#!/usr/bin/env python3
"""Build index.html: a searchable download page for the local book files.

Reads the book list from download.txt and links each entry to its local
copy in ./pdfs/ (the downloader preserves the "Book Name" filenames, with
" (2)", " (3)", ... suffixes for the few duplicate titles). Entries whose
local file is missing are shown as unavailable.

Usage:
    ./generate_index.py [download.txt] [pdfs_dir] [output.html]
"""
import csv
import datetime
import html
import json
import re
import sys
from collections import Counter
from pathlib import Path
from urllib.parse import quote

# Ordered (subject, keyword patterns). First match wins; matching is
# case-insensitive substring/regex on the book title.
SUBJECT_RULES = [
    ("Buddhism & Jainism", [r"buddh", r"bodhi", r"tibet", r"milarepa", r"jain", r"mahaveera", r"mahavira", r"deepankara", r"gatha", r"ashwaghos", r"ashwago", r"vaishaka poornima"]),
    ("Ayurveda & Medicine", [r"ayurveda", r"vaidya", r"medic", r"vriksha", r"mental health", r"vital force", r"\bhealth\b"]),
    ("Tantra & Sri Vidya", [r"tantra", r"tantrik", r"yantra", r"sri chakra", r"sri vidya", r"srividh?ya", r"lalit", r"bhavan"]),
    ("Yoga & Meditation", [r"dhyana", r"\bjhana\b", r"\bzen\b", r"\bom\b", r"\byoga\b"]),
    ("Vedanta & Philosophy", [r"vedanta", r"advaita", r"advaitha", r"upanishad", r"upanishat", r"shankara", r"sankara", r"ramanuja", r"madhwa", r"adhyasa", r"bhashya", r"brahm", r"jivanmukti", r"consciousness", r"thought", r"bhagavad", r"gitartha", r"geetha kosha", r"abhinavagupta", r"avadhoota", r"purnaprajna", r"hiriyanna", r"\bmanu\b", r"vichara", r"adhyatma", r"self.realization", r"\bdharma\b"]),
    ("Veda & Ritual", [r"rgveda", r"veda(?!nta)", r"vedic", r"yajna", r"smriti", r"sukta"]),
    ("Music, Dance & Theatre", [r"sangeeta", r"purandara", r"tyagaraja", r"subbulakshmi", r"chowdiah", r"vadyagalu", r"natya", r"nataraja", r"kriti", r"govinda", r"geete", r"kaipidi"]),
    ("Temples & Art", [r"temple", r"devalaya", r"devaalaya", r"agama", r"pratima", r"murti", r"vastu", r"shilpa", r"icon", r"architect", r"belur", r"hoysala", r"colossus", r"chitramala", r"\bchitra\b", r"gommateshwara", r"kshetra", r"tirupati", r"timmappa", r"vengadam", r"shrine", r"kumaraswamy", r"kalegalu", r"\bart\b", r"hanasoge", r"chunchunkatte"]),
    ("Saints & Biographies", [r"saint", r"swami", r"swamigalu", r"paramahamsa", r"maharaj", r"maharshi", r"dasaru", r"raghavendra", r"ramakrishna", r"sharadapeetha", r"tiruchi", r"vadiraj", r"venkatappa", r"kanaka"]),
    ("Gods & Devotion", [r"devata", r"hanuman", r"ganapati", r"ganesha", r"durga", r"vishnu", r"navagraha", r"salagrama", r"saligrama", r"gayatri", r"krishna", r"saraswathi", r"saraswati", r"aalvaar", r"alwar", r"shiva"]),
    ("Mind & Psychology", [r"psycholog", r"manah", r"manasa"]),
    ("Language & Culture", [r"sahitya", r"samskruti", r"culture", r"parampare", r"parampara", r"hindu", r"karaga", r"daiva", r"vivaaha", r"marriage", r"pranama", r"nidhi", r"padakosha", r"valmiki", r"sanskrit", r"avantisundari", r"dandi"]),
]

COMPILED_RULES = [
    (subj, [re.compile(p, re.IGNORECASE) for p in pats])
    for subj, pats in SUBJECT_RULES
]


def classify(name):
    """Return (language, subject) for a book title."""
    low = name.lower()
    lang = "Kannada" if "(kann" in low else "English"
    for subj, patterns in COMPILED_RULES:
        if any(p.search(name) for p in patterns):
            return lang, subj
    return lang, "General"


def split_title(name):
    """Split 'Title (Kannada).pdf' -> ('Title', 'Kannada')."""
    base = name[:-4] if name.lower().endswith(".pdf") else name
    m = re.search(r"\s*\((kann+n?ada|english)\)\s*$", base, re.IGNORECASE)
    if m:
        return base[: m.start()].strip(), m.group(1)
    return base, ""


