I'm honestly torn here. I think displaying un-obfuscated method names is valuable, however, my work has seen few samples where this is the case. More frequently, I see large, obfuscated names that would be a pain to handle (display) correctly.
We could take a route simliar to de4dot by attempting to identify and rename obfuscated method names but I see that potentially confusing users who pivot from capa to a tool like dnSpy.
Reading your original message I see you say "when possible" which could be a valid path forward, e.g., we only display method names when we can identify the name has not been obfuscated at all (or past an established threshold).
I agree with everything you say here. I'd only recommend including the name when its useful and helpful. I do not think we should do any renaming or show any obfuscated names. So, if its reasonably easy to determine if a name is non-obfuscated, then I think we should maybe do that.
How to determine if non-obfuscated? I guess ASCII only, something around entropy, not mixing casing or numbers too weirdly, ...?
We should probably move this into a separate feature request.
Originally posted by @williballenthin in #1326 (comment)
I agree with everything you say here. I'd only recommend including the name when its useful and helpful. I do not think we should do any renaming or show any obfuscated names. So, if its reasonably easy to determine if a name is non-obfuscated, then I think we should maybe do that.
How to determine if non-obfuscated? I guess ASCII only, something around entropy, not mixing casing or numbers too weirdly, ...?
We should probably move this into a separate feature request.
Originally posted by @williballenthin in #1326 (comment)