362 lines
13 KiB
Python
362 lines
13 KiB
Python
#!/usr/bin/env python3
|
||
# -*- coding: utf-8 -*-
|
||
"""
|
||
Data Modules - 配置文件
|
||
|
||
API 配置通过环境变量读取（支持 .env 文件）：
|
||
- EMBED_BASE_URL, EMBED_MODEL, EMBED_API_KEY
|
||
- RERANK_BASE_URL, RERANK_MODEL, RERANK_API_KEY
|
||
"""
|
||
|
||
import os
|
||
from pathlib import Path
|
||
from dataclasses import dataclass, field
|
||
from typing import Optional
|
||
|
||
from runtime_compat import normalize_windows_path
|
||
|
||
from .context_weights import TEMPLATE_WEIGHTS_DYNAMIC_DEFAULT
|
||
|
||
def _get_user_claude_root() -> Path:
|
||
raw = os.environ.get("NOMA_CLAUDE_HOME") or os.environ.get("CLAUDE_HOME")
|
||
if raw:
|
||
try:
|
||
return normalize_windows_path(raw).expanduser().resolve()
|
||
except Exception:
|
||
return normalize_windows_path(raw).expanduser()
|
||
return (Path.home() / ".claude").resolve()
|
||
|
||
|
||
def _load_dotenv_file(env_path: Path, *, override: bool = False) -> bool:
|
||
if not env_path.exists():
|
||
return False
|
||
try:
|
||
with open(env_path, "r", encoding="utf-8") as f:
|
||
for line in f:
|
||
line = line.strip()
|
||
if line and not line.startswith("#") and "=" in line:
|
||
key, _, value = line.partition("=")
|
||
key = key.strip()
|
||
value = value.strip()
|
||
if not key:
|
||
continue
|
||
# 默认不覆盖已有环境变量（保持“显式 > .env”优先级）
|
||
if override or key not in os.environ:
|
||
os.environ[key] = value
|
||
return True
|
||
except Exception:
|
||
return False
|
||
|
||
|
||
def _load_dotenv():
|
||
"""
|
||
加载 .env 文件（best-effort）。
|
||
|
||
约定：
|
||
- 项目级 `.env`（当前工作目录下）优先；
|
||
- 全局 `.env` 作为兜底：`~/.claude/novelmaster/.env`
|
||
"""
|
||
# 1) 当前目录（常见：用户从项目根目录执行）
|
||
_load_dotenv_file(Path.cwd() / ".env", override=False)
|
||
|
||
# 2) 用户级全局（常见：skills/agents 全局安装，API key 放这里最省心）
|
||
global_env = _get_user_claude_root() / "novelmaster" / ".env"
|
||
_load_dotenv_file(global_env, override=False)
|
||
|
||
|
||
def _load_project_dotenv(project_root: Path) -> None:
|
||
"""
|
||
加载某个项目根目录下的 `.env`（best-effort）。
|
||
优先加载顺序：.noma/config.env > .env
|
||
注意：不覆盖已存在环境变量，避免意外串台。
|
||
"""
|
||
project_root = Path(project_root)
|
||
|
||
# 优先加载 .noma/config.env（新版）
|
||
try:
|
||
_load_dotenv_file(project_root / ".noma" / "config.env", override=False)
|
||
except Exception:
|
||
pass
|
||
|
||
# 兼容旧版 .env
|
||
try:
|
||
_load_dotenv_file(project_root / ".env", override=False)
|
||
except Exception:
|
||
pass
|
||
|
||
_load_dotenv()
|
||
|
||
|
||
def _default_context_template_weights_dynamic() -> dict[str, dict[str, dict[str, float]]]:
|
||
return {
|
||
stage: {
|
||
template: dict(weights)
|
||
for template, weights in templates.items()
|
||
}
|
||
for stage, templates in TEMPLATE_WEIGHTS_DYNAMIC_DEFAULT.items()
|
||
}
|
||
|
||
|
||
@dataclass
|
||
class DataModulesConfig:
|
||
"""数据模块配置"""
|
||
|
||
# ================= 项目路径 =================
|
||
project_root: Path = field(default_factory=lambda: Path.cwd())
|
||
|
||
@property
|
||
def noma_dir(self) -> Path:
|
||
return self.project_root / ".noma"
|
||
|
||
@property
|
||
def state_file(self) -> Path:
|
||
return self.noma_dir / "state.json"
