Tsek character not functioning properly in Tibetan

[David in MA]

I'm using Publisher v 1.7.3 on Mac OS 10.15.2. I have a couple of Tibetan dictionaries installed in the Mac. When I use Unicode Tibetan, I do see the Tibetan option in the typology options. Ligatures and contextual alternates are working fine. I do not see my Tibetan dictionaries as options in the choice of dictionaries, but I don't care about that. 

The problem is that Publisher is treating the tsek ( or tseg: U+0F0B ) word separator as a letter. So a whole sentence becomes a single word and Publisher does not properly wrap lines after tseks. My workaround is to globally replace tseks with tsek followed by a zero width space. When I do that Publisher properly wraps lines and justifies paragraphs. 

Microsoft Word fixed this problem in Word for Mac 2019. Recent versions of LibreOffice and NeoOffice handle it properly as well. InDesign required enabling their World Ready paragraph formatter, but we are trying to get away from using InDesign in favor of using Publisher. It would help greatly if you can fix this problem so Tibetan tsek is treated as a word or syllable terminator to allow line wraps after them.

Thanks,

David Dvore

[walt.farrell]

3 hours ago, David in MA said:

The problem is that Publisher is treating the tsek ( or tseg: U+0F0B ) word separator as a letter.

Which it clearly shouldn't do, as U+0F0B is defined in the Unicode database as a punctuation character with the possibility of a line-break after it (gc=PO, lb=BA).

[David in MA]

I followed the directions for installing a spelling dictionary for Tibetan in the Mac OS. (Wow, that feature is really hidden.)  It now appears as an option for Spelling Language in Publisher, but does not solve the problem with Tseks not being recognized as word delimiters.

To be clear, a Tsek should not force a line break, but should allow a line break when needed, much like a space between words in English except that it appears as a dot. In Tibetan, words and syllables all end with a Tsek. As is, Publisher only wraps new lines where spaces exist between sentences and treats the whole sentence as a single word, breaking it at random places is there is no room on the line. If I add zero width spaces after every Tsek, then line wrapping and paragraph justification work as they should because the added zero-width space allows Publisher to wrap the line after a tsek when needed. But I shouldn't have to add those zero width spaces.

Otherwise, I'm finding Publisher to be a good replacement for InDesign and hope that you soon get the feature to import IMDL files working so we can import our old InDesign work without having to reflow all the text.

Thanks, David

EDIT - This post was merged from a separate topic as it best fits with your thread here.

Edited January 28, 2020 by Dan C

[thomaso]

@David, just curious in case Serif will start fixing it: Are there more glyphs in Tibetan which should allow/cause a line break?
According to the w3c article about "Requirements for Tibetan Text Layout" there seem to be some besides 'tseg' U+0F0B only.

https://w3c.github.io/tlreq/#line_breaking

Quote

 

>> Line breaks are also possible after:

• U+0F0D TIBETAN MARK SHAD ། - as long as the next line starts with a consonant (ie. not a second shad).

• U+0F14 TIBETAN MARK GTER TSHEG ༔

• U+0F7F TIBETAN SIGN RNAM BCAD ཿ (used to represent the visarga in transliterations). There is never a tsek after this character, eg. ཨོཾ་ཨཱཿཧཱུྃ་.  <<

 

Also I wonder whether a tseg always may cause a line break, or if there are situations where it should not, e.g.:

Quote

>> Line breaks do not occur after a tsek when it follows U+0F44 TIBETAN LETTER NGA ང (with or without a vowel sign) and precedes a shay (U+0F0D TIBETAN MARK SHAD །). <<

 

[David in MA]

It looks like this topic was renamed from my original post while I was replying to the previous comment. The new name doesn't exactly fit. There is no hyphenation in Tibetan as syllables are already delimited by Tseks. Words are either a single or multiple syllables which must be recognized as there is no additional grammatical character separating words. The point is that Publisher is treating the Tsek character as an alphabetical letter that is part of a word rather than delimiting words and allowing line breaks after them when needed. 

Thomaso - you did good research. Those points are all valid. In general, there can be multiple grammatical characters in a sequence that should not be broken up. Line breaks can occur after Tseks or spaces, but if there are additional grammatical characters adjacent to the Tsek, they should not be broken up.

I hope they rename this topic to something more fitting. Perhaps "Tibetan Tsek not recognized as a word delimiter" 

[walt.farrell]

33 minutes ago, David in MA said:

It looks like this topic was renamed from my original post while I was replying to the previous comment.

I suspect you clicked on a New Topic button rather than a Reply button. Your original topic is still there:

Perhaps a moderator will merge them. But you can rename this one, by using the Edit link at the bottom of the first post in this topic.

[thomaso]

>> It looks like this topic was renamed from my original post while I was replying to the previous comment.

Compare the "EDIT ..." note at the end of your second post in this thread, written by Serif moderator Dan C.
I agree, other users searching for "Tibetan" in the forum might miss it in the heading. You can try adding "Tibetan" and removing the misleading "Hyphenation" yourself by double-clicking the thread title above your first post. If it doesn't work, you can't edit it yet and may need to do more forum activity. - At least we know that Dan C has noticed your thread as a bug in Publisher.

To avoid confusion you shouldn't continue writing in your previous thread (Affinity on Desktop Questions (Mac and Windows)) but here, in this Publisher Bugs forum only.

[David in MA]

I was happy to see that Publisher v1.8 added the IDML import feature. This is a big help. But unfortunately, the Tibetan Tsek bug has not been fixed. I wish they would name this thread more appropriately. It strongly affects anyone working in Tibetan as it prevents lines from wrapping properly and paragraphs being justified. Better yet, I wish they would fix the bug. It can't be that difficult to adjust the properties of a single unicode character to work properly.

[walt.farrell]

31 minutes ago, David in MA said:

I wish they would name this thread more appropriately.

It's your thread. Look at the bottom of the first post in it and see if there's an Edit link. If there is, click it and you'll be able to change the topic subject.

[David in MA]

Walt, I saw your earlier message, but there is no Edit link in my initial post. Dan C edited my second post and renamed the thread in the process. I'm not able to override that, or make any edits to any of my posts.

Dan C, if you're listening, please change the name of this thread to "Tsek character not functioning properly in Tibetan", or at least something with Tibetan in the Title. thx.

[David in MA]

I just tried laying out some Tibetan text in Publisher 1.8.2 with hopes that the new version fixed the problem with lines not wrapping after the Tsek character. They are referring to it as hyphenation, which I suppose is accurate since Tsek is a syllable delimiter and there is no actual word delimiter for multi-syllabic words. At any rate, it still doesn't work properly. I'm hoping the next update will fix this.

[David in MA]

Once again my hopes were up that the latest release, v1.8.3 , would fix the Tibetan "Tseg" bug that doesn't allow lines to break after a tseg. Unfortunately, fixing that bug didn't make this cut. PLEASE fix this. Thanks.

[David in MA]

Yay! v1.8.4 fixed line-wrapping after Tsek ... almost. There is one exception that still needs fixing. When Tsek(U+0F0B)  is followed immediately by Shad ( either U+0F0D, U+0F14, or U+0F07F ), then the line wraps after the shad, not the tsek. This is a great improvement - thanks, but still needs one more tweak.

[thomaso]

24 minutes ago, David in MA said:

the line wraps after the shad, not the tsek. This is a great improvement

You may want to make this clear to developers:

Should a line break between a shad and a tsek – or shouldn't it break there, but only after the tsek?

[David in MA]

Thanks Thomaso. To be clear, when a Shad immediately follows a Tsek, the line break should be after the Shad, not after the Tsek. In all other cases I can think of, a line break can be after the Tsek. Prior to vs 1.8.4, Publisher did not allow line breaks after a Tsek. v1.8.4 fixed that, but there is this one exception that needs to be implemented. Thanks.

[CaitlinG]

I have a much more basic question, and based on the content here, I'm hoping that you all may be able to help. Thus far I haven't had any luck with typing in Tibetan with Affinity: the letters won't stack correctly. Any tips on how to do so would be much appreciated!