Regex which detect text at the end of url

Question

The title may be not appropriate but that I am trying to do is to cut url in certain condition. I want to create an regex which will throw away '/' at the end of the link.

For example

http://stackoverflow.com/questions/ask/
to
http://stackoverflow.com/questions/ask

I was able to archieve this with this ^(.*)\/(example) but now I needed a change, and I'm not sure how it supposed to look.

Basically, the change I need is to DON'T touch the '/' if it is main page:

http://stackoverflow.com/ stays http://stackoverflow.com/
http://stackoverflow.com/questions/ changes to ttp://stackoverflow.com/questions

From my point of view it supposes to look if url are ended with text between two '/' and without any '.', but so far I am failing to make this condition.

How it potentially supposed to look?

EDIT

Some of the url may not contain http or https, so for me it is totally ok if regex regex will work on a normal text to.

Please specify the language using another tag when using this tag, as documented in the [tag wiki](http://stackoverflow.com/tags/regex/info). — merlin2011, Dec 28 '16 at 07:46
My suggestion above uses the url as an example, too. The idea is working. — Wiktor Stribiżew, Dec 28 '16 at 07:51
If you mean there can be any protocol followed by any host, you may try `/^(?![^:]+:\/\/[^\/]+\/$)(.*)\/$/` — Wiktor Stribiżew, Dec 28 '16 at 07:58
I am not really sure what you need now, because how can you identify it is a URL you need? Following your current logic, the pattern you may use is `/^(?!(?:[^:]+:\/\/)?[^\/]+\/$)(.*)\/$/`. Or maybe exclude whitespaces: ``/^(?!(?:[^\s:]+:\/\/)?[^\/\s]+\/$)(.*)\/$/`` — Wiktor Stribiżew, Dec 28 '16 at 08:03
I would use something easier here: check number of occurences of '/' in url: if 3 - leave string as it is, if more, strip the last one. So, maybe you could do it without regex? — sinisake, Dec 28 '16 at 08:11
Wiktor, you didn't read my edit :(. It is very easy to do without regex, but sadly, because of some circumstances I needed one. — Olegs Jasjko, Dec 28 '16 at 08:22
Олег, I *read* your edit. `(?:[^\s:]+:\/\/)?` matches 1 or 0 occurrences of `xxx://` — Wiktor Stribiżew, Dec 28 '16 at 08:23
Then you could see that is't not a problem if regex don't care if there is a start of the link (url) or not. Basically for exampe, if `test.test/` stays as it is but `test.test/test/` becames `test.test/test` it is enough for me — Olegs Jasjko, Dec 28 '16 at 08:31
[Yes, I could see and saw it](https://regex101.com/r/eSxqj7/1). — Wiktor Stribiżew, Dec 28 '16 at 08:32

score 0 · Answer 1 · answered Dec 28 '16 at 07:45

0

If URL starts with http:// or https:// you can use following pattern.

(http[s]*://.*?/.*)/$

answered Dec 28 '16 at 07:45

Gokhan Celikkaya

537
1
5
17

Meninx - メネンックス · Answer 2 · 2016-12-28T09:31:45.100

0

This process can help:

^(http[s]?:\/\/[^\/]*)(.*)\/

Required URL = Group 1. + Group 2.

if Group 2 empty then
   1. It is main Url 
   2. Add "/" at the end
else 
   1. It is **not** main URL
   2. Keep it as it is

[Regex Demo]

edited Dec 28 '16 at 09:31

answered Dec 28 '16 at 07:56

Meninx - メネンックス

5,711
12
25

score 0 · Answer 3 · answered Dec 28 '16 at 08:17

You're looking to throw away the forward slash if it's just a trailing character and keep it if it denotes some kind of path in your website.

All that requires is checking whether the forward slash is the last character (I'm pretty sure that's the simplest way of going about it).

(?(?=\/$)(.+)\/)

This regex uses an if statement to check whether the forward slash is the last character, and in which case matches everything up until the forward slash.

score 0 · Answer 4 · answered Dec 28 '16 at 08:25

0

Use this :

(.*\.[a-z]+\/?(?:[\w]+(?:\/[\w]+)*)*)

The result will be stored in Group1, see example here https://regex101.com/r/egAQvX/6

answered Dec 28 '16 at 08:25

Boat Pathompon

42
3

score 0 · Answer 5 · answered Jan 29 '19 at 12:01

0

… Following your current logic, the pattern you may use is /^(?!(?:[^:]+:\/\/)?[^\/]+\/$)(.*)\/$/. Or maybe exclude whitespaces: /^(?!(?:[^\s:]+:\/\/)?[^\/\s]+\/$)(.*)\/$/ – Wiktor Stribiżew

answered Jan 29 '19 at 12:01

Armali

14,228
13
47
141

Regex which detect text at the end of url

5 Answers5