#!/usr/bin/env python3
"""Assemble 'The Sufi Who Fooled the Algorithm' PDF — Kiyosaki × Shah × Achebe style, Penguin 6×9in."""
import os, re, sys
from reportlab.lib.pagesizes import letter
from reportlab.lib.units import inch
from reportlab.lib.styles import ParagraphStyle
from reportlab.lib.enums import TA_CENTER, TA_LEFT, TA_JUSTIFY
from reportlab.lib.colors import HexColor
from reportlab.platypus import (
    Paragraph, Spacer, PageBreak, Frame, PageTemplate, BaseDocTemplate
)

OUTPUT_DIR = os.path.expanduser("~/odysseus_book/shah_kiyosaki")
CHAPTERS_DIR = os.path.join(OUTPUT_DIR, "chapters")
PDF_PATH = os.path.join(OUTPUT_DIR, "pdf", "The_Sufi_Who_Fooled_the_Algorithm.pdf")

MARGIN = 0.85 * inch
FOOTER_H = 0.4 * inch
PAGE_W = 6 * inch
PAGE_H = 9 * inch

DARK_TEAL = HexColor('#0D4F4F')
GOLD = HexColor('#C8A951')
DARK = HexColor('#1a1a1a')
GREY = HexColor('#666666')
LIGHT_GREY = HexColor('#f0f0f0')
ACCENT_ORANGE = HexColor('#E67E22')

style_title = ParagraphStyle('Title', fontName='Helvetica-Bold', fontSize=26,
    leading=32, alignment=TA_CENTER, textColor=DARK_TEAL, spaceAfter=8)
style_subtitle = ParagraphStyle('Subtitle', fontName='Helvetica-Oblique', fontSize=13,
    leading=17, alignment=TA_CENTER, textColor=GOLD, spaceAfter=20)
style_author = ParagraphStyle('Author', fontName='Helvetica', fontSize=12,
    leading=16, alignment=TA_CENTER, textColor=GREY, spaceAfter=28)
style_copyright = ParagraphStyle('Copyright', fontName='Helvetica', fontSize=9,
    leading=12, alignment=TA_CENTER, textColor=GREY)
style_ch_num = ParagraphStyle('ChNum', fontName='Helvetica-Bold', fontSize=10,
    leading=13, alignment=TA_CENTER, textColor=GOLD, spaceAfter=4, spaceBefore=18)
style_ch_title = ParagraphStyle('ChTitle', fontName='Helvetica-Bold', fontSize=20,
    leading=26, alignment=TA_LEFT, textColor=DARK_TEAL, spaceAfter=8, spaceBefore=4)
style_subtitle_ch = ParagraphStyle('SubtitleCh', fontName='Helvetica-Oblique', fontSize=12,
    leading=16, alignment=TA_LEFT, textColor=ACCENT_ORANGE, spaceAfter=14, spaceBefore=2)
style_framework = ParagraphStyle('Framework', fontName='Helvetica-Bold', fontSize=11,
    leading=15, alignment=TA_LEFT, textColor=DARK_TEAL, spaceAfter=6, spaceBefore=14,
    backColor=LIGHT_GREY, borderColor=GOLD, borderWidth=1, borderPadding=8)
style_body = ParagraphStyle('Body', fontName='Times-Roman', fontSize=10,
    leading=14.5, alignment=TA_JUSTIFY, textColor=DARK, spaceAfter=7, firstLineIndent=0)
style_dialogue = ParagraphStyle('Dialogue', fontName='Times-Roman', fontSize=10,
    leading=14.5, alignment=TA_LEFT, textColor=DARK, spaceAfter=5, leftIndent=24)
style_shah = ParagraphStyle('Shah', fontName='Times-Bold', fontSize=10,
    leading=14.5, alignment=TA_LEFT, textColor=DARK_TEAL, spaceAfter=3, leftIndent=24)
style_cfo = ParagraphStyle('CFO', fontName='Times-Bold', fontSize=10,
    leading=14.5, alignment=TA_LEFT, textColor=GOLD, spaceAfter=3, leftIndent=24)
style_lie = ParagraphStyle('Lie', fontName='Times-Bold', fontSize=10.5,
    leading=15, alignment=TA_LEFT, textColor=HexColor('#C0392B'), spaceAfter=6, spaceBefore=10,
    backColor=HexColor('#FDF2F2'), borderColor=HexColor('#C0392B'), borderWidth=1, borderPadding=10)
style_protocol = ParagraphStyle('Protocol', fontName='Times-Roman', fontSize=10,
    leading=14.5, alignment=TA_LEFT, textColor=DARK, spaceAfter=6, leftIndent=36)
style_reflection = ParagraphStyle('Reflection', fontName='Times-Italic', fontSize=10.5,
    leading=15, alignment=TA_CENTER, textColor=GREY, spaceAfter=16, spaceBefore=16,
    backColor=HexColor('#F8F9FA'), borderColor=GREY, borderWidth=0.5, borderPadding=12)
style_final = ParagraphStyle('Final', fontName='Times-Italic', fontSize=10,
    leading=14, alignment=TA_LEFT, textColor=GREY, spaceAfter=4, leftIndent=24)
style_toc_ch = ParagraphStyle('TOCCh', fontName='Helvetica-Bold', fontSize=11,
    leading=17, alignment=TA_LEFT, textColor=DARK_TEAL)
style_toc_sub = ParagraphStyle('TOCSub', fontName='Helvetica', fontSize=9.5,
    leading=13, alignment=TA_LEFT, textColor=GREY)
style_hook = ParagraphStyle('Hook', fontName='Times-Italic', fontSize=10.5,
    leading=15, alignment=TA_LEFT, textColor=DARK, spaceAfter=10, spaceBefore=6,
    backColor=HexColor('#EBF5FB'), borderColor=DARK_TEAL, borderWidth=1, borderPadding=10)

