|Re: Arabic Font In Export to PDF on Linux [message #1796147 is a reply to message #1795658]
||Sat, 06 October 2018 07:33
| Patrick Moran
Registered: March 2018
I used Lulu to publish some stuff with both English and Chinese and in the old days it used to be a horrifically frustrating job because you'd go through a whole series of things to do, and then it would tell you that the process failed because there was some weird font that was probably in my document only because the font I used most of the time did not have a couple of special characters. |
I had a look at your png files. The "coli" is just a heading, right. And the Arabic doesn't print at all. I think it's got to be a font problem. Usually Acrobat is pretty good at dealing with all kinds of input, but maybe it's behind the times with regard to Arabic. You're just going to have to dig things out for yourself.
I think your problem is with Oracle WebLogic and Acrobat. They may be limited in some cases because one our the other doesn't have usage rights to certain fonts. Usually Acrobat applications tell me it is substituting X for Y, but if X doesn't have something I need then I am screwed. I've never had it not produce anything and not give a warning. Could it be that your document is layered somehow and Acrobat is only seeing the "surround and coli" part?
Look at your WebLogic preferences pane. Can you set a different font? If so, do some research on the Internet to locate Unicode fonts that have a complete set of Arabic glyphs. Then try to match it with Acrobat-acceptable fonts. I suppose they have a list somewhere. Otherwise you just try to enter the file into Acrobat and it should tell you that it is having a font problem. Also, try to figure out where that frame is coming from. If you run the file through a hexdump you may be able to see how to isolate the frame from the content. You might even be able to erase the frame from the hexdump and turn the rest of the hex file back into printable text. If you can't find an application that can do that it wouldn't be hard to write a program, although, bummer, there is the text direction stuff to worry about too.
Do some experimentation. For one thing, take a look at WebLogic exported files. Do a hexdump on them and see how they may differ from the same brief text just written out directly into BBEdit or whatever you use for your hexdump. Over the last couple of decades they've taken care of most of the Chinese problems. (It used to be a complete mess.) Maybe they have some ways to go with Arabic.
Maybe you can get a correct-looking document exported and copied into a Word file, save it, and then re-export it/print it as a PDF file. Some of the word processors can really do amazing things, recognizing what fonts they are dealing with, what encodings, etc. and all without human intervention.
If you ever try to put any of this stuff into an email it would be interesting to see whether it survives the trip. (Sometimes files get "translated" back and forth between forms of encoding every time they go from one university hard drive to another. A college used to send nonsense "comic book swearing" texts from home via her university account, but when she sent the same thing from her office at the university it was no problem. Turned out there was one hard drive more in the chain between her home and the university and me than there was between the university and me. When people invented ASCII they had no idea where this would all end up.
You may find an application called iHex may help. It does a hexdump that includes the corresponding words in a second column, and it has an incredible set of Text Encoding/other.
Maybe the programmer who made iHex would have some advice on how to handle your problem.
[Updated on: Sat, 06 October 2018 17:16]
Report message to a moderator
Powered by FUDForum
. Page generated in 0.03704 seconds