Full implementation - Vision AI, config, improved pipeline
Major changes: - vision_analyzer.py: Real OpenAI/Anthropic vision API integration - Component detection with confidence scores - Atomizer hints extraction (objectives, constraints, parameters) - Material and feature identification - Timeline correlation with transcript - config.py: Full configuration system - API settings (provider, keys, models) - Processing settings (Whisper model, frame interval, scene detection) - Output settings (BOM, hints, PDF template) - Config file support (~/.cad-documenter.toml) - audio_analyzer.py: Enhanced transcription - Audio stream detection - Graceful fallback for missing audio - Keyword extraction - Technical term detection - Timeline correlation - video_processor.py: Smart frame extraction - Scene change detection via ffmpeg - Configurable thresholds - Best frame selection - doc_generator.py: Improved output - Better Markdown templates - BOM CSV export - Atomizer hints JSON - Component cards - cli.py: Rich CLI with progress indicators - Config file support - --init-config flag - Verbose mode - Better error messages - tests: Comprehensive test suite
This commit is contained in:
Mario Lavoie
179
src/cad_documenter/config.py
Normal file
179
src/cad_documenter/config.py
Normal file
@@ -0,0 +1,179 @@
|
||||
"""Configuration management for CAD-Documenter."""
|
||||
|
||||
import os
|
||||
from pathlib import Path
|
||||
from dataclasses import dataclass, field
|
||||
from typing import Literal
|
||||
|
||||
try:
|
||||
import tomllib
|
||||
except ImportError:
|
||||
import tomli as tomllib
|
||||
|
||||
|
||||
@dataclass
|
||||
class TranscriptionConfig:
|
||||
"""Transcription configuration."""
|
||||
model: str = "base" # tiny, base, small, medium, large
|
||||
language: str | None = None # None = auto-detect
|
||||
|
||||
|
||||
@dataclass
|
||||
class APIConfig:
|
||||
"""API configuration."""
|
||||
provider: Literal["openai", "anthropic"] = "openai"
|
||||
api_key: str | None = None
|
||||
vision_model: str | None = None # None = use provider default
|
||||
text_model: str | None = None
|
||||
|
||||
|
||||
@dataclass
|
||||
class ProcessingConfig:
|
||||
"""Video/audio processing configuration."""
|
||||
whisper_model: str = "base"
|
||||
frame_interval: float = 2.0
|
||||
use_scene_detection: bool = True
|
||||
max_frames: int = 15
|
||||
scene_threshold: float = 0.3
|
||||
|
||||
|
||||
@dataclass
|
||||
class OutputConfig:
|
||||
"""Output configuration."""
|
||||
include_bom: bool = True
|
||||
include_atomizer_hints: bool = True
|
||||
include_raw_transcript: bool = True
|
||||
include_frames: bool = True
|
||||
pdf_template: str = "default"
|
||||
|
||||
|
||||
@dataclass
|
||||
class Config:
|
||||
"""Main configuration."""
|
||||
api: APIConfig = field(default_factory=APIConfig)
|
||||
processing: ProcessingConfig = field(default_factory=ProcessingConfig)
|
||||
output: OutputConfig = field(default_factory=OutputConfig)
|
||||
|
||||
|
||||
def load_config(config_path: Path | None = None) -> Config:
|
||||
"""
|
||||
Load configuration from file and environment variables.
