Regular Expression GUI?

[major tom]

Is there any graphical tools available for evaluating (perl) regular expressions? Something a'la MS Word's "Search and replace" type program with regular expression support would be REALLY nice.. Then it would be possible to QUICKLY see what the regexp actually does on some test data.. and it would save me ALLOT of debugging time.. Thanks for all help. Regards, Tom

[impossiblerobot]

ActiveStates's Komodo IDE for Perl (and other languages) does this. It was the only Komodo feature I really liked.

They may still have a free version available for download.

Impossible Robot

[mattriff]

Dominus wrote the code that ActiveState used for their regex debugger. It's available for download on his website:

http://perl.plover.com/Rx/

Of course, as he says: "Rx is not yet a complete, end-user application; it is a kit for building such applications," so it's probably not what the original poster is looking for. Just an interesting side-note.

- Matt Riffle

[lachoy]

The pretty nifty (and free as in beer/cost) editor JEdit has a RETest plugin that works pretty well. Despite what you may think from the name, it's an all-purpose editor, not just a Java editor.

Chris
M-x auto-bs-mode

[boo_radley]

JEdit is a nifty editor, but the engine that JEdit's regex tool (RE Tester) employs has several differences from perl's -- in some cases, this makes it impossible (or just annoying :) ) to compare RETester's results with perl's interpretation of a given regex.

Here's some differences, right from the gnu.regexp documentation, included with the plugin :

Unsupported Syntax Some flavors of regular expression utilities support additional escape sequences, and this is not meant to be an exhaustive list. In the future, gnu.regexp may support some or all of the following:

(?:...) pure grouping operator (Perl5)
(?=...) positive lookahead operator (Perl5)
(?!...) negative lookahead operator (Perl5)
(?#...) embedded comment (Perl5)
(?mods) inlined compilation/execution modifiers (Perl5)
\G end of previous match (Perl5)
\b word break positional anchor (Perl5)
\B non-word break positional anchor (Perl5)
\< start of word positional anchor (egrep)
\> end of word positional anchor (egrep)
[.symbol.] collating symbol in class expression (POSIX)
[=class=] equivalence class in class expression (POSIX)




Additionally, I've seen it act rather odd when using .* without an EOL anchor.

[lachoy]

Very true. However, I'd imagine the plugin will be modified in the future to use the JDK 1.4 built-in java.util.regex instead of the external gnu.regexp library. The built-in library includes (according to the 4th ed. of Java in a Nutshell):

 (?:...)
 (?=...)
 (?!...)
 (?<flags>:...)
 \G
 \b
 \B

Since it's open-source, I'd imagine this would get done sooner rather than later :-)

Chris
M-x auto-bs-mode

[Fletch]

Don't forget the versatile perl -de 0. Throw a test string in $_ and then just throw succesive iterations of x /foo/ at it until you've got your regex. See perldoc perldebug for more info.

[belg4mit]

RegExplorer looks promising, if they ever support perl regular expressions (been keeping my on them for awhile). There are a few links to similar programs @ the bottom of the page.

--
perl -pe "s/\b;([mnst])/'\1/mg"

[jwest]

The visual-regexp package does this, and quite nicely. I'm very, very happy with it so far, and I've been using it intensively for the last six months or so.

It can be found at http://laurent.riesterer.free.fr/regexp/

--jwest


-><- -><- -><- -><- -><-
All things are Perfect
    To every last Flaw
    And bound in accord
         With Eris's Law
 - HBT; The Book of Advice, 1:7

[Anonymous Monk]

Yeah, this was EXACTLY what I was looking for.. What a great tool! Thanks for all help!

[Mongo]

As numerous people have mentioned, the Komodo IDE includes a regex debugger based on mark-jason dominus's backend. I'd like to toot my own horn and point out a purl perl alternative:

ReBug: A regular expression debugger in perl...I gave a talk on it at TPC5. The paper, and the application can be found here:

http://perl.jall.org/rebug/

[young_david]

ActiveState's Komodo has a cool regex toolkit. The toolkit will help you build, edit and debug your regular expressions. Check it out http://www.activate.com

-young_david

[boo_radley]

tooting my own horn :-) --
you may be interested in Wx Regex Tester.

