prism-religious Svelte Themes

Prism Religious

Bible study application with SvelteKit UI and Python importer for semantic Bible search

Prism Religious Studies

An academic research platform for religious texts, starting with Christianity. Provides semantic search, original language access, biblical geography, and AI-powered textual analysis for scholarly research.

Current Focus: Christianity (5 Bible translations, Hebrew/Greek texts, 1,342 biblical places) Future Expansion: Islam, Judaism, and other religious traditions

Features

Current (Christianity Module)

Semantic Search & Study Tools:

  • Multi-Translation Search: Semantic search across 5 English translations (KJV, ASV, BBE, YLT, Webster) using nomic-embed-text embeddings (768 dimensions)
  • Translation Comparison: View up to 4 translations side-by-side for detailed analysis
  • AI-Powered Insights: LLM-generated contextual commentary (historical context, literary analysis, theological themes)
  • Cross-References: AI-discovered related passages with explanatory connections
  • Translation Analysis: Detailed comparison of word choices, etymology, and textual variants
  • Citation Tools: One-click citation copying and export to TXT/JSON/CSV formats

Biblical Geography (1,342 Places):

  • Interactive Map: Leaflet.js-powered map with marker clustering for visual exploration
  • Confidence-Based Markers: Color-coded by archaeological confidence (Green ≥300, Yellow 80-300, Red <80)
  • Place Details: Coordinates, type (settlement/mountain/river/etc.), verse references, alternate names
  • Search & Filters: Semantic search for place names, filter by type and confidence level
  • Deep Linking: URL parameters support direct navigation from search results
  • Data Source: OpenBible.info (CC-BY 4.0) with comprehensive verse cross-references

Original Languages (Hebrew & Greek):

  • Hebrew Text: Westminster Leningrad Codex (WLC 4.20) for Old Testament
  • Greek Text: Society of Biblical Literature Greek New Testament (SBLGNT 1.0) for New Testament
  • Interlinear View: Word-by-word alignment with original script, transliteration, and English glosses
  • Strong's Numbers: Lexical references for deeper word study
  • Book Navigation: Browse by book, chapter, and verse with previous/next controls
  • Proper Typography: Unicode fonts (Noto Serif Hebrew, Noto Sans Greek) with RTL support

Visual Design

Color Palette (Mediterranean Academic Theme):

  • Terracotta (#E2725B): Primary accent, evokes archaeological sites
  • Olive Green (#6B8E23): Secondary accent, ancient landscape
  • Sand (#C2B280): Neutral backgrounds, parchment aesthetic
  • Deep Indigo (#1E3A5F): Scholarly depth, text emphasis

Typography:

  • Headings: Crimson Text (scholarly serif, 16-32px)
  • Body: Inter (modern sans-serif, 14-16px)
  • Hebrew: Noto Serif Hebrew (proper Unicode with RTL support)
  • Greek: Noto Sans Greek (classical letterforms)
  • Code/References: JetBrains Mono (monospace, 12-14px)

Layout:

  • Responsive: Mobile-first design (320px+)
  • Tablet: 2-column layouts (768px+)
  • Desktop: 3-column layouts (1024px+)
  • Navigation: Global top nav with active state indicators

Architecture

Frontend

  • Framework: SvelteKit 2.5 + TypeScript
  • Styling: Tailwind CSS 3.4 with custom Mediterranean theme
  • Mapping: Leaflet.js 1.9.4 with marker clustering
  • Icons: Lucide Svelte
  • Routing: Multi-page SPA with deep linking support

Backend

  • Search: Prism API (PostgreSQL + pgvector)
  • AI: Ollama (Qwen 2.5 14B, temperature 0.7)
  • Geography: Prism documents (domain: geography/biblical)
  • Original Texts: SWORD Project modules (via Python parser)
  • Caching:
    • AI responses: 1-hour in-memory cache
    • Geography: 10-minute cache
    • Verses: 30-minute cache

Multi-Page Structure

  • / - Landing page with features and statistics
  • /search - Semantic search with AI insights
  • /geography - Interactive map of biblical places
  • /languages - Hebrew/Greek original texts with interlinear view
  • /about - Methodology, data sources, and academic attributions

