Script_RunImmediatelyUntilEffect runs a script until either a specified
effect may occur or it reaches an end.
All existing script commands and natives, and some specials, call
Script_RequestEffects which allows us to analyze them.
Any downstream script commands/natives/specials will be statically known
not to call Script_RequestEffects and treated as if they have all
effects. Manually tagging them with requests_effects=1 and calling
Script_RequestEffects will make them analyzable.
Using these, we're able to execute scripts until they either exit with
no effect, or would possibly have an effect. This allows us to:
1. Not run on frame map scripts or triggers if they would have no
effect.
2. Immediately run triggers if they only affect flags/vars. This removes
the lag frames when biking into the Cycling Road, for example.
3. Migrate on load/on transition/on resume/on return to field/on dive
warp scripts onto the global script context if they would block
(approximated via SCREFF_HARDWARE).
4. Support arbitrary control flow in trainer scripts. The trainer does
not see the player if the script has no effect, and the trainer will
use whichever trainerbattle command is branched to.
5. Support arbitrary scripts in trainer scripts. cant_see and
cant_see_if_* commands have been introduced so that scripts are able
to do something when the player interacts with the trainer even if
that trainer wouldn't see them.
* Filename in list (no proper sorting yet)
* Line number and message in error list + removed sorting
* Style adjustment
* Added missing file/line number to "expected N passes/successes"
* Fixed Known Failing Passing test list
* Separated test list from totals
* Assumption fails list
* Better names
* Filename for KNOWN_FAILINGs passing
* Moved total back to the bottom
* Spaces correction
* Fixed test list count
* Source file for Alloc fails on tests
* Moved sourceLine from BattleTest to Test struct
* Fixed assumptions failed "and more" counter
* Fixed ASSUMPTION block not printing their line numbers
* Fixed when stopLine is printed
* Renamed stopLine to failedAssumptionsBlockLine
* consolidated gimmick checks, triggers, communication, and activation; updated test runner
* fixed improper use of .usableGimmick
* cleaning up battle_dynamax.c, changing function args to u32s
* fixed '#ifdef TESTING' causing errors
* updated z-moves to use gimmick interface, pared down redundancies; no AI/tests
* added support for z-moves in tests, consolidated gimmick fields
* removed ShouldUseMaxMove and .usingMaxMove
* renamed TryChangeZIndicator, updated z move display
* added several z-move tests and fixed various z-move interactions; fixed z-move category calc
* fixed useless battler arg in GetTypeBasedZMove
* added basic test check for bad Z-Move or Mega usage
* reworked test runner gimmick functionality; added support for Ultra Burst + Z-Move to test Light That Burns the Sky
* fixed gimmick test logic; fixed damage category override
* fixed mega rayquaza test fail
* consolidated gimmick indicator logic; added graphics to gGimmicksInfo
* removed TeraData struct
* reimplemented AI logic for Z-Moves; no changes
* updated Z-Move and Ultra Burst trigger gfx
* added testrunner check for multiple gimmick use
* fixed duplicate z-move call in test
* reorganized data/graphics/gimmicks.h
* added signature Z-Move ability tests; implemented Guardian of Alola
* fixed bad test update
* fixed Thousand Arrows not affecting Tera Flying; clean-up
* fixed -ate tests
* fixed tera tests
* fixed tera tests really
* fixed last batch of tests
* fixed -ate mega test again
* code review
* code review pt.2
* tweaked CanTera again
* dynamax flag only required for player
* Set TYPE_MYSTERY as type 0
* Remove redundant TYPE_NONE
* Moved Gen 1-3 type damage categories to gTypesInfo
* Set TYPE_NONE as 0 instead
* Grouped type info to a single file
* Fixed sTypeEffectivenessTable static name
* Adjusted MON_DATA_TERA_TYPE to account for shift in type IDs
* oops, missed the extern
* Moved Tera Type RGB values to gTypesInfo
* Fixed Tera Type test
* Added option test feature to set IVs
* Hidden Power type test
* Tera Type safeguard in givemon
* Removed isHiddenPowerType for a future PR that refactors Hidden Power
* Review changes
PARAMETRIZE_LABEL is like PARAMETRIZE, except that it allows the user to
provide a label which will be displayed in the test name line. This is
useful for tracking _which_ PARAMETRIZE case failed in the situation
where the numbers are unwieldy.
