Highlighting/parsing mid-paragraph display math

[CoryMcCartan]

Most of the time I use display math, it's mid-paragraph, and I'd like to continue the prose after the display without starting a new paragraph. In source mode, I would usually write like this:

We study the following partially linear structural model for an outcome $Y$, binary 
treatment $Z$, vector of covariates $X$, unobserved confounder $U$, and two 
additional variables $W_Y$ and $W_Z$.
$$
\begin{aligned}
Y &= \beta_u U + \beta_{w_y} W_Y + \beta_{w_z} W_Z + \tau Z + f_y(X) + \eps_y \\
Z &= \gamma_u U + \gamma_{w_y} W_Y + \gamma_{w_z} W_Z + f_z(X) + \eps_z \\
W_Y &= \alpha_u U + \alpha_{w_z} W_Z + f_{w_y}(X) + \eps_{w_y} \\
W_Z &=\varphi_u U + f_{w_z}(X) + \eps_{w_z}
\end{aligned}
$$ {#eq-struct} 
Above, $\eps_y, \eps_z, \eps_{w_z}, \eps_{w_y}$, and $\eps_u$ are independent noise 
terms, and $f_y$, $f_z$, $f_{w_y}$, and $f_{w_z}$ are generic functions of the covariates $X$.

This renders properly. However, in Positron it is highlighted as follows, where the sentence following the math is highlighted as an equation (using sentence wrapping):

Visual mode wants to reformat the equation like this:

We study the following partially linear structural model for an outcome $Y$, binary 
treatment $Z$, vector of covariates $X$, unobserved confounder $U$, and two 
additional variables $W_Y$ and $W_Z$. $$
\begin{aligned}
Y &= \beta_u U + \beta_{w_y} W_Y + \beta_{w_z} W_Z + \tau Z + f_y(X) + \eps_y \\
Z &= \gamma_u U + \gamma_{w_y} W_Y + \gamma_{w_z} W_Z + f_z(X) + \eps_z \\
W_Y &= \alpha_u U + \alpha_{w_z} W_Z + f_{w_y}(X) + \eps_{w_y} \\
W_Z &=\varphi_u U + f_{w_z}(X) + \eps_{w_z}
\end{aligned}
$$ {#eq-struct} Above, $\eps_y, \eps_z, \eps_{w_z}, \eps_{w_y}$, and $\eps_u$ are independent noise 
terms, and $f_y$, $f_z$, $f_{w_y}$, and $f_{w_z}$ are generic functions of the covariates $X$.

While this also renders properly, I find this much less readable, and it also breaks the document parsing. Positron now recognizes the text after the equation as the equation. This also causes all downstream headings to not be recognized in the Outline.

I basically can't use Visual mode for this reason. For very short display equations, its preferred inline format (text $$1 + 1 = 2$$ more text) is readable, but also not highlighted/recognized by Positron. But for multiline equations, it reformats the source in a way that breaks the highlighting/parsing, and in my opinion is less readable.

My questions are:

1. Is this the proper syntax for mid-paragraph display equations? Most of the Quarto documentation uses visual mode, and I can't find a multiline equation example.
2. Would it be possible to fix the parsing/highlighting here, and/or change visual mode to standardize syntax in a way that works with the parsing/highlighting? Happy to create an issue with smaller reprex if that's the case

[juliasilge]

Thank you so much for this report! This functionality actually comes from the VS Code Quarto extension (both the parsing as well as the whole visual editor). Would you be up for posting this on the repo for that extension?

I think the problem with identifying the headings and chunks is related to quarto-dev/quarto#521 and also quarto-dev/quarto#302; notice that quarto-dev/quarto#302 (comment) recognizes that editing in visual mode introduces the parsing problem.

[CoryMcCartan]

OK, I'd searched the quarto repo but somehow missed those issues, thanks!

I will post as an issue in the extension repo.