#!/usr/bin/env python3 """ generate_corpus_manifest.py — build the public exam-corpus manifest from OFFICIAL sources, verifying every source URL is live before it is written. Output: exam-corpus.yaml (consumed by run/initialization/seed_exam_corpus.py). Sources (all official exam-board hosts; public past-paper PDFs): AQA filestore.aqa.org.uk — fully templatable; enumerated + HEAD-verified here. Edexcel qualifications.pearson.com — date suffix non-derivable; confirmed URLs embedded. OCR www.ocr.org.uk/Images — opaque doc-id; confirmed URLs embedded. Every URL is HEAD/GET-checked (200 + application/pdf) before inclusion, so the committed manifest never carries a dead or wrong-cased link. Re-run to refresh as more sessions go public. Conventions (locked — see ~/cc/ideas/2026-06-07-exam-paper-ingestion.md): session = "YYYY-Mon" e.g. 2022-Jun exam_code = BOARD-award-PAPER-SESSIONCOMPACT-ROLE e.g. AQA-8463-1H-2022JUN-QP """ from __future__ import annotations import concurrent.futures as cf import os import sys import urllib.error import urllib.request from typing import Any, Dict, List, Optional, Tuple import yaml AQA_BASE = "https://filestore.aqa.org.uk/sample-papers-and-mark-schemes" ROLE_TOKEN = {"QP": "QP", "MS": "MS", "ER": "WRE"} # AQA filestore role tokens MONTHS = {"JUN": ("june", "Jun"), "NOV": ("november", "Nov")} FETCHED = "2026-06-07" def head_ok(url: str, timeout: int = 20) -> bool: """True iff the URL resolves to a real PDF (200 + application/pdf), following redirects. AQA soft-404s redirect to www.aqa.org.uk/req_path=... (text/html), so we check content-type. Uses a tiny Range GET (stdlib urllib) so we never pull the whole PDF just to verify it.""" req = urllib.request.Request(url, headers={"Range": "bytes=0-3", "User-Agent": "cc-corpus/1.0"}) try: with urllib.request.urlopen(req, timeout=timeout) as r: ctype = (r.headers.get("content-type") or "").lower() return r.status in (200, 206) and "pdf" in ctype except urllib.error.HTTPError as e: # A 206/200 PDF never lands here; 404/redirect-to-html will. ctype = (e.headers.get("content-type") or "").lower() if e.headers else "" return e.code in (200, 206) and "pdf" in ctype except Exception: return False # ─────────────────────────── AQA catalogue ─────────────────────────── # spec_code, subject, award, award_level, first_teach, [(filestore_papercode, paper_code, tier), ...] def _gcse_single(award: str) -> List[Tuple[str, str, Optional[str]]]: out = [] for paper in ("1", "2"): for tier in ("F", "H"): out.append((f"{award}{paper}{tier}", f"{award}/{paper}{tier}", tier)) return out def _trilogy(award: str) -> List[Tuple[str, str, Optional[str]]]: out = [] for subj in ("B", "C", "P"): for paper in ("1", "2"): for tier in ("F", "H"): out.append((f"{award}{subj}{paper}{tier}", f"{award}/{subj}/{paper}{tier}", tier)) return out def _alevel(award: str, papers=("1", "2", "3")) -> List[Tuple[str, str, Optional[str]]]: return [(f"{award}{p}", f"{award}/{p}", None) for p in papers] def _subj(award: str, papers, tiers=(None,)) -> List[Tuple[str, str, Optional[str]]]: """Generic GCSE/A-level builder. tiers=('F','H') for tiered subjects (Maths/Science), tiers=(None,) for untiered (English/Geography/CS/Business/Psychology).""" out = [] for p in papers: for t in tiers: tl = t or "" out.append((f"{award}{p}{tl}", f"{award}/{p}{tl}", t)) return out def _mfl(award: str) -> List[Tuple[str, str, Optional[str]]]: """AQA MFL: Listening/Reading/Writing papers, each Foundation/Higher (Speaking is teacher-conducted, no public QP). Filestore code encodes skill+tier, e.g. 8658LH = French Listening Higher.""" out = [] for skill in ("L", "R", "W"): for t in ("F", "H"): out.append((f"{award}{skill}{t}", f"{award}/{skill}{t}", t)) return out AQA_SPECS = [ # ── Sciences (round 1 — kept at full depth) ────────────────────────────────────── ("AQA-BIOL-8461", "BIOLOGY", "8461", "GCSE", "2016", _gcse_single("8461")), ("AQA-CHEM-8462", "CHEMISTRY", "8462", "GCSE", "2016", _gcse_single("8462")), ("AQA-PHYS-8463", "PHYSICS", "8463", "GCSE", "2016", _gcse_single("8463")), ("AQA-COMB-8464", "COMBINED SCIENCE