indianconstitution 1.3.0

A developer-first, research-grade Python framework for programmatic access to the Constitution of India — built for legal NLP, RAG, civic AI, and constitutional informatics.

IndianConstitution

A Developer-First, Research-Grade Python Framework for the Constitution of India

Sub-millisecond search · Strictly-typed API · Graph analysis · AI/RAG-ready · Zero external dependencies in core

📖 Documentation  ·  🚀 Quick Start  ·  🔬 Research Use  ·  📊 Benchmarks  ·  📜 Cite

---

Abstract

indianconstitution is a production-grade, research-ready Python library providing programmatic, structured, and type-safe access to the complete text of the Constitution of India — including all 448 articles, 12 schedules, the Preamble, and 106 amendments through the Constitution (One Hundred and Sixth Amendment) Act, 2023.

The library implements a zero-dependency inverted-index search engine (O(1) token lookup), a Pydantic v2 data model layer for type-safe constitutional data access, a NetworkX-backed relational graph for cross-article analysis, and a multi-format export engine — all designed for deployment in legal AI, retrieval-augmented generation (RAG), civic NLP, and constitutional informatics research.

This package is designed to meet the infrastructure standards expected by NeurIPS, ACL, and EMNLP research workflows — offering reproducibility, strict typing, and offline-first guarantees.

---

📑 Table of Contents

• Key Capabilities
• Architecture
• Quick Start
• Data Science Integration
• AI / Semantic Search
• CLI Reference
• Performance Benchmarks
• Research & Academic Use
• Citation
• Contributing
• Security
• License

---

✨ Key Capabilities

Capability	Description	Install Extra
Typed Article API	Fully annotated Article, Part, Schedule, Preamble Pydantic v2 models	core
Inverted-Index Search	Sub-millisecond lexical search via built-in inverted index — O(1) per token	core
Graph Analysis	NetworkX-backed relational graph of constitutional cross-references	[data]
Semantic / AI Search	Sentence-Transformers embeddings for contextual RAG retrieval	[ai]
Multi-Format Export	Export to JSON, CSV, and Markdown with a single call	core
pandas Integration	Direct DataFrame output of articles for data science workflows	[data]
Rich CLI	Terminal-native interface powered by Typer + Rich with syntax highlighting	core
Fully Offline	No API keys, no rate limits, no network calls required in core mode	core
Type Safety	100% mypy strict-mode compliance across all public APIs	core
Reproducible	Deterministic outputs; hermetic data layer pinned to 106th Amendment	core

---

📐 Architecture

┌─────────────────────────────────────────────────────────────────┐
│                        Public API Layer                         │
│          get_article()  ·  search()  ·  get_constitution()      │
└───────────────────────────────┬─────────────────────────────────┘
                                │
               ┌────────────────▼────────────────┐
               │    Constitution  (engine.py)     │
               │  Lazy-loading · Singleton cache  │
               └──┬──────────────┬───────────────┘
                  │              │
     ┌────────────▼───┐  ┌───────▼───────────┐  ┌──────────────────┐
     │  SearchEngine  │  │  ConstitutionGraph │  │    Exporter      │
     │ (inverted idx) │  │  (NetworkX graph)  │  │  JSON · CSV · MD │
     └────────────────┘  └────────────────────┘  └──────────────────┘
                  │
     ┌────────────▼──────────────────────────────────┐
     │             Pydantic v2 Data Layer             │
     │   Article · Part · Schedule · Preamble ·       │
     │   ConstitutionData · Amendment                 │
     └───────────────────────────────────────────────┘
                  │
     ┌────────────▼──────────────────────────────────┐
     │      constitution.json  (data/)                │
     │   Authoritative corpus — 106th Amendment 2023  │
     └────────────────────────────────────────────────┘

---

🚀 Quick Start

Installation

# ─── Core installation (zero external dependencies) ───────────────
pip install indianconstitution

# ─── Data science integrations (pandas, NetworkX, SciPy) ──────────
pip install "indianconstitution[data]"

# ─── AI/semantic search (sentence-transformers) ───────────────────
pip install "indianconstitution[ai]"

# ─── Full installation ────────────────────────────────────────────
pip install "indianconstitution[data,ai]"

Programmatic Access

from indianconstitution import get_article, search, get_constitution

# ─── Type-safe Article Retrieval ──────────────────────────────────
article = get_article("21A")
print(f"Article {article.number}: {article.title}")
# → Article 21A: Right to Education
print(f"  Part: {article.part}  |  Amendment: {article.inserted_by}")