Development

# Install dependencies
npm install

# Run dev server
npm run dev

# Build for production
npm run build

# Type checking
npm run check

Docker Deployment

# Build image
docker build -f docker/Dockerfile -t prism-religious-ui:latest .

# Run container
docker run -p 3000:3000 \
  -e VITE_PRISM_API_URL=http://prism:8100 \
  -e VITE_OLLAMA_API_URL=http://ollama:11434 \
  prism-religious-ui:latest

Environment Variables

  • VITE_PRISM_API_URL: Prism REST API endpoint (default: http://prism:8100)
  • VITE_OLLAMA_API_URL: Ollama API endpoint (default: http://ollama:11434)

Usage

  1. Navigate to Search page (or homepage search bar)
  2. Enter a query: word ("shepherd"), phrase ("love your enemies"), or reference ("John 3:16")
  3. Select translations to search (default: all 5)
  4. Click a verse in the results list to view details
  5. Center panel shows selected translations side-by-side
  6. Right panel provides AI-generated commentary, cross-references, and translation analysis
  7. Click Copy Citation or Export for academic workflows

Biblical Geography

  1. Navigate to Geography page
  2. Explore the interactive map with 1,342 biblical places
  3. Click markers to see place details (coordinates, verse references, alternate names)
  4. Use filters to narrow by place type (settlement, mountain, river, etc.) or confidence level
  5. Search for specific places using the semantic search bar
  6. Click Search Verses About This Place to find related passages
  7. Or navigate from search results using View on Map button (📍 icon indicates geography data)

Original Languages

  1. Navigate to Languages page
  2. Toggle between Hebrew (Old Testament) and Greek (New Testament) tabs
  3. Select book, chapter, and verse using dropdowns
  4. Switch to Interlinear view to see word-by-word alignment with transliteration and glosses
  5. Use Previous/Next Verse buttons to browse
  6. Strong's numbers displayed when available for lexical study
  7. Or navigate from search results using View in Hebrew/Greek button (א/Α icons indicate original text availability)

Performance

Search & AI:

  • Semantic search response: <500ms (including embedding + vector search)
  • Translation comparison: <200ms
  • AI commentary (first time): ~3-5s (Qwen 2.5 14B inference)
  • AI commentary (cached): <100ms (1-hour cache)

Geography Module:

  • Map load time: <3s (1,342 markers with clustering)
  • Place search: ~400ms (semantic search with filtering)
  • API cache: 10-minute TTL for frequently accessed data

Original Languages:

  • Verse fetch: <200ms (cached), <500ms (first load)
  • Verse cache: 30-minute TTL per verse
  • Interlinear rendering: <100ms (client-side)

Scalability: Supports 20+ concurrent users with in-memory caching

Data

Search Corpus:

  • Verses: 18,069 indexed verses across 5 English translations
  • Embeddings: nomic-embed-text (768 dimensions)
  • Vector Index: HNSW algorithm with 0.6-0.9 similarity scores

Biblical Geography:

  • Places: 1,342 locations with coordinates (latitude/longitude)
  • Confidence Scores: 0-500 scale (archaeological certainty)
  • Verse References: Comprehensive cross-references for each place
  • Place Types: Settlement, mountain, river, valley, region, body of water, wilderness, etc.

Original Texts:

  • Hebrew: Westminster Leningrad Codex (WLC 4.20) - complete Old Testament
  • Greek: Society of Biblical Literature Greek New Testament (SBLGNT 1.0) - complete New Testament
  • Interlinear Data: Word alignment, transliteration (scholarly standard), English glosses
  • Lexical References: Strong's numbers for deeper word study

Data Sources & Attribution

  • Hebrew Text: Westminster Leningrad Codex (WLC 4.20)
  • Greek Text: Society of Biblical Literature Greek New Testament (SBLGNT 1.0)
  • Geography: OpenBible.info (CC-BY 4.0) - 1,342 biblical places
  • Translations: SWORD Project modules (Public Domain)
  • Search Engine: Prism (Personal Semantic Data Layer)

Credits

  • Built with: Claude Code
  • LLM: Anthropic Claude Sonnet 4.5
  • Embedding Model: nomic-embed-text (768 dimensions)
  • Original Language Parsing: SWORD Project

Top categories

Loading Svelte Themes