281 lines
11 KiB
Python
281 lines
11 KiB
Python
# core/scripture_normalizer.py
|
||
from __future__ import annotations
|
||
import re
|
||
from typing import Dict, List, Tuple, Optional
|
||
|
||
# =========================
|
||
# Canonical book abbreviations
|
||
# =========================
|
||
# House style (edit as you like). Keys are normalized to lower-case without periods.
|
||
BOOK_CANON: Dict[str, str] = {
|
||
# ----- OT -----
|
||
"genesis": "Gen.", "gen": "Gen.", "ge": "Gen.", "gn": "Gen.", "genesisesis": "Gen.",
|
||
"exodus": "Ex.", "ex": "Ex.", "exo": "Ex.",
|
||
"leviticus": "Lev.", "lev": "Lev.", "le": "Lev.",
|
||
"numbers": "Num.", "num": "Num.", "nu": "Num.", "nums": "Num.",
|
||
"deuteronomy": "Deut.", "deut": "Deut.", "deu": "Deut.", "dt": "Deut.", "deutronomy": "Deut.",
|
||
"joshua": "Josh.", "josh": "Josh.", "jos": "Josh.",
|
||
"judges": "Judg.", "judg": "Judg.", "jdg": "Judg.",
|
||
"ruth": "Ruth", "ru": "Ruth",
|
||
"1 samuel": "1 Sam.", "i samuel": "1 Sam.", "1 sam": "1 Sam.", "1sam": "1 Sam.", "1sa": "1 Sam.",
|
||
"2 samuel": "2 Sam.", "ii samuel": "2 Sam.", "2 sam": "2 Sam.", "2sam": "2 Sam.", "2sa": "2 Sam.",
|
||
"1 kings": "1 Ki.", "i kings": "1 Ki.", "1 ki": "1 Ki.", "1kgs": "1 Ki.", "1 kgs": "1 Ki.", "1kings": "1 Ki.", "1ki": "1 Ki.",
|
||
"2 kings": "2 Ki.", "ii kings": "2 Ki.", "2 ki": "2 Ki.", "2kgs": "2 Ki.", "2 kgs": "2 Ki.", "2kings": "2 Ki.", "2ki": "2 Ki.",
|
||
"1 chronicles": "1 Chron.", "i chronicles": "1 Chron.", "1 chron": "1 Chron.", "1ch": "1 Chron.", "1chron": "1 Chron.",
|
||
"2 chronicles": "2 Chron.", "ii chronicles": "2 Chron.", "2 chron": "2 Chron.", "2ch": "2 Chron.", "2chron": "2 Chron.",
|
||
"ezra": "Ezra", "ezr": "Ezra",
|
||
"nehemiah": "Neh.", "neh": "Neh.",
|
||
"esther": "Esth.", "esth": "Esth.", "es": "Esth.",
|
||
"job": "Job",
|
||
"psalm": "Ps.", "psalms": "Ps.", "ps": "Ps.", "psm": "Ps.", "pss": "Ps.",
|
||
"proverbs": "Prov.", "prov": "Prov.", "pr": "Prov.",
|
||
"ecclesiastes": "Eccl.", "eccles": "Eccl.", "eccl": "Eccl.", "ecc": "Eccl.", "ec": "Eccl.",
|
||
"song of solomon": "Song", "song of songs": "Song", "song": "Song", "so": "Song", "sos": "Song",
|
||
"isaiah": "Isa.", "isa": "Isa.", "is": "Isa.",
|
||
"jeremiah": "Jer.", "jer": "Jer.", "je": "Jer.",
|
||
"lamentations": "Lam.", "lam": "Lam.", "la": "Lam.",
|
||
"ezekiel": "Ezek.", "ezek": "Ezek.", "eze": "Ezek.", "ezk": "Ezek.",
|
||
"daniel": "Dan.", "dan": "Dan.", "da": "Dan.",
|
||
"hosea": "Hos.", "hos": "Hos.", "ho": "Hos.",
|
||
"joel": "Joel", "joe": "Joel", "jl": "Joel",
|
||
"amos": "Amos", "am": "Amos",
|
||
"obadiah": "Obad.", "obad": "Obad.", "ob": "Obad.",
|
||
"jonah": "Jon.", "jon": "Jon.",
|
||
"micah": "Mic.", "mic": "Mic.",
|
||
"nahum": "Nah.", "nah": "Nah.",
|
||
"habakkuk": "Hab.", "hab": "Hab.",
|
||
"zephaniah": "Zeph.", "zeph": "Zeph.", "zep": "Zeph.",
|
||
"haggai": "Hag.", "hag": "Hag.",
|
||
"zechariah": "Zech.", "zech": "Zech.", "zec": "Zech.",
