valkey-cli: ensure output ends with a newline if missing when printing reply

[xbasel]

Example:

   127.0.0.1:6379> multi
   OK
   127.0.0.1:6379(TX)> client info
   QUEUED127.0.0.1:6379(TX)>

Fixes: #1778

[madolson]

I would guess it's more likely related to this code path,

valkey/src/valkey-cli.c

Line 2341 in 221ff44

if (!strcasecmp(command, "info") || !strcasecmp(command, "lolwut") ||

, which explicitly changes the output mode to raw if it's an info command. Probably need to check for multi there.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.01%. Comparing base (221ff44) to head (8d2b037).
Report is 4 commits behind head on unstable.

Additional details and impacted files

@@             Coverage Diff              @@
##           unstable    #1782      +/-   ##
============================================
- Coverage     71.18%   71.01%   -0.17%     
============================================
  Files           123      123              
  Lines         65619    65642      +23     
============================================
- Hits          46712    46618      -94     
- Misses        18907    19024     +117     

Files with missing lines	Coverage Δ
src/valkey-cli.c	54.39% <100.00%> (-1.69%)	⬇️

... and 12 files with indirect coverage changes

[enjoy-binbin]

can we correct the output_raw in here? something like this:

if (config.in_multi) output_raw = 0; /* multi must in xxx mode. */

[xbasel]

Are you suggesting this for readability, or to address a problem?

[madolson]

I assume readability, as it has the same result. I will say I would prefer we add a comment somewhere to include why multi does this. My hot take was going to be let's just do:

if (!config.in_multi) output_raw = 1; /* In a multi block, commands will return status instead of verbatim. */

I don't feel all that strongly though.

[xbasel]

You meant:

if (config.in_multi) output_raw = 0; /* In a multi block, commands will return status instead of verbatim. */

?

I want to skip the big if.
See 61559bf

This function could be more readable; I can refactor it in a separate PR.

[madolson]

I meant my suggestion inside the block of text, so it minimizes the number of lines changed. I'm OK with this though.

[madolson]

I meant my suggestion inside the block of text, so it minimizes the number of lines changed. I'm OK with this though.

[enjoy-binbin]

i am ok with this, but i still prefer the smaller diff way. The logic is pretty simple, we can argue that we correct it after the big if block since it is in multi.

Not a blocker to me.