CHAPTERS = [
    ("The Man Who Watched the Shadow", "Attention Economy, Doomscrolling, and the Shadow You Feed", "Attention Quadrant: Consumption / Curation / Creation / Contemplation"),
    ("Nasrudin's Lost Key", "Looking Where the Light Is — Data Bias, Streetlight Effect, and the Answers You'll Never Find", "Epistemic Quadrant: Known Knowns / Known Unknowns / Unknown Knowns / Unknown Unknowns"),
    ("The Man Who Ate the Menu", "Consuming Content vs. Digesting Wisdom — Information Obesity, Malnourished Minds", "Consumption Quadrant: Input / Filter / Digest / Apply"),
    ("The Commanding Self", "Social Media Ego, Performative Identity, and the Tyrant Inside", "Identity Quadrant: Persona / Shadow / Essence / Witness"),
    ("Learning How to Unlearn", "AI Era: Unlearning > Learning — Cognitive Flexibility as Survival Skill", "Unlearning Loop: Identify Assumption → Test → Discard → Replace → Test Again"),
    ("The Assumption of Knowledge", "Dunning-Kruger in the TikTok Era — The Illusion of Competence", "Competence Quadrant: Unconscious Incompetence → Conscious Incompetence → Conscious Competence → Unconscious Competence + Shah's Fifth"),
    ("The Teacher Who Refused to Teach", "Mentorship Crisis, Credentialism, and the Parrot Who Quoted Rumi", "Transmission Loop: Embody → Transmit → Verify → Release"),
    ("The Parrot and the Scholar", "Mimicry vs. Understanding — LLMs, Cargo Cult Science, and the Illusion of Intelligence", "Understanding Ladder: Pattern Match → Compress → Predict → Generate → Embody → Transmit"),
    ("The Dream That Woke the Dreamer", "Purpose, Midlife Crisis, and the Nightmare of a Life Not Lived", "Purpose Loop: Disorient → Question → Experiment → Integrate → Question Again"),
    ("The Story That Has No Ending", "Legacy, Algorithmic Immortality, and the Art of Leaving Incomplete", "Legacy Loop: Create → Release → Let Go → Trust → Begin Again"),
]

LIGHT_GREY = HexColor('#f0f0f0')
DARK_GREY = HexColor('#333333')

def clean_text(text):
    text = text.replace('&', '&').replace('<', '<').replace('>', '>')
    parts = text.split('**')
    for i in range(1, len(parts), 2):
        parts[i] = f'<b>{parts[i]}</b>'
    text = ''.join(parts)
    parts = text.split('*')
    for i in range(1, len(parts), 2):
        parts[i] = f'<i>{parts[i]}</i>'
    text = ''.join(parts)
    return text

def parse_chapter(md_text):
    lines = md_text.split('\n')
    sections = {'hook': [], 'dialogue': [], 'framework': [], 'lie': [], 'protocol': [], 'final': [], 'reflection': []}
    current = None
    for line in lines:
        stripped = line.strip()
        if not stripped:
            continue
        # Detect section headers
        low = stripped.lower()
        if low.startswith('# hook') or low.startswith('## hook') or low == 'hook':
            current = 'hook'
            continue
        elif low.startswith('# shah/cfo') or low.startswith('## shah/cfo') or low.startswith('shah/cfo') or low.startswith('dialogue'):
            current = 'dialogue'
            continue
        elif low.startswith('# framework') or low.startswith('## framework') or low.startswith('framework'):
            current = 'framework'
            continue
        elif low.startswith('# the lie') or low.startswith('## the lie') or low.startswith('the lie'):
            current = 'lie'
            continue
        elif low.startswith('# the protocol') or low.startswith('## the protocol') or low.startswith('protocol'):
            current = 'protocol'
            continue
        elif low.startswith('# shah/cfo final') or low.startswith('## shah/cfo final') or low.startswith('final word') or 'final word' in low:
            current = 'final'
            continue
        elif low.startswith('# reflection') or low.startswith('## reflection') or low.startswith('reflection'):
            current = 'reflection'
            continue
        if stripped.startswith('#'):
            continue
        if stripped == '---' or stripped == '***':
            continue
        if current:
            sections[current].append(stripped)
    for k in sections:
        sections[k] = '\n\n'.join(sections[k])
    return sections

