grep: Make exercise schema-compliant #712
Conversation
Move the 'files' object representing the three files that must be present when running the tests to the comments, because the new JSON schema doesn't allow shared data among test cases. This is the only remaining exercise having a shared data structure, and there are no plans at the moment to add support for this feature, so are we moving the data to preserve the information and allow compliance with the new standard.
|
Updated to fix some minor formatting issues. |
There was a problem hiding this comment.
all commits do what they claim to do
I wonder if any language track will complain that it's too hard to extract the files from the comments (find lines surrounded by |, strip trailing newlines), but until someone comes and argues for it I can't really argue on their behalf.
I was not even considering the possibility of algorithmically extracting the lyrics from the comments, @petertseng. 😄 I think this PR is just an immediate solution for preserving the lyrics and allow compliance. I would really like to see a rewriting of this exercise allowing more automatic test generation, but waiting for it would postpone #625, and I think that it would be good to have #658 merged soon to enforce compliance in Travis-CI. |
|
Found out I missed a word:
but until someone comes and argues |
Change: Test for incorrect greedy algorithm
This exercise has 3 text files encoded in the
canonical-data.json, which is incompatible with the new JSON schema.We are discussing in #699 what is the best way to deal with sharing test data among test cases, and it seems there isn't much support for adding this feature to the schema.
Anyway, that discussion doesn't need to suspend the migration of the current test data.
I proposed here to simply move the encoded files to the comments and proceed with the migration.
Related to #625.