# Exact Kannada titles for the 82 Kannada books in download.txt
# (including fixes for romanization typos like "Vyatitva" -> ವ್ಯಕ್ತಿತ್ವ
# and "Varshamana" -> ವರ್ಧಮಾನ).
KANNADA_TITLES = {
    "Aalvaarara Nudimuttugalu": "ಆಳ್ವಾರರ ನುಡಿಮುತ್ತುಗಳು",
    "Aayurvedadalli Naadivijnaana": "ಆಯುರ್ವೇದದಲ್ಲಿ ನಾಡಿವಿಜ್ಞಾನ",
    "Abhinavagupta": "ಅಭಿನವಗುಪ್ತ",
    "Adikavi Valmiki": "ಆದಿಕವಿ ವಾಲ್ಮೀಕಿ",
    "Annappa Daiva": "ಅಣ್ಣಪ್ಪ ದೈವ",
    "Ashwaghoshana Soundarananda": "ಅಶ್ವಘೋಷನ ಸೌಂದರಾನಂದ",
    "Avadhoota": "ಅವಧೂತ",
    "Badukige Belaku": "ಬದುಕಿಗೆ ಬೆಳಕು",
    "Bengalurina Karaga": "ಬೆಂಗಳೂರಿನ ಕರಗ",
    "Bharatada Devalaya nele hinnele": "ಭಾರತದ ದೇವಾಲಯ ನೆಲೆ ಹಿನ್ನೆಲೆ",
    "Bharatamuniya Natyashastra & Natarajana Kalpane": "ಭರತಮುನಿಯ ನಾಟ್ಯಶಾಸ್ತ್ರ ಮತ್ತು ನಟರಾಜನ ಕಲ್ಪನೆ",
    "Bharatiya Devaalayagala Jaanapada Moola": "ಭಾರತೀಯ ದೇವಾಲಯಗಳ ಜಾನಪದ ಮೂಲ",
    "Bodhidharma": "ಬೋಧಿಧರ್ಮ",
    "Bodhiya Belakinalli": "ಬೋಧಿಯ ಬೆಳಕಿನಲ್ಲಿ",
    "Brahmopasane": "ಬ್ರಹ್ಮೋಪಾಸನೆ",
    "Daasasahitya Mattu Samskruti": "ದಾಸಸಾಹಿತ್ಯ ಮತ್ತು ಸಂಸ್ಕೃತಿ",
    "Dandiya Avantisundari": "ದಂಡಿಯ ಅವಂತಿಸುಂದರಿ",
    "Deepankara": "ದೀಪಂಕರ",
    "Dharma Veda Vedanta": "ಧರ್ಮ ವೇದ ವೇದಾಂತ",
    "Dhyana Chitraavali": "ಧ್ಯಾನ ಚಿತ್ರಾವಳಿ",
    "Ganapatiya Kalpane": "ಗಣಪತಿಯ ಕಲ್ಪನೆ",
    "Ganapatiya Roopagalu": "ಗಣಪತಿಯ ರೂಪಗಳು",
    "Geetagovinda": "ಗೀತಗೋವಿಂದ",
    "Geetegondu Kaipidi": "ಗೀತೆಗೊಂದು ಕೈಪಿಡಿ",
    "Gommateshwara": "ಗೊಮ್ಮಟೇಶ್ವರ",
    "Hana Prapancha": "ಹಣ ಪ್ರಪಂಚ",
    "Hiriya Hejjegalu": "ಹಿರಿಯ ಹೆಜ್ಜೆಗಳು",
    "Jayadevana Geetha Govinda": "ಜಯದೇವನ ಗೀತಗೋವಿಂದ",
    "Jhana Jhana Hana": "ಝಣ ಝಣ ಹಣ",
    "K Venkatappa": "ಕೆ. ವೆಂಕಟಪ್ಪ",
    "Kalatapasvi Venkatappa": "ಕಲಾತಪಸ್ವಿ ವೆಂಕಟಪ್ಪ",
    "Kalatatvamaharshi Ananda Kumaraswamy": "ಕಲಾತತ್ತ್ವಮಹರ್ಷಿ ಆನಂದ ಕುಮಾರಸ್ವಾಮಿ",
    "Kanaka Dasaru": "ಕನಕದಾಸರು",
    "Kannada Naadina Dhaarmika Parampare": "ಕನ್ನಡ ನಾಡಿನ ಧಾರ್ಮಿಕ ಪರಂಪರೆ",
    "Kannada Naadinalli Ayurveda": "ಕನ್ನಡ ನಾಡಿನಲ್ಲಿ ಆಯುರ್ವೇದ",
    "Karnatakada janapada vadyagalu": "ಕರ್ನಾಟಕದ ಜನಪದ ವಾದ್ಯಗಳು",
    "Karnatakada Kalegalu": "ಕರ್ನಾಟಕದ ಕಲೆಗಳು",
    "Kshetra Darshana": "ಕ್ಷೇತ್ರ ದರ್ಶನ",
    "M S Subbulakshmi": "ಎಂ. ಎಸ್. ಸುಬ್ಬುಲಕ್ಷ್ಮಿ",
    "Mahakavi Ashwagosha": "ಮಹಾಕವಿ ಅಶ್ವಘೋಷ",
