# step: integer neural fields · wai-rs and field_verify.py
# order: 625
# features: field_int
# paths: wai/extensions/neural-fields.md wai/extensions/staged-delivery.md
# independent: int-fields media python3 -I volumetric-conformance/field_verify.py volumetric-conformance/fields
#
# wai.field.int_grid and its levels, wai.field.int_refine (extensions/neural-fields.md): the
# reference module's unit tests (the operator set's exponential, the hash's wrap and the
# direct index, a level not held read as zero, a level applied only as the next one, the
# domain's faces, a degenerate pose refused) and tests/int_field.rs
# (volumetric-conformance/fields/ regenerates byte for byte from its scenes and holds nothing
# else, every value, render, refusal, level application and layer outcome it states is
# reproduced at every prefix of the levels, every fault §11 plants in the decode is caught,
# and §8's table lists exactly the codes the corpus carries); and, in the step of every
# independent verifier above, the corpus's second implementation, field_verify.py, written
# from the extension's text and sharing no code with wai-rs: it parses every payload with
# §8's checks in their order, applies every level by §7.3, computes every point's values with
# its own grid, hash and network and every render and stage hash with its own camera and ray
# march, plants each fault of §11 in its own decode and requires the corpus to catch every
# one; standard library only, run isolated (-I).
#
# The paths this step brings to the workflow's trigger, and why:
# wai/extensions/neural-fields.md: tests/int_field.rs reads §8's table, and field_verify.py
#   implements the text.
# wai/extensions/staged-delivery.md: its level form (§6.2) is what FieldStages implements.
set -euo pipefail
(cd wai-rs && cargo test --no-default-features --features "$WAI_FEATURES" --lib int_field::)
(cd wai-rs && cargo test --no-default-features --features "$WAI_FEATURES" --test int_field)