Dark monitor plots do not respect readout pattern

[stscijgbot-jwql]

Issue JWQL-157 was created on JIRA by Bryan Hilbert:

The dark monitor keeps track of dark rates and stats on a per instrument, aperture, and readpattern basis. This information is all recorded correctly in the database tables.

However, when creating the dark monitor plots, the functions in monitor_dark_bokeh.py only pay attention to instrument and aperture. This means that statistics for different readpatterns are potentially being mixed, which will lead to confusion. This is only an issue for MIRI and NIRSpec. The other instruments collect darks with only a single readout pattern. But MIRI has darks using FASTR1 and SLOWR1 (as seen in PIDs 1517 and 1519. It looks like the other FAST* and SLOW* readout patterns are not used), while NIRSpec uses NRSRAPID and NRSIRS2RAPID. For these cases, the dark monitor should create separate plots for each readout pattern.