Comprehensive polyglot code repository, foundational algorithms, system kernels, compilers, and architectural patterns.
| Guide | What it covers |
|---|---|
docs/STRUCTURE.md |
A map of the repository: what every directory contains and how large it is |
docs/USAGE.md |
How to run, build, validate and study each part (with commands per module) |
docs/README.md |
Documentation index and links to every module README |
SOURCES.md |
Upstream projects, licenses and attribution |
New here? Read docs/STRUCTURE.md for the map, then
docs/USAGE.md for the exact commands.
This repository contains over 3.0 GB of curated codebases and 400,000+ files safely preserved on GitHub. To keep your local machine footprint minimal (< 50 MB), the repository utilizes Git's native Sparse-Checkout mechanism.
Use the helper utility scripts/sparse_checkout.sh:
# 1. Reset local checkout to minimal footprint (< 50 MB: algorithms, architectures, scripts)
./scripts/sparse_checkout.sh minimal
# 2. Check out a specific language or system to local disk on demand
./scripts/sparse_checkout.sh add languages/01_python
./scripts/sparse_checkout.sh add languages/02_c/linux_kernel
./scripts/sparse_checkout.sh add languages/09_rust
# 3. Remove a language or system from local disk (safely preserved on GitHub remote)
./scripts/sparse_checkout.sh remove languages/02_c/linux_kernel
# 4. View current local checked-out directories
./scripts/sparse_checkout.sh status
# 5. List all available modules in the repository
./scripts/sparse_checkout.sh list| Top-Level Directory | Description | See |
|---|---|---|
algorithms/ |
Polyglot algorithms across 25 languages plus CUDA, assembly, PTX, SASS, LLVM IR and WebAssembly | algorithms/README.md |
architectures/ |
Design patterns, DDD, enterprise integration, cloud, system design, principles and templates | architectures/README.md |
devops/ |
CI/CD, Docker, Kubernetes, Terraform, Ansible, Helm, monitoring and operations scripts | devops/README.md |
interview_prep/ |
Interview study vault: algorithms, C++, concurrency, Python, networking, OS, architecture, system design, behavioral, math | interview_prep/README.md |
formal_physics/ |
Lean 4 + Mathlib formalization of Taylor's Classical Mechanics (Ch. 6, 7, 13) | formal_physics/README.md |
formal_statistics/ |
Lean 4 + Mathlib formalization of Kutner's regression analysis (Ch. 1, 2) | formal_statistics/README.md |
web/ |
Web platform: HTML, CSS, Sass, Tailwind, JS/TS, React, Vue, Web Components, JSON, XML, GraphQL | web/README.md |
databases/ |
Graph/NoSQL query languages: Cypher, Gremlin, SPARQL, Datalog, MongoDB, Cassandra, Redis, Elasticsearch, Flux, PromQL, pgvector | databases/README.md |
scripts/ |
Computational scripts (math, physics, ML, finance) and utilities | scripts/ |
docs/ |
Repository documentation: structure map and usage guide | docs/README.md |
languages/ |
Upstream source code from major open-source projects (sparse checkout) | languages/ |
The algorithms/ directory hosts standalone, production-ready algorithm implementations across all 25 canonical languages, plus six systems/GPU tiers that go below the language level: CUDA, assembly (x86-64 / AArch64 / RISC-V), PTX, SASS, LLVM IR and WebAssembly. Each language directory (01_python/ through 25_lean4/, and 26_cuda/ through 31_wasm/) contains implementations organized by category: sorting, graphs, dynamic programming, strings, data structures, math, cryptography, and machine learning.
