vBulletin's unhtmlspecialchars() question

[Tommy Boy]

vBulletin's unhtmlspecialchars() function starts with the following line. What does it do, and what is its purpose?

$chars=preg_replace("/(&#)([0-9]*)(;)/esiU","chr(intval('\\2'))",$chars); 

[fury]

I may be wrong but I think htmlspecialchars() replaces unicode characters with their unicode number (&#xxxx; ) equivalent so that they can be stored in the database without a problem. Then when it selects the post it does an unhtmlspecialchars to display them correctly.

[Kier]

Correct. There are a large number of HTML character entities such as :, Where the integer represents an ASCII number. The code takes the HTML char entity and converts it into the appropriate character using the chr() command on the number within the entity.

[Tommy Boy]

Thank you very much!

[Tommy Boy]

I noticed that this function fails on large integers. I suspect it is a chr() limitation to accept only 8-bit integers. For instance, ™ which should represent a small TM (trademark) sign, will be converted to plain quotes ". Surprisingly (or not), their binary representations both start the same way:

10000100100010 - ™
00000000100010 - "

Obviously, IE recognizes ™ correctly, so I wonder how can I effectively convert all entities over 8-bit to their appropriate plain character to store into the database.

Reference: http://www.zytrax.com/tech/web/entities.html (or any google search for "html entities")