Diffing tables doesn't work when delta span multiple lines

[gphilipp]

We noticed this rather annoying issue :

Diffing

| Aslak | aslak@email.com | 123 |
| Joe   | joe@email.com   | 234 |
| Bryan | bryan@email.org | 456 |
| Ni    | ni@email.com    | 654 |

with

| Aslak | aslak@email.com | 123 |
| Ni    | ni@email.com    | 654 |

will return the following delta :

Tables were not identical:
      | Aslak | aslak@email.com | 123 |
    - | Joe   | joe@email.com   | 234 |
    - | Bryan | bryan@email.org | 456 |
      | Bryan | bryan@email.org | 456 |
      | Ni    | ni@email.com    | 654 |

Notice that the Bryan line is listed twice.

I want to tackle the issue but would like you to confirm this is unexpected ?

[gphilipp]

The culprit is the TableDiffer#createTableDiff method, the loop counter doesn't take into account the "size" of the delta. I'd like to update this PR with a fix but I can't manage to build cucumber-jvm : it fails the tests within .SubstitutionStepdefs.I_should_receive_an_email_with_the_body (even on v1.1.2).

[gphilipp]

Ok I figured out the build issue. Adding a mental note to self to not set core.autocrlf to true ever again. Working on a fix for this issue now.

[gphilipp]

Not passing yet, it fails the out of memory test...

[gphilipp]

Ok actually it's not failing because of an OutMemoryError, but rather because of a wrong expectation (in my opinion) in TableDifferTest.should_not_fail_with_out_of_memor(). The expected diff should not be :

+ | I just woke up |

but :

 + | I just woke up    |
   | I'm going to work |

I'll update this PR with the latest commit which makes all tests pass, including 3 new ones that I wrote to cover multiline deltas.

[aslakhellesoy]

Merged. Great patch!

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.