I have been trying to compile pdf in Indic language but output is blank!

[IDontKnow2Code]

Hello,
I have been trying to compile documents in Tamil (one of the Indian Languages). But neatroff outputs a blank pdf.
Font I used MuktaMalar
Here is the file Tamil.ms
I m new to troff, so please forgive me if it is something simple.
Thank you.

[aligrudi]

Hello, IDontKnow2Code ***@***.***> wrote:

I have been trying to compile documents in Tamil (one of the Indian Languages). But neatroff outputs a blank pdf. Font I used [MuktaMalar](https://fonts.google.com/specimen/Mukta+Malar?subset=tamil#about) Here is the file [Tamil.ms](http://0x0.st/-piB.ms) I m new to troff, so please forgive me if it is something simple.

First of all, note that macros like TL, AU, SH, and PP change the current font. Therefore, you should change the font after the invocation of each of these macros, redefine these macros, or use the following line to redefine R and B fonts: .fp - R MuktaMalar-Regular .fp - B MuktaMalar-Bold About the output, there seems to be a problem with Ghostscript. Try building Tamil.PDF in neatroff_make/demo directory (with uppercase letters in the extension); this uses the PDF postprocessor of Neatroff (produces the PDF output directly and does not use Ghostscript). Does this work? Ali

[IDontKnow2Code]

I removed TL,AU,SH,LP and previous font definition and redefined fonts as you said and compiled it using demo make with .PDF, I have a output but the letters are overlapping

[aligrudi]

IDontKnow2Code ***@***.***> wrote:

I removed TL,AU,SH,LP and previous font definition and redefined fonts as you said and compiled it using demo make with .PDF, I have a output but the words are overlapping

The problem with the font is that it does not include PostScript font name. I changed Neatpost to use its file name, if font name is missing. After this change, the PostScript backend of Neatpost should also work. About the overlapping words, can you include a very small example plus the way it should actually look. Ali

[IDontKnow2Code]

This is how it should look

But this is how it outputs

Title in both images are same please check that to know the differences.

[aligrudi]

IDontKnow2Code ***@***.***> wrote:

This is how it should look  But this is how it outputs  Title in both images are same please check that to know the differences.

Thanks. Actually rendering Tamil text is more complex than what I thought: https://docs.microsoft.com/en-us/typography/script-development/tamil OpenType GPOS and GSUB font features can be enabled with .ff request (Do you see any improvements?): .ff R +nukt +akhn +rphf +rkrf +pref +blwf +half +pstf +cjct +init +pres +abvs +blws +ptts +haln +psts +dist +abvm +blwm However, that does not seem to be enough for Tamil. Ali

[IDontKnow2Code]

Yeah I see some improvements.

But like you said its not complete, should I try using any other type of fonts like .pfb or .afm?

[aligrudi]

IDontKnow2Code ***@***.***> wrote:

Yeah I see some improvements.  But like you said its not complete, should I try using any other type of fonts like .pfb or .afm?

I have not read the details yet. However, it seems rather complex, especially for someone unfamiliar with this script; see the shaping engine section of the following page: https://docs.microsoft.com/en-us/typography/script-development/tamil Can you summarize what should be done besides applying the standard OpenType features? Thanks, Ali

[IDontKnow2Code]

I think matra is the problem here.
They are responsible for (from what I understand) க,கா,கி,கீ,கு,கூ,கெ,கே,கை,கொ,கோ,கௌ.(these letters are combination of vowels and a single consonant 'க்') (check files).
The problem here(in pdf we created) is க,கா,கி,கீ,கு,கூ seems to be forming correct, but others கெ,கே,கை,கொ,கோ,கௌ are not forming correctly.
Example
Here is the tamil.ms file
Here is the compiled tamil.PDF
Check the tamil.ms file in browser to see the correct form of the incorrectly formed letters.
Thank you.