The bug introduced in fcd0ceda is fixed by ensuring that movetype::merge() is
called with the strings "movement_costs", "vision_costs" and "jamming_costs"
instead of "movement", "vision" and "jamming".
Both [terrain_defaults][movement] and [terrain_defaults][movement_costs] are
supported and equivalent, as added to the master branch in 3522eb2c.
These now work again:
[resistance_defaults]
id="special_res_for_test"
default="30"
[/resistance_defaults]
[resistance_defaults]
id="copy_of_arcane"
default="(arcane)"
[/resistance_defaults]
[terrain_defaults]
id="special_terrain_for_test"
[movement]
default="(swamp_water + 1)"
orcishfoot="(swamp_water * 2)"
[/movement]
[/terrain_defaults]
Formulas can now access other parts of the movetype too, as this allowed the
code to be the same as in 1.15:
[terrain_defaults]
id="special_terrain_for_test"
[movement_costs]
default="(swamp_water + 1)"
orcishfoot="(vision_costs.swamp_water * 2)"
[/movement_costs]
[/terrain_defaults]
[terrain_defaults]
id="special_terrain_for_test"
[defense]
default="(20 + 7 * movement_costs.special_terrain_for_test)"
[/defense]
[/terrain_defaults]
The formula handling will recognise "resistance", "movement_costs",
"vision_costs", "jamming_costs" and "defense". For [resistance_defaults], the
formula will recognise both "(arcane)" and "(resistance.arcane)" as equivalent,
similarly for [terrain_defaults] "(swamp_water)" is a shorthand for whichever
subtag is being patched.
A [terrain_defaults] tag may use data added in a previous [terrain_defaults],
as in the examples above where the second tag's [defense] is based on the first
tag's [movement_costs], this gives orcish grunts on the special terrain a 62%
chance to be hit. However, relying on data in the same [terrain_defaults] that
creates or changes it is unsupported - if the [movement_costs] and [defense]
were in a single [terrain_defaults] tag then the result would be implementation
defined, because no guarantee is made of the order in which the children of the
tag are processed.
The unit tests for [terrain_defaults] and [resistance_defaults] must stay
out-of-tree until we support some method of testing them without affecting the
core units' stats during other tests. The test added here is limited to
checking the ways that WML can read the stats, it's also a framework for the
out-of-tree parts to use.
This is a squashed cherry pick of these commits from 1.15:
* 72863e578b (the unit test)
* 0ba433203e
* 3522eb2c2a (adding the aliases)
* 0035b776d7 (fix for empty children)
Backported to 1.14 so that this mechanism can be used instead of a793bce96's
method of running the cve_2018_1999023 test in non-strict mode, which means
that the test now passes without making equivalent-to-a793bce96 changes to
projectfiles/VC12/WML_tests.cmd and projectfiles/VC16/WML_tests.cmd.
This was part of working out whether a subset of the "fail" tests could be run
in one Wesnoth instance. To run a test that returns TEST_FAIL_BROKE_STRICT with
any other test would require a mechanism to reset lg::broke_strict()'s flag.
All tests that fail with an {ASSERT} will also set the lg::broke_strict() flag,
the tests with the new status are only those that would pass without the strict
flag.
In the SUF tests, change a test from fail-on-success to breaks-strict, rename
the formula tests and add some comments. The rename is because "fai" is
"Formula AI", an obsolete name for WFL.
(cherry-picked from commit 7359fddecb)
this test whether we succesfully prevent the exceution and parsing of
precompiled lua chunks, which can be used to break the lua sandbox.
This test does not cover all affected functions in particular not the ai
code that loads lua chunks.
Several keys were either missing or misnamed in the formula view.
These have been changed to match the Lua and WML names for the keys.
This affects canrecruit, moves, and max_moves, as well as status.
In addition, a large number of keys were missing entirely from the formula view.
Most of these keys have now been added and can be accessed from formulas.
A few that have little use in filters (such as image_mods or profile) are still missing.
The Lua view missed a few keys that the formula view did provide.
Although these could in principle have been obtained by using the special __cfg key,
they have now been added to the Lua view anyway.
(With unit tests to back it up!)
- Fixes [for]step=0
- Fixes [for]start,stop=0,0
- Fixes [for]start,end,step=0,1,-1
- Probably fixes several other things too
A new unit test, called 'test_role_lua' was added, to ensure that the [role]
tag could be used from Lua.
Also, in the first unit test the types list is stored in a variable, to
ensure that variable substitution works as intended
this commit also changed slightly the way that wml menu events
are synced. previously they were run via "run_in_synced_context",
now they are run via "run_in_synced_context_if_not_already", which
makes it safe to fire them from lua (a synced context).
Reverts commit e85c37e161 which was
an error.
I chose to use a scoped_pointer this time around because the
definition of play_controller should not actually depend on the
complete definition of unit_type.
Unlike [allied_with] and [enemy_of], these will match if only if
there is *any* ally or enemy matching the filter. The other versions
of the tag match only if all sides matching the filter are
allied / opposed.
This commit also adds a unit test that checks that it works as
expected.