Full coverage matrix: algorithms/README.md
| # | Directory | Language | Key Algorithms |
|---|---|---|---|
| 1 | algorithms/01_python/ |
Python | 5 sorts, 12 graph algos (BFS/DFS/Dijkstra/A*/Kosaraju/ConvexHull/TopoSort), 7 DP, FFT, Miller-Rabin, Strassen, KMP, Huffman, k-means, Caesar |
| 2 | algorithms/02_c/ |
C | 3 sorts, BFS/DFS/Dijkstra, DP, KMP, 4 data structures, Miller-Rabin, matmul_tiled |
| 3 | algorithms/03_cpp/ |
C++ | 2 sorts, BFS/Dijkstra/A*/Kosaraju/ConvexHull/TopoSort, FFT, Huffman, KMP, k-means |
| 4 | algorithms/04_java/ |
Java | 3 sorts, BFS/Dijkstra/A*, SegmentTree, Trie, KMP, Huffman, MathUtils |
| 5 | algorithms/05_csharp/ |
C# | 3 sorts, BFS/DFS/Dijkstra, DP, KMP, Trie, math |
| 6 | algorithms/06_javascript/ |
JavaScript | 3 sorts, BFS/Dijkstra, DP, KMP, math |
| 7 | algorithms/07_typescript/ |
TypeScript | 2 sorts, BFS, DP (LCS+Levenshtein), KMP, math |
| 8 | algorithms/08_r/ |
R | 3 sorts, BFS/DFS/Dijkstra, DP, KMP, k-means, math |
| 9 | algorithms/09_rust/ |
Rust | 3 sorts, BFS/DFS/Dijkstra/A*/ConvexHull/TopoSort, Trie, FFT, Huffman, Miller-Rabin, Caesar, Vigenere |
| 10 | algorithms/10_sql/ |
SQL | 6 analytical query patterns (recursive CTEs, windows, gaps & islands, pivots) |
| 11 | algorithms/11_golang/ |
Go | 3 sorts, BFS/DFS/Dijkstra/A*/Kosaraju/ConvexHull/TopoSort, Trie, Stack, Worker Pool, KMP |
| 12 | algorithms/12_php/ |
PHP | 2 sorts, BFS/Dijkstra, LCS, KMP, Trie, math |
| 13 | algorithms/13_swift/ |
Swift | 2 sorts, BFS, LCS, KMP, Trie, math |
| 14 | algorithms/14_julia/ |
Julia | 2 sorts, BFS, LCS, FFT, linear regression, math |
| 15 | algorithms/15_ruby/ |
Ruby | 2 sorts, BFS/Dijkstra, LCS, KMP, math |
| 16 | algorithms/16_kotlin/ |
Kotlin | 2 sorts, BFS, LCS, KMP, Trie, math |
| 17 | algorithms/17_matlab/ |
MATLAB | 2 sorts, BFS, LCS, KMP, math |
| 18 | algorithms/18_ocaml/ |
OCaml | Merge sort, Quick sort, BFS, LCS, KMP, math |
| 19 | algorithms/19_lua/ |
Lua | Bubble sort, BFS, LCS, Knapsack, KMP, math |
| 20 | algorithms/20_lisp/ |
Common Lisp | Bubble sort, BFS, LCS, KMP, math |
| 21 | algorithms/21_scala/ |
Scala | 2 sorts, BFS, LCS, math |
| 22 | algorithms/22_haskell/ |
Haskell | 2 sorts, BFS, LCS (naive + DP), math |
| 23 | algorithms/23_elixir/ |
Elixir | 2 sorts, BFS, LCS, math |
| 24 | algorithms/24_fortran/ |
Fortran | Bubble sort, BFS, Knapsack, GCD, matmul_blocked |
| 25 | algorithms/25_lean4/ |
Lean 4 | GCD (computational + formal proof), Fibonacci, Insertion sort, BFS (with termination proofs) |
| 26 | algorithms/26_cuda/ |
CUDA | 175 upstream NVIDIA samples: vectorAdd, reduction, scan, matrixMul, streams, cooperative groups, cuBLAS/cuFFT/Thrust, tensor-core kernels |
| 27 | algorithms/27_assembly/ |
Assembly | x86-64 (25), AArch64 (8), RISC-V 64 (6) syscall programs: sorts, GCD, Fibonacci, SIMD, CPUID, RDTSC |
| 28 | algorithms/28_ptx/ |
PTX | 20 kernels: vectorAdd, saxpy, reduction, tiled matmul, warp vote/shuffle, atomics, cp.async |
| 29 | algorithms/29_sass/ |
SASS | 9 annotated sm_90 (Hopper) disassembly listings plus reference notes |
| 30 | algorithms/30_llvm_ir/ |
LLVM IR | 47 modules lowered from the C reference suite |
