254 lines
9.5 KiB
Python
254 lines
9.5 KiB
Python
# core/scripture_normalizer.py
|
||
from __future__ import annotations
|
||
import re
|
||
from typing import Dict, List, Optional, Tuple
|
||
|
||
# --------------------------
|
||
# Canonical book abbreviations
|
||
# --------------------------
|
||
BOOK_CANON: Dict[str, str] = {
|
||
# OT
|
||
"genesis": "Gen.", "gen": "Gen.", "genesisesis": "Gen.",
|
||
"exodus": "Ex.", "ex": "Ex.",
|
||
"leviticus": "Lev.", "lev": "Lev.",
|
||
"numbers": "Num.", "num": "Num.", "nums": "Num.",
|
||
"deuteronomy": "Deut.", "deut": "Deut.", "deutronomy": "Deut.", "deu": "Deut.",
|
||
"joshua": "Josh.", "josh": "Josh.",
|
||
"judges": "Judg.", "judg": "Judg.",
|
||
"ruth": "Ruth",
|
||
"1 samuel": "1 Sam.", "i samuel": "1 Sam.", "1 sam": "1 Sam.", "1sam": "1 Sam.",
|
||
"2 samuel": "2 Sam.", "ii samuel": "2 Sam.", "2 sam": "2 Sam.", "2sam": "2 Sam.",
|
||
"1 kings": "1 Ki.", "i kings": "1 Ki.", "1 ki": "1 Ki.", "1kgs": "1 Ki.", "1 kgs": "1 Ki.",
|
||
"2 kings": "2 Ki.", "ii kings": "2 Ki.", "2 ki": "2 Ki.", "2kgs": "2 Ki.", "2 kgs": "2 Ki.",
|
||
"1 chronicles": "1 Chron.", "i chronicles": "1 Chron.", "1 chron": "1 Chron.",
|
||
"2 chronicles": "2 Chron.", "ii chronicles": "2 Chron.", "2 chron": "2 Chron.",
|
||
"ezra": "Ezra",
|
||
"nehemiah": "Neh.", "neh": "Neh.",
|
||
"esther": "Esth.", "esth": "Esth.",
|
||
"job": "Job",
|
||
"psalms": "Ps.", "psalm": "Ps.", "ps": "Ps.", "pss": "Ps.",
|
||
"proverbs": "Prov.", "prov": "Prov.",
|
||
"ecclesiastes": "Eccl.", "eccles": "Eccl.", "eccl": "Eccl.",
|
||
"song of solomon": "Song", "song of songs": "Song", "song": "Song",
|
||
"isaiah": "Isa.", "isa": "Isa.", # IMPORTANT: direct alias so 'Isa.' never becomes '1 Sam.'
|
||
"jeremiah": "Jer.", "jer": "Jer.",
|
||
"lamentations": "Lam.", "lam": "Lam.",
|
||
"ezekiel": "Ezek.", "ezek": "Ezek.",
|
||
"daniel": "Dan.", "dan": "Dan.",
|
||
"hosea": "Hos.", "hos": "Hos.",
|
||
"joel": "Joel",
|
||
"amos": "Amos",
|
||
"obadiah": "Obad.", "obad": "Obad.",
|
||
"jonah": "Jon.", "jon": "Jon.",
|
||
"micah": "Mic.", "mic": "Mic.",
|
||
"nahum": "Nah.", "nah": "Nah.",
|
||
"habakkuk": "Hab.", "hab": "Hab.",
|
||
"zephaniah": "Zeph.", "zeph": "Zeph.",
|
||
"haggai": "Hag.", "hag": "Hag.",
|
||
"zechariah": "Zech.", "zech": "Zech.",
|
||
"malachi": "Mal.", "mal": "Mal.",
|
||
# NT
|
||
"matthew": "Matt.", "matt": "Matt.", "mt": "Matt.",
|
||
"mark": "Mark", "mk": "Mark", "mrk": "Mark", "mr": "Mark",
|
||
"luke": "Luke", "lk": "Luke",
|
||
"john": "John", "jn": "John", "joh": "John",
|
||
"acts": "Acts",
|
||
"romans": "Rom.", "rom": "Rom.",
|
||
"1 corinthians": "1 Cor.", "i corinthians": "1 Cor.", "1 cor": "1 Cor.", "1co": "1 Cor.",
|
||
"2 corinthians": "2 Cor.", "ii corinthians": "2 Cor.", "2 cor": "2 Cor.", "2co": "2 Cor.",
|
||
"galatians": "Gal.", "gal": "Gal.",
|
||
"ephesians": "Eph.", "eph": "Eph.",
|
||
"philippians": "Phil.", "phil": "Phil.", "philippians 216": "Phil.", # import glitch seen