TRILOGY", "8464", "GCSE", "2016", _trilogy("8464")), ("AQA-BIOL-7401", "BIOLOGY", "7401", "AS", "2015", _alevel("7401", ("1", "2"))), ("AQA-BIOL-7402", "BIOLOGY", "7402", "A-level", "2015", _alevel("7402")), ("AQA-CHEM-7404", "CHEMISTRY", "7404", "AS", "2015", _alevel("7404", ("1", "2"))), ("AQA-CHEM-7405", "CHEMISTRY", "7405", "A-level", "2015", _alevel("7405")), ("AQA-PHYS-7407", "PHYSICS", "7407", "AS", "2015", _alevel("7407", ("1", "2"))), ("AQA-PHYS-7408", "PHYSICS", "7408", "A-level", "2015", _alevel("7408")), # ── Round 2 breadth — high-volume core (Maths, English) ─────────────────────────── ("AQA-MATH-8300", "MATHEMATICS", "8300", "GCSE", "2015", _subj("8300", ("1", "2", "3"), ("F", "H"))), ("AQA-MATH-7357", "MATHEMATICS", "7357", "A-level", "2017", _alevel("7357", ("1", "2", "3"))), ("AQA-MATH-7356", "MATHEMATICS", "7356", "AS", "2017", _alevel("7356", ("1", "2"))), ("AQA-ENGL-8700", "ENGLISH LANGUAGE", "8700", "GCSE", "2015", _subj("8700", ("1", "2"))), ("AQA-ENGLIT-8702", "ENGLISH LITERATURE", "8702", "GCSE", "2015", _subj("8702", ("1", "2"))), ("AQA-ENGL-7702", "ENGLISH LANGUAGE", "7702", "A-level", "2015", _alevel("7702", ("1", "2"))), ("AQA-ENGLIT-7712", "ENGLISH LITERATURE A", "7712", "A-level", "2015", _alevel("7712", ("1", "2"))), # ── Round 2 breadth — humanities / others ───────────────────────────────────────── ("AQA-GEOG-8035", "GEOGRAPHY", "8035", "GCSE", "2016", _subj("8035", ("1", "2", "3"))), ("AQA-GEOG-7037", "GEOGRAPHY", "7037", "A-level", "2016", _alevel("7037", ("1", "2"))), ("AQA-COMP-8525", "COMPUTER SCIENCE", "8525", "GCSE", "2020", _subj("8525", ("1", "2"))), ("AQA-COMP-7517", "COMPUTER SCIENCE", "7517", "A-level", "2015", _alevel("7517", ("1", "2"))), ("AQA-BUS-8132", "BUSINESS", "8132", "GCSE", "2017", _subj("8132", ("1", "2"))), ("AQA-BUS-7132", "BUSINESS", "7132", "A-level", "2015", _alevel("7132", ("1", "2", "3"))), ("AQA-PSYC-8182", "PSYCHOLOGY", "8182", "GCSE", "2017", _subj("8182", ("1", "2"))), ("AQA-PSYC-7182", "PSYCHOLOGY", "7182", "A-level", "2015", _alevel("7182", ("1", "2", "3"))), # ── Round 2 breadth — modern foreign languages (Listening/Reading/Writing, F+H) ─── ("AQA-FREN-8658", "FRENCH", "8658", "GCSE", "2016", _mfl("8658")), ("AQA-SPAN-8698", "SPANISH", "8698", "GCSE", "2016", _mfl("8698")), ("AQA-GERM-8668", "GERMAN", "8668", "GCSE", "2016", _mfl("8668")), ("AQA-FREN-7652", "FRENCH", "7652", "A-level", "2016", _alevel("7652", ("1", "2"))), ("AQA-SPAN-7692", "SPANISH", "7692", "A-level", "2016", _alevel("7692", ("1", "2"))), ("AQA-GERM-7662", "GERMAN", "7662", "A-level", "2016", _alevel("7662", ("1", "2"))), ] AQA_SESSIONS = ["JUN18", "JUN19", "NOV20", "NOV21", "JUN22", "JUN23", "JUN24"] AQA_ROLES = ["QP", "MS", "ER"] def aqa_url(papercode: str, role: str, session: str) -> Tuple[str, str]: mon = session[:3] yy = session[3:] folder, _ = MONTHS[mon] year = "20" + yy fname = f"AQA-{papercode}-{ROLE_TOKEN[role]}-{session}.PDF" return f"{AQA_BASE}/{year}/{folder}/{fname}", fname def session_pretty(session: str) -> Tuple[str, str]: mon = session[:3] # "JUN" | "NOV" yy = session[3:] # "22" _, pretty = MONTHS[mon] # ("2022-Jun" display session, "2022JUN" compact for exam_code — year-first, matches the # locked exam_code convention and the Edexcel/OCR entries). return f"20{yy}-{pretty}", f"20{yy}{mon}" def build_aqa() -> Dict[str, Any]: candidates: List[Tuple[str, str, str, str, str, str, Optional[str], str, str, str]] = [] # (spec_code, subject, award, paper_fc, paper_code, tier, role, session, url, fname) spec_meta = {} for spec_code, subject, award, level, first_teach, papers in AQA_SPECS: spec_meta[spec_code] = (subject, award, level, first_teach) for paper_fc, paper_code, tier in papers: for session in AQA_SESSIONS: for role in AQA_ROLES: url, fname = aqa_url(paper_fc, role, session) candidates.append((spec_code, subject, award, paper_fc, paper_code, tier, role, session, url, fname)) print(f"[AQA] HEAD-verifying {len(candidates)} candidate URLs...", file=sys.stderr) live: Dict[int, bool] = {} with cf.ThreadPoolExecutor(max_workers=24) as ex: futs = {ex.submit(head_ok, c[8]): i for i, c in enumerate(candidates)} done = 0 for fut in cf.as_completed(futs): i = futs[fut] live[i] = fut.result() done += 1 if done % 60 == 0: print(f" ...