Closed djw closed 11 years ago
I'm not sure I understand. How do we end up with two sets of fields. We should be using the fields in the schema rather than in the CSV (except when we create when we don't yet have schema ...)
@djw can you point to an example where this happens (gist + actual link to explorer.okfnlabs.org) - I can't see seem to reproduce this ...
Sure:
https://gist.github.com/djw/4fd629609cee9df60b13 http://explorer.okfnlabs.org/#djw/4fd629609cee9df60b13/view/grid
I don't have time to investigate, sorry!
No problem re time and thanks for reporting. I can now confirm and reproduce this issue and will fix :-)
FIXED. Issue in recline with interaction of CSV backend with recently added feature to use explicitly provided fields if present. https://github.com/okfn/recline/commit/5d24fd474d6e6bd065039d058ef64c6db4454df9
@djw you will need to fix your gist by hand if you now have a duped header data in the actual CSV :-/ (no way to auto fix that ...)
@djw in fact you do not have it duped in actual data so all good, hurrah!
As a result of #98, the header row in the table is being shown twice. The column headings were previously coming from the first row of the CSV data, but are now also in the schema section of the datapackage. When the project is loaded, both are picked up, and the table has duplicated header rows. If the CSV is persisted at any point, the duplicate rows are saved, and the problem multiplies.