identifiers are being parsed as feature labels in various visualizations

[ebolyen]

Originally by @gregcaporaso:

Various different visualizations, including q2-composition's da-barplot, q2-taxa's barplot, and q2-fmt's feature-peds are parsing feature identifiers as if they have a defined schema to make them more useful in the visualizations. We've had some discussion on this in https://github.com/qiime2/q2-taxa/pull/153 and some improvements to make less assumptions about the schema of identifiers were made in https://github.com/qiime2/q2-composition/pull/122 (aligning composition da-barplot with the approach for taxa barplot). However this is clearly something we need to better support generally.

[ebolyen]

This might end up looking like one of the following:

• A new method on a format which generates a dataframe/series
• A new object that we transform to which has a dataframe/series
• A typing.Annotation with some tag or expression indicating the dataframe/series format