{done}/{len(candidates)} ({sum(live.values())} live)", file=sys.stderr) specs: Dict[str, Dict[str, Any]] = {} for i, c in enumerate(candidates): if not live.get(i): continue spec_code, subject, award, paper_fc, paper_code, tier, role, session, url, fname = c sess_pretty, sess_compact = session_pretty(session) token = paper_fc[len(award):] # "1H" / "P1H" / "1" exam_code = f"AQA-{award}-{token}-{sess_compact}-{role}" spec = specs.setdefault(spec_code, {"papers": []}) spec["papers"].append({ "exam_code": exam_code, "paper_code": paper_code, "tier": tier, "session": sess_pretty, "doc_type": role, "file": { "source": f"url:{url}", "original_name": fname, "provenance": {"source_url": url, "fetched": FETCHED, "license": "AQA public past paper"}, }, }) spec_list = [] for spec_code, subject, award, level, first_teach, _papers in AQA_SPECS: if spec_code not in specs: continue papers = sorted(specs[spec_code]["papers"], key=lambda p: p["exam_code"]) spec_list.append({ "spec_code": spec_code, "exam_board_code": "AQA", "subject_code": subject, "award_code": award, "award_level": level, "first_teach": first_teach, "papers": papers, }) print(f"[AQA] {spec_code}: {len(papers)} live papers", file=sys.stderr) return {"exam_board_code": "AQA", "specifications": spec_list} # ─────────────── Edexcel / OCR — confirmed direct URLs (re-verified at build) ─────────────── # These boards aren't templatable (Edexcel has a non-derivable date suffix; OCR uses opaque # doc-ids), so confirmed URLs are listed as 6-tuples: (spec_code, paper_code, tier, session, role, # url). exam_code is DERIVED (see _mk_exam_code) so it always matches the locked convention. EXAM_CODE_PREFIX = {"EDEXCEL": "EDX", "OCR": "OCR"} def _ec_token(paper_code: str) -> str: t = paper_code.split("/")[-1] return str(int(t)) if t.isdigit() else t # "01"->"1", "1H"->"1H", "1CH"->"1CH", "11"->"11" def _mk_exam_code(prefix: str, award: str, paper_code: str, session: str, role: str) -> str: y, m = session.split("-") return f"{prefix}-{award}-{_ec_token(paper_code)}-{y}{m.upper()}-{role}" _PE = "https://qualifications.pearson.com/content/dam/pdf" _EDX = f"{_PE}/GCSE/Science/2016" _OCR = "https://www.ocr.org.uk/Images" EDEXCEL_SPECS = { "EDX-BIOL-1BI0": ("BIOLOGY", "1BI0", "GCSE", "2016"), "EDX-CHEM-1CH0": ("CHEMISTRY", "1CH0", "GCSE", "2016"), "EDX-PHYS-1PH0": ("PHYSICS", "1PH0", "GCSE", "2016"), "EDX-COMB-1SC0": ("COMBINED SCIENCE", "1SC0", "GCSE", "2016"), "EDX-MATH-1MA1": ("MATHEMATICS", "1MA1", "GCSE", "2015"), "EDX-ENGL-1EN0": ("ENGLISH LANGUAGE", "1EN0", "GCSE", "2015"), "EDX-ENGLIT-1ET0": ("ENGLISH LITERATURE", "1ET0", "GCSE", "2015"), "EDX-GEOG-1GA0": ("GEOGRAPHY A", "1GA0", "GCSE", "2016"), "EDX-HIST-1HI0": ("HISTORY", "1HI0", "GCSE", "2016"), "EDX-BUS-1BS0": ("BUSINESS", "1BS0", "GCSE", "2017"), "EDX-COMP-1CP2": ("COMPUTER SCIENCE", "1CP2", "GCSE", "2020"), "EDX-MATH-9MA0": ("MATHEMATICS", "9MA0", "A-level", "2017"), "EDX-ENGL-9EN0": ("ENGLISH LANGUAGE", "9EN0", "A-level", "2015"), "EDX-ENGLIT-9ET0": ("ENGLISH LITERATURE", "9ET0", "A-level", "2015"), "EDX-GEOG-9GE0": ("GEOGRAPHY", "9GE0", "A-level", "2016"), } EDEXCEL_PAPERS = [ # ── Sciences (round 1) ── ("EDX-BIOL-1BI0", "1BI0/1H", "H", "2024-Jun", "QP", f"{_EDX}/Exam-materials/1bi0-1h-que-20240511.pdf"), ("EDX-BIOL-1BI0", "1BI0/2F", "F", "2023-Jun", "QP", f"{_EDX}/Exam-materials/1bi0-2f-que-20230610.pdf"), ("EDX-BIOL-1BI0", "1BI0/2H", "H", "2023-Jun", "QP", f"{_EDX}/Exam-materials/1bi0-2h-que-20230610.pdf"), ("EDX-BIOL-1BI0", "1BI0/1F", "F", "2023-Jun", "MS", f"{_EDX}/Exam-materials/1bi0-1f-rms-20230824.pdf"), ("EDX-BIOL-1BI0", "1BI0/1H", "H", "2024-Jun", "MS", f"{_EDX}/Exam-materials/1bi0-1h-rms-20240822.pdf"), ("EDX-BIOL-1BI0", "1BI0/1H", "H", "2022-Jun", "MS", f"{_EDX}/exam-materials/1bi0-1h-rms-20220825.pdf"), ("EDX-CHEM-1CH0", "1CH0/1F", "F", "2023-Jun", "QP", f"{_EDX}/Exam-materials/1ch0-1f-que-20230523.pdf"), ("EDX-CHEM-1CH0", "1CH0/1H", "H", "2024-Jun", "QP", f"{_EDX}/Exam-materials/1ch0-1h-que-20240518.pdf"), ("EDX-CHEM-1CH0", "1CH0/2H", "H", "2024-Jun", "MS", f"{_EDX}/Exam-materials/1ch0-2h-rms-20240822.pdf"), ("EDX-PHYS-1PH