Tablefild custom elements processor

I have not getting any such issue. I have installed module and export data in csv format.

Exported table in csv

0,heading1,heading2,heading3,heading4,
1,heading11,heading12,heading13,heading14,
2,heading21,heading22,heading23,heading24,

@ajayyadav thank you for your response but when you render the field as custom elements there currently is no 'export as csv' option.

Anyways I had to adjust the processor to support multivalue fields as well, so there's another level of nesting added, but the output I think works out.

Thx! I reviewed the PR and added some comments.
I think we should also add a basic (kernel?) test which makes sure that the output is working as expected.

I removed the label and added the kernel test which was passing locally. I was checking the CI and I think that tests are not running there.

I was unsure about the header settings.

thank you, that looks pretty good!

I've added one question about the output-data-structure above, could you check that and make sure it's good? Maybe it is, but it seems not.

Besides that, it's great to see test-coverage here. I'd suggest to add a couple of cases though, so we cover the features of our formatter:
- run once without custom-header and once with
- set weights differently and make sure sorting is applied correctly

I added tests coverage for multi value fields - which explain the result structure and a small fix to the formatter code.

The header formatter configuration is only there for the FE to know if it should style the first row/column as header. On field configuration there's additional setting called `ignore_table_header` which again still passes through the first row - the config option is only used to determine whether the field is empty if contains only the first row.

thank you for the clarifications and better test, this seems pretty good already.

However, as commented above I think we shall still optimize for the 90% use-case to be better and keep that simple in the output, i.e. output only the single data value in the single case and avoid wrapping it in the multi-value data-structure when there are is only one value.

Thank you, that looks great now. Output is very nice for single-value fields and decent for multi-value fields + comes with test coverage! merged!

Automatically closed - issue fixed for 2 weeks with no activity.