|
||
"malachi": "Mal.", "mal": "Mal.",
|
||
# ----- NT -----
|
||
"matthew": "Matt.", "matt": "Matt.", "mt": "Matt.", "mat": "Matt.",
|
||
"mark": "Mark", "mrk": "Mark", "mk": "Mark", "mr": "Mark",
|
||
"luke": "Luke", "lk": "Luke",
|
||
"john": "John", "jn": "John", "jo": "John", "joh": "John",
|
||
"acts": "Acts", "act": "Acts", "ac": "Acts",
|
||
"romans": "Rom.", "rom": "Rom.", "ro": "Rom.", "rm": "Rom.",
|
||
"1 corinthians": "1 Cor.", "i corinthians": "1 Cor.", "1 cor": "1 Cor.", "1cor": "1 Cor.", "1co": "1 Cor.", "1 corinthians": "1 Cor.",
|
||
"2 corinthians": "2 Cor.", "ii corinthians": "2 Cor.", "2 cor": "2 Cor.", "2cor": "2 Cor.", "2co": "2 Cor.", "2 corinthians": "2 Cor.",
|
||
"galatians": "Gal.", "gal": "Gal.", "ga": "Gal.",
|
||
"ephesians": "Eph.", "eph": "Eph.", "eph.": "Eph.",
|
||
"philippians": "Phil.", "phil": "Phil.", "php": "Phil.", "phi": "Phil.", "philippians216": "Phil.",
|
||
"colossians": "Col.", "col": "Col.",
|
||
"1 thessalonians": "1 Thess.", "i thessalonians": "1 Thess.", "1 thess": "1 Thess.", "1 th": "1 Thess.", "1thess": "1 Thess.",
|
||
"2 thessalonians": "2 Thess.", "ii thessalonians": "2 Thess.", "2 thess": "2 Thess.", "2 th": "2 Thess.", "2thess": "2 Thess.",
|
||
"1 timothy": "1 Tim.", "i timothy": "1 Tim.", "1 tim": "1 Tim.", "1ti": "1 Tim.", "1 timothy": "1 Tim.",
|
||
"2 timothy": "2 Tim.", "ii timothy": "2 Tim.", "2 tim": "2 Tim.", "2ti": "2 Tim.", "2 timothy": "2 Tim.",
|
||
"titus": "Titus", "tit": "Titus", "ti.": "Titus",
|
||
"philemon": "Philem.", "philem": "Philem.", "phm": "Philem.",
|
||
"hebrews": "Heb.", "heb": "Heb.",
|
||
"james": "Jas.", "jas": "Jas.", "jam": "Jas.", "jms": "Jas.",
|
||
"1 peter": "1 Pet.", "i peter": "1 Pet.", "1 pet": "1 Pet.", "1pe": "1 Pet.",
|
||
"2 peter": "2 Pet.", "ii peter": "2 Pet.", "2 pet": "2 Pet.", "2pe": "2 Pet.",
|
||
"1 john": "1 John", "i john": "1 John", "1jn": "1 John", "1 jo": "1 John",
|
||
"2 john": "2 John", "ii john": "2 John", "2jn": "2 John", "2 jo": "2 John",
|
||
"3 john": "3 John", "iii john": "3 John", "3jn": "3 John", "3 jo": "3 John",
|
||
"jude": "Jude", "jud": "Jude",
|
||
"revelation": "Rev.", "rev": "Rev.", "re": "Rev.",
|
||
}
|
||
|
||
# also accept short two-word numbered patterns dynamically (e.g., "1 sam", "2 ki")
|
||
def _variants() -> Dict[str, str]:
|
||
base = dict(BOOK_CANON)
|
||
numbered = [
|
||
("samuel", "Sam."), ("kings", "Ki."), ("chronicles", "Chron."),
|
||
("corinthians", "Cor."), ("thessalonians", "Thess."),
|
||
("timothy", "Tim."), ("peter", "Pet."), ("john", "John"),
|
||
]
|
||
for n, prefix in (("1", "1"), ("i", "1"), ("2", "2"), ("ii", "2"), ("3", "3"), ("iii", "3")):
|
||
for name, abbr in numbered:
|
||
base[f"{n} {name}"] = f"{prefix} {abbr}"
|
||
base[f"{n}{name}"] = f"{prefix} {abbr}"
|
||
# ultra short no-space combos like "1co", "2ki"
|
||
base[f"{n}{name[:2]}"] = f"{prefix} {abbr}"
|
||
# very common shorthands that users type
|
||
base["ps"] = "Ps."
