These are chat archives for frictionlessdata/chat

4th
Sep 2017
jobarratt
@jobarratt
Sep 04 2017 08:10
@kleanthisk10 fantastic! Thank you for the update. We will take a look at this and get back to you with any comments.
AdriĆ  Mercader
@amercader
Sep 04 2017 09:41
@Stephen-Gates @pwalsh hi, reagarding ckanext-datapackage / ckanext-validation
The former in its current implementation is focused exclusively on importing and exporting Data Packages (so points 1 and 2 in your list Stephen). Main caveats are that as Paul mentioned this targeted the old version of the specs, and also IIRC it only imports/exports metadata but not data to the DataStore (I might be wrong)
sorry point 3 should be covered as well
ckanext-validation is currently mostly focused on validation of resources with goodtables. There is an internal API but reports are rendered on the UI (see frictionlessdata/ckanext-validation#6). I'm also currently working on integrating validation on the dataset creation form UI, with the creation process failing if errors are found, and a report being displayed.
AdriĆ  Mercader
@amercader
Sep 04 2017 09:47
This is tricky because CKAN does not provide many extension points at this level so I'm currently targeting CKAN master, but I hope to support at least a couple of current versions.
I need to revisit the time estimates and priorities today or tomorrow, I can link to them when finished
As to whether they should be merged in a single extension I don't think it's a priority right now as there is not a lot of overlap (tbh I haven't thought about it that much)
Stephen Gates
@Stephen-Gates
Sep 04 2017 11:22
Thanks @amercader For context Data Curator will do the validation locally using TableSchema.js before uploading (hence my priority list above) https://github.com/ODIQueensland/data-curator/blob/develop/README.md Based on http://okfnlabs.org/blog/2016/07/25/publish-data-packages-to-datahub-ckan.html I think the data and metadata is uploaded but only the metadata and links to the data are downloaded. So the data package that goes up, is not the same as the one that comes down as CSVs are replaced with URLs. I guess my Open Data Package function will have to handle data packages with local and remote resources. I suspect https://github.com/frictionlessdata/datapackage-js can help with that.
My bare minimum requirement is 1. and 2. above for the v1 spec if that helps.