Hi Ilya,
On 11.10.2004 21:12, Ilya Konstantinov wrote:
We already have the IME language used to input characters stored along with the characters. Do you think we should add control characters into our serialization as well, even when the direction is apparent from the IME language? (In contrast to cases when the user manually overrides the text's direction, when the automatic IME-based direction doesn't suit him.)
well, actually the language attribute does not suffice to solve the problem. For each character position, we have three language attributes simultaneous. One language is used if the character is Western text, one if the character is Asian text, and one if the character is CTL text (this is decided by checking the Unicode code point). So if we have a '(' in your paragraph, there is no way to tell from the character attributes, if this should be a RTL or a LTR bracket. If we intend to implement the IME based solution, we would need an additional internal Attribute "language of IME which has been used to enter the character".
Regards,
Frank
============================================================== To unsubscribe, send mail to [EMAIL PROTECTED] with "unsubscribe hebrew" in the message body.