    "Maharshi Daivarataru": "ಮಹರ್ಷಿ ದೈವರಾತರು",
    "ManahShastra Praveshika": "ಮನಃಶಾಸ್ತ್ರ ಪ್ರವೇಶಿಕಾ",
    "Manganoorina Buddhivantaru mattu Itara Kathegalu": "ಮಂಗನೂರಿನ ಬುದ್ಧಿವಂತರು ಮತ್ತು ಇತರ ಕಥೆಗಳು",
    "Murti Shilpa Nele Hinnele": "ಮೂರ್ತಿಶಿಲ್ಪ ನೆಲೆ ಹಿನ್ನೆಲೆ",
    "Nageya Nele": "ನಗೆಯ ನೆಲೆ",
    "Namma Sangeeta Mattu Vagyeyakararu": "ನಮ್ಮ ಸಂಗೀತ ಮತ್ತು ವಾಗ್ಗೇಯಕಾರರು",
    "Om Prathama Pada": "ಓಂ ಪ್ರಥಮ ಪದ",
    "Paramanasa Shastra": "ಪರಮಾನಸ ಶಾಸ್ತ್ರ",
    "Prachina Samskruti": "ಪ್ರಾಚೀನ ಸಂಸ್ಕೃತಿ",
    "Prakruthi Padakosha": "ಪ್ರಕೃತಿ ಪದಕೋಶ",
    "Purnaprajna Prashasti": "ಪೂರ್ಣಪ್ರಜ್ಞ ಪ್ರಶಸ್ತಿ",
    "Purusha Saraswathi": "ಪುರುಷ ಸರಸ್ವತಿ",
    "Raghavendra Swamigalu": "ರಾಘವೇಂದ್ರ ಸ್ವಾಮಿಗಳು",
    "Ramakrishna Paramahamsara Matukategalu": "ರಾಮಕೃಷ್ಣ ಪರಮಹಂಸರ ಮಾತುಕತೆಗಳು",
    "Sangeeta Itihasa Parampara": "ಸಂಗೀತ ಇತಿಹಾಸ ಪರಂಪರಾ",
    "Sangeeta Ratna Mysore T Chowdiah": "ಸಂಗೀತ ರತ್ನ ಮೈಸೂರು ಟಿ. ಚೌಡಯ್ಯ",
    "Shankaravaani": "ಶಂಕರವಾಣಿ",
    "Shantale Natyaroopaka": "ಶಾಂತಲೆ ನಾಟ್ಯರೂಪಕ",
    "Shiva Sankalpa": "ಶಿವಸಂಕಲ್ಪ",
    "Shubha Prarabdha": "ಶುಭ ಪ್ರಾರಬ್ಧ",
    "Sri Krishnana Vyatitva": "ಶ್ರೀ ಕೃಷ್ಣನ ವ್ಯಕ್ತಿತ್ವ",
    "Sri Madhwacharyaru": "ಶ್ರೀ ಮಧ್ವಾಚಾರ್ಯರು",
    "Sri Purandara Dasaru": "ಶ್ರೀ ಪುರಂದರದಾಸರು",
    "Sri Ramanuja Darshana": "ಶ್ರೀ ರಾಮಾನುಜ ದರ್ಶನ",
    "Sri Shankararu mattu Adhyasabhashya": "ಶ್ರೀ ಶಂಕರರು ಮತ್ತು ಅಧ್ಯಾಸಭಾಷ್ಯ",
    "Sri Sharadapeethada Manikya": "ಶ್ರೀ ಶಾರದಾಪೀಠದ ಮಾಣಿಕ್ಯ",
    "Sri Sukta": "ಶ್ರೀ ಸೂಕ್ತ",
    "Sri Vidya Upaasane": "ಶ್ರೀವಿದ್ಯಾ ಉಪಾಸನೆ",
    "Tibettina Yogi Milarepa": "ಟಿಬೆಟ್ಟಿನ ಯೋಗಿ ಮಿಲರೇಪ",
    "Tiruchi Swamigalu Baduku Belaku": "ತಿರುಚ್ಚಿ ಸ್ವಾಮಿಗಳು ಬದುಕು ಬೆಳಕು",
    "Tyagaraja": "ತ್ಯಾಗರಾಜ",
    "Vaidya Hitopadesha": "ವೈದ್ಯ ಹಿತೋಪದೇಶ",
    "Vaidya Sara Sangraha": "ವೈದ್ಯ ಸಾರ ಸಂಗ್ರಹ",
    "Vaishaka Poornima": "ವೈಶಾಖ ಪೂರ್ಣಿಮಾ",
    "Vaiwaswatha Manu": "ವೈವಸ್ವತ ಮನು",
    "Valmiki Prathibhe": "ವಾಲ್ಮೀಕಿ ಪ್ರತಿಭೆ",
    "Varshamana Mahaveera": "ವರ್ಧಮಾನ ಮಹಾವೀರ",
    "Veda Vangmaya mattu Upanishattugalu": "ವೇದ ವಾಙ್ಮಯ ಮತ್ತು ಉಪನಿಷತ್ತುಗಳು",
    "Vedadallina Kathegalu": "ವೇದದಲ್ಲಿನ ಕಥೆಗಳು",
    "Vicharalahari": "ವಿಚಾರಲಹರಿ",
    "Vivaaha Paddhatigalu": "ವಿವಾಹ ಪದ್ಧತಿಗಳು",
    "Yantragalu": "ಯಂತ್ರಗಳು",
}

