CODEWITHSUNDEEP
regexperl
Amit Paranjpe
Amit Paranjpe

Reputation: 25

Regular expression to get the first letter of a word after punctuation symbols in Perl

Can any body tell me a regular expression in Perl for getting the ucfirst letter of the word comming after a dot,question or exclamation sign...

My program reads string character by character.

Requirement :

input string : "abcd[.?!]\s*abcd"
output: "Abcd[.?!]\s*Abcd"

My program is as follows:

#!/usr/bin/perl

use strict;

my $str = <STDIN>;
my $len=length($str);
my $ch;

my $i;
for($i=0;$i<=length($str);$i++)
{
$ch = substr($str,$i,1);
print "$ch";
if($ch =~ 's/([.?!]\s*[a-z])/uc($1)/ge')
{
    $i=$i+1;
    $ch = substr($str, $i,1);
    my $ch = uc($ch);
    print "$ch";
}
#elsif($ch eq "?")
#{
#   $i=$i+1;
#   $ch = substr($str, $i,1);
#   my $ch = uc($ch);
#   print "$ch";
#}
#elsif($ch eq "!")
#{
#   $i=$i+1;
#   $ch = substr($str, $i,1);
#   my $ch = uc($ch);
#   print"$ch";
#}
#elsif($ch eq " ")
#{
#   $i=$i+1;
#   $ch = substr($str, $i,1);
#   my $ch = uc($ch);
#   print"$ch";
#}
#else
#{
#print "";
#}
}
print "\n";

Upvotes: 1

Views: 1482

Answers (4)

Toto
Toto

Reputation: 91430

If you have unicode string, you could use:

$str =~ s/(\pP|^)(\s*\pL)/$1\U$2/g;

Upvotes: 0

kevin cline
kevin cline

Reputation: 2736

can any body tell me the regex in perl for getting the ucfirst letter of the word comming after a dot,question or exclamation sign...

My program reads string character by character.

Requirement :

input string : "abcd[.?!]\s*abcd"

output: "Abcd[.?!]\s*Abcd"

Your output does not match your explanation. In the input, the initial "a" does not follow a period, question mark, or exclamation mark, but was changed to upper case.

You can and should do this sort of processing with a single substitution. To do exactly as you said:

s/[.?!]\K[[:lower:]]/uc($&)/ge

The \K discards the character matched by [.?!], leaving only the lower-case letter in the matched string. $& is the matched string. The e flag says to evaluate uc($&).

If you also want to make an initial letter uppercase:

s/(?:^|[.?!])\K[[:lower:]]/uc($&)/ge

Upvotes: 0

ikegami
ikegami

Reputation: 385897

Normally,

$s =~ s/(?<=[.?!]|^)\s*[a-z]/\U$1/g;

$s =~ s/(?<![^.?!])\s*[a-z]/\U$1/g;

$s =~ s/(?:^|[.?!])\s*\K[a-z]/\U$1/g;

But if you only read one character at a time,

my $after_punc = 1;
while (my $ch = ...) {
    if ($ch =~ /^[.?!]\z/) {
       $after_punc = 1;
    }
    elsif ($ch =~ /^[a-z]\z/) {
       $ch = uc($ch) if $after_punc;
       $after_punc = 0;
    }
    elsif ($ch =~ /^\s\z/) {
       # Ignore whitespace.
    }
    else {
       $after_punc = 0;
    }

    ...
}

Upvotes: 0

tripleee
tripleee

Reputation: 189477

Looping over the string, and then looping over the match, is completely redundant. Your entire program can be replaced with this:

perl -pe 's/(^|[.?!]\s*)([a-z])/$1\U\2/g' inputfile >outputfile

I added beginning of line to the first parenthesized expression, although your explanation doesn't include that (but your example does).

Upvotes: 1

Related Questions