| 31 | algorithms/31_wasm/ |
WebAssembly (WAT) | 18 modules: control flow, linear memory, SIMD, tables, multi-value |
| Directory | Language | Included Canonical Systems & Libraries |
|---|---|---|
languages/01_python/ |
Python | SciPy, AstroPy, CPython standard library, SymPy, scikit-learn |
languages/02_c/ |
C | Linux Kernel Core (mm, fs, net, crypto), FreeBSD Subsystems, PostgreSQL Engine, SQLite, curl, GCC Optimizer Core |
languages/03_cpp/ |
C++ | LLVM Optimizer, Clang AST, LLD Linker, DuckDB, ClickHouse DBMS, V8 Engine, PyTorch ATen, TensorFlow Core, Zig Std |
languages/04_java/ |
Java | OpenJDK java.base, Apache Kafka, Apache Flink, Google Guava, Commons Math |
languages/05_csharp/ |
C# | .NET Base Class Library |
languages/06_javascript/ |
JavaScript | Node.js Core Modules, Lodash, D3 |
languages/07_typescript/ |
TypeScript | TypeScript Compiler & Typechecker, RxJS |
languages/08_r/ |
R | R Core Interpretation Engine & Standard Statistical Packages |
languages/09_rust/ |
Rust | Rust Standard Library, Rustc Compiler Internals, Tokio, Ripgrep |
languages/10_sql/ |
SQL | PostgreSQL Schemas, Analytical dbt Models |
languages/11_golang/ |
Go | Go Standard Library & Runtime, Kubernetes, etcd Raft |
languages/12_php/ |
PHP | PHP Zend VM Engine & Core Standard Extensions |
languages/13_swift/ |
Swift | Swift Standard Library, Swift Compiler & SIL Optimizer |
languages/14_julia/ |
Julia | Julia Standard Library, OrdinaryDiffEq.jl, Distributions.jl |
languages/15_ruby/ |
Ruby | Ruby Standard Library & Core C Extensions |
languages/16_kotlin/ |
Kotlin | Kotlin Standard Library, Coroutine-ready Data Structures |
languages/17_matlab/ |
MATLAB / Octave | Numerical Linear Algebra, Signal Processing |
languages/18_ocaml/ |
OCaml | OCaml Standard Library, Lexer & Parser |
languages/19_lua/ |
Lua | Lua Interpreter, Virtual Machine, Incremental GC |
languages/20_lisp/ |
Common Lisp / Clojure | Clojure Standard Library, AIMA AI Implementations |
languages/21_scala/ |
Scala | Apache Spark, Cats |
languages/22_haskell/ |
Haskell | Glasgow Haskell Compiler (GHC), Haskell Base Library |
languages/23_elixir/ |
Elixir | Elixir Standard Library, Actor Concurrency Primitives |
languages/24_fortran/ |
Fortran | Fortran-lang Standard Library, BLAS / Matrix Routines |
languages/25_lean4/ |
Lean 4 | Mathlib Formal Mathematics, Category Theory, Topology |
Design patterns across the Gang of Four (GoF), concurrency primitives, and distributed systems patterns. Full index: architectures/README.md
| Directory | Content |
|---|---|
architectures/cpp_patterns/ |
24 GoF patterns in classic C++ |
architectures/go_patterns/ |
24 GoF patterns with tests in Go |
architectures/python_patterns/ |
GoF + pythonic patterns with linting and pytest |
architectures/java_patterns/ |
Builder and Observer demos |
architectures/concurrency_patterns/ |
Actor models, thread pools, worker pools |
architectures/distributed_patterns/ |
Consistent hashing, circuit breaker |
architectures/typescript_patterns/ |
Middleware pipeline, repository pattern |
architectures/enterprise_patterns/ |
Message router, splitter, aggregator, outbox, saga, dead letter channel |
