Before a CSV file is uploaded, the following parts of the file are analyzed:
The first 500 data rows (starting from the first data row)
500 rows in the middle of the file (for example, rows 750 - 1250 in a file with 2000 rows)
The last 500 rows
When uploading the file, the system might find a data type that does not match the analyzed rows (if the analyzed files only contain numbers, and one of the rows not analyzed contains a character string for example). The upload is aborted, and a message is displayed with a recommended course of action: You either make changes to the file (if it contains errors) and upload it again or you can change the data type of the column.
Column-Based File Analysis
The system checks every cell of each column in the file using the following priority:
STRING/DATS/TIMS is always proposed. If a column is a key figure, you have to select this. The system then determines whether it is a decimal, integer etc.
The date and date format of the cells are then checked. If the data in the content matches the data format selected in the application for uploading CSV files (or if the internal ABAP date format YYYYMMDD was selected with the format option YYYY.MM.DD), the cell content is identified as a date, and the data type DATS is selected.
Finally, the system checks whether the cell contains a time characteristic. If this is the case, the data type TIMS is selected.
In all other cases, the data type STRING is selected.
The 'leading zeros' flag can be set for characteristics with the data type STRING. In this case, an alpha conversion is performed (CHAR field).