Verify that tests actually implement the canonical tests

[fapdash]

It seems that some of the tests don't implement the tests that they claim to implement in their meta/tests.toml, see #228.
Also some exercises don't implement all canonical tests according to the toml.
Find out which ones with configlet sync --tests or configlet sync --tests -e testname
We should go through all the exercises and verify.

• crypto-square
• etl
• phone-number
• word-count
• acronym Sync acronym tests #350
• allergies Sync allergies tests #351
• anagram Sync anagram tests #352
• armstrong-numbers Sync armstrong-number tests #353
• binary N/A (deprecated)
• bob Sync test inputs for bob #385
• difference-of-squares N/A (good as-is)
• gigasecond Sync gigasecond tests and stub #367
• grains N/A (good as-is)
• hamming Sync hamming tests and update stub #368
• hello-world N/A (good as-is)
• leap Sync tests for 'leap' exercise with canonical tests #228
• luhn Sync tests for luhn and update stub #384
• nucleotide-count Sync nucleotide-count tests #357
• pangram Sync pangram tests #358
• perfect-numbers N/A (good as-is)
• raindrops Sync raindrops tests #354
• rna-transcription N/A (good as-is)
• robot-name N/A (no canonical tests)
• roman-numerals Sync roman-numerals tests #356
• run-length-encoding N/A (good as-is)
• trinary N/A (deprecated)
• two-fer N/A (good as-is)

[mkleehammer]

I would suggest the acronym tests be updated to include something with punctuation, such as the example from the readme: "Thank George It's Friday". Some of the community ones that passed would not pass this, which is not helpful for those learning from them.

[BNAndras]

I would suggest the acronym tests be updated to include something with punctuation, such as the example from the readme: "Thank George It's Friday". Some of the community ones that passed would not pass this, which is not helpful for those learning from them.

There's a canonical test for Hailey's Comet that would take care of this once synced. I can take a look later this week.

[BNAndras]

Atbash Cipher is missing the decoding half of the exercise so I'm separating that off into a separate issue for tracking.