Files
paperspec/tools/oracle/grade.py
T
alexmandClaude ea0a7efa78 feat: zipped packages, target set, checklists, audit fixes and a stress test
Features for the next six months (accessible HTML export, paperpack and
Lode on packages, the personal bundle):

- Zipped packages: the oracle writes and reads `.sct.zip` archives under
  the rules of spec/12 §12.2; validate.py and load.py accept them; a
  byte-identical zip of the tiny package joins the fixtures, and two
  broken archives (traversal, symbolic link) join fixtures/invalid.
- The target set of targeted F3 (spec/08 §8.10) is computed by the
  oracle and pinned by vectors/targets.jsonl under each digit rule; the
  rule now covers cells as well as sentences and the leaves of anchored
  containers, and defines a node's section.
- scienting-checklist/1 (provisional): a conformance check of an export,
  covered by a receipt as ["checklist", id, hash] (spec/08 §8.11), with a
  WCAG 2.1 AA fixture and its receipt.
- QA reports are signed like receipts (spec/08 §8.9).
- A `description` annotation body with role alt or long, so an
  accessible export can find a confirmed figure description; the layer
  fixture carries one, and the store's body-kind constraint allows it.
- scienting-spot-check/1, the seed payload, is registered with a schema.
- exercise, submission and grade are provisional until a course has
  used them.

Audit fixes, each with a vector or a broken package where one was
missing:

- Folder packages: symbolic links are refused on validate and load,
  manifest paths must not collide under case folding, listed files are
  matched by exact name, files are hashed in chunks under a size limit.
- Files are compared byte for byte, so CR LF, a byte order mark, NaN,
  a lone surrogate or invalid UTF-8 is refused instead of translated;
  integers beyond 2^53 are read as doubles, as canonical JSON writes them.
- Receipts: signature.key must equal by and the key must be valid when
  signed; a qa receipt's nodes must be the sample at the original's
  hashes, its scope must name qa.of, its errors must be sampled nodes;
  statuses are computed once, iteratively, so chains of spot-checks
  hundreds deep settle; Base64 and manifest shape errors become refusals
  instead of tracebacks; counts and the fidelity of roster-less packages
  are recomputed.
- The gate enforces rule 11 of spec/11 §11.4: changed nodes are
  validated against the node schema and the references between nodes
  are checked, so a set to a bad value, an insert with unknown
  properties or a delete of a referenced header cell is refused; a
  relabel to null drops the label; a reclassify drops foreign
  attributes; a split strips whitespace across runs; ids.jsonl order
  follows spec/04 §4.4 in every state; a move after itself is refused.
- Anchors: the relabel rule requires that nothing else changed below the
  node; an alias chain ending in a deletion no longer crashes resolve.
- Notation: a script after an empty group applies to the previous atom
  ({}^{14}C); letters are defined by Unicode block in spec/05 §5.4 and
  the oracle; \left) is complex.
- Page fingerprints delete whitespace before folding, so "10 -20" and
  "10-20" fold alike (spec/07 §7.1).
- Grades are summed with exact decimals; the reviewer error bound is
  computed with integer arithmetic.
- receipts and revocations are keyed by (doc_key, id) in SQL, since one
  revocation may be carried by several packages.

Checking:

- tools/stress.py: seeded randomized property tests over canonical JSON,
  normalization, notation, random correction walks, anchor resolution,
  Merkle roots, spot-check samples, error bounds, fingerprints, mutated
  archives and packages, rights and spot-check chains; CI runs it.
- check.py verifies every count and version the prose states against
  the repository, the zipped fixture and crafted archives, the checklist
  coverage, and nine more broken packages.
- CONTRIBUTING.md gains a release checklist.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
2026-10-14 20:07:22 -04:00

94 lines
3.7 KiB
Python

"""Grading a scienting-submission/1 against a scienting-exercise/1 key.
Non-normative; see spec/19-exercises.md.
"""
from decimal import Decimal
from fractions import Fraction
from . import anchors as A
from .tree import Refusal
def dec(d):
"""A decimal string, exactly as the sum came out (spec/19 §19.4: points are decimal strings, summed exactly)."""
return format(Decimal(d), "f")
def _zero_like(d):
"""Zero with the scale of `d`, so a missed element worth "0.5" scores "0.0"."""
return Decimal(0).quantize(Decimal(d))
def _resolved(state, anchor):
r = A.resolve(state, anchor)
if r["result"] == "stale":
return None
return r["node"], r.get("span")
def _match(state, rule, student, key):
s, k = _resolved(state, student), _resolved(state, key)
if not s or not k or s[0] != k[0]:
return False
if rule.get("match") == "overlap" and k[1] is not None:
if s[1] is None:
return True
lo, hi = max(s[1][0], k[1][0]), min(s[1][1], k[1][1])
size = k[1][1] - k[1][0]
return size > 0 and Fraction(max(0, hi - lo), size) >= Fraction(rule.get("min_overlap", "0.5"))
return True
def _any(state, rule, students, keys):
return any(_match(state, rule, s, k) for s in students for k in keys)
def key_anchors(body):
key = body["key"]
out = []
for el in key.get("elements", []) + key.get("chain", []) + key.get("errors", []):
out += el["anchors"]
return out
def grade(body, submission, state):
for a in key_anchors(body):
if A.resolve(state, a)["result"] == "stale":
raise Refusal("grade.key-stale", "an anchor in the key is stale")
kind, key, rule, ans = body["exercise"], body["key"], body.get("rule", {}), submission["answers"]
items = []
def item(name, pts, got, fb):
items.append({"element": name, "points": dec(pts if got else _zero_like(pts)), "max": dec(pts), "feedback": fb})
if kind in ("anchor", "check"):
if kind == "check":
v = key["verdict"]
ok = ans.get("verdict") == v["value"]
item("verdict", v["points"], ok, v.get("feedback", {}).get("hit" if ok else "miss", ""))
for i, el in enumerate(key["elements"]):
ok = _any(state, rule, ans.get("anchors", []), el["anchors"])
item("element:%d" % i, el["points"], ok, el.get("feedback", {}).get("hit" if ok else "miss", ""))
elif kind == "trace":
chain = ans.get("chain", [])
for i, link in enumerate(key["chain"]):
ok = i < len(chain) and _any(state, rule, chain[i], link["anchors"])
item("link:%d" % i, link["points"], ok, link.get("feedback", {}).get("hit" if ok else "miss", ""))
elif kind == "reproduce":
ok = abs(Decimal(ans["value"]) - Decimal(key["value"])) <= Decimal(key["tolerance"])
item("value", key["points"], ok, key.get("feedback", {}).get("hit" if ok else "miss", ""))
elif kind == "verify":
flags = ans.get("flags", [])
for i, err in enumerate(key["errors"]):
hit = [f for f in flags if _any(state, rule, [f["anchor"]], err["anchors"])]
ok = any(f["verdict"] == err["verdict"] for f in hit)
fb = err.get("feedback", {}).get("hit" if ok else "miss", "")
item("error:%d" % i, err["points"], ok, fb)
else:
raise Refusal("grade.kind", "unknown exercise kind " + kind)
score = sum((Decimal(x["points"]) for x in items), Decimal(0))
total = sum((Decimal(x["max"]) for x in items), Decimal(0))
return {"format": "scienting-grade/1", "submission": submission["id"], "exercise": submission["exercise"],
"score": dec(score), "max": dec(total), "items": items}