Encoding name normalization does not remove year suffix

[rossj]

This line is supposed to normalize encoding names by removing non-alphanumeric characters and stripping an appended year. The year is not being stripped with the current regex, causing encoding names with years to not match.

Example:
Input: iso_8859-5:1988
Output: iso885951988
Expected output: iso88595

I think the reason the current regex does not work is that the colon character is matched in the first part as a non-alphanumeric character, therefore causing the following year part to not match.

[erikkemperman]

Just noticed the same thing. Suggested fix:

.toLowerCase().replace(/:\d{4}$/, "").replace(/[^0-9a-z]/g, "");

[ashtuchkin]

Thanks guys, I think you're right, I'll add a fix soon.

…

-- Alexander Shtuchkin

On Wed, Jul 5, 2017 at 11:26 AM, Erik Kemperman ***@***.***> wrote: Just noticed the same thing. Suggested fix: .toLowerCase().replace(/:\d{4}$/, "").replace(/[^0-9a-z]/g, ""); — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#156 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAmVHYk30eR2uB93vvK3Kr06jEYV5yLcks5sK9VAgaJpZM4OMhuw> .

[erikkemperman]

Actually, maybe it should be

.toLowerCase().replace(/:\d{4}[^0-9a-z]*$/, "").replace(/[^0-9a-z]/g, "");

[rossj]

FWIW I just swapped the ORs in my local copy

/:\d{4}$|[^0-9a-z]/g

[erikkemperman]

I was thinking you might want to get rid of trailing (after the year) non-alphanumerics as well?

[erikkemperman]

Thanks!