Use empty history if search query fails

[ysthakur]

There's a bunch of places with .expect("todo: error handling"), and this PR tries to remove most of them. Here's the four different kinds:

• There are a couple inside examples/demo.rs that I didn't bother with.
• Some of these are for unwrapping the result of running a search query to look through the history. I figured that instead of panicking, it may be better to silently treat Errs as the history being empty.
• One of them comes after history.save(entry), in engine.rs. This has been edited to silently ignore an Err.
• The rest of the .expects are all inside engine.rs and are for handling the result of moving the history_cursor forward or backward. Unfortunately, all of them appear inside methods that return either () or an io::Result. These have also been edited to silently ignore Errs.

A disadvantage of silently handling Errs is that users won't know that something went wrong, but if the .expects are to be kept, they should at least be changed to have more useful messages to help users file bug reports.

There shouldn't be any user-facing changes from this.

[codecov]

Codecov Report

Attention: 14 lines in your changes are missing coverage. Please review.

Comparison is base (f396223) 49.26% compared to head (0dd3f7d) 49.34%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #690      +/-   ##
==========================================
+ Coverage   49.26%   49.34%   +0.08%     
==========================================
  Files          46       46              
  Lines        7911     7897      -14     
==========================================
  Hits         3897     3897              
+ Misses       4014     4000      -14     

Files	Coverage Δ
src/hinter/default.rs	0.00% <0.00%> (ø)
src/completion/history.rs	0.00% <0.00%> (ø)
src/engine.rs	5.06% <0.00%> (+0.06%)	⬆️

[amtoine]

silent errors sound worse to me than panicking 🤔
would be nice to give a proper error to the users

[ysthakur]

@amtoine Hmm, I guess so. I'm going to try fuzzing later to try to find cases that trigger panics, and if I can't find anything, I'll just change the expect messages to be more descriptive

[ysthakur]

I did some fuzzing in this branch and couldn't find any real defects, so I guess I'll just change the error messages to be more descriptive, there's probably no need to do actual error handling.

[ysthakur]

Actually, I have no idea what kind of message would be helpful there, so I think I'll just close this for now, lol

[fdncred]

darn, i was hoping this pr would get rid of some of the non-error-handling we have. maybe in the future sometime?