word_wrap in Email.php (library) does not handle multibyte characters well

[williamli]

word_wrap in Email.php (library) does not handle multibyte characters well. It made long line of plaintext Chinese characters look weird when it cuts multibyte characters in the middle.

[philsturgeon]

Could you try something out?

Go here http://php.net/manual/en/function.wordwrap.php and check out the contributed function iconv_wordwrap by "mail at dasprids dot de".

Put the same Chinese characters through that string, then let us know how it goes.

[williamli]

Sure. I will give it a try this weekend. Hope it is not too late.

[narfbg]

Some improvements have been made, but it's still not flawless ... I'm not sure it can ever be. If anybody has feedback - please comment.

[drew-hoffman]

So let me preface this by saying that my C knowledge is (almost) nonexistent.

PHP's wordwrap function uses strncmp, which operates on char's, which are 1 byte, as opposed to wstrcmp which can handle multi-byte characters... so the C code behind the wordwrap function sees your multi-byte character as 2 separate characters.

This can be fixed in 1 of 2 ways:

1. Re-writing the word_wrap function to not use php's wordwrap() and instead write our own version. This will be slower, but should work if we use mb_strlen() for calculations instead of strlen().

2. Add a mb_wordwrap() function to PHP, and then use that instead.

I don't have time to see which function works properly, but the wordwrap() function in the Email class could possibly be replaced by something here: http://stackoverflow.com/questions/3825226/multi-byte-safe-wordwrap-function-for-utf-8

[drivingmenuts]

Adding a multibyte wordwrap function to PHP is probably a bit beyond the scope of this project.

However, there are several examples of multibyte wordwrap functions at http://stackoverflow.com/questions/3825226/multi-byte-safe-wordwrap-function-for-utf-8

Any of those could be adapted/extended/coopted.