Understanding someone else's parser

John M. Dlugosz has asked for the wisdom of the Perl Monks concerning the following question:

Replies are listed 'Best First'.
Re: Understanding someone else's parser by rob_au (Abbot) on Jan 02, 2003 at 03:10 UTC
See YAPE::Regex::Explain from japhy - The release announcement for this module can be found on this site here. Also have a look at the node Learning Perl's Regular Expressions which includes links to Regex::Diagram and GraphViz::Regex which may also be of use. `perl -le 'print+unpack("N",pack("B32","00000000000000000000001000001101"))'`	[reply]
Re^2: Understanding someone else's parser by diotalevi (Canon) on Jan 02, 2003 at 04:59 UTC
I also recall japhy noting that the base module doesn't yet support all the really funky things[1] you can do with regexen. It's really a string parser and in general parsing regexes also reduces to requiring a Perl parser in Perl that isn't perl. [1] - my definition of funky starts at (?{}) / (??{}) so maybe it's good for everything normal programmers use. ;-) Fun Fun Fun in the Fluffy Chair	[reply]
Re: Re^2: Understanding someone else's parser by John M. Dlugosz (Monsignor) on Jan 02, 2003 at 08:11 UTC
Here's the kind of stuff I'm looking at: `while ( $line =~ s/([[:alpha:]][[:alpha:]\']{0,44})[_\-,\.\"\' +\)\?!:;\/&]{0,5}([ \t\n\r]\|$)/ / ) { update_word($self,$1, $encoded, '', '[_\-,\.\"\'\)\?!:;\/ +&\t\n\r]') if (length $1 >= 3);` [download] The first line is what POPfile considers a "word", after dealing with some other special cases first. I want to tinker with this program a bit. So, there is nothing funky about it. Just someone else's code to read and understand. —John	[reply] [d/l]
Re^b100: Understanding someone else's parser by diotalevi (Canon) on Jan 03, 2003 at 00:07 UTC
Re: Understanding someone else's parser by traveler (Parson) on Jan 02, 2003 at 16:16 UTC
I think you are after japhy's YAPE::Regex::Explain as noted above. In addition there is Rx. You can also try Komodo from Active State, which, IIRC, is based on Rx. I also recommend checking out GraphViz::Regex. HTH, --traveler	[reply]