# ─── Sub-millisecond Keyword Search ───────────────────────────────
results = search("right to equality", limit=5)
for r in results:
    print(f"  [{r.number}] {r.title} — Part {r.part}")

# ─── Full Constitution Object ──────────────────────────────────────
ic = get_constitution()
print(ic.preamble[:200])
print(f"Total Articles : {len(ic.data.articles)}")
print(f"Total Schedules: {len(ic.data.schedules)}")

Graph Analysis

from indianconstitution import get_constitution

ic = get_constitution()

# ─── Discover cross-article relational structure ──────────────────
related = ic.get_related_articles("32")
print("Article 32 references   :", related["references"])
print("Articles referencing 32 :", related["referenced_by"])

# ─── Compute centrality (which articles are most referenced?) ─────
import networkx as nx
G = ic.get_graph()
centrality = nx.degree_centrality(G)
top_5 = sorted(centrality, key=centrality.get, reverse=True)[:5]
print("Most referenced articles:", top_5)

Data Science Integration

from indianconstitution import get_constitution
import pandas as pd

ic = get_constitution()

# ─── Direct pandas DataFrame ──────────────────────────────────────
df = pd.DataFrame([a.model_dump() for a in ic.data.articles])
print(df[["number", "title", "part"]].head(10))

# ─── Multi-format export ──────────────────────────────────────────
ic.export("json",     "constitution_export.json")
ic.export("csv",      "constitution_export.csv")
ic.export("markdown", "constitution_export.md")

AI / Semantic Search

from indianconstitution import get_constitution

ic = get_constitution()

# ─── Contextual retrieval beyond keyword matching ─────────────────
# Requires: pip install "indianconstitution[ai]"
results = ic.semantic_search(
    "protection against arbitrary state action",
    top_k=5
)
for r in results:
    print(f"[{r.number}] {r.title}  (score: {r.score:.4f})")

RAG Pipeline Integration

from indianconstitution import get_constitution

ic = get_constitution()

def build_rag_context(query: str, top_k: int = 3) -> str:
    """Build a constitutional context block for LLM prompting."""
    results = ic.search(query, limit=top_k)
    context_blocks = []
    for article in results:
        context_blocks.append(
            f"**Article {article.number} — {article.title}**\n"
            f"{article.text}\n"
        )
    return "\n---\n".join(context_blocks)

# Usage with any LLM
context = build_rag_context("right to life and personal liberty")
print(context)

---

🖥️ Command-Line Interface

# ─── Retrieve and display an article with syntax highlighting ─────
indianconstitution get 21

# ─── Full-text search across all articles ────────────────────────
indianconstitution search "equality before law"

# ─── Display constitution statistics and metadata ─────────────────
indianconstitution stats

# ─── Export to JSON / CSV / Markdown ────────────────────────────
indianconstitution export --format json     --output constitution.json
indianconstitution export --format csv      --output constitution.csv
indianconstitution export --format markdown --output constitution.md

# ─── Show version ────────────────────────────────────────────────
indianconstitution --version

---

📊 Performance Benchmarks

Benchmarks measured on a commodity laptop (Intel Core i7-11th Gen, 16 GB RAM, Python 3.11, single thread, averaged over 1,000 iterations).

Operation	Latency	Notes
Initial data load	~45 ms	First call only; lazy-loaded from bundled JSON
Subsequent calls	~0 ms	In-process singleton cache — zero I/O
Keyword search (single token)	< 0.1 ms	Inverted-index O(1) lookup
Keyword search (multi-token, 3)	< 0.5 ms	Set intersection over index
Full CSV export (all articles)	~12 ms	Streaming writer
Full JSON export	~8 ms	orjson-compatible output
Graph construction (NetworkX)	~30 ms	One-time, lazy; cached thereafter
Semantic search (sentence-transformers)	~80 ms	GPU-accelerated with [ai] extra

Reproducibility note: All benchmarks are fully deterministic. The bundled constitution.json corpus is static and version-pinned. No external I/O is required in core mode.

