Skip to content

Excel import of time column incorrectly casts to double/gibberish if no leading 0 present #112

@Alex-Monahan

Description

@Alex-Monahan

Howdy! When using the nice workaround of loading Excel files into DuckDB using the spatial extension, time data is having issues if a leading zero is not present. I've attached 2 files: one that works correctly (which has leading zeroes), and one that fails.

Works: (01:00:00 format)

SELECT * 
FROM st_read('time_data_with_blanks.xlsx', layer='Sheet1');
time_column
00:26:00
NULL
02:26:00
03:26:00
04:26:00
05:26:00
06:26:00
07:26:00
08:26:00
09:26:00
10:26:00
11:26:00
12:26:00
13:26:00
14:26:00
15:26:00
16:26:00
17:26:00
18:26:00
19:26:00
20:26:00
21:26:00
22:26:00
23:26:00
00:26:00

Fails: (1:00:00 format)

SELECT * 
FROM st_read('time_data_with_blanks_no_leading_zero.xlsx', layer='Sheet1');
time_column
0.018055555555555554
NULL
0.10138888888888889
0.14305555555555555
0.18472222222222223
0.2263888888888889
0.26805555555555555
0.30972222222222223
0.35138888888888886
0.39305555555555555
0.43472222222222223
0.47638888888888886
0.5180555555555556
0.5597222222222222
0.6013888888888889
0.6430555555555556
0.6847222222222222
0.7263888888888889
0.7680555555555556
0.8097222222222222
0.8513888888888889
0.8930555555555556
0.9347222222222222
0.9763888888888889
1.0180555555555555

time_data_with_blanks.xlsx
time_data_with_blanks_no_leading_zero.xlsx

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions