外观
target_control API
模块说明
text
Fresh actor tracking and bounded, opt-in aim/use-input sequences.
This uses the existing durable action queue and cooperative controller lock.
It is not an atomic Host input lease, projectile predictor or hit confirmation.TargetTracker
python
class TargetTrackerSDK 说明：
text
Seal one actor lifecycle and continuously use packet-observed positions.
Once invalidated, construct a new tracker explicitly; a runtime ID reuse,
scope change, missing actor or stale position never silently rebinds it.
eye/target offsets are explicit environment-specific geometry, not inferred.TargetTracker.__init__
python
def __init__(
self,
*,
connection_id,
scope_revision,
unique_id,
runtime_id,
added_revision,
max_position_age = 0.5,
max_distance = 64,
)TargetTracker.update
python
def update(self, snapshot, origin, *, target_offset=(0, 0, 0))TargetTracker.aim
python
def aim(self, snapshot, origin, neutral_input, *, target_offset=(0, 0, 0))SDK 说明：
text
Return a full typed neutral AuthInput with angles toward current target.
Tick, modes and position come from independently fenced caller evidence;
this never increments a guessed tick or claims the frame was sent.TargetTracker.connection_id
python
self.connection_id公开实例属性，在 __init__ 中设置。
TargetTracker.scope_revision
python
self.scope_revision公开实例属性，在 __init__ 中设置。
TargetTracker.unique_id
python
self.unique_id公开实例属性，在 __init__ 中设置。
TargetTracker.runtime_id
python
self.runtime_id公开实例属性，在 __init__ 中设置。
TargetTracker.added_revision
python
self.added_revision公开实例属性，在 __init__ 中设置。
TargetTracker.max_position_age
python
self.max_position_age公开实例属性，在 __init__ 中设置。
TargetTracker.max_distance
python
self.max_distance公开实例属性，在 __init__ 中设置。
TargetTracker.closed
python
self.closed公开实例属性，在 __init__、update 中设置。
TargetTracker.last_revision
python
self.last_revision公开实例属性，在 __init__、update 中设置。
TargetTracker.last_position
python
self.last_position公开实例属性，在 __init__、update 中设置。
TargetInputSession
python
class TargetInputSessionSDK 说明：
text
One bounded tracking/aim or observed sustained item-use session.
observe() returns ready=True, observations_valid=True, sampled_monotonic,
connection_id, selected_slot, inventory, last_auth_input_tick, neutral_input,
and, when tracking, actors and aim_origin. Adapter must independently drain
ordered observations, reject gaps and derive each exact next tick from native
audit. Other input producers must be absent, as for existing MeleeSession.
use is an optional unstarted ItemActions.use_air attempt. It is sent ONCE;
subsequent fresh neutral frames sustain the existing environment-specific
use-input flow. Completion requires complete(sample) is True using actual
observations (e.g. typed crossbow chargedItem AND newer ammo decrement).
Cancel/time budget merely stops further submissions; no inferred release,
automatic second use-air, firing, neutral cleanup or completion by delay.TargetInputSession.__init__
python
def __init__(self, actions)TargetInputSession.record
python
def record(self)TargetInputSession.run
python
async def run(
self,
*,
observe,
journal,
control,
inventory,
slot,
tracker = None,
target_offset = (0, 0, 0),
use = None,
complete = None,
cancel = None,
timeout = 5,
interval = 0.05,
max_frames = 100,
max_sample_age = 0.5,
experimental_single_controller = False,
)TargetInputSession.actions
python
self.actions公开实例属性，在 __init__ 中设置。
TargetInputSession.started
python
self.started公开实例属性，在 __init__、run 中设置。
TargetInputSession.state
python
self.state公开实例属性，在 __init__、run 中设置。
TargetInputSession.attempts
python
self.attempts公开实例属性，在 __init__ 中设置。
TargetInputSession.last_solution
python
self.last_solution公开实例属性，在 __init__ 中设置。