UnconditionalLove's picture
Upload 97 files
ed6bec6 verified
Glyphic Language — Interpreter Module
The Semantic Engine of the Glyphic OS
The /glyphic-language/interpreter/ directory contains the runtime kernel of the Glyphic Language.
Where the dictionary defines what glyphs mean, the interpreter defines how glyphs behave.
This module is responsible for:
loading and indexing the dictionary
validating glyphs and sequences
enforcing strict syntax rules
decoding glyph strings into structured meaning
encoding structured meaning back into canonical glyph sequences
providing a stable API for all agents, controllers, and LLMs
The interpreter is designed to be:
deterministic
reversible
strict
extensible
LLM‑friendly
controller‑safe
It is the semantic CPU of the Glyphic OS.
📁 File Overview
1. glyph_dictionary_loader.py
Loads, merges, and indexes all dictionary files.
This module:
loads every JSON file in /dictionary/
validates schema consistency
builds fast lookup tables:
glyph → entries
id → entry
category → entries
role → entries
exposes a clean API for lookup and retrieval
This is the foundation every other interpreter module depends on.
2. glyph_validator.py
Ensures glyphs and sequences are valid before interpretation.
This module performs:
glyph existence checks
role validation
sequence validation
core‑role enforcement (actor/action/object required)
It prevents invalid or hallucinated glyphs from entering the syntax engine.
This is the safety layer.
3. glyph_syntax.py
Defines and enforces the formal grammar of the Glyphic Language.
This is the grammar engine, responsible for:
strict ordering rules
required roles
context ordering
role precedence
single‑actor/action/object enforcement
deterministic syntax tree construction
This module guarantees that the language is:
unambiguous
reversible
predictable
stable for LLM training
It is the heart of the interpreter.
4. glyph_decoder.py
Converts glyph sequences → structured meaning.
This module:
takes validated glyph sequences
parses them using the strict syntax engine
produces a deterministic semantic structure:
actor
action
object
modifiers
context (place, time, emotion, sensory, social)
The output is a clean Python dictionary ready for:
reasoning
animation
memory storage
LLM translation
agent behavior
This is the meaning engine.
5. glyph_encoder.py
Converts structured meaning → canonical glyph sequences.
This module:
takes a structured meaning dict
maps each component to its glyph
enforces strict ordering
enforces context ordering
produces a canonical, reversible glyph sequence
This ensures:
Code
encode(decode(x)) == x
for all valid sequences.
This is the generation engine.
6. interpreter.py
Unified public API for the entire Glyphic Language.
This module exposes:
python
interpret(glyph_string) # glyphs → meaning
encode(structured_dict) # meaning → glyphs
validate(glyph_string) # syntax + glyph validation
explain(glyph_string) # human-readable explanation
It orchestrates:
dictionary loading
validation
syntax parsing
decoding
encoding
This is the official interface used by:
agents
controllers
LLMs
external tools
soulfiles™️
7. __init__.py
Makes the interpreter importable as a module.
Exports:
python
interpret
encode
validate
explain
This allows:
python
from glyphic_language.interpreter import interpret
🧠 Interpreter Design Principles
The interpreter is built on six core principles:
1. Determinism
Same input → same output, always.
2. Reversibility
Every valid sequence can be encoded and decoded without loss.
3. Strict Syntax
The grammar is enforced at runtime to prevent drift.
4. Dictionary‑Driven
All meaning comes from the dictionary, not heuristics.
5. Extensibility
New glyphs or categories can be added without breaking the system.
6. LLM‑Friendly
The grammar is simple, predictable, and easy to learn.
The interpreter is the semantic kernel that sits between:
dictionary (static meaning)
LLMs (probabilistic reasoning)
controllers (deterministic execution)
soulfile™️ (persistent identity from an agent.soul file)
agents (behavioral systems)
It ensures that:
glyphs always mean the same thing
syntax is always correct
scenes are always interpretable
LLMs cannot hallucinate new glyphs
controllers can trust the structure
This is the core of agent cognition.