Format Math Markdown for Pandoc Pipeline

MathJax-delimited source → Pandoc-ready

Source from a MathJax wiki uses \(/\[, which Pandoc ignores as math. The dollar pass converts both so pandoc -o out.pdf typesets them.

SOURCE (Pandoc would pass these through as text):
The area \( A = \pi r^2 \) and the volume
\[ V = \tfrac{4}{3}\pi r^3 \]

TARGET: dollar

PANDOC-READY:
The area $ A = \pi r^2 $ and the volume

$$ V = \tfrac{4}{3}\pi r^3 $$

Already-dollar source needs no change

If the source already uses $/$$, it is Pandoc-ready and the pass leaves it untouched — safe to run as a guard step.

SOURCE (already Pandoc-ready):
Mean $\bar{x}$ and variance:
$$ s^2 = \frac{1}{n-1}\sum (x_i-\bar x)^2 $$

TARGET: dollar

OUTPUT (unchanged):
Mean $\bar{x}$ and variance:
$$ s^2 = \frac{1}{n-1}\sum (x_i-\bar x)^2 $$

Display math promoted to its own block for Pandoc

Pandoc prefers display math isolated. A mid-sentence \[...\] becomes a blank-line-surrounded $$ block.

SOURCE:
The transform is \[ \hat{f}(\xi) = \int f(x)e^{-2\pi i x\xi}\,dx \] in 1D.

TARGET: dollar

PANDOC-READY:
The transform is 
$$ \hat{f}(\xi) = \int f(x)e^{-2\pi i x\xi}\,dx $$
 in 1D.

Out of scope: AMS environment

Pandoc handles \begin{align} via its raw-LaTeX support, but this tool does NOT convert it. The environment passes through; if Pandoc needs delimiters around it, add them manually.

SOURCE:
\begin{align}
a &= b + c \\
d &= e - f
\end{align}

TARGET: dollar

OUTPUT (unchanged — handle in Pandoc directly):
\begin{align}
a &= b + c \\
d &= e - f
\end{align}

Verbatim LaTeX example preserved for the manual

A Pandoc how-to that quotes delimiter syntax inside a fenced block. The fence is skipped, so the documentation example survives the prep.

SOURCE:
Pandoc reads inline math as $...$, e.g.:
```text
Inline: $x^2$   Display: $$y = mx + b$$
```
Live math like \( z \) in prose is converted.

TARGET: dollar

PANDOC-READY:
Pandoc reads inline math as $...$, e.g.:
```text
Inline: $x^2$   Display: $$y = mx + b$$
```
Live math like $ z $ in prose is converted.

Pandoc's default markdown reader does not treat \(...\) or \[...\] as math, so it copies them through as literal text and they never become typeset equations in PDF/DOCX. Running the dollar target first converts them to $/$$, which Pandoc recognizes.

\begin{equation}, \begin{align}, etc. are not among the four recognized delimiter forms, so they pass through untouched. Pandoc can often handle them via raw LaTeX, but this tool will not wrap or transform them — manage those in Pandoc directly.

If copy-paste doubled \\frac to \\\\frac, Pandoc will choke on the bad command even after delimiters are correct. This tool does not unescape backslashes — fix the source or use a find/replace step before the Pandoc run.

Pandoc can mistake a stray $ in prose (a price) as opening math. This tool does not escape literal $ to \$. Escape them yourself, or keep the dollar target (which never inserts new $) and clean currency manually before converting.

Only triple-backtick fences are protected. An inline ` \(x\) shown in a Pandoc tutorial is rewritten. Put quoted delimiter syntax inside ``` fenced blocks to keep it literal in the Pandoc output.

Pandoc supports ~~~ fenced code, but THIS tool's skip logic only recognizes ``` fences. Math inside a ~~~` block will be converted. Use triple-backtick fences for verbatim examples destined for Pandoc.

Choosing the tex target outputs \(/\[, which Pandoc's default reader does NOT treat as math. For Pandoc prep, always use the dollar target. The tex target is for MathJax destinations, not Pandoc.

The tool changes delimiters only; \( x \) becomes $ x $ with inner spaces intact. Pandoc handles inner spacing fine, so this is harmless, but don't expect tidy-up.

Inputs over 500,000 characters on Free are blocked before prep. A book-length manuscript for Pandoc needs Pro (5,000,000) or a pre-split with md-splitter.