I've got a LaTeX project in a Git repository which is being worked on using both TeX Live (Linux) and MiKTeX (Windows). The project has a very large number of \includegraphics[...]{...}
statements for single files (both PDF and PNG), many of which cannot be combined because they are e.g. embedded in LaTeX tabular
environments. Everything compiles just fine, but my current hosting solution has a set limit for the number of individual files the Git repository can have. In order to keep my project under this limit, is there no way to add add a pre-compilation step which extracts a given archive to a given location? If this project was handled using only Linux, I could e.g. write a makefile which does this and then calls the relevant LaTeX/BibTeX commands, but doing this in a way which also works seamlessly with Windows would be extremely difficult and would probably take more time to implement than simply extracting the archive manually each time.
Target structure
For example, compilation currently depends on the following directory structure:
+-- content
| +-- specimens
| | +-- specimen-1.pdf
| | +-- specimen-2.pdf
| | +-- specimen-3.pdf
| | +-- specimen-4.pdf
| | +-- specimen-5.pdf
| | +-- specimen-6.pdf
| | +-- ...
| +-- experiment-setup.png
| +-- stimuli.png
+-- mybib.bib
+-- cool-experiment.tex
+-- fun-conference.sty
Source structure
The following structure is how it currently looks in the Git repository in order to keep the project under the file count limit:
+-- content
| +-- experiment-setup.png
| +-- specimens.zip
| +-- stimuli.png
+-- mybib.bib
+-- cool-experiment.tex
+-- fun-conference.sty