Reading a table of data

[ianadd (ian)]

I generate a csv file hourly with 4 columns and 10 rows.

1. ‘file’ integration reads the last line.
2. If I string all the rows together into a single line it is too long for ‘file’ integration to read.

Objective will be to have 10 sensors each having 4 attributes

Any suggestions.

[tom_l]

Assuming four of the values are not too big to fit, put it all on the one line again then create 10 file sensors, each taking a different split of the line.
e.g. here are the value templates for the first 4 sensors:

Gets first 4 values:

{{ value.split(',')[0:4] }}

Gets 5th to 8th values

{{ value.split(',')[4:8] }}

gets 9th to 12th values

{{ value.split(',')[8:12] }}

gets 13th to 16th values

{{ value.split(',')[12:16] }}

etc…

To be clear these value templates have to be used in separate file sensors. Each one taking a chunk of the single line csv values. You will need 10 files sensors in total. Then you can use 10 template sensors to read the values into attributes if you need to. As unfortunately the file sensor does not support attributes.

[ianadd (ian)]

Thanks, I hadn’t realised that I could read in part of a ‘too long’ line.