Feature request: support review comments

[Deng-yongbiao]

Thanks for your contribution, it's a very good job.

I would like to ask about the version of protobuf that generates files corresponding to python from proto files, because I need to use the corresponding c++ source files.

[masaccio]

I use Homebrew on a Mac to handle the binary dependencies and it used 28.2. If you are reasonably familiar with Python and command-line dependencies, you should be able to clone the repo and run the various parts of make bootstrap. This will generate the .proto files using a Python protodump. This is basically:

mkdir -p .bootstrap/protos
poetry run python3 src/build/protodump.py /Applications/Numbers.app .bootstrap/protos
poetry run python3 src/build/rename_proto_files.py .bootstrap/protos
mkdir -p src/numbers_parser/generated
for proto in .bootstrap/protos/*.proto; do \
	    /usr/local/bin/protoc -I=.bootstrap/protos --proto_path .bootstrap/protos --python_out=src/numbers_parser/generated $proto; \
	done

You probably don't need to go as far as using Poetry to run the dump scripts as I don't think they have unusual dependencies.

[Deng-yongbiao]

Thanks for your reply, that's awesome and really help me a lot. It seems that numbers-parser can perfectly support cell operations, but I did not find a way to handle text boxes and some annotation text. Is there any plan to support this?

[masaccio]

Text boxes are likely to be stored the same way as table captions, which I do support but it’s quite new and I suspect it might be a bit flakey. By annotations do you mean comments, or is this a different feature?

[Deng-yongbiao]

Text boxes are likely to be stored the same way as table captions, which I do support but it’s quite new and I suspect it might be a bit flakey. By annotations do you mean comments, or is this a different feature?

I mean some of the comments attached by other people. Can numbers-parser support this feature now?

[masaccio]

I'd need to research that so don't expect anything very quickly. Sometimes data is quite straightforward to extract, sometimes it requires reconstruction of a lot of data structures.