Refactoring keg_relocate to use ELFShim#interpreter

[rmNULL]

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same change?
• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your changes? Here's an example.
• Have you successfully run brew style with your changes locally?
• Have you successfully run brew tests with your changes locally?

---

Replaces all uses of ELFShim#with_interpreter? and removes it.
Introduce ELFShim#interpreter to return the interpreter in path
if present otherwise return nil.

@sjackman @woodruffw @MikeMcQuaid
there are style issues and concerns parsing interpreter from the output of shell command,
these are discussed in review comments.

[woodruffw]

LGTM!

[sjackman]

This PR drops support for getting the interpreter using readelf. I'm okay with that though.

$ readelf -l /bin/ls
…
  INTERP         0x0000000000000238 0x0000000000400238 0x0000000000400238
                 0x000000000000001c 0x000000000000001c  R      1
      [Requesting program interpreter: /lib64/ld-linux-x86-64.so.2]

[rmNULL]

would we need tests for this method ?

[MikeMcQuaid]

This PR drops support for getting the interpreter using readelf. I'm okay with that though.

How often is this being used by users? Can we figure this out and how disruptive it'll be before removing the code? There's not much code there and it seems weird to drop it unless it's unused/causing problems.

[sjackman]

Nearly there!

[sjackman]

This PR drops support for getting the interpreter using readelf. I'm okay with that though.

How often is this being used by users? Can we figure this out and how disruptive it'll be before removing the code? There's not much code there and it seems weird to drop it unless it's unused/causing problems.

I believe this code is never currently used, because this interpreter method is called only by keg_relocate, which guarantees that patchelf is installed, and so patchelf will be used instead of readelf. That said, it's easy enough to implement readelf, so may as well.

@rmNULL Parse this line using a regex.

$ readelf -l /bin/ls
…
 [Requesting program interpreter: /lib64/ld-linux-x86-64.so.2]
…

[rmNULL]

This PR drops support for getting the interpreter using readelf. I'm okay with that though.

How often is this being used by users? Can we figure this out and how disruptive it'll be before removing the code? There's not much code there and it seems weird to drop it unless it's unused/causing problems.

@MikeMcQuaid
there was no much code because it was just checking for presence of interpreter section.
however here, we have to parse that text from the output, which is a little arguable more code.

Another case, (as of now) this method is called only once(throughout the whole source),
and patchelf is guaranteed to exist when that call is made.
patchelf was a requirement before the change too, in that way we are not dropping any support.

[MikeMcQuaid]

Another case, (as of now) this method is called only once(throughout the whole source),
and patchelf is guaranteed to exist when that call is made.
patchelf was a requirement before the change too, in that way we are not dropping any support.

Ok, fine with me to remove that code!

[woodruffw]

Thanks, @rmNULL 🎉