mirror of
https://github.com/harvard-edge/cs249r_book.git
synced 2026-07-17 08:28:07 -05:00
5b6138b697
Adds book/tools/audit/ — a five-stage editorial audit pipeline (SCAN -> PLAN -> FIX -> VERIFY -> REPORT) for the MIT Press round 1 style rules in book-prose-merged.md. Implements Phase A of the plan at /Users/VJ/Desktop/MIT_Press_Feedback/15_audit_loop/PLAN.md. Architecture. Five-stage cycle with hard stopping conditions and verification gating. Five safety checks before any file is written: no null bytes, no leftover sentinels, expected byte delta matches actual, quarto structural delta is zero, no new issues introduced. The byte-delta check alone would have caught the discarded bulk-edit run that prompted Pass 15. Files. protected_contexts.py is the load-bearing safety layer with LineWalker and inline_protected_spans (14 adversarial tests covering every failure mode from the discarded bulk run). ledger.py is the JSON schema. scan.py, fix_script_lane.py, verify.py, loop.py are the stage runners. Seven check modules cover vs_period, compound_prefix, percent_symbol, lowercase_prose_references, acknowledgements_spelling, binary_units, and h3_titlecase. Validation. h3_titlecase finds 609 in vol1 (within 2 of plan's expected 611). vs_period finds 0 in vol1 (matching pass 10b) and 16 in vol2 (real residuals). compound_prefix finds 19 in vol1 (real misses pass 10b's regex did not catch) and 46 in vol2 (never swept).
84 lines
2.6 KiB
Python
84 lines
2.6 KiB
Python
"""Check: binary units (GiB, TiB, KiB, MiB) in body prose.
|
|
|
|
Rule: book-prose-merged.md section 1
|
|
"Never binary units in prose. Write `GB` and `TB`, not `GiB` or `TiB`."
|
|
|
|
Auto-fixable: NO. The fix depends on whether the source value was a binary
|
|
or decimal quantity. A GiB -> GB conversion changes the number by ~7%. The
|
|
correct fix requires either re-reading the spec source or leaving the
|
|
value as-is and just changing the unit label (which produces a slightly
|
|
incorrect number). We emit as 'needs_subagent=False, auto_fixable=False'
|
|
so the operator has to hand-review each case.
|
|
|
|
Protected contexts: code fences, YAML, math, HTML comments, inline code,
|
|
and tables (tables sometimes legitimately show GiB in spec-citation rows).
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import re
|
|
from pathlib import Path
|
|
|
|
from audit.ledger import Issue, make_issue_id
|
|
from audit.protected_contexts import (
|
|
LineWalker,
|
|
default_line_skip,
|
|
inline_protected_spans,
|
|
is_table_row,
|
|
position_in_spans,
|
|
)
|
|
|
|
CATEGORY = "binary-units-in-prose"
|
|
RULE = "book-prose-merged.md section 1"
|
|
RULE_TEXT = "Never use binary units (GiB, TiB, KiB, MiB) in body prose"
|
|
|
|
# Match KiB/MiB/GiB/TiB/PiB/EiB when preceded by a digit, possibly with a space
|
|
_BINARY_UNIT_RE = re.compile(r"\b\d+(?:\.\d+)?\s*([KMGTPE]iB)\b")
|
|
|
|
|
|
def check(
|
|
file_path: Path,
|
|
text: str,
|
|
scope: str,
|
|
start_counter: int = 0,
|
|
) -> tuple[list[Issue], int]:
|
|
"""Scan for binary unit suffixes in body prose."""
|
|
issues: list[Issue] = []
|
|
counter = start_counter
|
|
|
|
walker = LineWalker(text)
|
|
for line, state, line_num in walker:
|
|
if default_line_skip(line, state):
|
|
continue
|
|
if is_table_row(line):
|
|
continue
|
|
if "iB" not in line:
|
|
continue
|
|
|
|
spans = inline_protected_spans(line)
|
|
|
|
for m in _BINARY_UNIT_RE.finditer(line):
|
|
if position_in_spans(m.start(), spans):
|
|
continue
|
|
unit = m.group(1)
|
|
|
|
issues.append(
|
|
Issue(
|
|
id=make_issue_id(scope, CATEGORY, counter),
|
|
category=CATEGORY,
|
|
rule=RULE,
|
|
rule_text=RULE_TEXT,
|
|
file=str(file_path),
|
|
line=line_num,
|
|
col=m.start(),
|
|
before=line,
|
|
suggested_after="", # No automatic fix; needs review
|
|
auto_fixable=False,
|
|
needs_subagent=False, # Needs human judgment, not subagent
|
|
reason=f"{unit} in body prose (needs manual review)",
|
|
)
|
|
)
|
|
counter += 1
|
|
|
|
return issues, counter
|