# Latin -> Kannada fixes for known words, applied (in order) before the
# generic fallback engine runs for any unmapped titles.
KANNADA_PRE = [
    (r"\bsri\b", "ಶ್ರೀ"),
    (r"\bom\b", "ಓಂ"),
    (r"\bm\b", "ಎಂ."),
    (r"\bs\b", "ಎಸ್."),
    (r"\bk\b", "ಕೆ."),
    (r"\bd\b", "ಡಿ."),
    (r"\bt\b", "ಟಿ."),
    ("mysore", "ಮೈಸೂರು"),
    ("chowdiah", "ಚೌಡಯ್ಯ"),
    ("tibettina", "ಟಿಬೆಟ್ಟಿನ"),
    ("tiruchi", "ತಿರುಚ್ಚಿ"),
    ("milarepa", "ಮಿಲರೇಪ"),
    ("hanasoge", "ಹನಸೋಗೆ"),
    ("chunchun", "ಚುಂಚುನ್"),
    ("katte", "ಕಟ್ಟೆ"),
    ("muttu", "ಮುತ್ತು"),
    ("belur", "ಬೇಲೂರು"),
    ("belaku", "ಬೆಳಕು"),
    ("belakinalli", "ಬೆಳಕಿನಲ್ಲಿ"),
    ("aalvaarara", "ಆಳ್ವಾರರ"),
    ("nataraja", "ನಟರಾಜ"),
    ("natya", "ನಾಟ್ಯ"),
    ("krishna", "ಕೃಷ್ಣ"),
    ("kri", "ಕೃ"),
    ("kru", "ಕೃ"),
    ("devalaya", "ದೇವಾಲಯ"),
    ("devaalaya", "ದೇವಾಲಯ"),
    ("deva", "ದೇವ"),
    ("veda", "ವೇದ"),
    ("bharata", "ಭಾರತ"),
    ("dasa", "ದಾಸ"),
    ("kala", "ಕಲಾ"),
    ("sahitya", "ಸಾಹಿತ್ಯ"),
    ("rama", "ರಾಮ"),
    ("ananda", "ಆನಂದ"),
    ("adikavi", "ಆದಿಕವಿ"),
    ("sharada", "ಶಾರದಾ"),
    ("raghava", "ರಾಘವ"),
    ("matu", "ಮಾತು"),
    ("gosha", "ಘೋಷ"),
    ("ghoshana", "ಘೋಷನ"),
    ("purusha", "ಪುರುಷ"),
    ("varsha", "ವರ್ಷ"),
    ("upanishat", "ಉಪನಿಷತ್"),
    ("vangmaya", "ವಾಙ್ಮಯ"),
    ("yogi", "ಯೋಗಿ"),
    ("manikya", "ಮಾಣಿಕ್ಯ"),
    ("gommateshwara", "ಗೊಮ್ಮಟೇಶ್ವರ"),
    ("mopasane", "ಮೋಪಾಸನೆ"),
    ("hana", "ಹಣ"),
    ("saligrama", "ಸಾಲಿಗ್ರಾಮ"),
    ("sukta", "ಸೂಕ್ತ"),
    ("prachina", "ಪ್ರಾಚೀನ"),
    ("galu", "ಗಳು"),
]

