mirror of
https://github.com/xr843/Master-skill.git
synced 2026-05-10 13:26:25 +00:00
xianren
115 lines
3.5 KiB
Python
115 lines
3.5 KiB
Python
"""
|
|
Teacher Builder — orchestrates the full teacher skill generation pipeline.
|
|
|
|
Flow:
|
|
1. Collect data via sutra_collector
|
|
2. Analyze via prompt templates (sutra_analyzer + voice_analyzer)
|
|
3. Build via prompt templates (teaching_builder + voice_builder)
|
|
4. Write via skill_writer
|
|
"""
|
|
|
|
import json
|
|
import os
|
|
from typing import Optional
|
|
|
|
from fojin_bridge import FojinBridge, create_bridge
|
|
from sutra_collector import collect_teacher_data, collect_specific_texts
|
|
from skill_writer import create_teacher, DISCLAIMER
|
|
|
|
|
|
PROMPTS_DIR = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), "prompts")
|
|
|
|
|
|
def load_prompt(name: str) -> str:
|
|
"""Load a prompt template by name."""
|
|
path = os.path.join(PROMPTS_DIR, f"{name}.md")
|
|
with open(path, "r", encoding="utf-8") as f:
|
|
return f.read()
|
|
|
|
|
|
def build_analysis_prompt(
|
|
template_name: str,
|
|
teacher_name: str,
|
|
data: dict,
|
|
) -> str:
|
|
"""Build a complete analysis prompt by filling template with collected data."""
|
|
template = load_prompt(template_name)
|
|
|
|
entity_info = ""
|
|
if data.get("entity"):
|
|
e = data["entity"]
|
|
entity_info = f"名称: {e.get('name_zh', teacher_name)}\n"
|
|
entity_info += f"类型: {e.get('entity_type', '未知')}\n"
|
|
if e.get("description"):
|
|
entity_info += f"描述: {e['description']}\n"
|
|
if e.get("properties"):
|
|
for k, v in e["properties"].items():
|
|
entity_info += f"{k}: {v}\n"
|
|
|
|
lineage_info = ""
|
|
for rel in data.get("lineage", []):
|
|
lineage_info += f"- {rel['predicate']}: {rel.get('target_name', '未知')}\n"
|
|
|
|
texts_info = ""
|
|
for t in data.get("texts", [])[:20]:
|
|
texts_info += f"- 《{t.get('title_zh', '未知')}》({t.get('cbeta_id', '')})\n"
|
|
|
|
content_samples = ""
|
|
for sample in data.get("content_samples", []):
|
|
content_samples += f"\n### 《{sample['title']}》\n{sample['content'][:2000]}\n"
|
|
|
|
terms_info = ""
|
|
for term in data.get("terms", [])[:30]:
|
|
terms_info += f"- {term.get('headword', '')}: {term.get('definition', '')[:100]}\n"
|
|
|
|
prompt = template.replace("{teacher_name}", teacher_name)
|
|
prompt = prompt.replace("{entity_info}", entity_info)
|
|
prompt = prompt.replace("{lineage_info}", lineage_info)
|
|
prompt = prompt.replace("{texts_info}", texts_info)
|
|
prompt = prompt.replace("{content_samples}", content_samples)
|
|
prompt = prompt.replace("{terms_info}", terms_info)
|
|
|
|
return prompt
|
|
|
|
|
|
def build_teacher_prompt(
|
|
template_name: str,
|
|
teacher_name: str,
|
|
analysis_result: str,
|
|
) -> str:
|
|
"""Build a generation prompt from analysis results."""
|
|
template = load_prompt(template_name)
|
|
prompt = template.replace("{teacher_name}", teacher_name)
|
|
prompt = prompt.replace("{analysis_result}", analysis_result)
|
|
return prompt
|
|
|
|
|
|
def generate_teacher_skill(
|
|
name: str,
|
|
tradition: str,
|
|
school: str,
|
|
era: str,
|
|
languages: list,
|
|
teaching_content: str,
|
|
voice_content: str,
|
|
output_dir: str = "teachers",
|
|
fojin_entity_id: Optional[str] = None,
|
|
sources: Optional[list] = None,
|
|
) -> str:
|
|
"""Write the final teacher skill to disk."""
|
|
base_dir = os.path.join(
|
|
os.path.dirname(os.path.dirname(os.path.abspath(__file__))), output_dir
|
|
)
|
|
return create_teacher(
|
|
base_dir=base_dir,
|
|
name=name,
|
|
tradition=tradition,
|
|
school=school,
|
|
era=era,
|
|
languages=languages,
|
|
teaching_content=teaching_content,
|
|
voice_content=voice_content,
|
|
fojin_entity_id=fojin_entity_id,
|
|
sources=sources,
|
|
)
|