# Prabhupāda Says — Project Context
## What this project is
A Next.js app that converts legacy RTF files containing Vedic scriptures (Bhagavad-gītā, Śrīmad-Bhāgavatam, Caitanya-caritāmṛta, Kṛṣṇa Book, letters, magazines, etc.) into structured, browsable content with proper Unicode Sanskrit.
The original RTF files use **custom font encodings** (Indevr, ScaGoudy, Inbenr/Inbeni, RM Devanagari, Balaram, etc.) where ASCII characters map to Devanagari/Bengali glyphs and IAST diacritical marks.
---
## Tech Stack
- **Next.js 16** with Turbopack, React 19, TypeScript
- **Tailwind CSS v4** with `@tailwindcss/postcss`
- **shadcn/ui** (base-nova style, neutral base color, lucide icons) — components: Button, Card, Badge, Tooltip, Input
- **next-themes** for dark/light mode
- **Prisma v6** with MongoDB (provider `prisma-client-js`)
- **MongoDB** native driver (`mongodb` package) for seeding (Prisma requires replica set for writes)
- **Python 3** with BeautifulSoup4 + `indic-transliteration` for HTML extraction scripts
- **Database**: MongoDB at `DATABASE_URL` env variable with `?authSource=admin`
---
## Project Structure
```
spsays-training/
├── app/
│ ├── page.tsx → "Prabhupāda Says" landing page (Books, Glossary, RTFs)
│ ├── layout.tsx → ThemeProvider + TooltipProvider wrapper
│ ├── globals.css → shadcn/Tailwind v4 CSS with dark mode
│ ├── books/
│ │ ├── page.tsx → Books index (BG, SB, CC, SB1962, KB)
│ │ ├── chapter-viewer.tsx → Shared client component for verse-based books
│ │ ├── extracted-verse-count.tsx → (i) tooltip for verse counts
│ │ ├── bg/ → Bhagavad-gītā pages
│ │ │ ├── page.tsx → Chapter list
│ │ │ └── [chapter]/
│ │ │ ├── page.tsx → Redirects to first verse
│ │ │ └── [verse]/page.tsx → Single verse viewer
│ │ ├── sb/ → Śrīmad-Bhāgavatam pages
│ │ │ ├── page.tsx → Canto list
│ │ │ └── [canto]/
│ │ │ ├── page.tsx → Chapter list
│ │ │ └── [chapter]/
│ │ │ ├── page.tsx → Redirects to first verse
│ │ │ └── [verse]/page.tsx → Single verse viewer
│ │ ├── cc/ → Caitanya-caritāmṛta pages (same pattern, [lila]/[chapter]/[verse])
│ │ ├── sb1962/ → 1962 SB First Canto (same pattern as BG, [chapter]/[verse])
│ │ └── kb/ → Kṛṣṇa Book (prose, no verses)
│ │ ├── page.tsx → Volume list
│ │ └── [volume]/
│ │ ├── page.tsx → Chapter list + intro segments
│ │ └── [slug]/page.tsx → Reading page with sidebar
│ ├── glossary/
│ │ ├── page.tsx → Server component wrapper
│ │ └── glossary-client.tsx → Searchable client component
│ ├── rtfs/
│ │ ├── page.tsx → RTF file listing (gated by ALLOW_RTF env)
│ │ └── [slug]/page.tsx → RTF iframe viewer (gated by ALLOW_RTF env)
│ └── api/
│ ├── books/
│ │ ├── route.ts → GET all book metas
│ │ └── [book]/
│ │ ├── route.ts → GET single book meta
│ │ └── verses/route.ts → GET verses (?chapter=&verse=&division=&indexOnly=&page=&limit=)
│ └── glossary/route.ts → GET all glossary entries
├── components/
│ ├── ui/ → shadcn components (button, card, badge, tooltip, input)
│ ├── page-header.tsx → Shared header with ThemeToggle
│ ├── theme-provider.tsx → next-themes wrapper
│ └── theme-toggle.tsx → Sun/moon toggle
├── lib/
│ ├── db.ts → Prisma client singleton
│ ├── api.ts → Data access layer (getAllBooks, getBookMeta, getChapterVerses, etc.)
