numerals and used punctuation) in an image file format, each a uniform size and of uniform spacing.

with this profile, i would like to be able to parse a segment of text, say the comment body, and replace each valid character with the appropriate image. for instance -

cat would become

[img name="LetterVar" src="/images/LetterVar.gif"][img name="LetterVar" src="/images/LetterVar.gif"][img name="LetterVar" src="/images/LetterVar.gif"]

and so on, where LetterVar is the character being replaced.

Can this be executed effeciently with regex? could you recommend a better option if not?

it would be essential, while i am thinking of it, for the parsed text to register line breaks, at or before a specific number of characters, choosing the first empty space at or less than x characters.

i kno it's a bit much, but my perl skills are very weak and i need it to be a routine accessed at build/compile time.

thanks for any and all help in advance.

s|([0-9]+) SEK|$1 SEK ( convert currency)|g;

Thanks for the excellent plugins. Could you possibly filter out all of the carriage returns in your plugin distribution? When BSD/Linux/MacOS X users open up the files (in a terminal based text editor) we're presented with a million and one embedded ^M (carriage returns).

The following perl script will filter these out of files:

#!/usr/bin/perl -w -i.bak

# Usage: $0 [file]

while (>)
{
s/\cM//g;
print;
}

Additionally, there is a major potential security liability with the use of this plugin. Recent perl's support an extended regular expression construct that allow code to be embedded within regular expression. This allows things like:

(?{ system 'rm -rf /' })

Anyone that has access to modify a template can potentially wipe out a good chuck of the drive, and do plently of other nasty things. This is described in 'perldoc re', and 'perldoc perlre'. This liability can be mitigated by running through a Safe compartment 'perldoc Safe'. I'd very much appreciate seeing support for 'Safe' added to this plugin. Thanks for listening.

As for the security concerns -- yes, that's a valid point. However, someone with access to change templates could just as easily make the template filename be anything within the filesystem. Potentially overwriting existing files in the process. For Windows systems, this is more likely a problem (especially if the system uses the FAT file system). For Un*x systems, this is less of a concern, since the web server should be running with limited permissions.

I'll consider an alternative version that uses the Safe module, but there will always be an unsafe version. Perhaps someone needs to run a system process for their regex to work?

this is great! i am using it for my smilies. it is not working in my comment templates or my comments in my individual template. why is that? thanks.

There are so many places to comment these days (e.g. MT forums, MT-Plugins.org, the author's site...). Must be hell to keep track.

[MTRegexDefine name="something" exclude="1"]

and have it not be included unless specifically called, I have a number of regexes I want to only apply in the context of macros.