Commit: r1540: Change the signature of resolveType so that continue
Version 1540 was commited by steve on 2010-01-30 13:52:44 +0000 (Sat, 30 Jan 2010) BRANCH: style Change the signature of resolveType so that continue can be implemented. Continue works. But not without actions (ie it is as if without_actions is always given).
On 30.01.2010 14:52, svn commit wrote:
Version 1540 was commited by steve on 2010-01-30 13:52:44 +0000 (Sat, 30 Jan 2010) BRANCH: style
Change the signature of resolveType so that continue can be implemented.
Continue works. But not without actions (ie it is as if without_actions is always given).
a) Sorry, are you implementing now without_actions? In the trunk with_actions means, actions get carried forward, else no actions get carried forward. I did not know that "without_actions" exists too. b) Are more than the first two keys read? Will the style branch accecpt say 5x =.... and 100x !=... without cutting of some of the command?
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
On 30/01/10 13:58, Felix Hartmann wrote:
Continue works. But not without actions (ie it is as if without_actions is always given).
a) Sorry, are you implementing now without_actions? In the trunk with_actions means, actions get carried forward, else no actions get carried forward. I did not know that "without_actions" exists too.
No it was just a mistake, I meant it is as if with_actions is always given - I find the terminology very confusing! I would prefer that continue not change the normal effect of the actions and there be a without_actions (or no_carry_forward or no_propagate which would be more descriptive of what happens) command that prevented the carry forward. But no, I am not going to change anything unless you all want the change.
b) Are more than the first two keys read? Will the style branch accecpt say 5x =.... and 100x !=... without cutting of some of the command?
Yes it should work. Before the merge I will make a request for examples that can be made part of the testing cycle, so we can make sure that it does work before the merge. ..Steve
Hi Steve,
No it was just a mistake, I meant it is as if with_actions is always given - I find the terminology very confusing! I would prefer that continue not change the normal effect of the actions and there be a without_actions (or no_carry_forward or no_propagate which would be more descriptive of what happens) command that prevented the carry forward.
But no, I am not going to change anything unless you all want the change.
Although I "invented" with_actions, I have no problem with you inverting it if you think that would be better. At the time, it seemed like a reasonable solution to the problem of actions always being executed when using continue. Cheers, Mark
On 30.01.2010 16:58, Mark Burton wrote:
Hi Steve,
No it was just a mistake, I meant it is as if with_actions is always given - I find the terminology very confusing! I would prefer that continue not change the normal effect of the actions and there be a without_actions (or no_carry_forward or no_propagate which would be more descriptive of what happens) command that prevented the carry forward.
But no, I am not going to change anything unless you all want the change.
Although I "invented" with_actions, I have no problem with you inverting it if you think that would be better.
At the time, it seemed like a reasonable solution to the problem of actions always being executed when using continue.
If you invert it, please do it on style branch only (on trunk only once it is merged). Once style-branch comes, it'll mean a lot of rewriting of rules for me anyhow.
Cheers,
Mark _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Steve Ratcliffe schrieb am 30.01.2010 16:13:
No it was just a mistake, I meant it is as if with_actions is always given - I find the terminology very confusing! I would prefer that continue not change the normal effect of the actions and there be a without_actions (or no_carry_forward or no_propagate which would be more descriptive of what happens) command that prevented the carry forward.
I would approve such a change. In my opinion the "with_actions" is the more usual/natural case and so it should be the default. I think, the command for the other case should still include "actions", so that it is clear, what is not propagated. Or perhaps a wording without a negation, e.g. "actions_only_here". When I have the following lines in my style file key1=value1 {set first_action=true} key1=value1 {set second_action=true} [0x01 resolution 20 no_propagate continue] first_action=true & second_action!=true [0x02 resolution 20] I would assume, that two lines will be created. I.e. the actions of the first rule are propagated, and the actions of the second rule are only applied locally. Do you agree? Gruss Torsten
participants (5)
-
Felix Hartmann -
Mark Burton -
Steve Ratcliffe -
svn commit -
Torsten Leistikow