> >> > #!/usr/bin/perl
> >> > ## set up for output to be sent to browser
> >> > print "Content-type: text/html\n\n";
> >> > ## read in mp3store.html
> >> > if (open (INFILE, "<../../mp3store.htm"))
> >> > {
> >> > foreach $line (<INFILE>)
> >> > {
> >> > $line =~ tr/images\//..\/..\/images\//
> >>
> >> $line =~ s/images\//\.\.\/\.\.\/images/g;
> >
> > Leaning toothpicks! You don't need to escape the '.' on the
> > replacement
> > string side because they are literal '.' characters there. Changing
> > the
> > delimiter to something other than '/' will eliminate the need to
> > escape
> > the '/' characters as well.
>
> Yes, as I saw in your post. As you say though, if one does not need to
> escape the '.'s on the replacement side, because they are literal,
> should the '\' not be literal as well?
This should demonstrate what happens:
C:\>echo AAAAAAA | perl -pe"s/A/./g"
.......
C:\>echo AAAAAAA | perl -pe"s/A/\./g"
.......
C:\>echo AAAAAAA | perl -pe"s/A/\\./g"
\.\.\.\.\.\.\.
As far as "why", it's probably because \. isn't a valid escape sequence
so the parser (?) must just turn it into a '.'. Someone with a better
understanding of what's going on on the inside can better explain this.
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
<http://learn.perl.org/> <http://learn.perl.org/first-response>
