🐛 FIX: changes in source file for pdf build

[AakashGfude]

No description provided.

[github-actions]

🚀 Deployed on https://63ca017673a2ea7dfd1d0cae--vibrant-allen-f39ec9.netlify.app

[mmcky]

@AakashGfude the table of content and contents links don't seem to be clickable on the preview pdf

[AakashGfude]

@AakashGfude the table of content and contents links don't seem to be clickable on the preview pdf

ooh no !! I will rebuild and see, else need to debug.

[AakashGfude]

It looks okay when building locally. Wonder what the reason is.

[AakashGfude]

The CI is using sphinxv4 to build and is throwing errors because of this issue: QuantEcon/quantecon-book-theme#172 .
Will find a concrete solution for that issue and then merge with updates.

[mmcky]

@AakashGfude just checking in on this PR. Is the upstream issue is now resolved.

[mmcky]

@jlperla these are some small changes to get PDF builds to work on Julia. Once you update the Julia side of things for 1.8 we can loop around on this one and get PDF builds up and running for whole of site PDF builds.

[mmcky]

There is an issue with Jupyter-server that needs investigating but will remove this restriction once that is resolved.

jupyter-book/jupyter-book#1902

[mmcky]

@jlperla this is the latest pdf now that this branch is running.

quantecon-julia.pdf

@AakashGfude I noticed that the table of contents is missing from the pdf. Any ideas?

[mmcky]

@AakashGfude I have identified the issue.

It is from tools_and_techniques/geom_series.md

```{code-cell} julia
# Creates algebraic symbols that can be used in an algebraic expression
@variables g, r, x0
G = (1 + g)
R = (1 + r)
p0 = x0 / (1 - G * R ^ (-1))
print("Our formula is")
p0
```

When building the Julia lectures if there is generated math from code the ipynb json blob is

   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Our formula is"
     ]
    },
    {
     "data": {
      "text/latex": [
       "$$ \\begin{equation}\n",
       "\\frac{x0}{1 + \\frac{-1 - g}{1 + r}}\n",
       "\\end{equation}\n",
       " $$"
      ],
      "text/plain": [
       "x0 / (1 + (-1 - g) / (1 + r))"
      ]
     },

which gets wrapped in tex as

\begin{sphinxuseclass}{cell_output}
\begin{sphinxVerbatim}[commandchars=\\\{\}]
Our formula is
\end{sphinxVerbatim}
\begin{equation*}
\begin{split} \begin{equation}
\frac{x0}{1 + \frac{-1 - g}{1 + r}}
\end{equation}
 \end{split}
\end{equation*}
\end{sphinxuseclass}\end{sphinxVerbatimOutput}

and xelatex really does not like nested equation environments.

Is this something we could fix in sphinx-jupyterbook-latex?

maybe for Julia we should have nowrap for all math blocks that are generated by code

[mmcky]

sphinx / Jupyter-book must be interpreting the $$ that is found in returned latex objects in Julia

[mmcky]

This could be fixed using LaTexify which:

using Latexify
latexify(p0)

which returns in the json blobs

     "data": {
      "text/latex": [
       "\\begin{equation}\n",
       "\\frac{x0}{1 + \\frac{-1 - g}{1 + r}}\n",
       "\\end{equation}\n"

returns LaTeX and no markdown wrappers or delimiters

@jlperla do you know any other settings in Julia Symbolics that might turn off the markdown math delimiters?

[mmcky]

So the latexify will return the equation but it is wrapped in \begin{equation} such as:

      "text/latex": [
       "\\begin{equation}\n",
       "\\frac{x0}{1 + \\frac{-1 - g}{1 + r}}\n",
       "\\end{equation}\n"
      ],

so we still get double math environments. Will check latexify docs to see if we can not wrap the equation.

[mmcky]

@AakashGfude SymPy in the python world returns

...
      "text/latex": [
       "$\\displaystyle \\cos{\\left(x \\right)} + 1$"
      ],
      "text/plain": [
       "cos(x) + 1"
      ]

which also includes $ delimiters so I'm now not sure why we are getting double equation environments in the latex writer

[mmcky]

@jlperla the 7a9d2c7 should fix the issue in the short term but it isn't pretty. Let me know what you think?

[jlperla]

Looks good to me! Thanks for finding a workaround