The PDF itself (in english) takes around 8-9 seconds to complete for about 36 pages, some of which include bar charts. I'd also like to note that due to the format that the data is received in, I am using "writeHTML" to present it on the PDF.
My major issue came around when I started trying to generate the same PDFs with arabic translations. I'm using the "Almohanad" font as described in the TCPDF examples, and the processing time jumped up to 2 minutes.
Alas it did not seem to work (the ".CACHED" file didn't appear even though I gave the tcpdf file full user permissions)! I do have to mention that i needed to insert ob_clean() after the "require_once" section of my php file to avoid the following error:
AL-Mohanad Regular is a Regular TrueType Font. It has been downloaded 6652 times. 3 users have given the font a rating of 3.67 out of 5. Check out Character Map section to understand the Calligraphy of AL-Mohanad Regular.
I am having a similar issue, where the generated Word file is correct but not the PDF.
My client have a custom made font (they are very careful with their visual design) and it might be possible to ask for a re-made font file (with better unicode handling) if that is the issue.
Unfortunately this issue turns out to be more complex then we previously thought. This font does not require a simple encoding table but a variant of Arabic shaping. And there seems to be no documentation. I am afraid, the implementation of this issue has been postponed till a later date (no estimates are available at the moment). We will inform you via this thread as soon as this issue is resolved. We apologize for your inconvenience.
Need to create a PDF from page containing Urdu text in Joomla. However generated pages contain junk, even though the correct font is embedded in them. Steps followed:
Convert NafeesWeb (from Debian repository) from ttf to afm to .php, .z using ttf2pt1 (for generating the afm file) and makeweb.php (for .php and .z).
Set the pdfFontName to nafeeswebnaskh in the ur-PK (Urdu/Pakistan) language in Joomla. ur-PK is the default encoding for all pages in this instance of Joomla. All pages render fine in all browsers that have an Urdu font installed.
Now PDFs generate fine, but with junk instead of text. When the Urdu font was specified (incorrectly) as freesans earlier the PDFs used to render the latin text correctly and display the Urdu text as blank; however with the Urdu font set to nafeeswebnaskh the latin text displays as junk too.
Kpdf confirms that the NafeesWebNaskh font is embedded in the document.
tcpdf identifies itself as:
TCPDF 2.6.000_PHP4
The server is RHEL4 running Apache 2.0.52.
Sample generated PDF available on request.
Apparently the version of Joomla I'm using is incompatible with the latest version of TCPDF. So I went back to the old version of TCPDF with the almohanad font from the latest version and that seems to be working (though I still see a few blanks -- however I don't know Urdu so can't say whether the document is proper or not :)
Discovered that the ttf2ufm packaged with tcpdf is actually ttf2pt1, and used it with the -a -F options to generate a UFM file. After using makefont.php with the ttf and ufm files, latin characters display correctly in the generated PDFs but only a few (2-3) Urdu characers display -- the rest are blank.
The PDF itself (in english) takes around 8-9 seconds to complete for about 36
pages, some of which include bar charts. I'd also like to note that due to the
format that the data is received in, I am using "writeHTML" to present it on
the PDF.
My major issue came around when I started trying to generate the same PDFs
with arabic translations. I'm using the "Almohanad" font as described in the
TCPDF examples, and the processing time jumped up to 2 minutes.
For Windows: FontForge, CorelDRAW Graphics Suite X5-X7, CorelDRAW Graphics Suite 2017, FontCreator, Microsoft Windows Font Viewer, AMP Font Viewer.
For Mac OS: FontForge, Apple Font Book, Bohemian Coding Fontcase, Mac OS X Font Book.
For Linux: FontForge
All rights for the fonts given on this website reserved by their owners (authors, designers). The license given on the font page only represents received data. For detailed information, please, read the files (e.g., readme.txt) from archive or visit the website given by an author (designer) or contact with him if you have any doubt.
If there is no reported author (designer) or license, it means that there is no information on the given font, but it does not mean that the font is free.
