regex pattern for url with no ending slash and exclude certain text in url

Question

I'm looking for preg_match_all pattern to find all URL on a page that don't have trailing slash.

For example: if I have

a href="/testing/abc/">end with slash

a href="/testing/test/mnl">no ending slash

The result would be #2. Solution is posted at find pattern for url with no ending slash

I have tried to modify the provided pattern to exclude urls that have 'images' or '.pdf' but no luck yet.

Thanks.

Before using any regex, use a HTML parser (DOMDocument) to parse the page, then pick out the tags that you want to process and check the href/src. — nhahtdh, Mar 19 '13 at 17:06

sp00m · Accepted Answer · 2013-03-19T18:53:47.067

2

This one should suit your needs (demo):

href="(?:(?<!images).(?!(?:[.]pdf|/)"))*?"

(?:) = non-capturing groupe
(?<!images). = any char not preceded by images
.(?!(?:[.]pdf|/)") = any char not followed by .pdf" nor by /"
*? = match as short as possible

edited Mar 19 '13 at 18:53

answered Mar 19 '13 at 17:03

sp00m

47,968
31
142
252

Why the character class, `[.]`? Oh, I see. I've never seen anyone escape the `.` by a character class. Neat. – Andrew Cheong Mar 19 '13 at 17:05
@acheong87 It's a way to escape the `.` special char. Just the same as `\.`, but I find it more *readable* ;) – sp00m Mar 19 '13 at 17:07
Yeah, realized a second after I commented. Very nice regex; ingenious method. – Andrew Cheong Mar 19 '13 at 17:09
@acheong87 I realized that you realized a second after I commented too ;) – sp00m Mar 19 '13 at 17:10
Hi sp00m! this pattern will include url with trailing slash. How to work that one in? Thanks. – user2170712 Mar 19 '13 at 18:32
@sp00m, i'm testing by adding more conditions. Any suggestions: /href="(?: (?<!images2) (?<!images).(?!(?:[.]xml|[.]js|[.]pdf|/)"))*?"/ – user2170712 Mar 19 '13 at 19:49

Zack · Answer 2 · 2013-03-19T19:40:56.163

1

I found a way to exclude a link that has .pdf, by modifying the provided answer from the other question. Still looking at why it won't not match the images example though.

href=(['"])[^\s]+(?<![\/]|.pdf)\1

Link to a working test http://www.rubular.com/r/jmBVstpGZD

edited Mar 19 '13 at 19:40

answered Mar 19 '13 at 17:01

Zack

2,789
33
60

This regex would reject .bmp and .tif as well. Can you figure out why? :) – Andrew Cheong Mar 19 '13 at 17:04
hi zack, how about also exclude url that has 'images' (string) in it? – user2170712 Mar 19 '13 at 17:04
@acheong87 No I really have no idea oh cryptic master. :p I'm assuming you do know, and can provide some reasoning or possibly a link to why. – Zack Mar 19 '13 at 17:10
1

Haha, it's because `[.pdf]` is a character class that means "any _one_ of the characters, `.`, `p`, `d`, and `f`." You excluded anything that ended in one of those characters. What you meant is `\.pdf` without the square brackets. Sorry, didn't mean to leave you hangin'. – Andrew Cheong Mar 19 '13 at 17:59
@acheong87 I see what you mean now. I updated my answer just so if anyone saw it they wouldn't try to use it. And because it was incorrect. Thanks for your explanation! I don't use regex much and posted before I tested it I guess. – Zack Mar 19 '13 at 19:44
Nice. No worries; all here to learn. 'cept Jon Skeet, I mean. – Andrew Cheong Mar 20 '13 at 01:54

regex pattern for url with no ending slash and exclude certain text in url

2 Answers2

Linked