I am having troubles with a small bug. Maybe someone can help me. Whenever I set a numbered list to use Devanagari numbers and then close the document, when I reopen it, it will be using Farsi numbers. This gets to be annoying because every time I want make a pdf or something, I have to fix the numbers and set them back to use Devanagari.
How do you even “set a numbered list to use Devanagari numbers”? The only known locale supporting the Devanagari script is sd-IN or sd-Deva-IN “Sindhi, Devanagari” but it has no locale data that would define numbered lists. Could you also attach a small sample document to see what is actually stored?
In addition to @erAck’s request, mention OS name, LO version and save format (.odt or .doc(x)). Have you enabled Complex text layout (probably Hindi) in Tools>Options, Language Settings>Languages?
@erAck The numbering type can be set from “Customize” tab of the “Bullets and Numbering” dialog. There isn’t any issue with any of the other numbering types you can see in my screenshot. When they are selected, the selection is retained after closing and reopening the document. I have attached a screenshot and a sample document for your reference.
Bullets test.odt (10.9 KB)
@ajlittoz This is actually a cross platform issue. It is an issue that has existed for some years, but I have kept quiet until now hoping it would get fixed in a new version. As I remember, I first experienced it in Linux. The same issue is in the Mac version as I am now a Mac user. This is an issue on both Intel and M1 machines. I am pretty sure it was an issue on Windows when I had to borrow a friends computer for a while, but I am not 100% if I tested this specifically on that machine. I have enabled Complex text layout. This is an issue regardless of if I select Hindi, another language, or just select none for the language. This is an issue across all LO versions I have checked it on.
This way it is possible that nobody will ever take notice of your problem. Please upload a document demonstrating the problem. Upload a document that works for you on one platform but not on the other.
There is definitely a bug here. To eliminate all possible interferences, I created a specific list style I named Devanagari. I configured it for Devanagari numbering. The style is reloaded as Farsi.
I tested with other numbering from Hindic languages (so they look to me but I can’t tell which is which), even Tibetan. Only Devanagary behaves so. Since Farsi immediately precedes Devanagari, could it be a developer blunder where s/he forgot to fix the index (the eternal problem of 0- or 1-origin in arrays!)?
Report as a bug.
Ah thanks, so the entry after Farsi is supposed to be Devanagari, but when saving the document it already gets stored as
style:num-format="۱, ۲, ۳, ..."
which is the Farsi list. That’s clearly a bug and should be reported as such.
If bugs aren’t reported then they can be fixed only by accident.
Assumptions don’t help much. And no, it’s not an index problem but the eternal problem of copy-paste, here the definition list of characters to be used for the unique save-list.
Possibly also related: tdf#141565
Good spotting, yes that’s exactly the same style:num-format="۱, ۲, ۳, ..." for the <text:page-number> element. So at least we don’t need yet another bug just that the original reporter didn’t follow up with the request to provide details and people involved couldn’t reproduce.