[Anonymous Monk]

The best GUI for this kind of thing is IMHO a web page; I have one at locus.cz - the script behind it is about 1 page of Perl...

[zentara]

What about the old standard?

#!/usr/bin/perl
use warnings;

print "Please enter your test pattern \n";
chomp($pattern = <STDIN>);
print "Please enter your test strings, 1 at a time,\n";
print "if a match occurs, your pattern will be echoed.\n";

 while(<>){
# print if m!your pattern here!; 
  print if m!$pattern!;
          }

[rbc]

Take a look at the text_demo script that comes with Tk800.023 ...

#!/usr/bin/perl -w

use Tk;
require Tk::Text;

my $top = MainWindow->new;
$top->option('add','*Text.background'=>'white');

my $t = $top->Scrolled('Text',"-relief" => "raised",
                     "-bd" => "2",
                     "-setgrid" => "true");

my $m = $t->Menu();
$m->add("command", "-label" => "Open", "-underline" => 0, 
        "-command" => \&sayopen);
$m->add("command", "-label" => "Close", "-underline" => 0, 
        "-command" => \&sayclose);
$m->add("separator");
$m->add("command", "-label" => "Selection", "-underline" => 0, 
        "-command" => \&showsel);
$m->add("separator");
$m->add("command", "-label" => "Exit", "-underline" => 1, 
        "-command" => \&doexit);

$t->pack(-expand => 1, "-fill"   => "both");

$t->tag("configure", "underline","-underline","on");

$t->insert("0.0", "This window is a text widget.  It displays one or m
+ore lines of text
and allows you to edit the text.  Here is a summary of the things you
can do to a text widget:

1. Insert text. Press mouse button 1 to set the insertion cursor, then
type text.  What you type will be added to the widget.  You can backsp
+ace
over what you've typed using either the backspace key, the delete key,
or Control+h.

2. Resize the window.  This widget has been configured with the \"setG
+rid\"
option on, so that if you resize the window it will always resize to a
+n
even number of characters high and wide.  Also, if you make the window
narrow you can see that long lines automatically wrap around onto
additional lines so that all the information is always visible.

3. Scanning. Press mouse button 2 in the text window and drag up or do
+wn.
This will drag the text at high speed to allow you to scan its content
+s.

4. Select. Press mouse button 1 and drag to select a range of characte
+rs.
Once you've released the button, you can adjust the selection by press
+ing
button 1 with the shift key down.  This will reset the end of the
selection nearest the mouse cursor and you can drag that end of the
selection by dragging the mouse before releasing the mouse button.
You can double-click to select whole words, or triple-click to select
whole lines.

5. Delete. To delete text, select the characters you'd like to delete
and type Control+d.

6. Copy the selection. To copy the selection either from this window
or from any other window or application, select what you want, click
button 1 to set the insertion cursor, then type Control+v to copy the
selection to the point of the insertion cursor.

You can also bind commands to tags. Like press button 3 for menu ");

&insertwtag($t,"here","underline");

$t->tag("bind","underline","<3>", [sub { shift; shift->Post(@_)},$m,Ev
+(X),Ev(Y)] );

$t->bind("<Any-Enter>", sub { $t->focus });

$t->Subwidget('text')->OnDestroy(sub { print "Destroyed!\n"; print $t-
+>get('1.0','end') });


Tk::MainLoop;

sub insertwtag {
  local($w,$text,$tag) = @_;
  my $start = $w->index("insert");
  print "start=$start\n";
  $w->insert("insert",$text);
  $w->tag("add",$tag,$start,"insert");
}


sub sayopen { print "Open something\n" }
sub sayclose { print "Close something\n" }
sub showsel  { my @info = $t->tagRanges('sel');
               if (@info)
                {
                 print "start=$info[0] end=$info[1]\n" 
                }
             }
sub doexit 
{ 
 die "'die' no longer exits"
}

[simon.proctor]

You could write a Java applet/application using JDK1.4 and build a regex checker that way. The new regex package claims to be Perl compatible.