Plotly includes dates that are missing from the data in the x-axis

Question

Having financial data, there is no data over weekends/bank holidays/etc. In the example below I used this data set.

Assume having a dateframe where the index consists of timestamps then naively plotting the data using:

import pandas as pd

pd.options.plotting.backend = "plotly"

fig = df["open"].plot()

returns:

which is great... except the straight lines connecting the missing data. Adding:

fig.update_layout(xaxis=dict(type="category"))

Is almost perfect:

How can I now get the same tickers (Year, Month, day without the hours)? Once the index is treated as a category then I cannot use date formatter like %b and co.