fetch_overleaf_report.RdCopies the text of the report from your Overleaf project into a folder in your working directory, so that you - or an AI assistant such as Claude Code - can read the latest version of the report alongside your analysis. The copy is read-only and is never sent back to Overleaf.
fetch_overleaf_report(
dir = "report_cache",
main = "Report.tex",
chapters = "chapters"
)Folder to put the copy in, relative to your working directory.
Defaults to "report_cache". It is replaced each time
fetch_overleaf_report() runs.
The report's main file. Defaults to "Report.tex". If
there's no such file, the top-level .tex file containing
\documentclass is used instead.
Folder holding the chapter files. Defaults to
"chapters".
The path of the folder holding the copy, invisibly.
The copy holds:
the main file;
every .tex file in the chapters folder;
the bibliography (every .bib file in the project, such as
those in the bib folder);
every .tex file in atlas, such as the statistics
written by write_overleaf_tex.
Charts and other images aren't copied.
Each run fetches the latest version from Overleaf and replaces the whole
folder, so a chapter deleted in Overleaf disappears from the copy too.
The files are marked read-only, and the folder holds a .gitignore
so it's never committed if your analysis is in a git repository.
Nothing in the copy is connected to Overleaf:
fetch_overleaf_report() never commits or pushes anything, and
edits to the copy are lost the next time it runs.
For a project that syncs over git, this keeps a second copy of the
project in the package's cache, separate from the one used to send your
charts (see overleaf_cache_dir). The first run downloads it,
so is slower than later ones.
set_overleaf_project to choose the project
if (FALSE) { # \dontrun{
set_overleaf_project("https://www.overleaf.com/project/0123456789abcdef01234567")
fetch_overleaf_report()
} # }