│ ├── utils.ts → cn() helper
│ └── jsons/books/ → Extracted JSON data
│ ├── bg/ → BG chapters/verses/*.json + intro.json
│ ├── sb/ → SB cantos/chapters/verses/*.json + intro.json
│ ├── cc/ → CC lilas/chapters/verses/*.json + intro.json
│ ├── sb-1962/ → 1962 SB chapters/verses/*.json + intro.json
│ ├── kb/ → Kṛṣṇa Book
│ │ ├── intro.json → Book metadata + chapter listing (89 chapters)
│ │ ├── vol-1/intro/ → 5 intro segments (flap, dedication, George Harrison, preface, intro)
│ │ ├── vol-1/chapters/ → 54 chapter JSONs
│ │ └── vol-2/chapters/ → 35 chapter JSONs
│ └── glossary/index.json → 2,749 glossary entries
├── prisma/
│ └── schema.prisma → Verse, BookMeta, ProseChapter, GlossaryEntry models
├── scripts/
│ ├── bg_extract.py → BG extraction from bg.html
│ ├── sb_extract.py → SB extraction from sb.html
│ ├── cc_extract.py → CC extraction from cc.html
│ ├── sb1962_extract.py → 1962 SB extraction from others-before-1977.html
│ ├── match_sb1962.py → Match 1962 SB transliterations with standard SB IAST
│ ├── glossary_extract.py → Glossary extraction from glossary.html
│ ├── kb_extract_intro.py → KB book-level intro + chapter listing
│ ├── kb_extract_vol1_intro.py → KB Vol 1 front matter segments
│ ├── kb_extract_vol1_chapters.py → KB Vol 1 chapters
│ ├── kb_extract_vol2_chapters.py → KB Vol 2 chapters
│ ├── fix-bengali-encoding.py → Fix Bengali font encoding in cc.html
│ ├── inject-fonts.py → Inject @font-face rules into converted HTML
│ ├── seed-db.ts → Main MongoDB seeder (all books + glossary)
│ └── seed-kb-vol2.ts → KB Vol 2 seeder (uses $push workaround)
└── public/
├── rtfs/ → Original RTF files (~430MB, 17 files)
├── html/ → HTML converted from RTF (textutil + @font-face injection)
└── fonts/ → Custom TTF fonts (Balaram, ScaGoudy, Indevr, Inbenr, etc.)
```
---
## Database Schema (Prisma + MongoDB)
```prisma
type Synonym {
word String
meanings String[]
}
model Verse {
id String @id @default(auto()) @map("_id") @db.ObjectId
book String // "bg" | "sb" | "cc" | "sb1962"
division String? // null for BG/SB1962; canto for SB; lila for CC
chapter Int
verse Int
verseEnd Int?