K_VOWEL = {
    "a": "ಅ", "aa": "ಆ", "i": "ಇ", "ii": "ಈ", "ee": "ಈ",
    "u": "ಉ", "uu": "ಊ", "oo": "ಊ", "e": "ಎ", "ai": "ಐ",
    "o": "ಒ", "au": "ಔ", "ow": "ಔ", "ou": "ಔ",
}
K_SIGN = {
    "aa": "ಾ", "i": "ಿ", "ii": "ೀ", "ee": "ೀ", "u": "ು",
    "uu": "ೂ", "oo": "ೂ", "e": "ೆ", "ai": "ೈ", "o": "ೊ",
    "au": "ೌ", "ow": "ೌ", "ou": "ೌ",
}
K_CONS = {
    "k": "ಕ", "kh": "ಖ", "g": "ಗ", "gh": "ಘ",
    "c": "ಚ", "ch": "ಚ", "chh": "ಛ", "j": "ಜ", "jh": "ಝ",
    "t": "ತ", "th": "ಥ", "d": "ದ", "dh": "ಧ", "n": "ನ",
    "p": "ಪ", "ph": "ಫ", "f": "ಫ", "b": "ಬ", "bh": "ಭ", "m": "ಮ",
    "y": "ಯ", "r": "ರ", "l": "ಲ", "v": "ವ", "w": "ವ",
    "sh": "ಶ", "s": "ಸ", "h": "ಹ", "x": "ಕ್ಷ",
    "jn": "ಜ್ಞ", "ksh": "ಕ್ಷ",
}
HALANT = "್"


