[AZ-336] C2 VprStrategy: Protocol + DTOs + factory + composition

Foundational scaffolding for every concrete C2 backbone (UltraVPR,
NetVLAD, MegaLoc, MixVPR, SelaVPR, EigenPlaces, SALAD — AZ-337..AZ-340)
and the C2.5 ReRanker consumer side. No backbone is implemented here.

* VprStrategy Protocol (embed_query / retrieve_topk / descriptor_dim)
  + BackbonePreprocessor C2-internal Protocol (NOT in Public API per
  description.md § 6).
* DTOs in L1 _types/vpr.py — VprQuery, VprCandidate, VprResult; all
  frozen + slots; tuple-not-list for VprResult.candidates so the
  immutability invariant truly holds.
* Error family: VprError + VprBackboneError + VprPreprocessError +
  IndexUnavailableError; same-named but namespace-distinct from
  c6_tile_cache.IndexUnavailableError (the c2 family is the closed
  envelope C5 / C2.5 consume; concrete strategies rewrap the C6 form).
* C2VprConfig (strategy enum + backbone_weights_path + faiss_index_path)
  with strict validation at load; registered into Config.components on
  c2_vpr import.
* build_vpr_strategy factory with 7-strategy resolution table, lazy
  import, BUILD_VPR_<variant> gating, ImportError→
  StrategyNotAvailableError mapping, and pre-flight descriptor_dim
  match against DescriptorIndex.descriptor_dim() — mismatch fires
  ConfigError at startup, NOT at first frame.

Contract change vs the v1.0.0 draft: factory takes descriptor_index:
DescriptorIndex (not tile_store: TileStore) because descriptor_dim()
lives on DescriptorIndex per C6's Public API. The contract markdown
is updated to match.

Architecture: VprCandidate.tile_id is a plain (zoom, lat, lon) tuple,
keeping _types/ (L1) free of any c6_tile_cache (L3) import per
module-layout.md. Consumers reconstruct TileId at the C6 boundary.

Excluded per task spec:
* Concrete backbones (AZ-337..AZ-340).
* FAISS HNSW retrieve wiring (AZ-341).
* DescriptorNormaliser helper (AZ-283, already shipped).
* AC-9 single-thread binding — deferred per task spec Risk 4 until the
  generic compose_root thread-binding registry is in place (today
  each factory owns its own, e.g. fc_factory).

Tests: 45 ACs + NFRs in tests/unit/c2_vpr/test_protocol_conformance.py
covering AC-1..AC-8, the error family, the config validation, the
factory NFR (p99 ≤ 50 ms). The legacy smoke test is removed. Full
sweep 973 passed, 2 skipped (CI-only cmake / actionlint).

