XLSX parsing DateTime

General Usage
,
1

I am loading data from an Excel file that includes DateTime data. The DateTime string, when the time is 0:00 is parsed as a Date instead of a DateTime. An example of the data being loaded is:

Date	a
2023-07-11 0:00	45.12
2023-07-11 0:01	45.14
2023-07-11 0:02	45.09
2023-07-11 0:03	44.69
2023-07-12 0:00	44.89

If this data is read from the Excel spreadsheet it gives the output:

julia> test = DataFrame(XLSX.readtable("test.xlsx", "Sheet1"))
5×2 DataFrame
 Row │ Date                 a
     │ Any                  Any
─────┼──────────────────────────────
   1 │ 2023-07-11           45.1199
   2 │ 2023-07-11T00:01:00  45.1426
   3 │ 2023-07-11T00:02:00  45.0878
   4 │ 2023-07-11T00:03:00  44.6917
   5 │ 2023-07-12           44.8935

Another command will do better in terms of typing the non Date columns as follows:

julia> test = XLSX.openxlsx("test.xlsx") do xf
           DataFrame(XLSX.gettable(xf["Sheet1"]; infer_eltypes=true))
       end
5×2 DataFrame
 Row │ Date                 a
     │ Any                  Float64
─────┼──────────────────────────────
   1 │ 2023-07-11           45.1199
   2 │ 2023-07-11T00:01:00  45.1426
   3 │ 2023-07-11T00:02:00  45.0878
   4 │ 2023-07-11T00:03:00  44.6917
   5 │ 2023-07-12           44.8935

I am able to get the desired DateTime type for the first column using the command

julia> test.Date = DateTime.(test.Date)
5-element Vector{DateTime}:
 2023-07-11T00:00:00
 2023-07-11T00:01:00
 2023-07-11T00:02:00
 2023-07-11T00:03:00
 2023-07-12T00:00:00

However I don’t think that I should need to do this. I think the default behaviour for parsing a DateTime string like “2023-07-11 0:00” should return a DateTime type rather than a Date type.

Is this a bug?

2

Seems worth opening an issue on the XLSX repo.

3

Thanks done