def to_kannada(text):
    """Return accurate Kannada script title (via dictionary lookup + fallback)."""
    clean = text.strip()
    if clean in KANNADA_TITLES:
        return KANNADA_TITLES[clean]
    s = clean.lower()
    for pat, rep in KANNADA_PRE:
        s = re.sub(pat, rep, s)
    out = []
    pending = ""
    i, n = 0, len(s)
    while i < n:
        tok = None
        for ln in (3, 2, 1):
            c = s[i : i + ln]
            if c in K_VOWEL or c in K_CONS:
                tok = c
                break
        if tok is None:
            ch = s[i]
            if ch == "y" and (i + 1 >= n or not s[i + 1].isalpha()):
                tok = "i"  # word-final y is really an i (swamy -> ಸ್ವಾಮಿ)
            elif ch == "h" and (i + 1 >= n or s[i + 1] not in "aeiou"):
                if pending:  # visarga: manah, manah+consonant
                    out.append(pending)
                    pending = ""
                out.append("ಃ")
                i += 1
                continue
            else:  # Kannada from pre-map, punctuation, digits, spaces
                if pending:
                    out.append(pending)
                    pending = ""
                out.append(ch)
                i += 1
                continue
        if tok in K_VOWEL:
            if pending:
                base = pending[:-1]
                out.append(base if tok == "a" else base + K_SIGN[tok])
                pending = ""
            else:
                out.append(K_VOWEL[tok])
            i += len(tok)
        else:
            cons = K_CONS[tok]
            if pending == "ನ್":
                out.append("ನ್ನ" if tok == "n" else "ಂ")
                pending = ""
            elif pending == "ಮ್":
                out.append("ಮ್ಮ" if tok == "m" else "ಂ")
                pending = ""
            elif pending:
                out.append(pending)
            pending = cons + HALANT
            i += len(tok)
    if pending:
        out.append(pending)
    return "".join(out)

HERE = Path(__file__).resolve().parent
LIST = Path(sys.argv[1]) if len(sys.argv) > 1 else HERE / "download.txt"
PDFS = Path(sys.argv[2]) if len(sys.argv) > 2 else HERE / "pdfs"
OUT = Path(sys.argv[3]) if len(sys.argv) > 3 else HERE / "index.html"


def local_targets(name, seen):
    """Mirror download_all.sh naming: dup titles get ' (n)' suffixes."""
    n = seen.get(name, 0)
    seen[name] = n + 1
    if n == 0:
        return name
    p = Path(name)
    suffix = f" ({n + 1})"
    if p.suffix:
        return f"{p.stem}{suffix}{p.suffix}"
    return f"{name}{suffix}"


def human_size(nbytes):
    n = float(nbytes)
    for unit in ("B", "KB", "MB", "GB"):
        if n < 1024 or unit == "GB":
            return f"{n:.0f} {unit}" if unit == "B" else f"{n:.1f} {unit}"
        n /= 1024