class BookDoc(BaseDocTemplate):
    def __init__(self, *args, **kw):
        BaseDocTemplate.__init__(self, *args, **kw)
        self.page_count = 0

def footer_fn(canvas, doc):
    canvas.saveState()
    canvas.setFont('Helvetica', 8)
    canvas.setFillColor(HexColor('#999999'))
    canvas.drawCentredString(PAGE_W / 2, 0.35 * inch, f'— {doc.page} —')
    canvas.restoreState()

def build_pdf():
    os.makedirs(os.path.join(OUTPUT_DIR, 'pdf'), exist_ok=True)

    frame = Frame(MARGIN, MARGIN + FOOTER_H,
                  PAGE_W - 2 * MARGIN,
                  PAGE_H - 2 * MARGIN - FOOTER_H, id='normal')
    doc = BookDoc(PDF_PATH, pagesize=(PAGE_W, PAGE_H),
        leftMargin=MARGIN, rightMargin=MARGIN,
        topMargin=MARGIN, bottomMargin=MARGIN,
        title="The Sufi Who Fooled the Algorithm",
        author='Jauhari Che Wan')
    doc.addPageTemplates([PageTemplate(id='main', frames=frame, onPage=footer_fn)])

    story = []

    # Title page
    story.append(Spacer(1, PAGE_H * 0.18))
    story.append(Paragraph("The Sufi Who Fooled the Algorithm", style_title))
    story.append(Paragraph("Idries Shah for the Attention Age", style_subtitle))
    story.append(Paragraph("10 Teaching Stories Reimagined for AI, Anxiety, and the Art of Not Being Fooled", ParagraphStyle('Tagline',
        fontName='Helvetica-Oblique', fontSize=12, leading=15, alignment=TA_CENTER, textColor=HexColor('#666666'), spaceAfter=24)))
    story.append(Spacer(1, 14))
    story.append(Paragraph('Kiyosaki-Style × Shah\'s Teaching Stories × Achebe Oral Tradition', ParagraphStyle('StyleLine',
        fontName='Helvetica', fontSize=10, leading=13, alignment=TA_CENTER, textColor=HexColor('#888888'), spaceAfter=20)))
    story.append(Spacer(1, 14))
    story.append(Paragraph('Jauhari Che Wan', style_author))
    story.append(Spacer(1, PAGE_H * 0.1))
    story.append(Paragraph('SHAH × KIYOSAKI × ACHEBE',
        ParagraphStyle('Tag', parent=style_copyright, fontName='Helvetica-Bold', fontSize=9, textColor=GOLD)))
    story.append(PageBreak())

    # Copyright
    story.append(Spacer(1, PAGE_H * 0.28))
    for line in ['SHAH × KIYOSAKI × ACHEBE', '', 'The Sufi Who Fooled the Algorithm', 'Jauhari Che Wan',
                 '', 'Based on Idries Shah\'s Teaching Stories', 'and Robert Kiyosaki\'s Rich Dad Framework',
                 '', 'Copyright \u00a9 2025', 'All rights reserved',
                 '', 'Typeset in Times / Helvetica', 'Printed in Malaysia']:
        if line == '':
            story.append(Spacer(1, 4))
        else:
            story.append(Paragraph(line, style_copyright))
    story.append(PageBreak())

    # Dedication
    story.append(Spacer(1, PAGE_H * 0.28))
    story.append(Paragraph('For every mind that\'s ever been fooled by its own shadow<br/>and had the courage to stop running.', style_dedication := ParagraphStyle('Ded', fontName='Times-Italic', fontSize=11, leading=16, alignment=TA_CENTER, textColor=DARK, spaceAfter=24)))
    story.append(PageBreak())

    # TOC
    story.append(Paragraph('Contents', ParagraphStyle('TOCHeader',
        fontName='Helvetica-Bold', fontSize=18, leading=24, alignment=TA_LEFT, textColor=DARK_TEAL, spaceAfter=20)))
    for i, (title, subtitle, _) in enumerate(CHAPTERS, 1):
        story.append(Paragraph(f'<b>Chapter {i}</b>: {title}', style_toc_ch))
        story.append(Paragraph(f'    <i>{subtitle}</i>', style_toc_sub))
        story.append(Spacer(1, 4))
    story.append(PageBreak())