architectures/ddd_patterns/ |
Entity, value object, aggregate, repository, domain event, specification |
architectures/cloud_patterns/ |
Circuit breaker, bulkhead, retry/backoff, cache-aside, sidecar, strangler fig |
architectures/system_design/ |
Rate limiter, load balancer, LRU cache, WAL, sharding, CDN, CAP, Bloom guard |
architectures/principles/ |
SOLID, DRY/KISS/YAGNI, separation of concerns, Law of Demeter, twelve-factor |
architectures/clean_architecture/ |
Hexagonal, onion, layered, dependency inversion, ports and adapters |
architectures/templates/ |
ADR, RFC, design doc, runbook, C4 (Mermaid) |
architectures/api_design/ |
REST guidelines, versioning, pagination, error handling, GraphQL vs REST |
architectures/integration_patterns/ |
EIP: channels, pipes & filters, translator, wire tap, publish/subscribe, competing consumers |
architectures/data_patterns/ |
Unit of work, data mapper, active record, identity map, lazy load, DAO |
architectures/resilience_patterns/ |
Timeout, hedge, fail-fast, fallback, load shedding, backpressure, throttling, chaos |
architectures/microservices_patterns/ |
API gateway, BFF, discovery, mesh, database per service, CQRS, event sourcing, sagas |
architectures/distributed_patterns/ |
Leader election, gossip, vector clocks, CRDTs, 2PC, quorum, fencing tokens, Raft |
architectures/functional_patterns/ |
Monad, option/either, immutable value, currying, pipeline, memoization |
architectures/security_patterns/ |
Defense in depth, least privilege, zero trust, OAuth2/OIDC, JWT, RBAC/ABAC, threat modeling |
architectures/testing_patterns/ |
Test pyramid, doubles, property-based, contract, golden master, mutation |
architectures/deployment_patterns/ |
Blue/green, canary, rolling update, feature toggles, shadow, A/B testing |
architectures/event_driven_patterns/ |
Event notification, state transfer, event sourcing, event streams |
architectures/observability_patterns/ |
Logging, metrics, tracing, health checks, correlation IDs |
architectures/ui_patterns/ |
MVC, MVP, MVVM, Flux, Redux, VIPER |
architectures/architecture_styles/ |
Monolith, modular monolith, microservices, serverless, event-driven, SOA, space-based, mesh |
Beyond the high-level language suites, algorithms/ follows the GPU and systems toolchain from parallel source down to machine instructions.
| Directory | Level | Description |
|---|---|---|
algorithms/26_cuda/ |
parallel C++ | Curated NVIDIA/cuda-samples (BSD-3-Clause) organized by sample category |
algorithms/27_assembly/ |
machine code | Hand-written Linux syscall programs for x86-64 (AT&T/Intel), AArch64 and RISC-V 64 |
algorithms/28_ptx/ |
virtual ISA | Standalone PTX 8.x kernels (thread indexing, shared memory, warp intrinsics, atomics, cp.async) |
algorithms/29_sass/ |
device ISA | Annotated sm_90 (Hopper) SASS disassembly listings (nvdisasm reference) |
algorithms/30_llvm_ir/ |
compiler IR | LLVM IR lowered from the C reference algorithms; verify with opt, run with lli |
algorithms/31_wasm/ |
stack machine | WebAssembly text modules validated with wasmtime |
The web/ directory collects markup, styling, scripting, component and data-format examples.
