Enhance fold labels on folded beamer frames

[bk90]

Hello,

this PR enhances (IMHO) the labels on folded frames in LaTeX beamer documents.

1. The first commit changes the regex parsing comments on the \begin{frame} line. So far the comments where only found if they were separated from the \begin{frame} only by whitespace. Now optional [...] are allowed, as in \begin{frame}[noframenumbering] % Title page.
2. The second commit extends the search for \frametitle() by looking for \framesubtitle() on the following line and, if both are found, displaying them on the fold as frametitle: framesubtitle.
3. The third commit to me is optional due to commit (1). It changes the regex looking for frametitle, so that commented out frametitles are found as well. This originally was a workaround for the first commit, but still could be useful.

[lervag]

Thanks. Could you provide a couple of complete examples for the various cases?

[bk90]

Sure. Given this code

\begin{frame}[noframenumbering]  % Title page                                    
  \titlepage                                                                     
\end{frame}

\begin{frame}
  \frametitle{Title}
  \framesubtitle{Subtitle}
\end{frame}

\begin{frame}
  % \frametitle{Title}
\end{frame}

I get this folding:

\begin{frame}       Title page·····················································

\begin{frame}       Title: Subtitle·················································

\begin{frame}       Title··························································

Without the changes the folding is

\begin{frame}       ·······························································

\begin{frame}       Title···························································

\begin{frame}       ·······························································

The first example changes, because the regex now allows for [.*] in between \begin{frame} and the comment.

The second example changes, because the code now searches and extracts a \framesubtitle{Subtitle} in the line following \framtitle{Title}. This could be improved by searching not only the following line but all following lines. Do you think that's necessary?

The third example changes, because commented out \frametitle{Title} commands are parsed as well. I don't know if this is desired, what do you think?

[lervag]

Sure. Given this code ...

Thanks for the good examples and before vs after explanations.

The second example changes, because the code now searches and extracts a \framesubtitle{Subtitle} in the line following \framtitle{Title}. This could be improved by searching not only the following line but all following lines. Do you think that's necessary?

It seems possible, but I think this should already suffice for a useful improvement. If you want to look into even more advanced parsing, I'll be glad to consider an updated PR or a future PR for that!

The third example changes, because commented out \frametitle{Title} commands are parsed as well. I don't know if this is desired, what do you think?

I'm not sure. In a way, I think we should ignore the commented out title. It does not feel light the right way to control the fold text. Do you have a strong opinion about this?

[lervag]

I would simplify this to:

let frametitle .= ': ' . matchstr(getline(i+1),
      \ '^.*\\framesubtitle\(\[.*\]\)\?{\zs.\{-1,}\ze\(}\s*\)\?$')

[bk90]

I tried this and as far as I can tell this adds : 0 to the fold text if no subtitle is available, so the if condition seems necessary. But I could remove the extra variable and return statement, if you prefer.

[lervag]

I've looked through the code, and I think it is decent. I have two comments, and we'll have to decide about the commented out frametitles. I also want to add tests, but I can do that myself. If you want to try, look into test/tests/test-folding for examples of how to do it.

[lervag]

I would change \(\[.*\]\)\? to \(\[[^\]]*\]\)\?.

[bk90]

I can't get it to work with the suggestion. What would be the improvement? I copied the regex for detecting the brackets from the \frametitle regex a few lines above this.

[bk90]

The third example changes, because commented out \frametitle{Title} commands are parsed as well. I don't know if this is desired, what do you think?

I'm not sure. In a way, I think we should ignore the commented out title. It does not feel light the right way to control the fold text. Do you have a strong opinion about this?

I removed the code for the commented frame titles. I guess if someone comments the frame title its for a good reason, so we should respect this and not show it on the fold. If the title should be shown it can be added as a comment to the\begin{frame} line.

I also want to add tests, but I can do that myself. If you want to try, look into test/tests/test-folding for examples of how to do it.

I added a test, but have some trouble with it. The way it's currently set up the test breaks on recognizing the title comment on line 204. If I switch the order of the two added beamer frames and update the test accordingly (line 204 -> 211, 208 -> 204) the test works. Do you have any idea why? As far as I can tell the order of the frames should not have any effect. When I open the latex file in vim the fold texts are shown as expected either way.

[lervag]

I removed the code for the commented frame titles. I guess if someone comments the frame title its for a good reason, so we should respect this and not show it on the fold. If the title should be shown it can be added as a comment to the\begin{frame} line.

Great!

I also want to add tests, but I can do that myself. If you want to try, look into test/tests/test-folding for examples of how to do it.

I added a test, but have some trouble with it. The way it's currently set up the test breaks on recognizing the title comment on line 204. If I switch the order of the two added beamer frames and update the test accordingly (line 204 -> 211, 208 -> 204) the test works. Do you have any idea why? As far as I can tell the order of the frames should not have any effect. When I open the latex file in vim the fold texts are shown as expected either way.

Huh, this is very strange. For some reason, the v:foldend and v:foldlevel variables are not correct when we call foldtextresult(xxx). I looked into this a bit now and could not find anything wrong with your code, and I suspect it might actually be a bug in Vim. Not fully sure, though. The behaviour is consistent between Vim and neovim.

But I'll accept this regardless; let's just swap the tests to make them pass and add a comment (e.g. refer to the issue tread).

Nice work, thanks! If you are happy with this, then I can probably merge this in a day or two.

[bk90]

The test is updated and I think the PR is ready to be merged. Thank you for the great plugin, it has made my life easier for several years now. I'm glad I could make a small contribution :)

[lervag]

The test is updated and I think the PR is ready to be merged.

Merging now.

Thank you for the great plugin, it has made my life easier for several years now. I'm glad I could make a small contribution :)

I'm happy to hear it is useful to you! And thanks for your contribution, it is appreciated!