|
||
|
||
@property
|
||
def index_db(self) -> Path:
|
||
return self.noma_dir / "index.db"
|
||
|
||
# v5.1 引入: alias_index_file 已废弃，别名存储在 index.db aliases 表
|
||
|
||
@property
|
||
def chapters_dir(self) -> Path:
|
||
return self.project_root / "正文"
|
||
|
||
@property
|
||
def settings_dir(self) -> Path:
|
||
return self.project_root / "设定集"
|
||
|
||
@property
|
||
def outline_dir(self) -> Path:
|
||
return self.project_root / "大纲"
|
||
|
||
@property
|
||
def wiki_dir(self) -> Path:
|
||
return self.noma_dir / "wiki"
|
||
|
||
# ================= Embedding API 配置 =================
|
||
embed_api_type: str = "openai"
|
||
embed_base_url: str = field(default_factory=lambda: os.getenv("EMBED_BASE_URL", "https://api-inference.modelscope.cn/v1"))
|
||
embed_model: str = field(default_factory=lambda: os.getenv("EMBED_MODEL", "Qwen/Qwen3-Embedding-8B"))
|
||
embed_api_key: str = field(default_factory=lambda: os.getenv("EMBED_API_KEY", ""))
|
||
|
||
@property
|
||
def embed_url(self) -> str:
|
||
return self.embed_base_url
|
||
|
||
# ================= Rerank API 配置 =================
|
||
rerank_api_type: str = "openai"
|
||
rerank_base_url: str = field(default_factory=lambda: os.getenv("RERANK_BASE_URL", "https://api.jina.ai/v1"))
|
||
rerank_model: str = field(default_factory=lambda: os.getenv("RERANK_MODEL", "jina-reranker-v3"))
|
||
rerank_api_key: str = field(default_factory=lambda: os.getenv("RERANK_API_KEY", ""))
|
||
|
||
@property
|
||
def rerank_url(self) -> str:
|
||
return self.rerank_base_url
|
||
|
||
# ================= 并发配置 =================
|
||
embed_concurrency: int = 64
|
||
rerank_concurrency: int = 32
|
||
embed_batch_size: int = 64
|
||
|
||
# ================= 超时配置 =================
|
||
cold_start_timeout: int = 300
|
||
normal_timeout: int = 180
|
||
|
||
# ================= 重试配置 =================
|
||
api_max_retries: int = 3 # 最大重试次数
|
||
api_retry_delay: float = 1.0 # 初始重试延迟（秒），使用指数退避
|
||
|
||
# ================= 检索配置 =================
|
||
vector_top_k: int = 30
|
||
bm25_top_k: int = 20
|
||
rerank_top_n: int = 10
|
||
rrf_k: int = 60
|
||
|
||
vector_full_scan_max_vectors: int = 500
|
||
vector_prefilter_bm25_candidates: int = 200
|
||
vector_prefilter_recent_candidates: int = 200
|
||
|
||
# ================= Graph-RAG 配置 =================
|
||
graph_rag_enabled: bool = False
|
||
graph_rag_expand_hops: int = 1
|
||
graph_rag_max_expanded_entities: int = 30
|
||
graph_rag_candidate_limit: int = 150
|
||
graph_rag_boost_same_entity: float = 0.2
|
||
graph_rag_boost_related_entity: float = 0.1
|
||
graph_rag_boost_recency: float = 0.05
|
||
|
||
relationship_graph_from_index_enabled: bool = True
|
||
|
||
# ================= 实体提取配置 =================
|
||
extraction_confidence_high: float = 0.8
|
||
extraction_confidence_medium: float = 0.5
|
||
|
||
# ================= 列表截断限制 =================
|
||
max_disambiguation_warnings: int = 500
|
||
max_disambiguation_pending: int = 1000
|
||
max_state_changes: int = 2000
|
||
|
||
context_recent_summaries_window: int = 3
|
||
context_recent_meta_window: int = 3
|
||
context_alerts_slice: int = 10
