meta data for SB curated data

[sunbeomso]

Hi,

would you plan to provide meta data for SB curated dataset?

For example, contract_name for each file; there may exist multiple contracts in a single file, but only one contract should be designated as a main(top-level) contract when deployed.

In addition, compiler_version for each file; some analyzers takes this information as input to obtain code that will actually be analyzed (e.g., EVM bytecode or their own intermediate languages). Using a single version of solc may fail when a given file is not compilable with that solc.

[jff]

Thanks for the suggestion. So far, we didn't have the need to have that information condensed in a single file. What sort of format did you have on mind?

If this is implemented, I suggest creating a script under utils/scripts that generates the file.

[sunbeomso]

@jff
I think csv format such as a below one might be good:

filename	main_name	lines	compiler_version
0x627fa62ccbb1c1b04ffaecd72a53e37fc0e17839	TokenBank	100	v0.4.19+commit.c4cbbb05

...