0", "1PH0/1H", "H", "2023-Jun", "QP", f"{_EDX}/Exam-materials/1ph0-1h-que-20230526.pdf"), ("EDX-PHYS-1PH0", "1PH0/2F", "F", "2023-Jun", "QP", f"{_EDX}/Exam-materials/1ph0-2f-que-20230617.pdf"), ("EDX-PHYS-1PH0", "1PH0/1H", "H", "2024-Jun", "QP", f"{_EDX}/Exam-materials/1ph0-1h-que-20240523.pdf"), ("EDX-PHYS-1PH0", "1PH0/2H", "H", "2023-Jun", "MS", f"{_EDX}/Exam-materials/1ph0-2h-rms-20230824.pdf"), ("EDX-PHYS-1PH0", "1PH0/2H", "H", "2022-Jun", "MS", f"{_EDX}/exam-materials/1ph0-2h-rms-20220825.pdf"), ("EDX-COMB-1SC0", "1SC0/1CH", None, "2023-Jun", "MS", f"{_EDX}/Exam-materials/1sc0-1ch-rms-20230824.pdf"), # ── Maths 1MA1 (round 2) ── ("EDX-MATH-1MA1", "1MA1/1H", "H", "2023-Jun", "QP", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1h-que-20230520.pdf"), ("EDX-MATH-1MA1", "1MA1/1H", "H", "2023-Jun", "MS", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1h-rms-20230824.pdf"), ("EDX-MATH-1MA1", "1MA1/1F", "F", "2023-Jun", "MS", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1f-rms-20230824.pdf"), ("EDX-MATH-1MA1", "1MA1/1F", "F", "2024-Jun", "QP", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1f-que-20240517.pdf"), ("EDX-MATH-1MA1", "1MA1/1H", "H", "2024-Jun", "QP", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1h-que-20240517.pdf"), ("EDX-MATH-1MA1", "1MA1/1F", "F", "2024-Jun", "MS", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1f-rms-20240822.pdf"), ("EDX-MATH-1MA1", "1MA1/1H", "H", "2023-Nov", "MS", f"{_PE}/GCSE/Mathematics/2015/Exam-materials/1ma1-1h-rms-20240111.pdf"), ("EDX-MATH-1MA1", "1MA1/1H", "H", "2022-Jun", "MS", f"{_PE}/GCSE/mathematics/2015/exam-materials/1ma1-1h-rms-20220825.pdf"), ("EDX-MATH-1MA1", "1MA1/3H", "H", "2022-Jun", "MS", f"{_PE}/GCSE/mathematics/2015/exam-materials/1ma1-3h-rms-20220825.pdf"), # ── English Language 1EN0 / Literature 1ET0 (round 2) ── ("EDX-ENGL-1EN0", "1EN0/01", None, "2024-Jun", "QP", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-01-que-20240524.pdf"), ("EDX-ENGL-1EN0", "1EN0/01", None, "2023-Nov", "QP", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-01-que-20231108.pdf"), ("EDX-ENGL-1EN0", "1EN0/01", None, "2024-Jun", "MS", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-01-rms-20240822.pdf"), ("EDX-ENGL-1EN0", "1EN0/02", None, "2024-Jun", "MS", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-02-rms-20240822.pdf"), ("EDX-ENGL-1EN0", "1EN0/01", None, "2023-Jun", "MS", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-01-rms-20230824.pdf"), ("EDX-ENGL-1EN0", "1EN0/02", None, "2023-Jun", "MS", f"{_PE}/GCSE/English-Language/2015/Exam-materials/1en0-02-rms-20230824.pdf"), ("EDX-ENGLIT-1ET0", "1ET0/01", None, "2023-Jun", "QP", f"{_PE}/GCSE/English-Literature/2015/Exam-materials/1et0-01-que-20230518.pdf"), ("EDX-ENGLIT-1ET0", "1ET0/02", None, "2023-Jun", "QP", f"{_PE}/GCSE/English-Literature/2015/Exam-materials/1et0-02-que-20230525.pdf"), ("EDX-ENGLIT-1ET0", "1ET0/02", None, "2024-Jun", "QP", f"{_PE}/GCSE/English-Literature/2015/Exam-materials/1et0-02-que-20240521.pdf"), ("EDX-ENGLIT-1ET0", "1ET0/01", None, "2023-Jun", "MS", f"{_PE}/GCSE/English-Literature/2015/Exam-materials/1et0-01-rms-20230824.pdf"), ("EDX-ENGLIT-1ET0", "1ET0/01", None, "2024-Jun", "MS", f"{_PE}/GCSE/English-Literature/2015/Exam-materials/1et0-01-rms-20240822.pdf"), # ── A-level Maths 9MA0 / English 9EN0 / 9ET0 (round 2) ── ("EDX-MATH-9MA0", "9MA0/01", None, "2023-Jun", "QP", f"{_PE}/A-Level/Mathematics/2017/Exam-materials/9ma0-01-que-20230607.pdf"), ("EDX-MATH-9MA0", "9MA0/31", None, "2023-Jun", "QP", f"{_PE}/A-Level/Mathematics/2017/Exam-materials/9ma0-31-que-20230621.pdf"), ("EDX-MATH-9MA0", "9MA0/02", None, "2024-Jun", "QP", f"{_PE}/A-Level/Mathematics/2017/Exam-materials/9ma0-02-que-20240612.pdf"), ("EDX-MATH-9MA0", "9MA0/31", None, "2023-Jun", "MS", f"{_PE}/A-Level/Mathematics/2017/Exam-materials/9ma0-31-rms-20230817.pdf"), ("EDX-MATH-9MA0", "9MA0/01", None, "2024-Jun", "MS", f"{_PE}/A-Level/Mathematics/2017/Exam-materials/9ma0-01-rms-20240815.pdf"), ("EDX-ENGL-9EN0", "9EN0/01", None, "2024-Jun", "MS", f"{_PE}/A-Level/English-Language/2015/Exam-materials/9en0-01-rms-20240815.pdf"), ("EDX-ENGL-9EN0", "9EN0/02", None, "2024-Jun", "MS", f"{_PE}/A-Level/English-Language/2015/Exam-materials/9en0-02-rms-20240815.pdf"), ("EDX-ENGLIT-9ET0", "9ET0/01", None, "2024-Jun", "QP", f"{_PE}/A-Level/English-Literature/2015/Exam-materials/9et0-01-que-20240525.pdf"), ("EDX-ENGLIT-9ET0", "9ET0/01", None, "2023-Jun", "MS", f"{_PE}/A-Level/English-Literature/2015/Exam-materials/9et0-01-rms-20230817.pdf"), ("EDX-ENGLIT-9ET0", "9ET0/03", None, "2023-Jun", "MS", f"{_PE}/A-Level/English-Literature/2015/Exam-materials/9et0-03-rms-20230817.pdf"), # ── Humanities (round 2) ── ("EDX-GEOG-1GA0", "1GA0/01", None, "2023-Jun", "QP", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-01-que-20230523.pdf"), ("EDX-GEOG-1GA0", "1GA0/01", None, "2023-Jun", "MS", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-01-rms-20230824.pdf"), ("EDX-GEOG-1GA0", "1GA0/02", None, "2023-Jun", "QP", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-02-que-20230610.pdf"), ("EDX-GEOG-1GA0", "1GA0/02", None, "2023-Jun", "MS", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-02-rms-20230824.pdf"), ("EDX-GEOG-1GA0", "1GA0/01", None, "2024-Jun", "MS", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-01-rms-20240822.pdf"), ("EDX-GEOG-1GA0", "1GA0/03", None, "2024-Jun", "QP", f"{_PE}/GCSE/Geography-A/2016/Exam-materials/1ga0-03-que-20240615.pdf"), ("EDX-HIST-1HI0", "1HI0/10", None, "2023-Jun", "QP", f"{_PE}/GCSE/History/2016/Exam-materials/1hi0-10-que-20230519.pdf"), ("EDX-HIST-1HI0", "1HI0/10", None, "2023-Jun", "MS", f"{_PE}/GCSE/History/2016/Exam-materials/1hi0-10-rms-20230824.pdf"), ("EDX-HIST-1HI0", "1HI0/12", None, "2023-Jun", "MS", f"{_PE}/GCSE/History/2016/Exam-materials/1hi0-12-rms-20230824.pdf"), ("EDX-HIST-1HI0", "1HI0/13", None, "2024-Jun", "MS", f"{_PE}/GCSE/History/2016/Exam-materials/1hi0-13-rms-20240822.pdf"), ("EDX-HIST-1HI0", "1HI0/33", None, "2023-Jun", "MS", f"{_PE}/GCSE/History/2016/Exam-materials/1hi0-33-rms-20230824.pdf"), ("EDX-BUS-1BS0", "1BS0/01", None, "2023-Jun", "QP", f"{_PE}/GCSE/Business/2017/Exam-materials/1bs0-01-que-20230519.pdf"), ("EDX-BUS-1BS0", "1BS0/02", None, "2023-Jun", "QP", f"{_PE}/GCSE/Business/2017/Exam-materials/1bs0-02-que-20230613.pdf"), ("EDX-BUS-1BS0", "1BS0/02", None, "2023-Jun", "MS", f"{_PE}/GCSE/Business/2017/Exam-materials/1bs0-02-rms-20230824.pdf"), ("EDX-BUS-1BS0", "1BS0/02", None, "2024-Jun", "QP", f"{_PE}/GCSE/Business/2017/Exam-materials/1bs0-02-que-20240606.pdf"), ("EDX-BUS-1BS0", "1BS0/01", None, "2024-Jun", "MS", f"{_PE}/GCSE/Business/2017/Exam-materials/1bs0-01-rms-20240822.pdf"), ("EDX-COMP-1CP2", "1CP2/01", None, "2023-Jun", "QP", f"{_PE}/GCSE/Computer-science/2020/Exam-materials/1cp2-01-que-20230520.pdf"), ("EDX-COMP-1CP2", "1CP2/01", None, "2023-Jun", "MS", f"{_PE}/GCSE/Computer-science/2020/Exam-materials/1cp2-01-rms-20230824.pdf"), ("EDX-COMP-1CP2", "1CP2/02", None, "2023-Jun", "QP", f"{_PE}/GCSE/Computer-science/2020/Exam-materials/1cp2-02-que-20230526.pdf"), ("EDX-COMP-1CP2", "1CP2/01", None, "2024-Jun", "QP", f"{_PE}/GCSE/Computer-Science/2020/Exam-materials/1cp2-01-que-20240702.pdf"), ("EDX-COMP-1CP2", "1CP2/01", None, "2024-Jun", "MS", f"{_PE}/GCSE/Computer-science/2020/Exam-materials/1cp2-01-rms-20240822.pdf"), ("EDX-GEOG-9GE0", "9GE0/01", None, "2023-Jun", "QP", f"{_PE}/A-Level/Geography/2016/Exam-materials/9ge0-01-que-20230518.pdf"), ] OCR_SPECS = { "OCR-BIOL-J247": ("BIOLOGY", "J247", "GCSE", "2016"), "OCR-CHEM-J248": ("CHEMISTRY", "J248", "GCSE", "2016"), "OCR-PHYS-J249": ("PHYSICS", "J249", "GCSE", "2016"), "OCR-COMB-J250": ("COMBINED SCIENCE", "J250", "GCSE", "2016"), "OCR-MATH-J560": ("MATHEMATICS", "J560", "GCSE", "2015"), "OCR-ENGL-J351": ("ENGLISH LANGUAGE", "J351", "GCSE", "2015"), "OCR-ENGLIT-J352": ("ENGLISH LITERATURE", "J352", "GCSE", "2015"), "OCR-COMP-J277": ("COMPUTER SCIENCE", "J277", "GCSE", "2020"), "OCR-GEOG-J383": ("GEOGRAPHY A", "J383", "GCSE", "2016"), "OCR-BUS-J204": ("BUSINESS", "J204", "GCSE", "2017"), "OCR-HIST-J411": ("HISTORY B (SHP)", "J411", "GCSE", "2016"), "OCR-MATH-H240": ("MATHEMATICS A", "H240", "A-level", "2017"), "OCR-ENGLIT-H472": ("ENGLISH LITERATURE", "H472", "A-level", "2015"), "OCR-ENGL-H470": ("ENGLISH