OT: PERL

What you obviously need is "Google For Dummies."

John

That book didn't sell because nobody could find it :-)

I found the Perl "dummies" book right away... I just wanted confirmation from PERL'ies ;-) ...Jim Thompson

Why buy anything

The Perl tutorials are pretty good. I use Perl occassionally, and have never deeded anything beyond perldoc.perl.org.

Just discovered the site myself. Found some nice tutorials on just the part I need, find and replace strings (where the strings are variable length and content of mixed text and numerics). Provoked this morning by a need to process some PSpice netlists into a format readable by Cadence Virtuoso. ...Jim Thompson

Of course real men use AWK for that. ;)

Cheers

Phil Hobbs

I was going to say, linux has a lot of simpler solutions than going for a full blown Perl. Even SED can do a lot. Lex, yacc, lots of choices. IIRC Lex was Erik Schmidt's (google) contribution to the world of software.

My issue with Perl is it has a lot of cryptic shortcuts. Fine for experts, but it makes studying existing code difficult.

I'm catching on. I can do a find with (text rather than Perl to illustrate)...

space'multiple-characters-text-and-numbers'space

(Note the single quote marks)

I want to replace with...

space{multiple-characters-text-and-numbers}space

(Note curly brackets)

When I try it in UltraEdit I get the curly brackets alright, but between them is replaced with the Perl "equation", rather than the original "multiple-characters-text-and-numbers".

Is there a way to "capture" that, or is there another way to approach the problem? ...Jim Thompson

Thanks, Les, good lead! I'm using UltraEdit which has built-in regular expression support, Perl, Unix and their own flavor. I'll poke around for "regsub". ...Jim Thompson

In Tcl, the thing to use is regsub. Is there a regsub equivalent in Perl? Googling a bit, it looks like regsub *came from* Perl.

In Tcl ( I haven't used Perl in decades) , it would be:

C:\tcl>type h.tcl set s { blorp 'This has single quotes and text 43' arpf }

puts $s regsub -all " '" $s " {" s regsub -all "' " $s "} " s

puts $s

C:\tcl>tclsh85 h.tcl blorp 'This has single quotes and text 43' arpf blorp {This has single quotes and text 43} arpf

C:\tcl>

-- Les Cargill

"regsub" is Tcl-specific :-( ...Jim Thompson

Thanks, I'll give that a whirl. ...Jim Thompson

Drat! Maybe this will work ( I cannot test it; sorry )

$string =~ s/regex/replacement/g;

So try $string =~ s/ '/ {/g;

followed by

$string =~ s/' /{ /g;

My Perl is too weak to know whether this is good or not.

-- Les Cargill

Here's a stand alone perl program (between the --- cut lines )to do what you want:

--- cut #! /usr/bin/perl

while() { if( m/^ '(.*)' $/o ) { print ' {'.$1.'} '."\n"; } else { print; } }

--- cut

Copy this to some_file.pl. Change the #! /usr/bin/perl to match your system's perl executable, set the execute permission and run:

some_file.pl input.txt > output.txt

On Windows, you might have to run:

perl some_file.pl input.txt > output.txt

This looks explicitly for lines with one space, a quote ('), any text, another (') and a final space. All other lines are dumped as-is.

Try this: Find: \s'(\w+)'\s

The \s matches with any whitespace (space, tab), parentheses define a matching region which can be referred in the replace. You can define multiple match regions to swap things, for example.

Replace: \{\1\} The replacement string should have space character in the beginning and end.

Thanks, Mikko! ...Jim Thompson