«Don’t export CSVs from Excel»
https://ftfy.readthedocs.io/en/latest/avoid.html #Python #Encoding #Unicode #Mojibake #Excel #CSV
Edited 1y ago
Biggest issue I have for Excel creating CSV files, is that Excel has a habit of adding empty rows at the bottom if the Excel file had rows deleted. The CSV will end up with rows with just commas on them.
The ERP we use will choke on those empty rows. I then need to remove them up in an editor.
Web advice says it won't if you delete all the blank rows from the range selected.
Fairly easy little VBA to write - find the blank rows, delete, do the CSV creation.
Send an example, l like a (small) Excel challenge.
@RussEby Interesting use case. What happens if you save to xlsx, open in #LibreOffice and save to csv/tsv?
Edited 1y ago
@nemobis That's an interesting suggestion. I'll try that when I run into another file with the issue.
@mdione @RussEby Marcos, don't be that guy... People have to work around stuff they have little control over and had no choice having to use most of the time. I can imagine why an old ERP is not expecting "empty" comma filled lines, and then, some "dev around ERPs" stuff I've seen is like adversarial adaptations, because the ERP vendor doesn't want you to integrate freely with it.
@vdd @mdione @RussEby ERP != CRM... and "should" to our benefit, surely... As I said, ERP vendors usually don't want interoperability, they want to sell you "another module"
Curious, what is ETL?
Edited 1y ago
Most of my CSVs, I create using Python and don't have issues.
Where the problem comes up is when others are trying to load a CSV and it shows 400 'failures' out of 900 and they were only trying to load 500 rows.
I have to explain that the 'failures' are the empty rows and can be ignored. Unless some of the records did fail and then I need to filter the results for them so they can corrected and reuploaded.