| Directory | Content |
|---|---|
web/html/ |
Semantic HTML5, forms, ARIA, tables, canvas, inline SVG, responsive images, <dialog> |
web/css/ |
Reset, flexbox, grid, custom properties, animations, container queries, scroll-driven effects |
web/scss/ |
Sass variables, nesting, mixins, functions, placeholders, loops, maps, @use modules |
web/tailwind/ |
Tailwind configuration, @apply component layers and responsive utility components |
web/webcomponents/ |
Custom elements, shadow DOM, templates and adopted stylesheets |
web/javascript/ |
Modern JavaScript: DOM, fetch, promises, ESM, closures, classes, observers, workers |
web/typescript/ |
Generics, discriminated unions, utility/mapped/conditional types, decorators |
web/react/ |
React 19 function components, hooks, context, reducers and error boundaries |
web/vue/ |
Vue 3 Composition and Options APIs, props/emit, v-model |
web/json/ |
JSON/JSONL, JSON Schema, JSON-LD, GeoJSON, JSON Patch, OpenAPI |
web/xml/ |
Sitemaps, RSS/Atom, SVG, XSLT, XSD, SOAP and configuration |
web/graphql/ |
GraphQL SDL, queries, mutations, subscriptions, fragments, federation |
databases/ covers graph and NoSQL engines plus time-series and vector search; the relational suite remains in algorithms/10_sql/.
| Directory | Engine | Language |
|---|---|---|
databases/graph/cypher/ |
Neo4j | Cypher |
databases/graph/gremlin/ |
Apache TinkerPop | Gremlin |
databases/graph/sparql/ |
RDF stores | SPARQL 1.1 |
databases/graph/datalog/ |
Souffle / DDlog | Datalog |
databases/document/mongodb/ |
MongoDB | MQL / mongosh |
databases/wide_column/cassandra/ |
Apache Cassandra | CQL |
databases/key_value/redis/ |
Redis | RESP / Lua |
databases/search/elasticsearch/ |
Elasticsearch | Query DSL |
databases/timeseries/ |
InfluxDB / Prometheus | Flux / PromQL |
databases/vector/ |
PostgreSQL + pgvector | SQL |
The devops/ directory holds runnable pipeline, infrastructure and operational
configuration.
| Directory | Tooling |
|---|---|
devops/ci/github_actions/ |
CI, matrix builds, CD with OIDC, release please, CodeQL, Docker publish, reusable workflows |
devops/ci/gitlab/ |
staged pipelines, manual deploy, per-MR review apps |
devops/ci/jenkins/ |
declarative pipeline, multibranch, shared library |
devops/docker/ |
Node/Python/Go images, multi-stage and distroless builds, Compose |
devops/kubernetes/ |
Deployment, Service, ConfigMap, Secret, Ingress, HPA, StatefulSet, DaemonSet, Job/CronJob, RBAC, NetworkPolicy, PDB, probes |
devops/terraform/ |
provider/backend, variables, outputs, VPC, EC2, S3 + ALB |
devops/ansible/ |
playbook, inventory, groups and roles |
devops/helm/ |
chart metadata, values and templated workloads |
devops/monitoring/ |
Prometheus, alert rules, Loki, Grafana, OpenTelemetry, blackbox |
devops/scripts/ |
deploy, backup, healthcheck, rolling restart, log rotation |
devops/make/ |
application, Python and self-documenting task runners |
The interview_prep/ directory is a study vault built around a concurrent
algorithms-plus-systems plan, with runnable examples and focused notes.