    # Chapters
    for ch_num, (title, subtitle, framework) in enumerate(CHAPTERS, 1):
        fname = os.path.join(CHAPTERS_DIR, f"Chapter_0{ch_num}.md") if ch_num < 10 else os.path.join(CHAPTERS_DIR, f"Chapter_{ch_num}.md")
        if not os.path.exists(fname):
            print(f"  WARNING: {fname} not found")
            continue

        with open(fname, 'r', encoding='utf-8') as f:
            md = f.read()

        print(f"  Adding Chapter {ch_num}: {title}...")
        sections = parse_chapter(md)

        # Chapter header
        story.append(Paragraph(f'Chapter {ch_num}', style_ch_num))
        story.append(Paragraph(title, style_ch_title))
        story.append(Paragraph(subtitle, style_subtitle_ch))
        story.append(Spacer(1, 10))

        # HOOK
        if sections['hook']:
            hook_text = clean_text(' '.join(sections['hook'].split('\n')))
            story.append(Paragraph(hook_text, style_hook))
            story.append(Spacer(1, 6))

        # SHAH/CFO DIALOGUE
        if sections['dialogue']:
            for para in sections['dialogue'].split('\n\n'):
                clean = clean_text(para.strip())
                if not clean:
                    continue
                low = clean.lower()
                if low.startswith('shah:'):
                    story.append(Paragraph(f'<b>SHAH:</b> {clean[5:].strip()}', style_shah))
                elif low.startswith('cfo:'):
                    story.append(Paragraph(f'<b>CFO:</b> {clean[4:].strip()}', style_cfo))
                else:
                    story.append(Paragraph(clean, style_dialogue))
            story.append(Spacer(1, 6))

        # FRAMEWORK
        if sections['framework']:
            story.append(Paragraph(f'FRAMEWORK: {framework}', style_framework))
            story.append(Spacer(1, 4))
            for para in sections['framework'].split('\n\n'):
                clean = clean_text(para.strip())
                if clean:
                    story.append(Paragraph(clean, style_body))
            story.append(Spacer(1, 6))

        # THE LIE
        if sections['lie']:
            lie_lines = sections['lie'].split('\n')
            lie_title = lie_lines[0] if lie_lines else "THE LIE WE'RE TOLD"
            story.append(Paragraph(lie_title, style_lie))
            lie_body = '\n'.join(lie_lines[1:]) if len(lie_lines) > 1 else ''
            for para in lie_body.split('\n\n'):
                clean = clean_text(para.strip())
                if clean:
                    story.append(Paragraph(clean, style_body))
            story.append(Spacer(1, 6))

        # PROTOCOL
        if sections['protocol']:
            story.append(Paragraph('THE PROTOCOL — Do This This Week', ParagraphStyle('ProtoHeader',
                fontName='Helvetica-Bold', fontSize=12, leading=16, alignment=TA_LEFT, textColor=DARK_TEAL, spaceAfter=6, spaceBefore=10)))
            for i, para in enumerate(sections['protocol'].split('\n\n'), 1):
                clean = clean_text(para.strip())
                if clean:
                    if clean.lower().startswith(f'step {i}') or clean.lower().startswith(f'step {i}:'):
                        clean = clean.split(':', 1)[-1].strip()
                    story.append(Paragraph(f'<b>STEP {i}:</b> {clean}', style_protocol))
            story.append(Spacer(1, 6))

        # SHAH/CFO FINAL WORD
        if sections['final']:
            for para in sections['final'].split('\n\n'):
                clean = clean_text(para.strip())
                low = clean.lower()
                if low.startswith('shah:'):
                    story.append(Paragraph(f'<b>SHAH:</b> {clean[5:].strip()}', style_shah))
                elif low.startswith('cfo:'):
                    story.append(Paragraph(f'<b>CFO:</b> {clean[4:].strip()}', style_cfo))
                else:
                    story.append(Paragraph(clean, style_final))
            story.append(Spacer(1, 6))

        # REFLECTION
        if sections['reflection']:
            clean = clean_text(' '.join(sections['reflection'].split('\n')))
            story.append(Paragraph(clean, style_reflection))

        story.append(PageBreak())

    print('Building PDF...')
    doc.build(story)
    print(f'Done: {PDF_PATH}')

    # Stats
    import subprocess
    result = subprocess.run(['wc', '-c', PDF_PATH], capture_output=True, text=True)
    size_kb = int(result.stdout.split()[0]) / 1024
    with open(PDF_PATH, 'rb') as f:
        data = f.read()
    pages = len(re.findall(rb'/Type\s*/Page[^s]', data))
    print(f'Pages: {pages}, Size: {size_kb:.0f} KB')

if __name__ == '__main__':
    build_pdf()