LANGUAGE", "H470", "A-level", "2015"), } OCR_PAPERS = [ # ── Sciences (round 1) ── ("OCR-BIOL-J247", "J247/01", "F", "2024-Jun", "QP", f"{_OCR}/727713-question-paper-paper-1.pdf"), ("OCR-BIOL-J247", "J247/01", "F", "2024-Jun", "MS", f"{_OCR}/727745-mark-scheme-paper-1.pdf"), ("OCR-BIOL-J247", "J247/03", "H", "2024-Jun", "QP", f"{_OCR}/727715-question-paper-paper-3.pdf"), ("OCR-BIOL-J247", "J247/03", "H", "2024-Jun", "MS", f"{_OCR}/727747-mark-scheme-paper-3.pdf"), ("OCR-BIOL-J247", "J247/01", "F", "2023-Jun", "QP", f"{_OCR}/704945-question-paper-paper-1.pdf"), ("OCR-BIOL-J247", "J247/03", "H", "2023-Jun", "MS", f"{_OCR}/704979-mark-scheme-paper-3.pdf"), ("OCR-BIOL-J247", "J247/03", "H", "2022-Jun", "QP", f"{_OCR}/678031-question-paper-paper-3.pdf"), ("OCR-BIOL-J247", "J247/01", "F", "2022-Jun", "MS", f"{_OCR}/678076-mark-scheme-paper-1.pdf"), ("OCR-CHEM-J248", "J248/01", "F", "2024-Jun", "QP", f"{_OCR}/727718-question-paper-paper-1.pdf"), ("OCR-CHEM-J248", "J248/03", "H", "2024-Jun", "MS", f"{_OCR}/727751-mark-scheme-paper-3.pdf"), ("OCR-CHEM-J248", "J248/01", "F", "2023-Jun", "QP", f"{_OCR}/704950-question-paper-paper-1.pdf"), ("OCR-CHEM-J248", "J248/03", "H", "2022-Jun", "QP", f"{_OCR}/678036-question-paper-paper-3.pdf"), ("OCR-PHYS-J249", "J249/01", "F", "2024-Jun", "QP", f"{_OCR}/727724-question-paper-paper-1.pdf"), ("OCR-PHYS-J249", "J249/03", "H", "2024-Jun", "MS", f"{_OCR}/727755-mark-scheme-paper-3.pdf"), ("OCR-PHYS-J249", "J249/01", "F", "2023-Jun", "QP", f"{_OCR}/704956-question-paper-paper-1.pdf"), ("OCR-PHYS-J249", "J249/03", "H", "2022-Jun", "MS", f"{_OCR}/678086-mark-scheme-paper-3.pdf"), ("OCR-COMB-J250", "J250/01", "F", "2024-Jun", "QP", f"{_OCR}/727730-question-paper-paper-1.pdf"), ("OCR-COMB-J250", "J250/07", "H", "2024-Jun", "MS", f"{_OCR}/727763-mark-scheme-paper-7.pdf"), # ── Maths J560 (round 2) ── ("OCR-MATH-J560", "J560/01", "F", "2024-Jun", "QP", f"{_OCR}/727817-question-paper-paper-1.pdf"), ("OCR-MATH-J560", "J560/01", "F", "2024-Jun", "MS", f"{_OCR}/727824-mark-scheme-paper-1.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2024-Jun", "QP", f"{_OCR}/727820-question-paper-paper-4.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2024-Jun", "MS", f"{_OCR}/727827-mark-scheme-paper-4.pdf"), ("OCR-MATH-J560", "J560/01", "F", "2023-Jun", "QP", f"{_OCR}/705050-question-paper-paper-1.pdf"), ("OCR-MATH-J560", "J560/01", "F", "2023-Jun", "MS", f"{_OCR}/705057-mark-scheme-paper-1.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2023-Jun", "QP", f"{_OCR}/705053-question-paper-paper-4.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2023-Jun", "MS", f"{_OCR}/705060-mark-scheme-paper-4.pdf"), ("OCR-MATH-J560", "J560/01", "F", "2022-Jun", "QP", f"{_OCR}/678149-question-paper-paper-1.pdf"), ("OCR-MATH-J560", "J560/01", "F", "2022-Jun", "MS", f"{_OCR}/678156-mark-scheme-paper-1.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2022-Jun", "QP", f"{_OCR}/678152-question-paper-paper-4.pdf"), ("OCR-MATH-J560", "J560/04", "H", "2022-Jun", "MS", f"{_OCR}/678159-mark-scheme-paper-4.pdf"), # ── English Language J351 / Literature J352 (round 2) ── ("OCR-ENGL-J351", "J351/01", None, "2024-Jun", "QP", f"{_OCR}/727556-question-paper-communicating-information-and-ideas.pdf"), ("OCR-ENGL-J351", "J351/01", None, "2024-Jun", "MS", f"{_OCR}/727658-mark-scheme-communication-information-and-ideas.pdf"), ("OCR-ENGL-J351", "J351/02", None, "2024-Jun", "QP", f"{_OCR}/727558-question-paper-exploring-effects-and-impact.pdf"), ("OCR-ENGL-J351", "J351/02", None, "2024-Jun", "MS", f"{_OCR}/727659-mark-scheme-exploring-effects-and-impact.pdf"), ("OCR-ENGL-J351", "J351/01", None, "2023-Jun", "QP", f"{_OCR}/704782-question-paper-communicating-information-and-ideas.pdf"), ("OCR-ENGL-J351", "J351/01", None, "2023-Jun", "MS", f"{_OCR}/704888-mark-scheme-communication-information-and-ideas.pdf"), ("OCR-ENGL-J351", "J351/01", None, "2022-Jun", "QP", f"{_OCR}/677852-question-paper-communicating-information-and-ideas.pdf"), ("OCR-ENGL-J351", "J351/01", None, "2022-Jun", "MS", f"{_OCR}/677967-mark-scheme-communication-information-and-ideas.pdf"), ("OCR-ENGLIT-J352", "J352/01", None, "2024-