Alban Gruin [off-list ref] writes:
This rewrites complete_action() from shell to C.
A new mode is added to rebase--helper (`--complete-action`), as well as
a new flag (`--autosquash`).
Finally, complete_action() is stripped from git-rebase--interactive.sh.
The original complete_action() would return the code 2 when the todo
list contained no actions. This was a special case for rebase -i and
-p; git-rebase.sh would then apply the autostash, delete the state
directory, and die with the message "Nothing to do". This cleanup is
rewritten in C instead of returning 2. As rebase -i no longer returns
2, the comment describing this behaviour in git-rebase.sh is updated to
reflect this change.
Hmph. It is easy enough to do the clean-up ourselves in this code,
instead of asking the caller to do so. On the other hand, stashing
of local changes is done by the caller, so it feels a bit strange
way to divide the labor between the two parts.
Other than that design choice, the patch itself looks reasonable and
a fairly faithful reimplementation of what the scripted one did.
Hi Junio,
Le 10/08/2018 à 21:25, Junio C Hamano a écrit :
Alban Gruin [off-list ref] writes:
quoted
This rewrites complete_action() from shell to C.
A new mode is added to rebase--helper (`--complete-action`), as well as
a new flag (`--autosquash`).
Finally, complete_action() is stripped from git-rebase--interactive.sh.
The original complete_action() would return the code 2 when the todo
list contained no actions. This was a special case for rebase -i and
-p; git-rebase.sh would then apply the autostash, delete the state
directory, and die with the message "Nothing to do". This cleanup is
rewritten in C instead of returning 2. As rebase -i no longer returns
2, the comment describing this behaviour in git-rebase.sh is updated to
reflect this change.
Hmph. It is easy enough to do the clean-up ourselves in this code,
instead of asking the caller to do so. On the other hand, stashing
of local changes is done by the caller, so it feels a bit strange
way to divide the labor between the two parts.
Other than that design choice, the patch itself looks reasonable and
a fairly faithful reimplementation of what the scripted one did.
This was the behaviour of the old complete_action(). The new one cleans
up by itself; this is the
apply_autostash()/sequencer_remove_state()/todo_list_release() dance
done at three places inside complete_action(), and has not changed since
v3[0].
Maybe I misunderstood what you said?
[0]
https://public-inbox.org/git/20180710121557.6698-11-alban.gruin@gmail.com/
Cheers,
Alban