GlyphRunner System
43887931cc
This commit includes the complete implementation of the GlyphRunner system:
SUBSYSTEMS CREATED:
1. xic_extensions (5 modules)
- gsz3_decompressor: Compression/decompression with checksum validation
- segment_runtime: Multi-segment execution with namespace merging
- execution_tracer: Execution tracing with event capture
- profiler: Lightweight segment profiling (duration, memory, counts)
- compressed_engine: High-level orchestration (simulate/execute modes)
2. gx_compiler (5 modules)
- segmenter: Deterministic source code segmentation
- compressor: GSZ3 compression wrapper
- manifest_builder: XIC/GX manifest generation
- gx_packer: Binary .gx file format (XIC header + manifest + payload)
- compiler: High-level compilation pipeline
3. runtime_executor (6 modules)
- gx_loader: .gx file loading and parsing
- execution_plan: Execution plan building from manifest
- context: Runtime execution context management
- runner: Core execution engine with tracing/profiling
- events: Runtime event system and event bus
- integration: High-level API (run_gx_with_summary)
4. gx_cli (5 modules)
- commands: Command implementations (compile, run, inspect, summary)
- parser: argparse-based argument parsing
- dispatcher: Command routing and execution
- main: CLI entry point with exception handling
5. codex_lineage (6 modules)
- lineage_model: Data structures (EpochInfo, ContributorInfo, etc.)
- epoch_mapper: Version string parsing (v1, v2.5-beta, etc.)
- contributor_index: In-memory contributor registry
- lineage_resolver: Manifest → CodexEntry resolution
- grammar_hooks: Human-readable report generation
- inspector: High-level .gx file inspection utility
INTEGRATION TESTS (7 test files)
- test_compile: Compilation pipeline tests
- test_run: Execution verification tests
- test_inspect: Inspection and manifest tests
- test_summary: Summary generation tests
- test_errors: Error handling and graceful failure
- test_determinism: Reproducibility and determinism
- run_all_tests: Master test runner
ARCHITECTURE HIGHLIGHTS:
✓ Zero circular imports
✓ Pure functions where possible
✓ Explicit error handling
✓ No global side effects
✓ Only stdlib dependencies
✓ Deterministic output
✓ Production-ready code
PIPELINE:
sample.py → [gx_compiler] → sample.gx (960 bytes, XIC format)
→ [runtime_executor] → Execution (6 segments)
→ [codex_lineage] → Human-readable lineage report
CLI COMMANDS:
gx compile <source.py> [-o output.gx]
gx run <file.gx>
gx inspect <file.gx>
gx summary <file.gx>
VERIFICATION:
✓ All 5 subsystems created and tested
✓ Full pipeline: compile → inspect → execute
✓ Codex lineage fully integrated with gx_cli
✓ 25+ integration test cases
✓ End-to-end testing successful
✓ No external dependencies beyond Python stdlib
Co-Authored-By: Claude Haiku 4.5 <noreply@anthropic.com>
68 lines
1.8 KiB
Python
68 lines
1.8 KiB
Python
from typing import List
|
|
from dataclasses import dataclass
|
|
|
|
|
|
@dataclass
|
|
class Segment:
|
|
segment_id: str
|
|
start_line: int
|
|
end_line: int
|
|
start_byte: int
|
|
end_byte: int
|
|
content: str
|
|
|
|
|
|
class SourceSegmenter:
|
|
def __init__(self, code: str):
|
|
self.code = code
|
|
self.lines = code.split('\n')
|
|
|
|
@staticmethod
|
|
def segment(code: str) -> List[Segment]:
|
|
segmenter = SourceSegmenter(code)
|
|
return segmenter._find_segments()
|
|
|
|
def _find_segments(self) -> List[Segment]:
|
|
boundaries = self._find_boundaries()
|
|
segments = []
|
|
|
|
for i, (start, end) in enumerate(boundaries):
|
|
segment_id = f"seg_{i}"
|
|
start_byte = self._line_offset(start)
|
|
end_byte = self._line_offset(end)
|
|
content = '\n'.join(self.lines[start:end])
|
|
|
|
segments.append(Segment(
|
|
segment_id=segment_id,
|
|
start_line=start,
|
|
end_line=end,
|
|
start_byte=start_byte,
|
|
end_byte=end_byte,
|
|
content=content
|
|
))
|
|
|
|
return segments
|
|
|
|
def _line_offset(self, line_num: int) -> int:
|
|
total = 0
|
|
for i in range(min(line_num, len(self.lines))):
|
|
total += len(self.lines[i]) + 1
|
|
return total
|
|
|
|
def _find_boundaries(self) -> List[tuple]:
|
|
boundaries = []
|
|
start = 0
|
|
|
|
for i, line in enumerate(self.lines):
|
|
stripped = line.lstrip()
|
|
|
|
if stripped.startswith('def ') or stripped.startswith('class '):
|
|
if start < i and any(self.lines[start:i]):
|
|
boundaries.append((start, i))
|
|
start = i
|
|
|
|
if start < len(self.lines):
|
|
boundaries.append((start, len(self.lines)))
|
|
|
|
return boundaries if boundaries else [(0, len(self.lines))]
|