Jun", "QP", f"{_OCR}/727830-question-paper-exploring-modern-and-literary-heritage-texts.pdf"), ("OCR-ENGLIT-J352", "J352/01", None, "2024-Jun", "MS", f"{_OCR}/727832-mark-scheme-exploring-modern-and-literary-heritage-texts.pdf"), ("OCR-ENGLIT-J352", "J352/02", None, "2024-Jun", "QP", f"{_OCR}/727831-question-paper-exploring-poetry-and-shakespeare.pdf"), ("OCR-ENGLIT-J352", "J352/02", None, "2024-Jun", "MS", f"{_OCR}/727833-mark-scheme-exploring-poetry-and-shakespeare.pdf"), ("OCR-ENGLIT-J352", "J352/01", None, "2023-Jun", "QP", f"{_OCR}/705069-question-paper-exploring-modern-and-literary-heritage-texts.pdf"), ("OCR-ENGLIT-J352", "J352/01", None, "2023-Jun", "MS", f"{_OCR}/705075-mark-scheme-exploring-modern-and-literary-heritage-texts.pdf"), # ── A-level Maths H240 / English Lit H472 / Lang H470 (round 2) ── ("OCR-MATH-H240", "H240/01", None, "2024-Jun", "QP", f"{_OCR}/726654-question-paper-pure-mathematics.pdf"), ("OCR-MATH-H240", "H240/01", None, "2024-Jun", "MS", f"{_OCR}/726795-mark-scheme-pure-mathematics.pdf"), ("OCR-MATH-H240", "H240/02", None, "2024-Jun", "QP", f"{_OCR}/726656-question-paper-pure-mathematics-and-statistics.pdf"), ("OCR-MATH-H240", "H240/02", None, "2024-Jun", "MS", f"{_OCR}/726796-mark-scheme-pure-mathematics-and-statistics.pdf"), ("OCR-MATH-H240", "H240/01", None, "2023-Jun", "QP", f"{_OCR}/703866-question-paper-pure-mathematics.pdf"), ("OCR-MATH-H240", "H240/01", None, "2023-Jun", "MS", f"{_OCR}/704008-mark-scheme-pure-mathematics.pdf"), ("OCR-MATH-H240", "H240/01", None, "2022-Jun", "QP", f"{_OCR}/676845-question-paper-pure-mathematics.pdf"), ("OCR-MATH-H240", "H240/01", None, "2022-Jun", "MS", f"{_OCR}/677005-mark-scheme-pure-mathematics.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2024-Jun", "QP", f"{_OCR}/726602-question-paper-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2024-Jun", "MS", f"{_OCR}/726762-mark-scheme-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2023-Jun", "QP", f"{_OCR}/703813-question-paper-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2023-Jun", "MS", f"{_OCR}/703974-mark-scheme-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2022-Jun", "QP", f"{_OCR}/676783-question-paper-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGLIT-H472", "H472/01", None, "2022-Jun", "MS", f"{_OCR}/676965-mark-scheme-drama-and-poetry-pre-1900.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2024-Jun", "QP", f"{_OCR}/726595-question-paper-exploring-language.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2024-Jun", "MS", f"{_OCR}/726764-mark-scheme-exploring-language.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2023-Jun", "QP", f"{_OCR}/703806-question-paper-exploring-language.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2023-Jun", "MS", f"{_OCR}/703976-mark-scheme-exploring-language.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2022-Jun", "QP", f"{_OCR}/676772-question-paper-exploring-language.pdf"), ("OCR-ENGL-H470", "H470/01", None, "2022-Jun", "MS", f"{_OCR}/676967-mark-scheme-exploring-language.pdf"), # ── Humanities (round 2) ── ("OCR-COMP-J277", "J277/01", None, "2024-Jun", "QP", f"{_OCR}/727534-question-paper-computer-systems.pdf"), ("OCR-COMP-J277", "J277/01", None, "2024-Jun", "MS", f"{_OCR}/727652-mark-scheme-computer-systems.pdf"), ("OCR-COMP-J277", "J277/02", None, "2024-Jun", "QP", f"{_OCR}/727535-question-paper-computational-thinking-algorithms-and-programming.pdf"), ("OCR-COMP-J277", "J277/02", None, "2024-Jun", "MS", f"{_OCR}/727653-mark-scheme-computational-thinking-algorithms-and-programming.pdf"), ("OCR-GEOG-J383", "J383/01", None, "2024-Jun", "QP", f"{_OCR}/727564-question-paper-living-in-the-uk-today.pdf"), ("OCR-GEOG-J383", "J383/01", None, "2024-Jun", "MS", f"{_OCR}/727661-mark-scheme-living-in-the-uk-today.pdf"), ("OCR-GEOG-J383", "J383/02", None, "2024-Jun", "QP", f"{_OCR}/727566-question-paper-the-world-around-us.pdf"), ("OCR-GEOG-J383", "J383/02", None, "2024-Jun", "MS", f"{_OCR}/727662-mark-scheme-the-world-around-us.pdf"), ("OCR-BUS-J204", "J204/01", None, "2024-Jun", "QP", f"{_OCR}/727519-question-paper-business-1-business-activity-marketing-and-people