Support steps[*].defer: steps.<key>.defer: for deferring steps in a runbook

[k1LoW]

Support defer: for deferring steps in a runbook.

[github-actions]

BenchmarkManyRunbooks-4

	main (-)	#1119 (c15d75b)	+/-
Number of iterations	1	1	0
Nanoseconds per iteration	3,649,723,834 ns/op	3,666,577,067 ns/op	16,853,233 ns/op
Bytes allocated per iteration	1,382,016,856 B/op	1,382,780,400 B/op	763,544 B/op
Allocs per iteration	18,998,611 allocs/op	19,000,726 allocs/op	2,115 allocs/op

Metadata

	main (-)	#1119 (c15d75b)
goos	linux	linux
goarch	amd64	amd64
pkg	github.com/k1LoW/runn	github.com/k1LoW/runn
cpu	AMD EPYC 7763 64-Core Processor	AMD EPYC 7763 64-Core Processor

BenchmarkOpenAPI3-4

	main (-)	#1119 (c15d75b)	+/-
Number of iterations	1	1	0
Nanoseconds per iteration	5,071,139,633 ns/op	4,904,953,899 ns/op	-166,185,734 ns/op
Bytes allocated per iteration	2,290,114,064 B/op	2,289,635,544 B/op	-478,520 B/op
Allocs per iteration	31,559,087 allocs/op	31,562,032 allocs/op	2,945 allocs/op

Metadata

	main (-)	#1119 (c15d75b)
goos	linux	linux
goarch	amd64	amd64
pkg	github.com/k1LoW/runn	github.com/k1LoW/runn
cpu	AMD EPYC 7763 64-Core Processor	AMD EPYC 7763 64-Core Processor

BenchmarkSingleRunbook-4

	main (-)	#1119 (c15d75b)	+/-
Number of iterations	22	22	0
Nanoseconds per iteration	46,740,542 ns/op	46,116,816 ns/op	-623,726 ns/op
Bytes allocated per iteration	23,176,495 B/op	23,154,610 B/op	-21,885 B/op
Allocs per iteration	183,428 allocs/op	183,412 allocs/op	-16 allocs/op

Metadata

	main (-)	#1119 (c15d75b)
goos	linux	linux
goarch	amd64	amd64
pkg	github.com/k1LoW/runn	github.com/k1LoW/runn
cpu	AMD EPYC 7763 64-Core Processor	AMD EPYC 7763 64-Core Processor

---

Reported by octocov

[github-actions]

Code Metrics Report

	main (e1b024f)	#1119 (c15d75b)	+/-
Coverage	64.5%	64.6%	+0.0%
Code to Test Ratio	1:0.7	1:0.7	+0.0
Test Execution Time	6m39s	6m11s	-28s

Details

  |                     | main (e1b024f) | #1119 (c15d75b) |  +/-  |
  |---------------------|----------------|-----------------|-------|
+ | Coverage            |          64.5% |           64.6% | +0.0% |
  |   Files             |             77 |              77 |     0 |
  |   Lines             |           8958 |            9003 |   +45 |
+ |   Covered           |           5783 |            5821 |   +38 |
+ | Code to Test Ratio  |          1:0.7 |           1:0.7 |  +0.0 |
  |   Code              |          16403 |           16487 |   +84 |
+ |   Test              |          12498 |           12586 |   +88 |
+ | Test Execution Time |          6m39s |           6m11s |  -28s |

Code coverage of files in pull request scope (74.1% → 74.2%)

Files	Coverage	+/-
bind.go	82.2%	+0.2%
cdp.go	76.8%	0.0%
db.go	76.2%	0.0%
dbg.go	4.4%	-0.1%
dump.go	68.9%	-0.7%
exec.go	73.1%	0.0%
grpc.go	77.4%	0.0%
http.go	78.0%	0.0%
include.go	90.6%	+0.1%
operator.go	80.5%	+0.3%
parse.go	74.1%	0.0%
runner_runner.go	63.6%	0.0%
ssh.go	66.1%	0.0%
step.go	91.3%	0.0%
test.go	93.1%	+0.5%

---

Reported by octocov

[katsumi-kato-LITALICO]

I would like to dockhood it and check it out.
I have a feeling that there will be some elements I would like to add to stores with the addition of the defer.
I have a feeling that there will be cases where I will want to evaluate steps that are already stacked in the LIFO of the defer.
There may be cases where an event that requires post-processing occurs several times, but the actual post-processing only needs to be done once.

[katsumi-kato-LITALICO]

Does the defer section support formula expansion?
I haven't come up with a use case that is dynamic.

[k1LoW]

Does the defer section support formula expansion?

No

runn/operator.go

Line 703 in e647346

st.deferred, ok = v.(bool)

I haven't come up with a use case that is dynamic.

Me too 👍