[texhax] Any crazy math formulas for testing a TeX language interpreter

[Karl Berry]

Doug - just ideas, sorry for not doing any research:

- maybe MathJax or Wikipedia have test suites of TeX-math-syntax-input
you could steal.  Well, it would probably be LaTeX syntax, technically,
but should be simple enough to convert in that particular case.

- Perhaps the AMS has some kind of math testing documents from amstex
development, that would then depend on plain tex (+amstex), but not latex.
(Their CTAN/macros/latex/required/amslatex/math/testmath.tex document
does use a fair amount of LaTeX.)  Barbara?

- Perhaps in the course of developing the OpenType math fonts, either
STIX or GUST or anyone made test documents that could be adapted.
Barbara, Khaled, Jerzy, Apostolos, anyone?

- It's true there are examples of most math constructs in the TeX for
the Impatient book, as Justin suggests, and we used (effectively) plain
TeX for it.  Not very much real math, though.
CTAN/info/impatient/math.tex

- Justin's other suggestions also seem sound to me, perhaps also going
back to Knuth's original mf.web->mf.tex.  The MF program has far more
math than TeX, as you probably know.

- However, I fear nothing extant is going to explicitly test all the
myriad features of TeX math typesetting in the way a test suite "should"
(= separately, with known-good results and variations).

Best,
Karl