Introduce const_sds for const-content sds

[zuiderkwast]

sds is a typedef of char *.

const sds means char * const, i.e. a const-pointer to non-const content.

More often, you would want const char *, i.e. a pointer to const-content. Until now, it's not possible to express that. This PR adds const_sds which is a pointer to const-content sds.

To get a const-pointer to const-content sds, you can use const const_sds.

In this PR, some uses of const sds are replaced by const_sds. We can use it more later.

Fixes #1542

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 70.94%. Comparing base (d13aad4) to head (2ba610a).
Report is 3 commits behind head on unstable.

Additional details and impacted files

@@            Coverage Diff            @@
##           unstable    #1553   +/-   ##
=========================================
  Coverage     70.94%   70.94%           
=========================================
  Files           120      120           
  Lines         65044    65046    +2     
=========================================
+ Hits          46143    46147    +4     
+ Misses        18901    18899    -2     

Files with missing lines	Coverage Δ
src/sds.c	86.85% <100.00%> (+0.16%)	⬆️
src/sds.h	79.03% <100.00%> (+1.61%)	⬆️
src/server.c	87.64% <100.00%> (+0.03%)	⬆️

... and 11 files with indirect coverage changes

[rjd15372]

LGTM

[SoftlyRaining]

I like this! Looks good to me as-is, but any thoughts on proactively spreading const_sds to other areas of code?

[zuiderkwast]

any thoughts on proactively spreading const_sds to other areas of code

When we have it, it's possible to use it in new code, so we can suggest it in code reviews.

Constness has to be added bottom-up, i.e. start in sds.c with functions like sdslen. A non-const value can always be used as a const value, but not vice versa.

We could add const in many other situations too, not related to sds. I don't think it's worth doing a lot of refactoring though just for that.

[ranshid]

LGTM even though I still think that the const should have been applied on the pointer as well like I commented in #1542
IMO we should not allow "assignment" to a const_sds.

We can always evolve to doing this as well I guess