I have a mercurial hook like so:
[hooks]
pretxncommit.myhook = python:path/to/file:myhook
with the code looking like this:
def myhook(ui, repo, **kwargs):
#do some stuff
but this hook runs on commands that use the commit logic to do something else, in my case hg shelve. is there a way to get the command that the user has input to avoid running the hook on that command?
perhaps something like this:
def myhook(ui, repo, command, **kwargs):
if command is "hg shelve"
return 0
#do some stuff
Unfortunately the answer seems to be no. I just debugged into the hook mechanism of hg 3.1, and the information about which command was issued is not propagated into the hook function. The only way I can think of is to hack something ugly with the debugger api to extract informations from the call stack.
Another hack would be to inspect sys.argv, but I fear that this is also very unreliable (as it can't detect if something was executed via the Command Server).
BTW I used this snippet to attach a debugger:
def myhook(ui, repo, **kwargs):
print kwargs
from pdb import set_trace
set_trace()