Incorrect dates with numfmt h:mm AM/PM

[stla]

Hello,

The xlsx file is the same as tidyverse/readxl#264

> y <- tidy_xlsx("test02.xlsx", sheet=1)
> y$data$Sheet1
  address row col content formula formula_type formula_ref formula_group
1      A1   1   1       0    <NA>         <NA>        <NA>            NA
2      A2   2   1       1    <NA>         <NA>        <NA>            NA
3      A3   3   1     1.5    <NA>         <NA>        <NA>            NA
4      A4   4   1     1.8    <NA>         <NA>        <NA>            NA
  type data_type error logical numeric                date character
1    s character  <NA>      NA      NA                <NA>         A
2    n      date  <NA>      NA      NA 1904-01-02 00:00:00      <NA>
3    n      date  <NA>      NA      NA 1904-01-02 12:00:00      <NA>
4    n      date  <NA>      NA      NA 1904-01-02 19:12:00      <NA>
  comment height width style_format local_format_id
1    <NA>     15  8.38       Normal               1
2    <NA>     15  8.38       Normal               2
3    <NA>     15  8.38       Normal               2
4    <NA>     15  8.38       Normal               2

[nacnudus]

This is a famous bug in Excel.

There should have been a warning from tidy_xlsx(), and it should have parsed the date using the 1904 system, but it was misreading the file because openxlsx uses the date1904 tag differently from my copy of Excel. I have fixed this in 6e1da16.

Please can you try it?

[stla]

Could you give me the command to install this specific commit ?
I don't remember it. And also, last time I installed a package from a specific commit, the installation failed, I didn't understand why. But I can try.

[nacnudus]

It is in the master branch, so you can use devtools::install_github("nacnudus/tidyxl") as normal.

[stla]

Done. Now I get the same dates as readxl. I think I can close.