When trying to add Ruby text to a phonetic transcription, the word (as selected) is broken up into individual letters and diacritics, making it impossible to actually add the text. The word I was trying to add text to is a transcription of a reconstructed Proto-Indo-European word for horse: *h₁ek̑u̯os. In the Ruby dialog this is split into these separate words: * | h | ₁ | ek | ̑ | u | ̑ | os. Expected behaviour: Writer should accept the selection as a word or allow some way to define the area that is to be glossed with Ruby text.
Thank you Seán. Could you please provide a series of steps to reproduce the issue, and possible an example document to make it easier? What is the "Ruby dialog"? Is this from an extension? Was this ever possible in LibreOffice, or is it a defect that's recently appeared after updating? And could you please paste here the detailed version information copied from Help > About LibreOffice?
Created attachment 186579 [details] Samples of problems with ruby text
Sorry, ruby is the generic term for what LibreOffice calls Asian Phonetic Guide. I can’t say if you could do exactly this before as I’ve not tried using phonetics as base text before. I’d guess that however a word is detected works OK for ordinary text, but it doesn’t for text containing combining diacritic marks. The workaround is to select some arbitrary text, add ruby to it then change the base text in the dialog. What’s really needed is a way for the user to define the extent of the base text that the ruby text is applied to and the most obvious way to do that is through selection. Looking at Bug 107184, it seems the automatic selection can go wrong for Asian text too. A continuing problem mentioned in that bug is that if you add to the base text, it is possible for the addition to become new base text. Attempting to fix this by moving text between base text fields in the dialog can produce bizarre results. Some of the base text may disappear or be re-ordered. There is also no obvious way to _remove_ the ruby text. A copy + paste unformatted will do it, but that’s not too intuitive. All of these issues seem linked to the user not having easy control over what is defined as the base text and no visual indicator at all of the extent of the base text, though clearly it must be working in a similar way to HMTL ruby tags behind the scenes. E.g. if you select a phrase Writer will divide it into words so the ruby text will be attached to each word, but if you select some arbitrary character, then replace it with a phrase in the dialog, the ruby text is attached/aligned to the whole phrase. In the first case, inserting in the middle of the phrase might cause problems if you try to fix it in the dialog, but the second the ruby text stays aligned to the whole phrase. Attached is a document with some samples of base and ruby text and more details. Version: 7.5.2.2 (X86_64) / LibreOffice Community Build ID: 50(Build:2) CPU threads: 8; OS: Linux 6.2; UI render: default; VCL: gtk3 Locale: ga-IE (ga_IE.UTF-8); UI: ga-IE 7.5.2-1 Calc: threaded
[Automated Action] NeedInfo-To-Unconfirmed