ENH: #40231 does not allow writing multiple dataframes to a single sheet

The problem

The following code works well, both in 1.4.0 and 1.2.5:

df = pd.DataFrame(["=1+7"])
with pd.ExcelWriter("test.xlsx",engine="openpyxl") as writer:
    df.to_excel(writer,sheet_name="Sheet2")
    pd.DataFrame(["=1+1"]).to_excel(writer,sheet_name="Sheet2",header=None,index=False,startcol=4,startrow=3)

however, the following code raises an error in 1.4.0, but works in 1.2.5 (it creates a new sheet and writes both dataframes to it):

pd.DataFrame(["=1+1"]).to_excel("test.xlsx")
df = pd.DataFrame(["=1+7"])
with pd.ExcelWriter("test.xlsx",engine="openpyxl",mode="a") as writer:
    df.to_excel(writer,sheet_name="Sheet2")
    pd.DataFrame(["=1+1"]).to_excel(writer,sheet_name="Sheet2",header=None,index=False,startcol=4,startrow=3)

Also other options cannot replicate the old behaviour.

Describe the solution you’d like

An added option to if_sheet_exists="write_to" that justs selects the sheet and writes into it, without first deleting its contents. This also allows to append a dataframe within a sheet.

EDIT: in #40230 this was also suggested behaviour, but it was not implemented:

Originally I considered adding a flag like overwrite=True to pd.ExcelWriter but I don’t think this would be sufficient to cover all the potential behaviours, so would suggest an enum along the lines of the if_exists option in df.to_sql, with the following options:

* new_sheet: Create a new sheet with a different name.

* overwrite_sheet: Delete the contents of the sheet, then write to it.

* overwrite_cells: Write directly to the named sheet without deleting the previous contents.

[feefladder]

This will be solved by #42222 which will be merged soon and out in 1.4.0 😃

[rhshadrach]

@simonjayhawkins - #42222 is adding back in the feature that was (unknowingly, I think) removed. If that’s regarded as an enhancement, how can this regression get fixed for e.g. 1.3.2?

That said, there was opposition to implementing the option in question back in the PR causing this regression. If there is still sufficient opposition to this, then it seems to me the regression should be closed as a won’t-fix.