Regex: How can I find those html files with links that are not identical in different places?

Robin Cruise

I have this link at the beginning of html page:

<link rel="canonical" href="https://xxx.com/en/page-AAA.html" />

also I have another link on the middle of the file:

<a href="https://xxx.com/en/page-AAA.html"><img src="index_files/flag_lang_en.jpg" title="en" alt="en" width="28" height="19" /></a>

You see that there are the same links, but in different contexts and places. Compare it.

But how can I find those html files with links that are not identical in those different places?

Suppose the first link will be: <link rel="canonical" href="https://xxx.com/en/page-CCC.html" /> in this case are not identical those too, so regex should find that file that contains different links.

How can I do this with Regex?

guy038

Hi, @robin-Cruise and All,

Let’s suppose you have, at least, two links of the form https://xxx.com/en/••••••••••.••••, where the part ••••••••••.•••• is different.

Then, the regex (?s)(https://xxx.com/en/)([^"]+)".+?\1(?!\2").+?" will match the range between these two links, included !

Thus, the regex does not match anything if all the https://xxx.com/en/••••••••••.•••• , of current file, have the same ••••••••••.•••• part.

Best Regards,

guy038

Robin Cruise

This post is deleted!

Robin Cruise

@guy038 thanks a lot. You are the best !

Robin Cruise

by the way, @guy038 Can you explain what does this part of your regex do?

\1(?!\2")

guy038

Hello, @robin-cruise and All,

In the search regex (?s)(https://xxx.com/en/)([^"]+)".+?\1(?!\2").+?" :

The regex part https://xxx.com/en/ looks for the literal string https://xxx.com/en/, stored as group 1
The regex part ([^"]+)" represents the remainder of the internet address ( for instance the string page-AAA.html ), followed with a double-quote, because [^"]+ is a non-null range of consecutive chars, all different from ", stored as group 2
Now, the part .+? stands for the shortest range of any char till…
- The group 1 ( \1 ). So an other string https://xxx.com/en/
- Which must be followed by .+?", which represents the shortest non-null range of any char before a double-quote…
- But ONLY IF this range is different from \2 ( i.e. different, for instance, from the string page-AAA.html and a " char )

Note also that the [^"]+" syntax, without the parentheses, is more restrictive than .+?" and must be preferred because of the negative look-ahead (?!\2")

Besst Regards,

guy038