Rendering bug about asciimathml

[pkra]

Not quite a rendering but rather a conversion issue -- the generated MathML is

    <math>
      <mstyle mathcolor="blue" fontfamily="serif" displaystyle="true">
        <mrow>
          <mo>{</mo>
          <mi>w</mi>
          <mrow>
            <mo>|</mo>
            <mi>w</mi>
            <mo>&isin;</mo>
            <msup>
              <mrow>
                <mo>{</mo>
                <mi>a</mi>
                <mo>,</mo>
                <mi>b</mi>
                <mo>}</mo>
              </mrow>
              <mo>+</mo>
            </msup>
            <mo>,</mo>
            <mo>|</mo>
          </mrow>
          <mi>w</mi>
          <mrow>
            <mo>|</mo>

            <mo>_</mo>

            <mi>a</mi>
            <mo>=</mo>
            <mo>|</mo>
          </mrow>
          <mi>w</mi>
          <msub>
            <mrow>
              <mo>|</mo>
            </mrow>
            <mi>b</mi>
          </msub>
          <mo>}</mo>
        </mrow>
      </mstyle>
    </math>

which seems wrong. The matching of the | seems to be the problem.

[dpvc]

Peter is right, the problem is that AsciiMath matches the | that you are using as "such that" with the first one for the absolute values.

One way to prevent that is to put the initial | into invisible delimiters, so that it won't match anything. E.g.

{w {:|:} w in {a, b}^+, |w|_a = |w|_b}

[drlippman]

Yeah, sadly the absolute value / "such that" handling in AsciiMath is a bit ambiguous, due to the ambiguous nature of the notation itself.  I've been thinking about it off and on for years, and I still haven't come up with any ideas for how to better parse an expression like that.

Recently we added functionality that causes abs(x) to be rendered as |x|, and that provides a way to avoid the mismatched symbol issue.  But that change hasn't made it into MathJax yet.

[pkra]

@drlippman you wrote

Recently we added functionality that causes abs(x) to be rendered as |x|, and that provides a way to avoid the mismatched symbol issue. But that change hasn't made it into MathJax yet.

Ah! When did that happen? We should think about a workflow to pull in such updates. Darn, we could have done that with 2.4 recently...

[drlippman]

It happened in March.  Peter made several changes.  There are a couple others I still haven't gotten around to committing. Hopefully we can figure a process and get them into 2.5 :)

[pkra]

Thanks, @drlippman . I've created an issue on MathJax's issue tracker to keep track of this. Moving forward, it might make sense to add release tags to this repository and simply post an issue on our thread with each release. It would also be very helpful to add tests to the MathJax test suite for new features and bug fixes.