Numbers in pages are just disappeared after PDF conversion

Hi Everyone!
I just have a problem with Libre Office Writer 7.1.8 on Fedora 34.
Well, It just somehow looks a little bit weird to explain;
Everytime I try to export a document into a pdf, the pdf itself after conversion don’t show numbers, I use the Droid Arabic Kufi font, also I tried to adjust the settings a lots until I restore it to the defaults :joy:
Also I searched the web a lot
And finally, I deleted the whole Libre office package and Install it again, and nothing changed when I try to export the .odt documents.

Guys, I need your help!

Note: Pages Header and Footer show the numbers as it should to be after the conversion, only numbers in the page paragraphs not showing!

الدواجن.odt (76.0 KB)

Update: I realized it happens for all my documents using Libre Office

Update 2: I also tried to print my .odt to file, and it also doesn’t want to show numbers in the pdf

Update 3: If I copy the area where the numbers supposed to be in the pdf, and paste it in another place, it just paste the numbers, but what the hell it is invisible in the pdf!

Attach an .odt sample file.

1 Like

Okay my friend, I uploaded the whole .odt document right now.

I have installed a “Droid Arabic Kufi” that I found somewhere (v.1.00 from Dec 2012) on my Windows 10; and I can’t repro the problem exporting (using toolbar), and viewing in Adobe Acrobat Reader (or I do not know where to look). Please provide a screenshot showing the specific problematic place in ODT, and comparison to the same place in PDF, to know exactly what is problematic.

My original suspicion was that it might be the font’s permissions, that might prevent embedding it into the PDF. That may depend on specific version of the font (cf. tdf#145967, where an older version of “Bebas Neue” was restricted, but a newer was not), so posting the font version that you use may be needed.

Additionally, mention how did you export to PDF (was it Writer’s Export Directly as PDF toolbar icon / File|Export, or maybe you printed to PDF?), and what software you used to view the PDF.

1 Like

Well, I also use version 1.00 of the font .

And here is the difference between the .odt format which shows the numbers as it should:

And the pdf format after conversion:

Also about the font permission, I always use it since a couple of year ago, and there isn’t any problem with exporting to PDF, but the problem with the exporting started happening recently.
So I don’t think there’s a problem with it since I didn’t have adjust anything.

And for PDF viewer I use the one that Gnome-shell ships, it’s “Evince”.

About exporting to pdf, I actually used every option available, I tried exporting it directly, and regularly with adjusting every single option also I tried to printing it to File using the print option, and nothing changed.

Then please provide the resulting PDF (you may need to change its extension to e.g. ODT before uploading here, because of the site’s restrictions). Maybe it’s something specific to the viewer? Which could appear after some update of the software.

1 Like

Reproduced on my Fedora 35 box with LO Apparently Droid Arabic Kufi is installed by default.

It seems to be related to the Droid Arabic Kufi font. If I replace it with Liberation Sans, all the numbers export without problem.
If you look carefully at the numbers, you’ll notice they have a somewhat non consistent shape and colour (like some transparency has been applied). This seems to me akin to some picture embedded in the font. So, I would incriminate the font itself.

Also, you won’t be able to fix easily your document: it is an unmanageable mess of direct formatting. Use styles.
Vertical spacing in done with empty paragraphs. Horizontal spacing is done with space characters making you sensitive to the font used (face and size) or with default tabs.
You are not consistent with your font face usage. For example, Lohit Devanagari (adapted for one of India languages) is in effect in the footer.

I won’t give further advice until you tidy up your document formatting-wise.

1 Like

Is there is a way to convert the .odt into a bunch of JPEG photos, then converts into a pdf again?

I just need to convert it into a pdf, I don’t care if it’s a picture pdf, haha!

Well, I appreciate that, I am working on it on other documents!
Thanks for the advice!

Is there is a way to convert the .odt into a bunch of JPEG photos, then converts into a pdf again?

I selected the text containing the numbers in the generated PDF (as said, it’s shown OK in my case), and copy-pasted it to Writer and Word, to inspect the formatting of the text. And while the Arabic characters had the “Droid Arabic Kufi” font, the numbers were in Lucida Sans Unicode, which hints that there happened a substitution for characters missing from the font. So which font was used for the export in your case?

Several things here.

  1. The character formatting in this case should assign “Droid Arabic Kufi” only to CTL font settings; Western section should use some other font - to control the formatting precisely.
  2. There seems to be a difference in font substitution between showing in Writer vs. exporting to PDF (I seem to recall some discussion about that) - that would be a real bug.
1 Like

Hey there, sorry for taking some time, that’s because I have some final exams these days…

Well, that is the PDF file
الدواجن.odt (438.2 KB)

I just renamed it as .odt, so you have to rename it as .pdf!

In my case the numbers were in “Liberation Sans”!!
But the rest of the text were in “Droid Arabic Kufi”.

I did it, and sadly the numbers remain invisible!

Hope you find a solution for the bug, I really appreciate LibreOffice developers’ work and what they are doing of a series of contributions for such a great project! :eagle::eagle::eagle:

No they were in NotoColorEmoji:

We really need to file a bug.

1 Like

Well, a huge thank for that conclusion, also I appreciate that a looot!
Before talking about a reproducing a bug I need a quick solve to my problem
So I understand that I have to change the font into another one, tho the pages alignment will change.
At all that is not the real problem, the problem is I have another 3 document with about 300 pages at total.
So, is there is a single way to convert the numeric objects “which have some transparencies” as embedded images so it become visible after conversion to PDF?
Or is there is a way to select all “numbers only” and changes it to another font?