data structure problem in feature molecular details

[sibyl229]

Hi @a8wright in the d2c implementation for feature molecular details, the sequence_context.data.sequences attribute returns an array for some features and an object for other features. This isn't a good practice, as it easily confuses anything program that consumes the API.

Test case:
The following should return the same data structure, currently they are different:

• /rest/widget/feature/WBsf019129/molecular_details
• /rest/widget/feature/WBsf000519/molecular_details

Possible changes:

• Always return the sequence and flanks on the proper strand (and only that strand), in which case you can always use the "array implementation", OR
• Always return the sequence and flanks on both strands and indicate the default strand, in which case you can always use the "object implementation", OR
• Some other data structure that you think is appropriate and is consistent across all features.

relates to and blocks(!!!) feature: WormBase/website#6871