|
||||
|
||||
Priority (highest to lowest):
|
||||
1. Environment variables
|
||||
2. Config file
|
||||
3. Defaults
|
||||
"""
|
||||
config = Config()
|
||||
|
||||
# Try to load config file
|
||||
if config_path is None:
|
||||
# Check common locations
|
||||
locations = [
|
||||
Path.cwd() / "cad-documenter.toml",
|
||||
Path.cwd() / ".cad-documenter.toml",
|
||||
Path.home() / ".cad-documenter.toml",
|
||||
Path.home() / ".config" / "cad-documenter" / "config.toml",
|
||||
]
|
||||
for loc in locations:
|
||||
if loc.exists():
|
||||
config_path = loc
|
||||
break
|
||||
|
||||
if config_path and config_path.exists():
|
||||
with open(config_path, "rb") as f:
|
||||
data = tomllib.load(f)
|
||||
|
||||
# API config
|
||||
if "api" in data:
|
||||
api_data = data["api"]
|
||||
config.api.provider = api_data.get("provider", config.api.provider)
|
||||
config.api.api_key = api_data.get("api_key", config.api.api_key)
|
||||
config.api.vision_model = api_data.get("vision_model", config.api.vision_model)
|
||||
config.api.text_model = api_data.get("text_model", config.api.text_model)
|
||||
|
||||
# Processing config
|
||||
if "processing" in data:
|
||||
proc_data = data["processing"]
|
||||
config.processing.whisper_model = proc_data.get("whisper_model", config.processing.whisper_model)
|
||||
config.processing.frame_interval = proc_data.get("frame_interval", config.processing.frame_interval)
|
||||
config.processing.use_scene_detection = proc_data.get("use_scene_detection", config.processing.use_scene_detection)
|
||||
config.processing.max_frames = proc_data.get("max_frames", config.processing.max_frames)
|
||||
config.processing.scene_threshold = proc_data.get("scene_threshold", config.processing.scene_threshold)
|
||||
|
||||
# Output config
|
||||
if "output" in data:
|
||||
out_data = data["output"]
|
||||
config.output.include_bom = out_data.get("include_bom", config.output.include_bom)
|
||||
config.output.include_atomizer_hints = out_data.get("include_atomizer_hints", config.output.include_atomizer_hints)
|
||||
config.output.include_raw_transcript = out_data.get("include_raw_transcript", config.output.include_raw_transcript)
|
||||
config.output.include_frames = out_data.get("include_frames", config.output.include_frames)
|
||||
config.output.pdf_template = out_data.get("pdf_template", config.output.pdf_template)
|
||||
|
||||
# Override with environment variables
|
||||
if os.environ.get("CAD_DOC_PROVIDER"):
|
||||
config.api.provider = os.environ["CAD_DOC_PROVIDER"]
|
||||
|
||||
if os.environ.get("OPENAI_API_KEY"):
|
||||
if config.api.provider == "openai" and not config.api.api_key:
|
||||
config.api.api_key = os.environ["OPENAI_API_KEY"]
|
||||
|
||||
if os.environ.get("ANTHROPIC_API_KEY"):
|
||||
if config.api.provider == "anthropic" and not config.api.api_key:
|
||||
config.api.api_key = os.environ["ANTHROPIC_API_KEY"]
|
||||
|
||||
if os.environ.get("CAD_DOC_WHISPER_MODEL"):
|
||||
config.processing.whisper_model = os.environ["CAD_DOC_WHISPER_MODEL"]
|
||||
|
||||
return config
|
||||
|
||||
|
||||
def create_default_config(path: Path) -> None:
|
||||
"""Create a default config file."""
|
||||
content = '''# CAD-Documenter Configuration
|
||||
|
||||
[api]
|
||||
# Vision API provider: "openai" or "anthropic"
|
||||
provider = "openai"
|
||||
|
||||
# API key (or set OPENAI_API_KEY / ANTHROPIC_API_KEY environment variable)
|
||||
# api_key = "sk-..."
|
||||
|
||||
# Model overrides (optional - uses provider defaults if not set)
|
||||
# vision_model = "gpt-4o"
|
||||
# text_model = "gpt-4o-mini"
|
||||
|
||||
[processing]
|
||||
# Whisper model for transcription: tiny, base, small, medium, large
|
||||
whisper_model = "base"
|
||||
|
||||
# Seconds between frame extractions (if not using scene detection)
|
||||
frame_interval = 2.0
|
||||
|
||||
# Use scene change detection for smarter frame selection
|
||||
use_scene_detection = true
|
||||
|
||||
# Maximum frames to send to vision API
|
||||
max_frames = 15
|
||||
|
||||
# Scene detection sensitivity (0.0-1.0, lower = more sensitive)
|
||||
scene_threshold = 0.3
|
||||
|
||||
[output]
|
||||
# Include Bill of Materials in documentation
|
||||
include_bom = true
|
||||
|
||||
# Include Atomizer FEA hints
|
||||
include_atomizer_hints = true
|
||||
|
||||
# Include raw transcript at end of documentation
|
||||
include_raw_transcript = true
|
||||
|
||||
# Include extracted frames in output directory
|
||||
include_frames = true
|
||||
|
||||
# PDF template name (for --pdf option)
|
||||
pdf_template = "default"
|
||||
'''
|
||||
path.parent.mkdir(parents=True, exist_ok=True)
|
||||
path.write_text(content)
|
||||
Reference in New Issue
Block a user