Hi Yosef, I think that the planning of the solution to iussues 18024 and 27174 needs to go through Falko Tesch and needs to include input from the Arabic users as well.
One way to solve the problem, as you suggest, is to have some user initiated signal such as a button on the toolbar that the user can press to start an embedded LTR or RTL run. Pressing the button a second time inserts a PDO to stop the run. The problem with this solution is that it requires a lot of user knowledge about bidi. It also can result in the user starting a run but only pressing the button after the begining of the run, which could result in some confusion. I would prefer a more automatic solution. An alternative solution might be to automatically insert an RLO or LRO whenever the user types a character whose direction is opposite that of the paragraph base direction, and then automatically insert a PDO when the user types a character whose direction is the same ase the paragraph base direction. This would solve the problem with neutral puctuation marks at the end of embedded runs that are in the opposite direction from the direction of the paragraph. In either solution, there are cursor control problems that need some thought. LRO, RLO and PDO are all zero-width characters. Dispite the fact that they are not displayed, we would need to have the cursor stop on them so that the user can erase them when need be. It might be best to have a double flag on the cursor after an RLO or LRO so the user is aware that a there is a directional override character behind the cursor. Regards, - yba On Thu, 1 Apr 2004, Yosef Leibovich wrote: > Thank you for your quick answer. I proposed this two methods, I thought > of the first method at first, but now I think the second method is better. > What I actually wanted is your reference to my specific sollution, of > which I think highly, but not sure it is really a good one and is it > easy enough to implement so that the developers will eventually > implement it. > So if you're having time (although Pesach is coming) I'll be glad for a > comment through the mailing list or through the issuezilla (if you'd > like I'll paste your message from the mailing list to the issuezilla and > vice versa). > Thanks Again. > Oh and I'm fixing the classification of the bugs as you explained. > Jonathan Ben Avraham wrote: > > >Hi Chen and YL, > >It looks to me like both issues 18024 and 27174 are reporting the same > >problem but propose different solutions. Both of these issues are reported > >as DEFECT which is incorrect. They might be "defects" at the level of > >product requirements, but at the engineering level they are certainly > >a missing feature, so they should be either reported as either FEATURE or > >ENHANCEMENT. > > > >To paraphrase the problem: There is currently no way to embed RTL text in > >LTR paragraphs or LTR text in RTL paragraphs when the embedded run of text > >ends with neutral characters such as a period, question mark or other > >punctuation mark, such that the visual placement of the ending neutral > >character is at the visual end of the embedded run of text. > > > >This is certainly an essential feature, whithout which the product could > >reasonably be considered defective for bidi users. > > > > - yba > > > > > >On Thu, 1 Apr 2004, Chen Levy wrote: > > > > > > > >>-----BEGIN PGP SIGNED MESSAGE----- > >>Hash: SHA1 > >> > >>Well, the main bug that sucks in, as a black hole, all the other bugs > >>regarding combination of L2R, R2L and neutral characters is: "[Issue 18024] - > >>Direction of weak characters: A new method for dealing with text direction > >>without using keyboard layout" > >>see: http://www.openoffice.org/issues/show_bug.cgi?id=18024 > >> > >>The resolution of this issue suppose to solve the control over > >>directionallity, if I understand it correctly. > >> > >>Unfortunately, the solution that is going to be implemented relies on the > >>OS IME (like in Win2K and beyond), and it leaves the rest of us (*nix, Win9x) > >>out in the cold. > >> > >>IMHO you raised a valid point, because control over text directionality isn't > >>a nicety. Its a necessity, and I for any solution that will solve this > >>problem on all platforms. > >> > >> > >> > >>>ביום חמישי, 1 באפריל 2004, 12:45, כתבת: > >>>Hi, > >>>Nowadays there's a big problem to combine English and Hebrew text in the > >>>OpenOffice. As of Now, in an R2L paragraph, all punctuations are using CTL > >>>fonts and considered R2L, another side-effect of this phenomena is, once an > >>>English sentence is typed, and it ends with closing braces they'll be shown > >>>in the end of the sentence ("I LIKE MY DOG STYE (cockerspynel) is good" > >>>will render as "cockespynel) is good) ELYTS GOD YM EKIL I" see attachment). > >>>As of now there's no way for the naive user (who don't know how to insert > >>>LRM/RLM signs) to overcome this two troubles, eLaTex's feature \L{}\R{} > >>>should solve this issue and will allow the user 100% controll on the > >>>document. > >>> > >>> > >>> > >>>>Hi YL, > >>>>Why do you think that this feature is necessary? > >>>> > >>>> - yba > >>>> > >>>>On Thu, 1 Apr 2004, YL wrote: > >>>> > >>>> > >>>>>What do you think? I'd like to hear comments from hebrew speaking > >>>>>users. http://qa.openoffice.org/issues/show_bug.cgi?id=27174 > >>>>>In short: > >>>>>This feature will, once an English character is typed, encapsulate it > >>>>>within a square which will denote an in-paragraph English sentence. > >>>>>When a special key is pressed the English sentence will end, and reguar > >>>>>writing mode will return. Obviously all text inside English box will be > >>>>>treated like L2R text, (solving us problems of punctuations font, > >>>>>directionality, and many other problems). > >>>>> > >>>>>BTW this can be implemented using LRM signs in the end of each > >>>>>English-sentence box, so that the implementation doesn't need to > >>>>>interfere with BIDi and copy/pasting the text will be possible. > >>>>> > >>>>>============================================================== > >>>>>To unsubscribe, send mail to [EMAIL PROTECTED] > >>>>>with "unsubscribe hebrew" in the message body. > >>>>> > >>>>> > >>>=============================== > >>>To unsubscribe, send mail to [EMAIL PROTECTED] > >>>with "unsubscribe hebrew" in the message body. > >>> > >>> > >>- -- > >>Levy, Chen (mailing lists mail) <mailist at chenlevy dot com> > >>Fingerprint: E547 54F9 0246 6533 66B1 8A58 5D9D CF61 2322 0E21 > >>-----BEGIN PGP SIGNATURE----- > >>Version: GnuPG v1.2.3 (GNU/Linux) > >> > >>iD8DBQFAbCFJXZ3PYSMiDiERAj8OAJ9YD43GExJRrhXNPhZY4cF7tsQHIwCfQOVs > >>NioDtljAPLgoqj9lz1mfzoQ= > >>=vDkL > >>-----END PGP SIGNATURE----- > >>=============================== > >>To unsubscribe, send mail to [EMAIL PROTECTED] > >>with "unsubscribe hebrew" in the message body. > >> > >> > >> > >> > > > > > > > > -- EE 77 7F 30 4A 64 2E C5 83 5F E7 49 A6 82 29 BA ~. .~ Tk Open Systems =}------------------------------------------------ooO--U--Ooo------------{= - [EMAIL PROTECTED] - tel: +972.2.679.5364, http://www.tkos.co.il - ============================================================== To unsubscribe, send mail to [EMAIL PROTECTED] with "unsubscribe hebrew" in the message body.
