Why does grep return nothing?

Question

I'm somewhat at ease with regex, but not with grep particularly, and can't figure out why the following regex returns nothing:

wget -qO- 'http://www.acme.com/index.html' | grep -iPo '(?s)(^<div class="titlebar">.+?<div class="colleft">)'

I prepended (?s) because the catch-all ".+?" includes carriage-returns (either CRLF, CR, or LF, depending on how the text was saved).

Any idea why it doesn't work as expected?

Thank you.

Because neither of those class names appears anywhere in the page that `wget` returned? What is `(?s)` doing? — Etan Reisner, Jun 05 '15 at 19:14
`(?s)` enables the feature that `.` can match a newline. http://perldoc.perl.org/perlre.html -- not useful here since grep is a line-oriented tool. — glenn jackman, Jun 05 '15 at 19:19
Thanks for the info. I used ?s because this answer says that grep supports multiline with -P : http://stackoverflow.com/questions/152708/how-can-i-search-for-a-multiline-pattern-in-a-file — Gulbahar, Jun 05 '15 at 20:27
Show sample input (i.e. output from wget) and expected output. — Ed Morton, Jun 05 '15 at 23:46

score 0 · Answer 1 · answered Jun 05 '15 at 19:24

0

grep is line-oriented, so if there are newlines between the tags, grep can't find it. You'll want:

wget -qO- 'http://website.invalid/index.html' |
perl -0777 -nE 'say for /(^<div class="titlebar">.+?<div class="colleft">)/msg'

answered Jun 05 '15 at 19:24

glenn jackman

Thanks. To avoid installing Perl (appliance with little memory), I tried pcregrep, but it returns nothing : pcregrep -M "(.+)" input.html – Gulbahar Jun 05 '15 at 20:44
What tools do you have available? sed/awk/... ? – glenn jackman Jun 05 '15 at 20:44

1 Answers1