Commit
8a33777063862817f6f9b06eedc39d326a95ca25
by Michael Brooks
Fix analysis error in string-input testing.
Add comments correlating .in file content with line of code that reads it.
This test's correctness expectation is bootstrapped from the actual behaviour of scanf. A mistake in knowing "which character are we on at this point" led to an inaccurate understanding of what scanf does on an edge case. The test was showing that the CFA analogs behave the same as scanf. But the test was not exercising the case that it suggested it was. This issue is now fixed.
Old position, mistaken: case 14 begins on line ccccuuuucccc Old position, corrected: cases 12/13 consume leading cccc; case 14 begins on uuuucccc New positions: as commented in test .cfa
prior scanf understanding, mistaken: include skips unwanted characters before capturing wanted characters, while exclude fails on unwanted characters
scanf understanding, corrected: include and exclude fail on unwanted characters