christianp / asciimath2tex Goto Github PK

I have the following AsciiMath expression:

P(B|A) = (P(A|B) * P(B))/(P(A))

Which this library converts to:

P \left( B \left \lvert A \right ) = \left( P \frac{A \left \lvert B \right ) \cdot P \left( B \right )}{P \left( A \right )} \right. \right.

Then rendering it via KaTeX I get:

While instead I should get:

I'm no TeX expert, I guess the generated TeX isn't quite right? 🤔

When I'm trying to parse this string:
{\displaystyle \frac{{m}}{{n}},\quad\text{gdzie}\quad{m},{\;n}\in{{{\mathbb{Z}}}}\quad\text{oraz}\quad{n}\ne{0}}

the output is:
\left\lbrace d i s p l a y s t y \le \frac{\left\lbrace m \right \rbrace}{\left\lbrace n \right \rbrace} , \quad \text{gdzie} \quad \left\lbrace m \right \rbrace , \left\lbrace ; n \right \rbrace \in \left\lbrace \left\lbrace \left\lbrace \mathbb{Z} \right \rbrace \right \rbrace \right \rbrace \quad \text{oraz} \quad \left\lbrace n \right \rbrace \ne \left\lbrace 0 \right \rbrace \right \rbrace

Am I doing something wrong, or the string is wrong?

To use asciimath2tex in a TypeScript project, we need type definitions. The following works for me and may be a solution for you too:

In packages.json, specify the type definitions file:

"typings": "type.d.ts",

Add file type.d.ts:

export as namespace asciimath2tex;

class AsciiMathParser {
  constructor();
  parse(str): string;
}

export default AsciiMathParser;

With the type definitions in place, I can do:

import AsciiMathParser from 'asciimath2tex';
...
...
const parser = new AsciiMathParser();
const tex = parser.parse("int_(i=1)^10 x^2/2 dx");

Hello!
I use your parser as an ECMAScript module and I see that I cannot pass an expression that contains ',,' (or simply ',' as the whole expression). The exception occurs:

asciimath2tex.js:455 Uncaught Error: Didn't EOF: 
    at AsciiMathParser.consume (asciimath2tex.js:455)
    at AsciiMathParser.parse (asciimath2tex.js:420)

else if(!this.eof(pos)) {
	throw(new Error("Didn't EOF: ",this.source(pos)));
}

However, if I include the parser as a CommonJS module taking the source code directly from the repository, the expressions are parsed fine, as in asciimath2tex.js the exception is not thrown for the same case. Instead, there are:

else if (!this.eof(t)) {
	var h = this.source(t, t + 1);
	a.push({tex: h, pos: t, ttype: "character"}), (e += h), (t += 1);
}

Could you explain, please, why there is a difference between the scripts and what behavior of the parser is correct?

!in renders as \not\in. There should be similar commands for \not\exists and \not\subset.

In fact, maybe ! before any relation should render as \not. Ideally, this should go in the official AsciiMath too.

Hey I'd like to give this a try, could you publish it to NPM and add a proper license (MIT perhaps)?

I suspect that in the method strip_space the regex should be corrected in /^(\s|\\(?!\\\s))*/. As it is now, the sequence \ is simply stripped, instead of being translated in \, (that is the thin space). The rendering in MathML seem to confirm that this is what should happen, and also MathJax does so. The examples in the tests where \ is used seem on the contrary not correct. Am I right?

It has new Parser(); instead of new AsciiMathParser();

Christian, could you please bump the version number of the npm package? That way the recent packages.json with typing definitions will be installed and the build performed with the types definition file. Working on the npm package is above my pay grade :)

It would be useful to have a CommonJS export, so that this library can be easily imported in CommonJS environments (like Node.js).

Hi all. I'm working on implement LaTeX output support support for asciidoctor/asciimath, which is quite similar to what you guys are doing in here.

As stated in here, KaTeX does not support the \twoheadrightarrowtail macro (neither do vanilla LaTeX, amsmath or amssymb). In fact, I was surprised that I couldn't find a TeX package that supports this specific symbol.

Did you guys have any luck with this? I believe we should exchange information about this issue.

There's another similar implementation here, have you tried? If so could we write a short comparison between the two?

I'm using [email protected]. There seems to be something funky going on if I wrap a matrix in parentheses. I noticed this when trying to mark up a matrix determinant with the following input:

parser.parse('det([[a,b],[c,d]])');

Expected output:

'\det{\left ( \left [ \begin{matrix} a & b \\ c & d \end{matrix} \right ] \right )}'

Actual output:

'\det{\left ( \begin{matrix} \left [ a , b \right ] & \left [ c , d \right ] \end{matrix} \right )}'

It seems that the minimal test case that fails is

parser.parse('([[a],[b]])');

Expected output:

'\left ( \left [ \begin{matrix} a \\ b \end{matrix} \right ] \right )'

Actual output:

'\left ( \begin{matrix} \left [ a \right ] & \left [ b \right ] \end{matrix} \right )'