---

🔬 Research & Academic Use

indianconstitution is engineered for research-grade deployment. It is suitable as a corpus infrastructure layer for:

• Constitutional NLP — structured retrieval for legal reasoning models, clause boundary detection
• RAG pipelines — grounding LLM outputs with authoritative, citation-traceable constitutional text
• Civic data science — network analysis of rights inter-dependencies and amendment history
• Legal education technology — interactive constitutional exploration platforms and quiz engines
• Multilingual legal AI — Hindi/English constitutional analysis (see [Unreleased] roadmap)
• Comparative constitutional law — structured data enabling cross-jurisdictional ML studies

Data Provenance & Corpus Integrity

The constitutional corpus (constitution.json) is derived from the official text of the Constitution of India as published by the Ministry of Law and Justice, Government of India. The data is:

• Curated and validated to the Constitution (One Hundred and Sixth Amendment) Act, 2023
• Structured against the Pydantic v2 schema — every field is validated on load
• Versioned alongside the library — data updates are tracked via the CHANGELOG.md
• Reproducible — the corpus is deterministic and hermetically bundled in the wheel

Reproducibility Checklist

For NeurIPS / ACL / EMNLP paper authors using this library:

• Pin to a specific release: pip install indianconstitution==1.3.0
• Record the __version__ in your experiment scripts
• Cite via the BibTeX entry below
• Archive the data corpus via Zenodo DOI (see Citation section)

---

📜 Citation

If you use indianconstitution in academic research, a thesis, or any published work, please cite it as follows:

BibTeX (Preferred)

@software{vikhram2026indianconstitution,
  author       = {S, Vikhram},
  title        = {{IndianConstitution: A Developer-First, Research-Grade
                   Python Framework for the Constitution of India}},
  year         = {2026},
  version      = {1.3.0},
  publisher    = {PyPI},
  url          = {https://github.com/Vikhram-S/IndianConstitution},
  doi          = {10.5281/zenodo.XXXXXXX},
  note         = {Available on PyPI: \url{https://pypi.org/project/indianconstitution/}
                  Corpus pinned to the Constitution (106th Amendment) Act, 2023.},
  license      = {Apache-2.0},
}

APA 7th Edition

S, Vikhram. (2026). IndianConstitution: A Developer-First, Research-Grade Python Framework for the Constitution of India (Version 1.3.0) [Software]. PyPI. https://doi.org/10.5281/zenodo.XXXXXXX

IEEE

V. S, "IndianConstitution: A Developer-First, Research-Grade Python Framework for the Constitution of India," version 1.3.0, 2026. [Online]. Available: https://github.com/Vikhram-S/IndianConstitution. DOI: 10.5281/zenodo.XXXXXXX.

ACL Anthology Format

Vikhram S. 2026. IndianConstitution: A Developer-First, Research-Grade Python Framework 
for the Constitution of India. Software release v1.3.0. 
Available: https://github.com/Vikhram-S/IndianConstitution

A machine-readable CITATION.cff is provided at the repository root for use with GitHub's "Cite this repository" feature and Zenodo DOI minting.

---

🛡️ Security

Security vulnerabilities should be reported privately via the GitHub Security Advisory mechanism. Do not open public issues for security reports.

• Supply-chain security: All GitHub Actions are pinned to immutable SHA hashes (OSSF Scorecard compliant)
• Dependency hygiene: Automated Dependabot PRs for all dependency updates
• Static analysis: CodeQL scanning on every push to main
• Vulnerability disclosure: See SECURITY.md for the full policy

---

🤝 Contributing

We welcome contributions from researchers, legal professionals, and developers. See CONTRIBUTING.md for guidelines on:

• Setting up the development environment
• Running the test suite (pytest + Hypothesis property-based testing)
• Code quality standards (Ruff + Mypy strict mode)
• Documentation contributions (MkDocs Material)
• Submitting pull requests and the review process

---

🙏 Acknowledgements

This library is developed and maintained by Vikhram S at Saveetha Engineering College, Chennai, India. We gratefully acknowledge:

• The Ministry of Law and Justice, Government of India for maintaining the authoritative constitutional text
• The developers of Pydantic, Typer, Rich, NetworkX, and sentence-transformers — the foundational libraries that power this framework
• The open-source community for their invaluable feedback and contributions

---

📄 License

Copyright © 2026 Vikhram S. Released under the Apache License 2.0.

You may use this software freely for academic, commercial, and government purposes with proper attribution. See LICENSE for the full text.

---

Built with ❤️ for the Indian legal tech ecosystem · Engineered to research-grade standards