Hi, I'm writing a docutils Transform for the Breathe extension for Sphinx and struggling as I would like to collect information across all the files and then use sum total info to transform each file. From what I can tell the Transforms seems to be run on one file at a time without any awareness of the other files. I guess this is by design from docutils but can anyone suggest other approaches here which might help?
I'm essentially looking to extract external filenames out of each source file and then process all the files in one big chuck and then use the results of that processing to transform the source files.
I was hoping there might be a "read-all-source-files" type event where I could go in and use the information I've gathered to do some doctree tweaking but there only seems to be per-file events at the moment.