|
||
base["prov"] = "Prov."
|
||
base["eccles"] = "Eccl."
|
||
base["deut."] = "Deut."
|
||
base["mt"] = "Matt."
|
||
base["mk"] = "Mark"
|
||
base["lk"] = "Luke"
|
||
base["jn"] = "John"
|
||
base["ti"] = "Tim." # used only with a leading number, handled below
|
||
base["co"] = "Cor." # used only with a leading number, handled below
|
||
return base
|
||
|
||
BOOK_MAP = _variants()
|
||
|
||
# Words to strip like "Read", "chapter"
|
||
CRUFT_RE = re.compile(r"\b(read|see|chap(?:ter)?|ch)\b\.?", re.I)
|
||
|
||
# Book prefix (allows "1Co", "2 Pet.", "Rom", etc.)
|
||
BOOK_RE = re.compile(
|
||
r"""
|
||
^\s*
|
||
(?:
|
||
(?P<num>[1-3]|i{1,3})\s* # optional 1/2/3 or roman
|
||
)?
|
||
\s*
|
||
(?P<book>[A-Za-z\.]+(?:\s+[A-Za-z\.]+){0,2}) # book words (1-3 words)
|
||
""",
|
||
re.X,
|
||
)
|
||
|
||
# chapter/verse piece like "4:6,7-9" or "21"
|
||
C_V_RE = re.compile(
|
||
r"""
|
||
(?:
|
||
(?P<ch>\d+)
|
||
(?:
|
||
:(?P<vs>[\d,\-\u2013\u2014\s]+)
|
||
)?
|
||
)
|
||
""",
|
||
re.X,
|
||
)
|
||
|
||
def _clean_text(s: str) -> str:
|
||
s = s.replace("\xa0", " ").replace("\u2009", " ").replace("\u202f", " ")
|
||
s = CRUFT_RE.sub("", s)
|
||
s = s.replace("..", ".").replace("—", "-").replace("–", "-")
|
||
s = re.sub(r"\s{2,}", " ", s)
|
||
return s.strip(" ;,.\t\r\n ")
|
||
|
||
def _canon_key(raw: str) -> str:
|
||
key = raw.lower().strip()
|
||
key = key.replace(".", "")
|
||
key = re.sub(r"\s+", " ", key)
|
||
return key
|
||
|
||
def _canon_book(book_raw: str, num: Optional[str] = None) -> Optional[str]:
|
||
"""
|
||
Try to map a raw book token (with optional leading number) to our canonical abbr.
|
||
"""
|
||
key = _canon_key(book_raw)
|
||
# if the token was like "co" or "ti" and a number exists, bias to Cor./Tim.
|
||
if num and key in ("co", "cor"):
|
||
return f"{num} Cor."
|
||
if num and key in ("ti", "tim"):
|
||
return f"{num} Tim."
