Fix for panic when submitting non-existent version for job history CLI command

[chelseakomlo]

Fixes a panic where running nomad job history -full -version=500 example would result in:

panic: runtime error: invalid memory address or nil pointer dereference                                  
[signal SIGSEGV: segmentation violation code=0x1 addr=0x100 pc=0x15834a5]                                

goroutine 1 [running]:    
github.com/hashicorp/nomad/command.(*JobHistoryCommand).formatJobVersion(0xc420319a20, 0x0, 0x0, 0x0, 0x1, 0x0, 0x0)
        /opt/gopath/src/github.com/hashicorp/nomad/command/job_history.go:252 +0x45                      
github.com/hashicorp/nomad/command.(*JobHistoryCommand).Run(0xc420319a20, 0xc4200c6030, 0x3, 0x3, 0xc4200a9000)
        /opt/gopath/src/github.com/hashicorp/nomad/command/job_history.go:185 +0xc1d                     
github.com/hashicorp/nomad/vendor/github.com/mitchellh/cli.(*CLI).Run(0xc4202f83c0, 0xc4202f83c0, 0xc42026d4e0, 0x10)
        /opt/gopath/src/github.com/hashicorp/nomad/vendor/github.com/mitchellh/cli/cli.go:255 +0x1eb     
main.RunCustom(0xc4200c6010, 0x5, 0x5, 0xc4204d3f68)                                                     
        /opt/gopath/src/github.com/hashicorp/nomad/main.go:127 +0x4b4                                    
main.Run(0xc4200c6010, 0x5, 0x5, 0xc420098058)      
        /opt/gopath/src/github.com/hashicorp/nomad/main.go:72 +0x3f                                      
main.main()               
        /opt/gopath/src/github.com/hashicorp/nomad/main.go:68 +0x63 

Fixes #4542

[nickethier]

I wonder if we should be doing a nil check on the job from the caller as well? Seems like a caller error if it's wanting to format a nil job.

I'm ok with this getting merged though.

[preetapan]

could you add a test case for this in job_history_test.go?

[chelseakomlo]

There are not any success case tests for job_history or similar commands such as job_inspect or job_revert- I tried a couple different test setups but wasn't able to get a successful test passing for this. I'm going to merge as is but we can look at what a success test case could look like.

[preetapan]

would be nice to verify with a unit test, otherwise LGTM

[Jeff-Hanson]

I know this is merged already but was wondering if it would have been better to put the nil check on line 173? That is the first place you know job is nil. That would avoid possibly passing a nil job to the Format call on line 175.

[github-actions]

I'm going to lock this pull request because it has been closed for 120 days ⏳. This helps our maintainers find and focus on the active contributions.
If you have found a problem that seems related to this change, please open a new issue and complete the issue template so we can capture all the details necessary to investigate further.