def main():
    with LIST.open(encoding="utf-8-sig") as f:
        rows = list(csv.DictReader(f, delimiter="\t"))
    try:
        pdf_dir = PDFS.relative_to(HERE)
    except ValueError:
        pdf_dir = PDFS

    meta_path = HERE / "metadata.json"
    dir_by_filename: dict[str, str] = {}
    if meta_path.is_file():
        for item in json.loads(meta_path.read_text(encoding="utf-8")):
            src_name = Path(item["source"]).name
            dir_by_filename[src_name] = item["directory"]

    seen: dict = {}
    entries = []
    for row in rows:
        name = (row.get("Book Name") or "").strip()
        if not name:
            continue
        target = local_targets(name, seen)
        path = PDFS / target
        size = human_size(path.stat().st_size) if path.is_file() else None
        lang, subj = classify(name)
        book_dir = dir_by_filename.get(target) or dir_by_filename.get(name)
        entries.append((name, target, size, lang, subj, book_dir))

    present = sum(1 for e in entries if e[2] is not None)
    stamp = datetime.datetime.now().strftime("%Y-%m-%d %H:%M")
    langs = Counter(e[3] for e in entries)
    subjs = Counter(e[4] for e in entries)
    general = sorted({e[0] for e in entries if e[4] == "General"})

    def pill(group, value, label, active=False):
        cls = "pill active" if active else "pill"
        return (
            f'<button type="button" class="{cls}" data-group="{group}" '
            f'data-value="{html.escape(value, quote=True)}">{label}</button>'
        )

    lang_pills = [pill("lang", "", "All languages", True)]
    for lang in ("Kannada", "English"):
        if langs.get(lang):
            lang_pills.append(
                pill("lang", lang, f"{lang} ({langs[lang]})")
            )
    subj_order = [s for s, _ in SUBJECT_RULES if subjs.get(s)]
    if general:
        subj_order.append("General")
    subj_pills = [pill("subj", "", "All subjects", True)]
    for subj in subj_order:
        subj_pills.append(
            pill("subj", subj, f"{html.escape(subj)} ({subjs[subj]})")
        )

    trs = []
    kn_leftover = []
    for name, target, size, lang, subj, book_dir in entries:
        base = name[:-4] if name.lower().endswith(".pdf") else name
        if lang == "Kannada":
            core, _tag = split_title(name)
            kn = to_kannada(core)
            if re.search(r"[a-zA-Z]", kn):
                kn_leftover.append((core, kn))
            title_html = (
                f'<div class="kn">{html.escape(kn)}</div>'
                f'<div class="translit">{html.escape(core)}</div>'
            )
        else:
            title_html = html.escape(base)
        if book_dir:
            dir_href = f"{quote(book_dir)}/"
            title_html = f'<a class="book-link" href="{dir_href}">{title_html}</a>'
        if size is not None:
            href = f"{quote(str(pdf_dir))}/{quote(target)}"
            dl_link = f'<a class="dl" href="{href}" download="{html.escape(target, quote=True)}">PDF</a>'
            if book_dir:
                read_link = f'<a class="read" href="{quote(book_dir)}/">Read</a> '
            else:
                read_link = ""
            link = f"{read_link}{dl_link}"
            status = f'<span class="size">{size}</span>'
        else:
            link = '<span class="missing">Not downloaded yet</span>'
            status = '<span class="missing">—</span>'
        trs.append(
            f'<tr data-lang="{lang}" data-subj="{html.escape(subj, quote=True)}">'
            f"<td>{title_html}</td>"
            f"<td>{lang}</td>"
            f'<td><span class="chip">{html.escape(subj)}</span></td>'
            f'<td class="col-size">{status}</td>'
            f'<td class="col-dl">{link}</td></tr>'
        )