.pdf"), ("OCR-BUS-J204", "J204/01", None, "2024-Jun", "MS", f"{_OCR}/727634-mark-scheme-business-1-business-activity-marketing-and-people.pdf"), ("OCR-BUS-J204", "J204/02", None, "2024-Jun", "QP", f"{_OCR}/727520-question-paper-business-2-operations-finance-and-influences-on-business.pdf"), ("OCR-BUS-J204", "J204/02", None, "2024-Jun", "MS", f"{_OCR}/727635-mark-scheme-business-2-operations-finance-and-influences-on-business.pdf"), ("OCR-BUS-J204", "J204/01", None, "2023-Jun", "QP", f"{_OCR}/704745-question-paper-business-1-business-activity-marketing-and-people.pdf"), ("OCR-BUS-J204", "J204/01", None, "2023-Jun", "MS", f"{_OCR}/704864-mark-scheme-business-1-business-activity-marketing-and-people.pdf"), ("OCR-HIST-J411", "J411/11", None, "2024-Jun", "QP", f"{_OCR}/727590-question-paper-the-people-s-health-c.1250-to-present-with-the-norman-conquest-1065-1087.pdf"), ("OCR-HIST-J411", "J411/11", None, "2024-Jun", "MS", f"{_OCR}/727678-mark-scheme-the-people-s-health-c.1250-to-present-with-the-norman-conquest-1065-1087.pdf"), ] def build_board(board_code: str, specs_meta: Dict, papers: List) -> Dict[str, Any]: prefix = EXAM_CODE_PREFIX[board_code] print(f"[{board_code}] re-verifying {len(papers)} confirmed URLs...", file=sys.stderr) live: Dict[int, bool] = {} with cf.ThreadPoolExecutor(max_workers=24) as ex: futs = {ex.submit(head_ok, p[5]): i for i, p in enumerate(papers)} for fut in cf.as_completed(futs): live[futs[fut]] = fut.result() by_spec: Dict[str, List[Dict[str, Any]]] = {} for i, (spec_code, paper_code, tier, session, role, url) in enumerate(papers): if not live.get(i): print(f" DROP (not live): {url}", file=sys.stderr) continue award = specs_meta[spec_code][1] by_spec.setdefault(spec_code, []).append({ "exam_code": _mk_exam_code(prefix, award, paper_code, session, role), "paper_code": paper_code, "tier": tier, "session": session, "doc_type": role, "file": {"source": f"url:{url}", "original_name": os.path.basename(url), "provenance": {"source_url": url, "fetched": FETCHED, "license": f"{board_code} public past paper"}}, }) spec_list = [] for spec_code, (subject, award, level, first_teach) in specs_meta.items(): if spec_code not in by_spec: continue spec_list.append({ "spec_code": spec_code, "exam_board_code": board_code, "subject_code": subject, "award_code": award, "award_level": level, "first_teach": first_teach, "papers": sorted(by_spec[spec_code], key=lambda p: p["exam_code"]), }) print(f"[{board_code}] {spec_code}: {len(by_spec[spec_code])} live papers", file=sys.stderr) return {"exam_board_code": board_code, "specifications": spec_list} def main() -> None: out_path = os.path.join(os.path.dirname(__file__), "exam-corpus.yaml") boards = [ build_aqa(), build_board("EDEXCEL", EDEXCEL_SPECS, EDEXCEL_PAPERS), build_board("OCR", OCR_SPECS, OCR_PAPERS), ] n_specs = sum(len(b["specifications"]) for b in boards) n_papers = sum(len(s["papers"]) for b in boards for s in b["specifications"]) manifest = { "version": 1, "defaults": {"bucket": "cc.examboards"}, "provenance": { "collected_by": "kcar", "collected_at": FETCHED, "license_posture": ("Public exam-board past papers downloaded from each board's own " "official site (AQA filestore, Pearson DAM, OCR Images). Stored in " "the private dev cc.examboards bucket for internal exam-marker dev/test. " "Each item records its source_url. Review redistribution rights before " "any public exposure."), "sources": { "AQA": "https://filestore.aqa.org.uk/sample-papers-and-mark-schemes/", "EDEXCEL": "https://qualifications.pearson.com/en/support/support-topics/exams/past-papers.html", "OCR": "https://www.ocr.org.uk/qualifications/past-paper-finder/", }, }, # Optional: uncomment + set on dev .94 to exercise user-side flows / first-sweep. # "test_subset": {"user_email": "teacher@kevlarai.test", "papers": 2}, # "system_identity": {"user_email": "teacher@kevlarai.test"}, "boards": boards, } with open(out_path, "w") as fh: yaml.safe_dump(manifest, fh, sort_keys=False, default_flow_style=False, width=120) print(f"\nWROTE {out_path}: {n_specs} specs, {n_papers} papers across {len(boards)} boards", file=sys.stderr) if __name__ == "__main__": main()