Thanks! I am wondering if I have misunderstood what you are referring to. Do you have a problem with nastaliq, as shown below? Or just with the letter forms of the standard naskh font. My Urdu comments only relate to nastaliq.
please ask through to change the font to Geeza Pro rather than the dumb stupid SF Arabic - which suitable for 3rd grade elementary student - that can't handle UTF-8 charachter as result enlarge the charachter and shows more like Persian or Urdu but not Arabic.
Just found out after all the efforts and research: it is not the Arabic fonts but iOS 11 is using New San Francisco as default font and it affects the outcome of Arabic text and perhaps any other language that use non-Latin fonts too. I guess for us that use Arabic fonts will get stuck with a hope that will consider changing the fonts like it was.
OK, I understand now. Some users have ALL their text looking like those Arabic UTF characters, and I thought maybe you were one of them. When that happens, it is because a nastaliq Arabic font gets used instead of a naskh Arabic font. And a nastaliq Arabic font gets used when the machine thinks the user prefers Urdu over Arabic language.
Also, solution to this problem on most of the earlier posts is to install ubuntu latex recommended package, however please note that I am not using LaTeX from Ubuntu's software center (which is old TeX Live 2009) but rather have manually installed basic version of TeX Live 2012 as per instructions here:
In this case, the missing file is ecrm1000.tfm according to the error message. To know which package it belongs to, you can use the command tlmgr search --file ecrm1000.tfm --global which finds it in package ec. So you have to install it with tlmgr install ec.
Then you may have other errors, other files to look up and other packages to install. Since the process may be cumbersome, you may want to install at least collection-latexrecommended and collection-fontsrecommended if you really don't want a full install.
Edit: I found out (in this post) and by running tlmgr info pzdr that the error caused by the hyperref package can be fixed with the zapfding package because there is a pzdr.afm and a pzdr.tfm in the zapfding package. After installing zapfding everyting was working. The cbfonts like I mention below is not needed for me anymore.
Maybe cbfonts was needed for the other errors. I was able to fix the mentioned error Font \XeTeXLink@font=pzdr at 0.00002pt not loadable by using only zapfding. I cannot say something about the other "multiple errors for several fonts" anymore because I am working on a different TeX setup now. I am leaving the original answer below in case that the other errors were really caused by textgreek.
Another problem is that the new font is too condensed; The height of the characters is considerably less than the other Arabic fonts, and space between words is too small when you are reading huge blocks of texts. I noticed this while viewing my Arabic blog posts that I wrote:
Another issue, is that the space after some punctuation marks is quite huge comparing to the normal characters, giving them an unnatural feeling and making them harder to read (The white space after these marks are almost twice the white space after normal words, notice the areas I highlighted and compare them to the other white spaces):
If we are to fix the problem, then I believe a public poll for a number of suggested Arabic fonts should be created before such change is made. This could be helpful for 20.10 or other future releases.
Another problem is that the new font is too condensed; The height of the characters is considerably less than the other Arabic fonts, and space between words is too small when you are reading huge blocks of texts.
This has now been implemented, and I have confirmed that it works as intended via the latest groovy and focal daily builds. So, provided that you are connected to internet when installing, it will work with the Ubuntu 20.10 ISO. As regards focal it will work only with the ISO for the Ubuntu 20.04.2 point release.
After you install a font into the Fonts folder in the operating system and start Microsoft Word for Mac, the font unexpectedly is not available in the Font dialog box, in the drop-down list, or in the Formatting Palette.
Third-party fonts are not directly supported in Microsoft Office for Mac applications. Some third-party fonts may work in one application and not in another. Other third-party fonts are installed in a "family". A family usually consists of the third-party font itself together with some or all of its variations (bold, italic, and so forth). Sometimes, a font may be displayed in Microsoft Word, Microsoft PowerPoint, Microsoft Excel, or Microsoft Entourage, but you may be unable to use one of its variations, such as italic.
First, restart your computer, and then test the font again. Some installations are not complete until the computer is restarted. This also makes sure that all applications are restarted after the installation.
For best performance in Word, try to run with all your fonts enabled all the time. Each time that Word starts, it compares its font cache with the system font cache. If the two don't match, Word will regenerate its own font cache, which can take a few seconds. If you have dynamically enabled fonts, the system font cache will appear different nearly every time that Word runs this comparison.
03c5feb9e7