|
||
"colossians": "Col.", "col": "Col.",
|
||
"1 thessalonians": "1 Thess.", "i thessalonians": "1 Thess.", "1 thess": "1 Thess.",
|
||
"2 thessalonians": "2 Thess.", "ii thessalonians": "2 Thess.", "2 thess": "2 Thess.",
|
||
"1 timothy": "1 Tim.", "i timothy": "1 Tim.", "1 tim": "1 Tim.", "1ti": "1 Tim.",
|
||
"2 timothy": "2 Tim.", "ii timothy": "2 Tim.", "2 tim": "2 Tim.", "2ti": "2 Tim.",
|
||
"titus": "Titus", "tit": "Titus",
|
||
"philemon": "Philem.", "philem": "Philem.",
|
||
"hebrews": "Heb.", "heb": "Heb.",
|
||
"james": "Jas.", "jas": "Jas.",
|
||
"1 peter": "1 Pet.", "i peter": "1 Pet.", "1 pet": "1 Pet.",
|
||
"2 peter": "2 Pet.", "ii peter": "2 Pet.", "2 pet": "2 Pet.",
|
||
"1 john": "1 John", "i john": "1 John",
|
||
"2 john": "2 John", "ii john": "2 John",
|
||
"3 john": "3 John", "iii john": "3 John",
|
||
"jude": "Jude",
|
||
"revelation": "Rev.", "rev": "Rev.", "1 ch": "1 Chron.", "1 chr": "1 Chron.", "1 Ch": "1 Chron.", "1 Chr": "1 Chron.",
|
||
"2 ch": "2 Chron.", "2 chr": "2 Chron.", "2 Ch": "2 Chron.", "2 Chr": "2 Chron.",
|
||
"ch": "Chron.", "chr": "Chron.",
|
||
}
|
||
|
||
# Short bare tokens that must *never* be treated as “roman numeral + book”
|
||
# (prevents 'Isa.' => '1 Sam.' and similar).
|
||
NEVER_PREFIX_NUMERAL = {"isa", "isaiah", "job", "joel", "amos", "nah", "hag", "mal", "rom", "gal", "eph", "tit", "heb", "jas", "jude"}
|
||
|
||
def _canon_key(s: str) -> str:
|
||
s = (s or "").strip().lower()
|
||
s = s.replace(".", " ")
|
||
s = re.sub(r"\s+", " ", s)
|
||
return s
|
||
|
||
def _direct_lookup(key: str) -> Optional[str]:
|
||
"""Try direct lookups with and without spaces (e.g., 'i samuel', '1co')."""
|
||
if key in BOOK_CANON:
|
||
return BOOK_CANON[key]
|
||
nospace = key.replace(" ", "")
|
||
if nospace in BOOK_CANON:
|
||
return BOOK_CANON[nospace]
|
||
return None
|
||
|
||
# Split helpers
|
||
CRUFT_RE = re.compile(r"\b(read|see|chap(?:ter)?|ch)\b\.?", re.I)
|
||
def _clean_text(s: str) -> str:
|
||
s = s.replace("\xa0", " ").replace("\u2009", " ").replace("\u202f", " ")
|
||
s = CRUFT_RE.sub("", s)
|
||
s = s.replace("—", "-").replace("–", "-")
|
||
s = re.sub(r"\s+", " ", s)
|
||
return s.strip(" ;,.\t\r\n ")
|
||
|
||
BOOK_PREFIX_RE = re.compile(
|
||
r"""
|
||
^\s*
|
||
(?:(?P<num>[1-3]|i{1,3})\s*)? # optional leading number (1/2/3 or i/ii/iii)
|
||
(?P<book>[A-Za-z\.]+(?:\s+[A-Za-z\.]+){0,2}) # 1-3 words for the book
|
||
""",
|
||
re.X,
|
||
)
|
||
|
||
C_V_RE = re.compile(r"(?P<ch>\d+)(?::(?P<vs>[\d,\-\s]+))?")
|
||
|
||
def _canon_book(book_raw: str, num: Optional[str]) -> Optional[str]:
|
||
"""
|
||
Decide the canonical book abbreviation.
|
||
Rules:
|
||
1) Always try direct alias matches first (with/without spaces).
|
||
2) If the *book* piece looks like a short token that could be misread as a roman
|
||
numeral + another book (e.g., 'Isa'), *never* do numeral heuristics.
|
||
3) Otherwise, allow '1/2/3 + co/cor/ti/tim/pet/peter/jn/john' style heuristics.
