small misunderstandings in your haproxy reqrep markdown

[sese]

You assume in the bellow expression:

reqrep ^([^\ :]*)\ /api/v1/api-name/(.*) \1\ /staging/path-name/\

the first expression in search is the host, also you don't give an explanation for spaces before first and second reg expression and also there is a space at replacing path.

You consider as sample the expression:

https://example.com/api/v1/api-name/

But this is wrong because the expression is something like:

GET /api/v1/api-name/

So, the first match corresponds to a method, not to a host

You can figure this out if you read the tutorial related to reqrep, where tells:

The statement reqrep applies a regex to each line of the request buffer in a case sensitive manner. As does reqirep, which is case insensitive

So, the request buffer looks like bellow:

GET /api/v1/api-name HTTP/1.1 User-Agent: Mozilla/4.0 (compatible; MSIE5.01; Windows NT) Host: www.tutorialspoint.com Accept-Language: en-us Accept-Encoding: gzip, deflate Connection: Keep-Alive

Now I believe make more sense why there is a space after the first reg expression

^([^\ :]*)\ /api/v1/api-name/(.*)

will match the expression

GET /api/v1/api-name HTTP/1.1

In conclusion, there are no 3 parts in regex, but only 2, as the haproxy tutorial explain:

reqrep <search> <replace> [<cond>]

Thank you for your effort, you point me to right direction, also that space bother me and I investigate more in order to understand better. So I'd like to share my effort in order to let you fix your explanation.