    body = "\n".join(trs)
    page = f"""<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Prof S K Ramachandra Rao Books</title>
<style>
  :root {{
    --maroon: #72211f; --maroon-dark: #571915; --gold: #c9a227;
    --ink: #2e2620; --muted: #6d6259; --line: #e3d9c8;
    --chip-bg: #f6e8c9; --chip-ink: #6b4a12;
  }}
  * {{ box-sizing: border-box; }}
  body {{ font-family: Georgia, 'Palatino Linotype', 'Times New Roman', serif;
         margin: 0; padding: 0; color: var(--ink); line-height: 1.55;
         background: #fff; }}
  header {{ background: var(--maroon); color: #f7ecd4; padding: 2.2rem 1rem 1.6rem;
            border-bottom: 4px solid var(--gold); }}
  .wrap {{ max-width: 980px; margin: 0 auto; padding: 0 1rem; }}
  header h1 {{ font-size: 2rem; margin: 0 0 0.4rem; letter-spacing: 0.01em; }}
  header p {{ margin: 0.25rem 0; color: #ecd9ae; font-size: 0.95rem; }}
  header a {{ color: #ffe9b0; }}
  main {{ padding-bottom: 2rem; }}
  #search {{ width: 100%; font-size: 1.05rem; padding: 0.6rem 0.8rem;
             border: 1px solid #b3a48c; border-radius: 6px; margin: 1.2rem 0 0.4rem;
             font-family: inherit; }}
  .grouplabel {{ display: block; font-size: 0.85rem; color: var(--muted);
                 margin: 0.9rem 0 0.35rem; }}
  .pills {{ display: flex; flex-wrap: wrap; gap: 0.4rem; }}
  .pill {{ font-family: -apple-system, 'Segoe UI', Verdana, sans-serif;
           font-size: 0.85rem; padding: 0.3rem 0.85rem; border-radius: 999px;
           border: 1px solid #a08b6d; background: #fff; color: var(--ink);
           cursor: pointer; }}
  .pill.active {{ background: var(--maroon); border-color: var(--maroon); color: #fff; }}
  #count {{ color: var(--muted); font-size: 0.9rem; }}
  .note {{ color: var(--muted); font-size: 0.85rem; font-style: italic; }}
  .kn {{ font-size: 1.05rem; }}
  .translit {{ font-size: 0.78rem; color: var(--muted);
               font-family: -apple-system, 'Segoe UI', Verdana, sans-serif; }}
  table {{ width: 100%; border-collapse: collapse; margin-top: 0.4rem; }}
  th, td {{ text-align: left; padding: 0.5rem 0.6rem; border-bottom: 1px solid var(--line); }}
  th {{ border-bottom: 2px solid var(--maroon); font-size: 0.9rem; }}
  .chip {{ display: inline-block; font-family: -apple-system, 'Segoe UI', Verdana, sans-serif;
           font-size: 0.75rem; background: var(--chip-bg); color: var(--chip-ink);
           border-radius: 4px; padding: 0.1rem 0.5rem; white-space: nowrap; }}
  .col-cat, .col-size, .col-dl {{ white-space: nowrap; }}
  a.book-link {{ color: inherit; text-decoration: none; display: inline-block; }}
  a.book-link:hover {{ color: var(--maroon); text-decoration: underline; }}
  a.read {{ display: inline-block; background: #fff; color: var(--maroon);
            border: 1px solid var(--maroon); text-decoration: none;
            font-family: -apple-system, 'Segoe UI', Verdana, sans-serif;
            font-size: 0.82rem; padding: 0.22rem 0.75rem; border-radius: 6px;
            margin-right: 0.25rem; }}
  a.read:hover {{ background: var(--chip-bg); }}
  a.dl {{ display: inline-block; background: var(--maroon); color: #fff;
          text-decoration: none; font-family: -apple-system, 'Segoe UI', Verdana, sans-serif;
          font-size: 0.82rem; padding: 0.25rem 0.8rem; border-radius: 6px; }}
  a.dl:hover {{ background: var(--maroon-dark); }}
  .missing {{ color: #8a8a8a; font-style: italic; }}
  footer {{ border-top: 1px solid var(--line); color: var(--muted);
            font-size: 0.85rem; padding: 1.2rem 1rem 1.8rem; text-align: center; }}
  footer p {{ margin: 0.35rem 0; }}
  footer a {{ color: var(--muted); text-decoration: underline; text-decoration-color: #c8bca8;
              text-underline-offset: 2px; }}
  footer a:hover {{ color: var(--maroon); text-decoration-color: var(--maroon); }}
  .src-section {{ max-width: 780px; margin: 1rem auto 0; padding-top: 0.75rem;
                  border-top: 1px dashed var(--line); text-align: left;
                  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
                  font-size: 0.75rem; line-height: 1.45; color: #7b7067; }}
  .src-heading {{ font-size: 0.7rem; text-transform: uppercase; letter-spacing: 0.06em;
                  color: #8c8076; margin: 0 0 0.4rem; text-align: center; }}
  .src-grid {{ display: grid; grid-template-columns: repeat(auto-fit, minmax(300px, 1fr));
               gap: 0.28rem 1.5rem; margin: 0; padding: 0; list-style: none; }}
  .src-grid a {{ font-family: ui-monospace, SFMono-Regular, Menlo, Monaco, Consolas, monospace;
                 font-size: 0.73rem; color: #5c5147; text-decoration: none;
                 border-bottom: 1px dotted #b8ab99; }}
  .src-grid a:hover {{ color: var(--maroon); border-bottom-color: var(--maroon); }}
</style>
</head>
<body>
  <header>
    <div class="wrap">
      <h1>Prof S K Ramachandra Rao Books</h1>
      <p>Backup of the book collection from <a href="https://profskr.com/">profskr.com</a>.
      All rights belong to &ldquo;The Prof S K Ramachandra Rao Memorial Trust&rdquo;.</p>
    </div>
  </header>
  <main class="wrap">
    <input id="search" type="search" placeholder="Search titles…" aria-label="Search titles">
    <span class="grouplabel">Language</span>
    <div class="pills" id="lang-pills">
      {" ".join(lang_pills)}
    </div>
    <span class="grouplabel">Subject</span>
    <div class="pills" id="subj-pills">
      {" ".join(subj_pills)}
    </div>
    <p id="count"></p>
    <table>
      <thead><tr><th>Title</th><th>Language</th><th>Category</th><th>Size</th><th>File</th></tr></thead>
      <tbody>
{body}
      </tbody>
    </table>
  </main>
<script>
  var box = document.getElementById('search');
  var rows = document.querySelectorAll('tbody tr');
  var count = document.getElementById('count');
  var activeLang = '', activeSubj = '';
  function filter() {{
    var q = box.value.toLowerCase();
    var n = 0;
    rows.forEach(function (r) {{
      var hit = r.cells[0].textContent.toLowerCase().includes(q)
        && (!activeLang || r.dataset.lang === activeLang)
        && (!activeSubj || r.dataset.subj === activeSubj);
      r.style.display = hit ? '' : 'none';
      if (hit) n++;
    }});
    count.textContent = n + ' of ' + rows.length + ' shown';
  }}
  document.querySelectorAll('.pill').forEach(function (p) {{
    p.addEventListener('click', function () {{
      var group = p.dataset.group;
      document.querySelectorAll('.pill[data-group="' + group + '"]').forEach(function (x) {{
        x.classList.remove('active');
      }});
      p.classList.add('active');
      if (group === 'lang') activeLang = p.dataset.value;
      else activeSubj = p.dataset.value;
      filter();
    }});
  }});
  box.addEventListener('input', filter);
  filter();
</script>
  <footer>
    <div class="wrap">
      <p class="note">Language and subject labels are assigned automatically from the book
      titles, so a few books may be misclassified. Kannada titles are transliterated
      automatically and may contain spelling errors.</p>
      <p>Downloaded on {stamp} &middot; backup of <a href="https://profskr.com/">profskr.com</a> &middot; <a href="changelog.txt">Changelog</a></p>
      <div class="src-section">
        <p class="src-heading">Archive scripts (source)</p>
        <ul class="src-grid">
          <li><a href="download_all.sh.html">download_all.sh</a> &mdash; Fetch all book PDFs from Google Drive links</li>
          <li><a href="retry_download.py.html">retry_download.py</a> &mdash; Verify PDFs &amp; retry Drive virus-scan interstitials</li>
          <li><a href="convert_pages.py.html">convert_pages.py</a> &mdash; Render PDF pages to thumb, full &amp; clean B&amp;W WebP</li>
          <li><a href="generate_readers.py.html">generate_readers.py</a> &mdash; Generate per-book web readers with page navigation</li>
          <li><a href="generate_index.py.html">generate_index.py</a> &mdash; Build this catalog with Kannada titles &amp; filters</li>
          <li><a href="upload.sh.html">upload.sh</a> &mdash; Sync static archive to remote storage via rclone</li>
        </ul>
      </div>
    </div>
  </footer>
</body>
</html>
"""
    OUT.write_text(page, encoding="utf-8")
    print(f"Wrote {OUT} ({present}/{len(entries)} local files found).")
    missing = [e[1] for e in entries if e[2] is None]
    if missing:
        print(f"Missing locally (in entries but not in {PDFS}):")
        for m in missing:
            print(f"  - {m}")
    entry_targets = {e[1] for e in entries}
    extra = sorted(
        p.name for p in PDFS.iterdir()
        if p.is_file() and not p.name.startswith(".") and p.name not in entry_targets
    ) if PDFS.is_dir() else []
    if extra:
        print(f"Extra locally (in {PDFS} but not in entries):")
        for x in extra:
            print(f"  - {x}")
    print("Languages:", dict(langs))
    print("Subjects:")
    for subj in subj_order:
        print(f"  {subjs[subj]:3d}  {subj}")
    if general:
        print("General titles:")
        for t in general:
            print(f"  - {t}")
    if kn_leftover:
        print("Kannada titles with untransliterated letters:")
        for core, kn in kn_leftover:
            print(f"  - {core}  =>  {kn}")


if __name__ == "__main__":
    main()