|
||
"""
|
||
key = _canon_key(book_raw)
|
||
# Step 1: direct lookups
|
||
direct = _direct_lookup(key)
|
||
if direct:
|
||
return direct
|
||
|
||
# Guard: tokens like 'isa' must not be split into 'i' + 'sa'
|
||
bare = key.replace(" ", "")
|
||
if bare in NEVER_PREFIX_NUMERAL:
|
||
return _direct_lookup(bare)
|
||
|
||
# Step 3: numeral + short token combos
|
||
if num:
|
||
n = {"i": "1", "ii": "2", "iii": "3"}.get(num.lower(), num)
|
||
short = key.replace(" ", "")
|
||
# common short targets where number matters
|
||
map_num = {
|
||
"co": "Cor.", "cor": "Cor.",
|
||
"ti": "Tim.", "tim": "Tim.",
|
||
"pe": "Pet.", "pet": "Pet.", "peter": "Pet.",
|
||
"jo": "John", "jn": "John", "joh": "John", "john": "John",
|
||
"sa": "Sam.", "sam": "Sam.", "samuel": "Sam.",
|
||
"ki": "Ki.", "kgs": "Ki.", "kings": "Ki.",
|
||
"chronicles": "Chron.", "chron": "Chron.",
|
||
"thessalonians": "Thess.", "thess": "Thess.",
|
||
}
|
||
if short in map_num:
|
||
return f"{n} {map_num[short]}"
|
||
|
||
# Fallback
|
||
return None
|
||
|
||
def _parse_segment(seg: str, last_book: Optional[str]) -> Tuple[Optional[str], Optional[str], bool]:
|
||
"""
|
||
Parse one semicolon-delimited piece.
|
||
Returns (book_canon, cv, preserve_original_if_unparsed).
|
||
"""
|
||
raw = _clean_text(seg)
|
||
if not raw:
|
||
return (None, None, False)
|
||
|
||
m = BOOK_PREFIX_RE.match(raw)
|
||
book_part = None
|
||
num = None
|
||
rest = raw
|
||
|
||
if m:
|
||
# What the regex thinks is a numeric prefix and a book word blob:
|
||
num = (m.group("num") or "").strip() or None
|
||
book_text = (m.group("book") or "").strip()
|
||
rest = raw[m.end():].strip(" :,;.")
|
||
|
||
# Try to resolve the book using robust rules
|
||
book_part = _canon_book(book_text if not num else f"{num} {book_text}", num=num)
|
||
|
||
# If regex split out a bogus 'i' from something like "Isa.", fix by trying full token directly.
|
||
if not book_part:
|
||
whole = _canon_book(book_text, None)
|
||
if whole:
|
||
book_part = whole
|
||
rest = raw[m.end():].strip(" :,;.")
|
||
|
||
# no book found → inherit previous if we have verses; otherwise preserve as-is
|
||
if not book_part:
|
||
# If there's obvious chapter/verse, but no book, we cannot link → preserve as-is.
|
||
if C_V_RE.search(rest):
|
||
return (None, None, True)
|
||
# Or the whole thing might already be a known book alone:
|
||
whole = _canon_book(raw, None)
|
||
if whole:
|
||
return (whole, None, False)
|
||
return (None, None, True)
|
||
|
||
# normalize chapter/verse
|
||
rest = rest.replace(" ", "")
|
||
if not rest:
|
||
return (book_part, None, False)
|
||
|
||
mcv = C_V_RE.search(rest)
|
||
if mcv:
|
||
cv = rest
|
||
cv = cv.replace(" ", "")
|
||
return (book_part, cv, False)
|
||
|
||
# not a recognizable cv → keep original piece untouched
|
||
return (None, None, True)
|
||
|
||
def normalize_scripture_field(text: str) -> Tuple[str, List[str]]:
|
||
"""
|
||
Normalize a whole scripture_raw string.
|
||
Returns (normalized_text, warnings).
|
||
Unknown/unparseable chunks are preserved verbatim.
|
||
"""
|
||
warnings: List[str] = []
|
||
if not text:
|
||
return ("", warnings)
|
||
|
||
pieces = [p for p in re.split(r"\s*;\s*", text) if p and p.strip()]
|
||
out: List[str] = []
|
||
last_book: Optional[str] = None
|
||
|
||
for piece in pieces:
|
||
book, cv, preserve = _parse_segment(piece, last_book)
|
||
|
||
if preserve:
|
||
# Keep the original chunk
|
||
out.append(_clean_text(piece))
|
||
continue
|
||
|
||
if book and not cv:
|
||
out.append(book)
|
||
last_book = book
|
||
continue
|
||
|
||
if not book and cv:
|
||
warnings.append(f"Missing book for '{piece.strip()}'")
|
||
continue
|
||
|
||
if book and cv:
|
||
out.append(f"{book} {cv}")
|
||
last_book = book
|
||
|
||
norm = "; ".join(s for s in (o.strip(" ;,") for o in out) if s)
|
||
norm = re.sub(r"\s+", " ", norm).strip(" ;,")
|
||
return (norm, warnings) |