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
Oleksandr Bezdieniezhnykh
2026-05-12 05:25:35 +03:00
parent 823c0f1b2e
commit 3665acef66
12 changed files with 1202 additions and 56 deletions
@@ -0,0 +1,214 @@
"""C2 VPR strategy composition-root factory (AZ-336).
:func:`build_vpr_strategy` selects exactly one strategy by
``config.components['c2_vpr'].strategy`` and respects compile-time
``BUILD_VPR_<variant>`` gating: requesting a strategy whose flag is
OFF raises :class:`StrategyNotAvailableError` at composition time
(NOT at first frame).
Concrete strategy modules
(``ultra_vpr``, ``net_vlad``, ``mega_loc``, ``mix_vpr``,
``sela_vpr``, ``eigen_places``, ``salad``) are imported lazily —
a Tier-0 workstation build with ``BUILD_VPR_ULTRA_VPR=OFF`` MUST
NOT load ``c2_vpr.ultra_vpr`` (ADR-002 / I-5; verifiable via
``sys.modules``).
Pre-flight validation: after constructing the strategy, the factory
queries :meth:`VprStrategy.descriptor_dim` and asserts it matches
the C6 ``DescriptorIndex`` sidecar's ``descriptor_dim()``. Mismatch
→ :class:`ConfigurationError` at startup, NOT at first frame.
Factory signature deviates from the v1.0.0 contract draft in one
place: the contract spec named the second parameter ``tile_store:
TileStore``, but ``descriptor_dim()`` lives on
:class:`DescriptorIndex` per C6's actual Public API. We inject
``descriptor_index`` directly; the contract markdown is updated to
match.
"""
from __future__ import annotations
import logging
import os
from typing import TYPE_CHECKING
from gps_denied_onboard.config.schema import ConfigError
from gps_denied_onboard.runtime_root.errors import StrategyNotAvailableError
if TYPE_CHECKING:
from gps_denied_onboard.components.c2_vpr import C2VprConfig, VprStrategy
from gps_denied_onboard.components.c6_tile_cache import DescriptorIndex
from gps_denied_onboard.components.c7_inference import InferenceRuntime
from gps_denied_onboard.config.schema import Config
__all__ = ["build_vpr_strategy"]
_LOG = logging.getLogger("gps_denied_onboard.c2_vpr")
# Strategy resolution table — mirrors the contract's
# ``vpr_strategy_protocol.md`` v1.0.0 § Composition-Root Factory table
# verbatim. ANY mutation of this dict MUST be mirrored in the contract.
_STRATEGY_TO_BUILD_FLAG: dict[str, str] = {
"ultra_vpr": "BUILD_VPR_ULTRA_VPR",
"net_vlad": "BUILD_VPR_NETVLAD",
"mega_loc": "BUILD_VPR_MEGALOC",
"mix_vpr": "BUILD_VPR_MIXVPR",
"sela_vpr": "BUILD_VPR_SELAVPR",
"eigen_places": "BUILD_VPR_EIGENPLACES",
"salad": "BUILD_VPR_SALAD",
}
_STRATEGY_TO_MODULE: dict[str, tuple[str, str]] = {
"ultra_vpr": (
"gps_denied_onboard.components.c2_vpr.ultra_vpr",
"UltraVprStrategy",
),
"net_vlad": (
"gps_denied_onboard.components.c2_vpr.net_vlad",
"NetVladStrategy",
),
"mega_loc": (
"gps_denied_onboard.components.c2_vpr.mega_loc",
"MegaLocStrategy",
),
"mix_vpr": (
"gps_denied_onboard.components.c2_vpr.mix_vpr",
"MixVprStrategy",
),
"sela_vpr": (
"gps_denied_onboard.components.c2_vpr.sela_vpr",
"SelaVprStrategy",
),
"eigen_places": (
"gps_denied_onboard.components.c2_vpr.eigen_places",
"EigenPlacesStrategy",
),
"salad": (
"gps_denied_onboard.components.c2_vpr.salad",
"SaladStrategy",
),
}
def _is_build_flag_on(flag_name: str) -> bool:
"""Read a compile-time ``BUILD_*`` flag from the environment.
``ON`` / ``1`` / ``true`` / ``yes`` (case-insensitive) → ``True``;
anything else (including unset) → ``False``. Defaults to OFF so
test environments must opt-in explicitly per strategy.
"""
raw = os.environ.get(flag_name, "")
return raw.strip().lower() in {"on", "1", "true", "yes"}
def _c2_config(config: "Config") -> "C2VprConfig":
"""Pull the registered C2 config block.
``c2_vpr.__init__`` registers it on import; a missing
registration is a developer error and surfaces as ``KeyError``
rather than a silent fallback.
"""
return config.components["c2_vpr"]
def build_vpr_strategy(
config: "Config",
*,
descriptor_index: "DescriptorIndex",
inference_runtime: "InferenceRuntime",
) -> "VprStrategy":
"""Construct the :class:`VprStrategy` impl selected by config.
1. Reads ``config.components['c2_vpr'].strategy``.
2. Checks the matching ``BUILD_VPR_<variant>`` flag — if OFF,
raises :class:`StrategyNotAvailableError` BEFORE any import.
3. Lazily imports the concrete strategy module.
4. Constructs the strategy via its module-level
``create(config, descriptor_index, inference_runtime)``
factory function (each concrete strategy module exports
``create`` as its public entry-point; concrete constructors
stay private).
5. Pre-flight ``descriptor_dim`` match: ``strategy.descriptor_dim()``
vs ``descriptor_index.descriptor_dim()``. Mismatch raises
:class:`ConfigError`; ONE ERROR log
``kind="c2.vpr.dim_mismatch"`` is emitted; the strategy is
NOT bound.
6. On success, ONE INFO log ``kind="c2.vpr.strategy_loaded"``
with ``strategy`` + ``descriptor_dim``.
Raises:
StrategyNotAvailableError: compile-time flag OFF or
concrete module not yet built (AZ-337..AZ-340 pending).
ConfigError: ``descriptor_dim`` mismatch between strategy
and corpus index.
"""
block = _c2_config(config)
strategy = block.strategy
flag_name = _STRATEGY_TO_BUILD_FLAG.get(strategy)
module_info = _STRATEGY_TO_MODULE.get(strategy)
if flag_name is None or module_info is None:
# Defensive — config validation rejects unknown strategy labels
# at load (C2VprConfig.__post_init__), so this branch is only
# reachable if the resolution table and the validation set
# drift apart.
_LOG.error(
"c2.vpr.build_flag_off",
extra={"strategy": strategy, "reason": "unknown_strategy"},
)
raise StrategyNotAvailableError(
f"VprStrategy {strategy!r} is not buildable in this binary."
)
if not _is_build_flag_on(flag_name):
_LOG.error(
"c2.vpr.build_flag_off",
extra={"strategy": strategy, "flag": flag_name},
)
raise StrategyNotAvailableError(
f"BUILD_VPR_{strategy.upper()} is OFF for this binary; "
f"cannot select strategy={strategy}."
)
module_name, class_name = module_info
try:
module = __import__(module_name, fromlist=[class_name])
except ModuleNotFoundError as exc:
raise StrategyNotAvailableError(
f"VprStrategy {strategy!r} is configured but its concrete impl "
f"module {module_name!r} has not been built into this binary "
"yet (AZ-337 / AZ-338 / AZ-339 / AZ-340 pending)."
) from exc
create_fn = getattr(module, "create", None)
if create_fn is None:
strategy_cls = getattr(module, class_name)
instance = strategy_cls(
config,
descriptor_index=descriptor_index,
inference_runtime=inference_runtime,
)
else:
instance = create_fn(
config,
descriptor_index=descriptor_index,
inference_runtime=inference_runtime,
)
strategy_dim = instance.descriptor_dim()
corpus_dim = descriptor_index.descriptor_dim()
if strategy_dim != corpus_dim:
_LOG.error(
"c2.vpr.dim_mismatch",
extra={
"strategy": strategy,
"strategy_dim": strategy_dim,
"corpus_dim": corpus_dim,
},
)
raise ConfigError(
f"descriptor_dim mismatch: strategy={strategy_dim}, "
f"corpus={corpus_dim}"
)
_LOG.info(
"c2.vpr.strategy_loaded",
extra={"strategy": strategy, "descriptor_dim": strategy_dim},
)
return instance