At present, if you run a regex command (e.g. the b command) you will get two history entries, one for the actual command the user typed, and one for the resolved command:
(lldb) b foo.c:12 Breakpoint 1: no locations (pending). Breakpoint set in dummy target, will get copied into future targets. (lldb) history 0: b foo.c:12 1: breakpoint set --file 'foo.c' --line 12 2: history
That seems wrong in principle, there shouldn't be double entry like this. But it also has some subtle side effects that are even less desirable. For instance, if you write a Python based command that calls "HandleCommand" of a regex command, then the resolved command gets inserted into the history list AFTER the python command the user actually ran. That in turn means that up-arrow doesn't rerun the command the user typed, but instead runs the resolved regex command.
Since regex command resolution isn't stateful - the same input string will always resolve to the same command - there's no actual reason to do this double entry, just storing the command the user typed will suffice. And there's already a way to see resolved commands, so you don't need it for this purpose either.
Is this comment out of date now? It mentions passing in "true"?