wrong count on textearea counter about products.ploneformgen HOT 5 CLOSED

I'm guessing you're on an operating system where a new line actually is represented as two characters (linefeed + carriage return)

from products.ploneformgen.

I'm on ubuntu but i've the same problem on windows. Probably i've bad exposed my problem. If i have a textarea with a limit of 100 characters the js counter count every newline as one, but the form validator count it by two.

from products.ploneformgen.

I have also received complaints about this from users.
I've tested this with Plone running on OSX and CentOS. I've also tested this from Chrome on OSX and IE9 on Windows. I am always able to reproduce the issue. The javascript character count in the browser always counts newlines as one character but REQUEST.form['my-field'] always contains \r\n.

My solution was to change len(value) to len(value.replace('\r\n', '\n')) in the validation script. I will submit a pull request with this change.

from products.ploneformgen.

It's more seriously broken: it also counts non-ascii characters as multiple in the validator -- I guess it's the usual 7-bit-string issue: something like ä will have two bytes in UTF8, and the js correctly counts it as one character, whereas the validator sees it as two bytes.

from products.ploneformgen.

Yes I recently ran across the issue with non-ascii characters as well and have not found time to fix it. I'm not sure if I can reliably convert the string from the form to unicode. The below is an example of the problem and how converting to unicode strings fixes the problem. The pull request #70 is still needed to fix the newlines issue though.

emdash = '\xe2\x80\x94'
degree = '\xc2\xb0'
emdash
'\xe2\x80\x94'
print emdash
—
len(emdash)
3
degree
'\xc2\xb0'
print degree
°
len(degree)
2

If you convert the string to unicode it is 1 character:

len(unicode(emdash, "UTF-8"))
1
len(unicode(degree, "UTF-8"))
1

from products.ploneformgen.