|
||
# direct lookup in our expanded map
|
||
if key in BOOK_MAP:
|
||
val = BOOK_MAP[key]
|
||
# If lookup returned a numbered abbr like "1 Cor." (already good)
|
||
return val
|
||
# sometimes users omit space: "1co", "2ki"
|
||
m = re.match(r"(?P<n>[1-3]|i{1,3})(?P(rest>[A-Za-z].*))$", key)
|
||
if m:
|
||
n = m.group("n")
|
||
rest = m.group("rest")
|
||
if rest in BOOK_MAP:
|
||
base = BOOK_MAP[rest]
|
||
# normalize roman to arabic
|
||
n = {"i": "1", "ii": "2", "iii": "3"}.get(n, n)
|
||
# if base already includes number, replace it
|
||
if base[0] in "123":
|
||
return f"{n} {base.split(' ', 1)[1]}"
|
||
return f"{n} {base}"
|
||
# last chance: try the core canon directly
|
||
return BOOK_CANON.get(key)
|
||
|
||
def _parse_segment(seg: str, last_book: Optional[str]) -> Tuple[Optional[str], Optional[str], bool]:
|
||
"""
|
||
Parse a semicolon-delimited segment.
|
||
Returns (book_canon, cv_string, preserve_raw).
|
||
- If preserve_raw is True, cv_string contains the original (cleaned) segment to keep verbatim.
|
||
"""
|
||
original = seg.strip()
|
||
s = _clean_text(seg)
|
||
if not s:
|
||
return (None, None, False)
|
||
|
||
# Try to detect a leading book token
|
||
m = BOOK_RE.match(s)
|
||
book = None
|
||
rest = s
|
||
if m:
|
||
num = (m.group("num") or "").strip()
|
||
raw_book = (m.group("book") or "").strip()
|
||
raw_joined = f"{num} {raw_book}".strip()
|
||
canon = _canon_book(raw_joined or raw_book, num=num or None)
|
||
if canon:
|
||
book = canon
|
||
rest = s[m.end():].strip(",;: .")
|
||
else:
|
||
# Not recognized as a book -> keep whole thing verbatim
|
||
return (None, original, True)
|
||
|
||
if not book:
|
||
# Inherit previous recognized book if we have one
|
||
if last_book:
|
||
book = last_book
|
||
else:
|
||
# There is no book context — keep segment verbatim to avoid data loss
|
||
return (None, original, True)
|
||
|
||
# Normalize the chapter/verse part
|
||
rest = re.sub(r"\s+", "", rest)
|
||
rest = re.sub(r":\s*", ":", rest)
|
||
|
||
if not rest:
|
||
return (book, None, False)
|
||
|
||
if C_V_RE.search(rest):
|
||
cv = rest.replace(" ", "")
|
||
else:
|
||
m2 = re.search(r"\d+(?::[\d,\-]+)?", rest)
|
||
cv = m2.group(0).replace(" ", "") if m2 else None
|
||
|
||
if not cv:
|
||
# Nothing parseable after a valid book; just keep the book
|
||
return (book, None, False)
|
||
|
||
# Normalize dashes, commas
|
||
cv = cv.replace("–", "-").replace("—", "-")
|
||
cv = re.sub(r",\s*", ",", cv)
|
||
|
||
return (book, cv, False)
|
||
|
||
def normalize_scripture_field(text: str) -> Tuple[str, List[str]]:
|
||
"""
|
||
Normalize an entire scripture_raw string.
|
||
- Unknown pieces are preserved verbatim.
|
||
- Known pieces are standardized and each segment repeats the book.
|
||
Returns (normalized_text, warnings).
|
||
"""
|
||
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:
|
||
out.append(cv or piece.strip())
|
||
# do not update last_book when we couldn't recognize a book
|
||
continue
|
||
|
||
if book and cv:
|
||
out.append(f"{book} {cv}")
|
||
last_book = book
|
||
continue
|
||
|
||
if book and not cv:
|
||
# book only (e.g., "Acts")
|
||
out.append(book)
|
||
last_book = book
|
||
continue
|
||
|
||
# If we get here: (no book, no preserve) — nothing useful to add
|
||
if piece.strip():
|
||
# As a final safeguard, keep original
|
||
out.append(piece.strip())
|
||
|
||
norm = "; ".join(x.strip() for x in out if x.strip())
|
||
norm = re.sub(r"\s+", " ", norm).strip(" ;,")
|
||
return (norm, warnings) |