Regex to get a filename from a url - regex

Question

I am trying to write a regex to get the filename from a url if it exists.
This is what I have so far:
(?:[^/][\d\w\.]+)+$
So from the url http://www.foo.com/bar/baz/filename.jpg, I should match filename.jpg
Unfortunately, I match anything after the last /.
How can I tighten it up so it only grabs it if it looks like a filename?

Answer 1

The examples above fails to get file name "file-1.name.zip" from this URL:
"http://sub.domain.com/sub/sub/handler?file=data/file-1.name.zip&v=1"
So I created my REGEX version:
[^/\\&\?]+\.\w{3,4}(?=([\?&].*$|$))
Explanation:
[^/\\&\?]+ # file name - group of chars without URL delimiters
\.\w{3,4} # file extension - 3 or 4 word chars
(?=([\?&].*$|$)) # positive lookahead to ensure that file name is at the end of string or there is some QueryString parameters, that needs to be ignored

Answer 2

This one works well for me.
(\w+)(\.\w+)+(?!.*(\w+)(\.\w+)+)

Answer 3

(?:.+\/)(.+)
Select all up to the last forward slash (/), capture everything after this forward slash. Use subpattern $1.

Answer 4

Non Pcre
(?:[^/][\d\w\.]+)$(?<=\.\w{3,4})
Pcre
(?:[^/][\d\w\.]+)$(?<=(?:.jpg)|(?:.pdf)|(?:.gif)|(?:.jpeg)|(more_extension))
Demo
Since you test using regexpal.com that is based on javascript(doesnt support lookbehind), try this instead
(?=\w+\.\w{3,4}$).+

Answer 5

I'm using this:
(?<=\/)[^\/\?#]+(?=[^\/]*$)
Explanation:
(?<=): positive look behind, asserting that a string has this expression, but not matching it.
(?<=/): positive look behind for the literal forward slash "/", meaning I'm looking for an expression which is preceded, but does not match a forward slash.
[^/\?#]+: one or more characters which are not either "/", "?" or "#", stripping search params and hash.
(?=[^/]*$): positive look ahead for anything not matching a slash, then matching the line ending. This is to ensure that the last forward slash segment is selected.
Example usage:
const urlFileNameRegEx = /(?<=\/)[^\/\?#]+(?=[^\/]*$)/;
const testCases = [
"https://developer.mozilla.org/en-US/docs/Web/API/MutationObserverInit#yo",
"https://developer.mozilla.org/static/fonts/locales/ZillaSlab-Regular.subset.bbc33fb47cf6.woff2",
"https://developer.mozilla.org/static/build/styles/locale-en-US.520ecdcaef8c.css?is-nice=true"
];
testCases.forEach(testStr => console.log(`The file of ${testStr} is ${urlFileNameRegEx.exec(testStr)[0]}`))

Answer 6

It might work as well:
(\w+\.)+\w+$

Answer 7

You know what your delimiters look like, so you don't need a regex. Just split the string. Since you didn't mention a language, here's an implementation in Perl:
use strict;
use warnings;
my $url = "http://www.foo.com/bar/baz/filename.jpg";
my #url_parts = split/\//,$url;
my $filename = $url_parts[-1];
if(index($filename,".") > 0 )
{
print "It appears as though we have a filename of $filename.\n";
}
else
{
print "It seems as though the end of the URL ($filename) is not a filename.\n";
}
Of course, if you need to worry about specific filename extensions (png,jpg,html,etc), then adjust appropriately.

Answer 8

> echo "http://www.foo.com/bar/baz/filename.jpg" | sed 's/.*\/\([^\/]*\..*\)$/\1/g'
filename.jpg

Answer 9

Assuming that you will be using javascript:
var fn=window.location.href.match(/([^/])+/g);
fn = fn[fn.length-1]; // get the last element of the array
alert(fn.substring(0,fn.indexOf('.')));//alerts the filename

Answer 10

Here is the code you may use:
\/([\w.][\w.-]*)(?<!\/\.)(?<!\/\.\.)(?:\?.*)?$
names "." and ".." are not considered as normal.
you can play with this regexp here https://regex101.com/r/QaAK06/1/:

Answer 11

In case you are using the JavaScript URL object, you can use the pathname combined with the following RegExp:
.*\/(.[^(\/)]+)
Benefit:
It matches anything at the end of the path, but excludes a possible trailing slash (as long as there aren't two trailing slashes)!

Answer 12

Try this one instead:
(?:[^/]*+)$(?<=\..*)

Answer 13

This is worked for me, no matter if you have '.' or without '.' it take the sufix of url
\/(\w+)[\.|\w]+$