Modify functions emulator to invoke RTDB emulator triggers

[jmwski]

Description

This change is part of the effort to integrate the RTDB emulator with the functions emulator.

Scenarios Tested

I have tested starting up a local RTDB emulator and a functions emulator that registers a single function to be called onWrite. I then wrote to the corresponding location in the RTDB emulator an observed that the function is called.

[coveralls]

Coverage decreased (-0.2%) to 61.154% when pulling 783d764 on wyszynski/rtdb-emulator-functions into 08f053e on master.

[coveralls]

Coverage decreased (-0.4%) to 61.533% when pulling f4010dd on wyszynski/rtdb-emulator-functions into 57101c4 on next.

WANT_LGTM=abeisgoat

[samtstern]

LGTM. The regex stuff is hairier than I'd like but looks about as clean as it can be!

[samtstern]

@IanWyszynski should we put this on master or start a rtdb-emulator branch? I think this needs to be paired with a PR to update the RTDB jar at least and I wouldn't want this to accidentally get partially released.

[yuchenshi]

@samtstern Let's keep this PR open until we have a new RTDB jar and this PR can then be amended to also include the change to the jar URL and checksum. In this way, the whole PR is atomic and roll back safe.

[jmwski]

@yuchenshi I merged b0c58ab into the next before I saw your suggestion about keeping all the changes in the PR. It seemed like have a separate PR for releases was consistent with the practice for other emulators. However, I've added the changes in that PR to this one so at least we have the whole diff in one place.

[yuchenshi]

@IanWyszynski Your call :) I think we're in a good spot to merge this to next. Hit it?

[thechenky]

@IanWyszynski make sure to squash and merge (instead of just merging) so it's all contained in one commit.