reference String
bookTitle String
devanagari String
bengali String? // only CC has Bengali
transliteration String
synonyms Synonym[]
translation String
purport String[]
@@unique([book, division, chapter, verse])
@@index([book])
@@index([book, division])
@@index([book, division, chapter])
@@map("verses")
}
model BookMeta {
id String @id @default(auto()) @map("_id") @db.ObjectId
book String @unique // "bg" | "sb" | "cc" | "sb1962" | "kb"
data Json
@@map("book_meta")
}
model ProseChapter {
id String @id @default(auto()) @map("_id") @db.ObjectId
book String // "kb"
volume Int
chapter Int? // null for intro segments
title String
type String // "chapter" | "intro-segment"
slug String
order Int
paragraphs String[]
@@unique([book, volume, slug])
@@index([book])
@@index([book, volume])
@@index([book, volume, type])
@@map("prose_chapters")
}
model GlossaryEntry {
id String @id @default(auto()) @map("_id") @db.ObjectId
term String
definition String
definitionHtml String
@@unique([term])
@@map("glossary")
}
```
---
## Books Extracted
| Book | Code | Type | Verses/Chapters | Division |
|------|------|------|-----------------|----------|
| Bhagavad-gītā As It Is | `bg` | verse | 653 verses, 18 chapters | — |
| Śrīmad-Bhāgavatam | `sb` | verse | 8,260 verses, 213 chapters | 10 cantos |
| Śrī Caitanya-caritāmṛta | `cc` | verse | 11,348 verses, 62 chapters | 3 līlās (adi/madhya/antya) |
| SB First Canto (1962 Edition) | `sb1962` | verse | 762 verses, 19 chapters | — |
| Kṛṣṇa Book | `kb` | prose | 89 chapters (54+35), 2 volumes | 2 volumes |
| Glossary | — | glossary | 2,749 entries | — |
---
## Font Encoding System
### ScaGoudy → IAST mapping (used in BG, SB, CC, glossary)
```
ä→ā Ä→Ā å→ṛ Å→Ṛ ñ→ṣ Ñ→Ṣ ö→ṭ Ö→Ṭ ë→ṇ Ë→Ṇ ò→ḍ Ò→Ḍ
ç→ś Ç→Ś ì→ṅ ï→ñ ü→ū Ü→Ū é→ī É→Ī à→ṁ ù→ḥ è→ṝ û→ṁl ÿ→ḷ
```
### Bengali fonts (Inbenr/Inbeni in CC)
Custom-encoded fonts where ASCII codepoints render as Bengali script. The cc.html had **Apple Symbol encoding corruption** — `textutil` mapped bytes through Symbol encoding instead of PUA. Fixed by `fix-bengali-encoding.py` which remaps 356,396 characters back to PUA range U+F0xx.
### Devanagari generation
IAST transliteration → Devanagari via `indic-transliteration` library. Key fixes:
- `join_virama_words()` — joins virāma + consonant AND virāma + vowel (fixed 12,000+ rendering issues in SB)
- `normalize_devanagari_anusvara()` — ङ् → ं before velars, ञ् → ं before palatals
---
## Verse Navigation (verse-based books)
- **One verse per page** with path-based URLs: `/books/bg/1/5`, `/books/sb/6/6/1`, `/books/cc/adi/1/3`
- **ChapterViewer** (`app/books/chapter-viewer.tsx`): Client component with:
- Sidebar showing ALL verse numbers (from lightweight index query)
- Prev/Next buttons at top and bottom
- Keyboard navigation (← →)
- Client-side verse caching + prefetching adjacent verses
- Click any verse in sidebar → fetches from API and navigates
- **API route** supports: `?verse=N` (single), `?indexOnly=true` (lightweight index), `?page=&limit=` (paginated)
- Server component SSRs the first verse, client handles navigation
---
## KB (Kṛṣṇa Book) — Prose Book
Different from verse books — chapters are prose paragraphs, no verse structure:
- `ProseChapter` model with `paragraphs: String[]`
- 5 intro segments per volume (flap, dedication, George Harrison intro, preface, introduction)
- Sanskrit verses within prose are grouped into `...
` blocks with `
` tags
- Reading page has left sidebar with chapter index, prev/next navigation
---
## Data Access Layer (`lib/api.ts`)
```typescript
// Book meta
getAllBooks(): Promise
getBookMeta(book: string): Promise
// Verse index (lightweight — for sidebar)
getChapterVerseIndex(book, division, chapter): Promise
// Single verse
getSingleVerse(book, division, chapter, verse): Promise
// Paginated verses
getChapterVerses(book, division, chapter, page?, limit?): Promise
// Prose chapters (for KB)
getProseChapters(book, volume, type?): Promise
getProseChapter(book, volume, slug): Promise
// Glossary
getAllGlossaryEntries(): Promise
```
The data access layer includes `normalizeProse()` which handles mixed storage (some chapters stored with native arrays, some with stringified JSON due to BSON corruption workaround). It also includes `sanitizeHtml()` which strips `