Flattening a list of lists

[duncandewhurst]

I'm trying to flatten this file using the following command: flatten-tool flatten -f xlsx example.json. However, the resulting xlsx file is empty.

Flattening the nodes list with flatten-tool flatten -f xlsx --root-list-path nodes example.json works OK. However, flattening the links list with flatten-tool flatten -f xlsx --root-list-path links example.json returns an error: 'list' object has no attribute 'items'.

I think this is because links.location.geometry.coordinates is a list of lists (the list in the actual file is longer):

{
  "links": [
    {
       "location": {
          "coordinates": [
              [
                26.081,
                -24.406
              ],
              [
                26.081,
                -24.406
              ]
           ]
        }
     }
  ]
}

[Bjwebb]

Yes, it looks like we lack support for flattening of lists of lists. Here's a TODO in the code https://github.com/OpenDataServices/flatten-tool/blob/main/flattentool/json_input.py#L422 !

It should be fairly easy to implement this, to produce a cell with the contents 26.081,-24.406;26.081,-24.406. @duncandewhurst is that what you expect?

[duncandewhurst]

Thanks, @Bjwebb. I was expecting something like [26.081;-24.405]; [26.09; -24.416] but your suggestion looks easier to work with.

We've yet to decide whether we're going to use Flatten Tool for Open Fibre (see Open-Telecoms-Data/open-fibre-data-standard#14) so no need to implement this yet.

[Bjwebb]

For info, my suggestion is based on what we already support for unflattening (which I think requires a schema).

[duncandewhurst]

@Bjwebb we've decided to use Flatten Tool for the Open Fibre alpha, would you have time to implement this before you go on leave?

[Bjwebb]

I've made a PR for this #401

[duncandewhurst]

Closing as the PR is merged.