|
||
context_max_appearing_characters: int = 10
|
||
context_max_urgent_foreshadowing: int = 5
|
||
context_story_skeleton_interval: int = 20
|
||
context_story_skeleton_max_samples: int = 5
|
||
context_story_skeleton_snippet_chars: int = 400
|
||
context_extra_section_budget: int = 800
|
||
context_ranker_enabled: bool = True
|
||
context_ranker_recency_weight: float = 0.7
|
||
context_ranker_frequency_weight: float = 0.3
|
||
context_ranker_hook_bonus: float = 0.2
|
||
context_ranker_length_bonus_cap: float = 0.2
|
||
context_ranker_alert_critical_keywords: tuple[str, ...] = (
|
||
"冲突",
|
||
"矛盾",
|
||
"critical",
|
||
"break",
|
||
"违规",
|
||
"断裂",
|
||
)
|
||
context_ranker_debug: bool = False
|
||
context_reader_signal_enabled: bool = True
|
||
context_reader_signal_recent_limit: int = 5
|
||
context_reader_signal_window_chapters: int = 20
|
||
context_reader_signal_review_window: int = 5
|
||
context_reader_signal_include_debt: bool = False
|
||
context_genre_profile_enabled: bool = True
|
||
context_genre_profile_max_refs: int = 8
|
||
context_genre_profile_fallback: str = "shuangwen"
|
||
context_compact_text_enabled: bool = True
|
||
context_compact_min_budget: int = 120
|
||
context_compact_head_ratio: float = 0.65
|
||
context_writing_guidance_enabled: bool = True
|
||
context_writing_guidance_max_items: int = 6
|
||
context_writing_guidance_low_score_threshold: float = 75.0
|
||
context_writing_guidance_hook_diversify: bool = True
|
||
context_methodology_enabled: bool = True
|
||
context_methodology_genre_whitelist: tuple[str, ...] = ("*",)
|
||
context_methodology_label: str = "digital-serial-v1"
|
||
context_writing_checklist_enabled: bool = True
|
||
context_writing_checklist_min_items: int = 3
|
||
context_writing_checklist_max_items: int = 6
|
||
context_writing_checklist_default_weight: float = 1.0
|
||
context_writing_score_persist_enabled: bool = True
|
||
context_writing_score_include_reader_trend: bool = True
|
||
context_writing_score_trend_window: int = 10
|
||
context_rag_assist_enabled: bool = True
|
||
context_rag_assist_top_k: int = 4
|
||
context_rag_assist_min_outline_chars: int = 40
|
||
context_rag_assist_max_query_chars: int = 120
|
||
context_dynamic_budget_enabled: bool = True
|
||
context_dynamic_budget_early_chapter: int = 30
|
||
context_dynamic_budget_late_chapter: int = 120
|
||
context_dynamic_budget_early_core_bonus: float = 0.08
|
||
context_dynamic_budget_early_scene_bonus: float = 0.04
|
||
context_dynamic_budget_late_global_bonus: float = 0.08
|
||
context_dynamic_budget_late_scene_penalty: float = 0.06
|
||
context_template_weights_dynamic: dict[str, dict[str, dict[str, float]]] = field(
|
||
default_factory=_default_context_template_weights_dynamic
|
||
)
|
||
context_genre_profile_support_composite: bool = True
|
||
context_genre_profile_max_genres: int = 2
|
||
context_genre_profile_separators: tuple[str, ...] = (
|
||
"+",
|
||
"/",
|
||
"|",
|
||
",",
|
||
"，",
|
||
"、",
|
||
)
|
||
|
||
export_recent_changes_slice: int = 20
|
||
export_disambiguation_slice: int = 20
|
||
|
||
# ================= 查询默认限制 =================
|
