Convert i/rz-bin calls inside Rizin to the API calls

[XVilka]

librz/main/rz-find.c
409:		rz_sys_cmdf("rizin"
421:			rz_sys_cmdf("rizin -qc \"/E %s\" \"%s\"", kw, efile);

librz/egg/egg_Cfile.c
289:    output = rz_sys_cmd_strf("rz-bin -o '%s.text' -O d/S/'%s' '%s'.o",

librz/core/cmd_debug.c
1673:                           res = rz_sys_cmd_strf("env RZ_BIN_PREFIX=\"%s\" rz-bin %s-B 0x%08" PFMT64x " -S \"%s\" | grep \"%s\"", name, mode, baddr, filesc, sect);
1676:                           res = rz_sys_cmd_strf("env RZ_BIN_PREFIX=\"%s\" rz-bin %s-B 0x%08" PFMT64x " -S \"%s\"", name, mode, baddr, filesc);

Originally posted by @XVilka in #1238 (comment)

I have removed the following: ``` librz/core/cmd_info.c 703: rz_sys_cmdf("rz-bin -O \"%s\" \"%s\"", rz_str_trim_head_ro(input + 1), desc->name); 706: rz_sys_cmdf("rz-bin -O help"); ```

because those are not i commands.

[ret2libc]

I didn't see any work on this during this time, so I doubt it will be done for 0.4.0.
I'm removing the 0.4.0 milestone for now until someone starts to actively looking at this.

cc @XVilka

[XVilka]

Only partially resolved.

[Rot127]

Fixing the cmd_debug.c case it not possible easily.

The issue is fundamentally that Core and Bin are not separated well enough.
RzCoreBin implements the whole printing of sections and other stuff.
The RzCoreBin API always requires a RzCore pointer passed and uses
'core->io', 'core->bin', 'core->hash' everywhere.

RzCore is built with the assumption, that the binary registered with RzCore
(via core->io, core->bin etc.) is the only binary in the whole universe of interest.

But the command dmS should not just list the sections of the debugged binary,
but also of the libraries dynamically loaded.
So this assumption becomes wrong.

We can't just close the current binary and open the library file,
just for listing sections. Because we are in live debug mode.

The current dmS implementation avoids this by simply calling rz-bin
and indirectly spawning a RzCore in a new process for each binary we want the sections of.

Creating a new RzCore for each binary of interest is not possible,
because RzCore is not thread save.

In short: To fix this RzCoreBin needs to be separated properly.
Which is a little bigger task.

[Rot127]

Same problem for egg_Cfile.c unfortunately.

The other one is a search. It makes more sense to refactor the search before removing it.

[Rot127]

So the cmd_debug.c could be resolved apparently. I found a case where they simply backed up core->bin, set it to the library object, read symbols and other stuff and then set it again to the currently debugged binary. I guess it works. But who knows if this has consequences (e.g. due to a state change in RzIO). So it could be fixed, but wouldn't feel confortable with it.
If we do this hacky solution now, we maybe forget fixing it properly in the future? Would you agree?