Add a context= or info= parameter to test() for looped/functional tests #7119
Description
Whenever we need to loop test(), the error becomes hard to determine from the failure itself:
for (ii in 0:4 / 5) test(1 + ii, ii < 0.5)
# Test 1.6 ran without errors but failed check that x equals y:
# > x = ii < 0.5
# First 1 of 1 (type 'logical'):
# [1] FALSE
# > y = TRUE
# First 1 of 1 (type 'logical'):
# [1] TRUE
# 1 element mismatch
# Test 1.8 ran without errors but failed check that x equals y:
# > x = ii < 0.5
# First 1 of 1 (type 'logical'):
# [1] FALSE
# > y = TRUE
# First 1 of 1 (type 'logical'):
# [1] TRUE
# 1 element mismatchIt would be a lot better to be able to insert some string into the test() call to capture the context of the loop, e.g.
for (ii in 0:4 / 5) test(1 + ii, ii < 0.5, context=sprintf("ii=%.1f", ii))Which could then be read directly off of the error.