Check the previous character and print it - TeX

निरंजन

I have submitted a package named `unitipa` which will soon be visible to all. Thanks. I have cited you answer in the documentation :)

निरंजन

@samcarter I got an answer on SE https://tex.stackexchange.com/a/563447/174620

निरंजन

@samcarter I also tried this -
  
```  
% !TEX TS-program = lualatex
\documentclass{article}
\usepackage{luacode}
\begin{luacode}
function myligs ( s )
   s = unicode.utf8.gsub ( s , '(\\?)([%a@]+)' , function( backslash, text )
       if backslash=='' then 
           text = unicode.utf8.gsub(text, '(.)(æ)', '%1%\\textbf{%1}' )  
       end
       return backslash .. text
   end)
   return s
end
\end{luacode}

\begin{document}

\directlua{luatexbase.add_to_callback("process_input_buffer", myligs, "myligs")}


am

aæ

bæ

ax
\end{document}  
```  
  
This code is a tweaked version of yours. In this one I successfully managed to pass the previous character to the argument of some command (whenever `æ` comes). Now I want something similar for `̪` this diacritic, but it seems to not work :'(  
  
```  
% !TEX TS-program = lualatex
\documentclass{article}
\usepackage{tipa}
\usepackage{luacode}
\begin{luacode}
function myligs ( s )
   s = unicode.utf8.gsub ( s , '(\\?)([%a@]+)' , function( backslash, text )
       if backslash=='' then 
           text = unicode.utf8.gsub(text, '(.)(̪)', '%1%\\textsubbridge{%1}' )  
       end
       return backslash .. text
   end)
   return s
end
\end{luacode}

\begin{document}

\directlua{luatexbase.add_to_callback("process_input_buffer", myligs, "myligs")}


am

aæ

bæ

ax

t̪  
  
\textsubbridge{t}
\end{document}  
```  
  
I am very close to what I want to achieve, but couldn't complete it.

निरंजन replying to samcarter

Thanks Ulrike, if you are reading this comment, kindly note that I have no issues when I compile char+diacritic with regular LuaLaTeX. The problem that I am facing is `tipa` (a package for typing IPA) provides a command (eg. `\textsubbridge{a}`) which needs the argument of the diacritic bearing character, but in Unicode setting the character comes before the diacritic. Hence I want to make that diacritic look back and use the previous character as the argument of the according tipa command. Please see the code posted in the above comment. Thanks

निरंजन

Friends, this  seems to work provided I use it in a command or an environment, is there **any** other way to make this automatic? (automatic in the sense wherever `t̪` comes `\textsubbridge{t}` should replace it.)
  
```  
%!TeX program = lualatex
\documentclass{article}
\usepackage{tipa}
\usepackage{luacode} 
\begin{luacode}
function myreplace(s)
  s = s:gsub ( "t̪" , "\\textsubbridge{t}" )
  tex.sprint ( s )
end
\end{luacode}
\newcommand\replace[1]{\directlua{myreplace( \luastring{#1}) }}

\begin{document}
\replace{t̪}
\end{document}  
```  
  
@samcarter You might not see that character because of font reasons probably or in some fonts it is shown with some extra kerning. So please check the output of this one.

samcarter

@निरंजन Ulrike commented about your problem https://chat.stackoverflow.com/transcript/message/50495806#50495806

samcarter replying to निरंजन

For the t-thingy I can't help - when I copy that from your comment, I get only one letter in my editor

samcarter replying to निरंजन

For `aa` and `bb` replace `%1a` with `%1%1`

निरंजन

@samcarter I want `t̪` to be replaced by `\textsubbridge{t}`. Is it possible? I tried doing it in your code, but didn't work. See this -  
  
```  
%!TeX Program = lualatex
\documentclass{article}
\usepackage{luacode}
\usepackage{tipa}
\begin{luacode}
    function myligs ( s )
    s = unicode.utf8.gsub ( s , '(\\?)([%a@]+)' , function( backslash, text )
    -- no substitutions inside (La)TeX macros
    if backslash=='' then 
    text = unicode.utf8.gsub(text, '(t̪)(\\textsubbridge{t})', '%1a' )  
    end
    return backslash .. text
    end)
    return s
    end
\end{luacode}

\begin{document}
    \directlua{luatexbase.add_to_callback("process_input_buffer", myligs, "myligs")}
    
    t̪
    
    \textsubbridge{t}
\end{document}
```

निरंजन

@samcarter, re: Thanks for [your answer](#a1577), but unfortunately I don't understand what you have done. Ideally I would expect output `am, aa, bb, ax` but the active character seems to save the first preceding character,

samcarter

https://tex.stackexchange.com/a/294154/36296 might be a starting point

samcarter replying to निरंजन

This sounds like ligatures, doesn't it?

निरंजन

but can I trigger the metafont shape of my choice when I issue a unicode character?

निरंजन replying to user 3.14159

Nope, parsing is not the option unfortunately :(

user 3.14159 replying to निरंजन

If you are willing to parse the full expression, you can do that. Otherwise it will be hard to impossible.

निरंजन

Is there any other way of accessing the `.mf` fonts and using some characters from there?

निरंजन replying to joulev

Alas :'(

निरंजन replying to samcarter

Linguists use IPA for transcription of different sounds. Traditionally in LaTeX `tipa` provided a metafont which had its own syntax (eg. `\textrtailt` for `ʈ`) After Unicode keyboards have been made, it's very easy to type these characters, hence this old syntax is considered to be obsolete. I am trying to turn Unicode characters into `tipa` commands directly by using `newunicodechar` package. See a simple example.  
  
```
\documentclass{article}  
\usepackage{tipa}  
\usepackage{newunicodechar}
\newunicodechar{ɔ}{\textopeno}  

\begin{document}  
ɔ 
\end{document}
```  
  
This code prints `ɔ` in the `tipa` metafont.  
  
but there are some diacritic marks in IPA which are used **after** some characters, but the tipa syntax for them is somewhat like `\diacritic{x}` which will print the diacritic around `x`. I have successfully turned most of the IPA characters into `tipa` commands, but I am stuck at these diacritics. I want a macro which could do something like `\diacritic{\lastchar}` which means the preceding character will be passed as an argument to the diacritic command.

joulev

@निरंजन https://tex.stackexchange.com/a/13334

निरंजन replying to user 3.14159

What if we activate certain characters and do that? eg. In the MWE provided, if `æ` is active can it save the preceding character?

निरंजन

It's quiet complicated, I'll try to create an MWE for the same. Give me some time.

user 3.14159

I do not think that there are mechanisms that allow you to access the preceding character in "normal" text. It is possible if you scan the whole text with a parser or the like, but this might lead to other problems.

samcarter

Can you give a bit more context where you want to use this? This might be an xy-problem.

1 Answer