| Directory | Track | Contents |
|---|---|---|
interview_prep/algorithms/ |
Algorithms (see also algorithms/) |
LeetCode patterns: arrays, strings, linked lists, trees, graphs, DP, heap/trie/bit, techniques |
interview_prep/cpp/ |
C++ | STL components from scratch, A Tour of C++, Effective C++ examples, 100 questions + answer programs |
interview_prep/concurrency/ |
Concurrency | threads, locks, condition variables, atomics, lock-free stacks/queues, futures, thread pools |
interview_prep/python/ |
Python | Fluent Python: data model, generators, decorators, descriptors, metaclasses, asyncio |
interview_prep/networking/ |
Networking | top-down Ch. 1-5: application, transport, network, link, sockets and simulations |
interview_prep/operating_systems/ |
Operating Systems | OSTEP: CPU/memory virtualization, concurrency, persistence |
interview_prep/computer_architecture/ |
Architecture | caches, branch prediction, memory hierarchy, pipelining |
interview_prep/system_design/ |
System Design | scalability, low-level/high-throughput, DDIA notes, worked designs |
interview_prep/behavioral/ |
Behavioral | STAR method, story banks, leadership and question bank |
interview_prep/math_stats/ |
Math & Stats | probability, statistics, linear algebra, Markov chains, Monte Carlo |
interview_prep/deep_dives/ |
Deep dives | numbered C++20, STL, smart-pointer, reimplementation, networking (15-19) and OS (20-31) topics |
Run the self-checks with make test-interview.
The formal_physics/ directory contains a machine-checked formalization of
John R. Taylor's Classical Mechanics:
- Chapter 6 (Calculus of Variations): functionals, the first variation as a Fréchet derivative, stationarity, the Euler–Lagrange equation, the Beltrami identity, the brachistochrone and the cycloid, the second variation with Legendre's condition and conjugate points.
- Chapter 7 (Lagrange's Equations): configuration space, the generalized
momentum, Newton ⇔ Euler–Lagrange, curvilinear coordinates, rheonomic
T₂ + T₁ + T₀, Hamilton's principle, holonomic constraints, Noether's theorem. - Chapter 13 (Hamiltonian Mechanics): phase space, Hamilton's equations, energy conservation, the Legendre transform, the Poisson bracket, the symplectic operator and Liouville's theorem.
Every theorem is checked by the Lean kernel against Mathlib (v4.16.0); there are
no sorrys. Build with formal_physics/scripts/verify.sh or
make test-formal-physics.
Computational scripts across multiple domains:
| Domain | Scripts |
|---|---|
mathematics/ |
Euclidean algorithms, finite geometry (Fano plane), multilinear algebra, perfect numbers |
physics/ |
Curvilinear tensors, rigid body inertia, symbolic Euler-Lagrange |
machine_learning/ |
k-means clustering, logistic regression |
finance/ |
Corporate valuation (DCF), dividend growth models |
numerical_analysis/ |
Root finding (bisection, Newton-Raphson, Brent) |
computer_science/ |
K4 planar graphs, Major system cipher |
distributed_systems/ |
Consistent hashing ring |
Utilities: sparse_checkout.sh (git sparse checkout manager), ingest_10x.py (bulk upstream ingestion).
# Run everything
make test
# Run specific suites
make test-python # Python algorithms
make test-c # C algorithms
make test-cpp # C++ algorithms
make test-java # Java algorithms
make test-rust # Rust algorithms
make test-go # Go algorithms
make test-scripts # Computational scripts
make test-algorithms # All polyglot algorithm suites
make test-architectures # Design patterns and concurrency
make test-asm # x86-64 assembly (assemble + run)
make test-llvm-ir # Verify generated LLVM IR
make test-wasm # Validate WebAssembly text modules
make test-web # HTML/CSS/SCSS/JS/TS/JSON/GraphQL validation
make test-cuda # NVIDIA samples (requires nvcc; skipped when absent)
make test-devops # Validate DevOps YAML and shell scripts
make test-interview # Run interview-prep self-checks (Python + C++)
make test-formal-physics # Lean 4 / Mathlib formal physics (needs a local Mathlib)
make test-formal-statistics # Lean 4 / Mathlib formal statistics (needs a local Mathlib)
# Clean build artifacts
make cleanSee docs/USAGE.md for per-module commands. Targets whose
toolchains are unavailable are skipped rather than failed (for example CUDA
without nvcc, or the Lean modules without a local Mathlib — use
formal_physics/scripts/verify.sh to fetch it once).
All imported code preserves original author attributions, notices, and copyright statements. Complete upstream repository links, author organizations, and license terms are detailed in SOURCES.md.
© 2026 Igor Kan