||
query_recent_chapters_limit: int = 10
|
||
query_scenes_by_location_limit: int = 20
|
||
query_entity_appearances_limit: int = 50
|
||
query_recent_appearances_limit: int = 20
|
||
|
||
# ================= 伏笔紧急度 =================
|
||
foreshadowing_urgency_pending_high: int = 100
|
||
foreshadowing_urgency_pending_medium: int = 50
|
||
foreshadowing_urgency_target_proximity: int = 5
|
||
foreshadowing_urgency_score_high: int = 100
|
||
foreshadowing_urgency_score_medium: int = 60
|
||
foreshadowing_urgency_score_target: int = 80
|
||
foreshadowing_urgency_score_low: int = 20
|
||
foreshadowing_urgency_threshold_show: int = 60
|
||
|
||
foreshadowing_tier_weight_core: float = 3.0
|
||
foreshadowing_tier_weight_sub: float = 2.0
|
||
foreshadowing_tier_weight_decor: float = 1.0
|
||
|
||
# ================= 角色活跃度 =================
|
||
character_absence_warning: int = 30
|
||
character_absence_critical: int = 100
|
||
character_candidates_limit: int = 800
|
||
|
||
# ================= Strand Weave 节奏 =================
|
||
strand_quest_max_consecutive: int = 5
|
||
strand_fire_max_gap: int = 10
|
||
strand_constellation_max_gap: int = 15
|
||
|
||
strand_quest_ratio_min: int = 55
|
||
strand_quest_ratio_max: int = 65
|
||
strand_fire_ratio_min: int = 20
|
||
strand_fire_ratio_max: int = 30
|
||
strand_constellation_ratio_min: int = 10
|
||
strand_constellation_ratio_max: int = 20
|
||
|
||
# ================= 爽点节奏 =================
|
||
pacing_segment_size: int = 100
|
||
pacing_words_per_point_excellent: int = 1000
|
||
pacing_words_per_point_good: int = 1500
|
||
pacing_words_per_point_acceptable: int = 2000
|
||
|
||
# ================= RAG 存储 =================
|
||
@property
|
||
def rag_db(self) -> Path:
|
||
return self.noma_dir / "rag.db"
|
||
|
||
@property
|
||
def vector_db(self) -> Path:
|
||
return self.noma_dir / "vectors.db"
|
||
|
||
def ensure_dirs(self):
|
||
self.noma_dir.mkdir(parents=True, exist_ok=True)
|
||
|
||
@classmethod
|
||
def from_project_root(cls, project_root: str | Path) -> "DataModulesConfig":
|
||
root = normalize_windows_path(project_root).expanduser().resolve()
|
||
# 在构造配置前加载项目级 `.env`，以确保 EMBED_*/RERANK_* 等字段可生效
|
||
_load_project_dotenv(root)
|
||
return cls(project_root=root)
|
||
|
||
|
||
_default_config: Optional[DataModulesConfig] = None
|
||
|
||
|
||
def get_config(project_root: Optional[Path] = None) -> DataModulesConfig:
|
||
global _default_config
|
||
if project_root is not None:
|
||
return DataModulesConfig.from_project_root(project_root)
|
||
if _default_config is None:
|
||
# 默认不要盲目以 CWD 作为 project_root（很容易写到错误目录）。
|
||
# 使用统一的 project_locator 自动探测：
|
||
# - 支持 NOMA_PROJECT_ROOT
|
||
# - 支持 `.claude/.noma-current-project` 指针文件
|
||
# - 支持从当前目录/父目录寻找 `.noma/state.json`
|
||
from project_locator import resolve_project_root
|
||
|
||
root = resolve_project_root()
|
||
_default_config = DataModulesConfig.from_project_root(root)
|
||
return _default_config
|
||
|
||
|
||
def set_project_root(project_root: str | Path):
|
||
global _default_config
|
||
_default_config = DataModulesConfig.from_project_root(project_root)
|