Documentation

AST Analysis Phase

The first gate in the Pyvorin pipeline: what the AST walk looks for, what disqualifies a function, and how support tiers are assigned.

Compiler Pipeline Overview

How Pyvorin turns a Python function into native machine code — every stage from parsing to the wrapper that makes the result callable.

Guards and Deoptimisation

The contract behind Pyvorin compiled code: what guards protect, what happens when one fails, and why a failed guard can never corrupt your program semantics.

How Local Compilation Works

The full path from import through compilation to native execution in pyvorin-native 1.0.9 — where the cache lives, what the fallback does, and exactly what does and does not leave the machine.

10 min read

LLVM IR Generation

How pyvorin-native 1.0.9 lowers a typed Python function to LLVM IR through llvmlite — with a real before-and-after example taken from the compiler's own output.

Native Code Generation

How pyvorin-native 1.0.9 turns LLVM IR into a loadable shared object: object emission, linking against the runtime, and when SIMD vectorisation fires.

Network Requirements for Pyvorin

Every endpoint pyvorin-native 1.0.9 contacts, when each fires, offline grace windows, air-gap implications, TLS, and the env overrides that redirect traffic.

7 min read

Protected Component Delivery and Verification

Every file in pyvorin-native 1.0.9 ships in the wheel and is SHA256 hash-checked at import. The Ed25519 manifest and cache signatures are not yet active in the shipped build — here is exactly what is and is not verified today.

8 min read

Runtime Libraries

What pyvorin-native 1.0.9's runtime libraries do: typed containers, the Python–native ABI, CPython interop, and where the runtime itself limits speed.

Type Inference System

Dynamic Python in, static native code out: how Pyvorin infers types, what it assumes, and how guards keep wrong assumptions harmless.