1207 lines
51 KiB
Python
1207 lines
51 KiB
Python
"""
|
|
OASIS Agent Profile generator
|
|
Convert entities from the Zep graph into the Agent Profile format required by the OASIS simulation platform
|
|
|
|
Optimisations:
|
|
1. Use Zep retrieval to enrich node information
|
|
2. Refine the prompt to produce very detailed personas
|
|
3. Distinguish between individual and abstract-group entities
|
|
"""
|
|
|
|
import json
|
|
import random
|
|
import time
|
|
from typing import Dict, Any, List, Optional
|
|
from dataclasses import dataclass, field
|
|
from datetime import datetime
|
|
|
|
from openai import OpenAI
|
|
try:
|
|
from zep_cloud.client import Zep # noqa: F811
|
|
except ImportError:
|
|
class Zep: # type: ignore[no-redef]
|
|
def __init__(self, *a, **kw): pass
|
|
class graph:
|
|
def search(self, **kw): raise NotImplementedError("zep-cloud not installed; use graphiti_service")
|
|
|
|
from ..config import Config
|
|
from ..utils.logger import get_logger
|
|
from ..utils.locale import get_language_instruction, get_locale, set_locale, t
|
|
from .zep_entity_reader import EntityNode, ZepEntityReader
|
|
|
|
logger = get_logger('mirofish.oasis_profile')
|
|
|
|
|
|
@dataclass
|
|
class OasisAgentProfile:
|
|
"""OASIS Agent Profile data structure"""
|
|
# Common fields
|
|
user_id: int
|
|
user_name: str
|
|
name: str
|
|
bio: str
|
|
persona: str
|
|
|
|
# Optional fields — Reddit style
|
|
karma: int = 1000
|
|
|
|
# Optional fields — Twitter style
|
|
friend_count: int = 100
|
|
follower_count: int = 150
|
|
statuses_count: int = 500
|
|
|
|
# Extra persona information
|
|
age: Optional[int] = None
|
|
gender: Optional[str] = None
|
|
mbti: Optional[str] = None
|
|
country: Optional[str] = None
|
|
profession: Optional[str] = None
|
|
interested_topics: List[str] = field(default_factory=list)
|
|
|
|
# Source entity information
|
|
source_entity_uuid: Optional[str] = None
|
|
source_entity_type: Optional[str] = None
|
|
|
|
created_at: str = field(default_factory=lambda: datetime.now().strftime("%Y-%m-%d"))
|
|
|
|
def to_reddit_format(self) -> Dict[str, Any]:
|
|
"""Convert to the Reddit platform format"""
|
|
profile = {
|
|
"user_id": self.user_id,
|
|
"username": self.user_name, # OASIS expects the field name 'username' (no underscore)
|
|
"name": self.name,
|
|
"bio": self.bio,
|
|
"persona": self.persona,
|
|
"karma": self.karma,
|
|
"created_at": self.created_at,
|
|
}
|
|
|
|
# Add extra persona information (if any)
|
|
if self.age:
|
|
profile["age"] = self.age
|
|
if self.gender:
|
|
profile["gender"] = self.gender
|
|
if self.mbti:
|
|
profile["mbti"] = self.mbti
|
|
if self.country:
|
|
profile["country"] = self.country
|
|
if self.profession:
|
|
profile["profession"] = self.profession
|
|
if self.interested_topics:
|
|
profile["interested_topics"] = self.interested_topics
|
|
|
|
return profile
|
|
|
|
def to_twitter_format(self) -> Dict[str, Any]:
|
|
"""Convert to the Twitter platform format"""
|
|
profile = {
|
|
"user_id": self.user_id,
|
|
"username": self.user_name, # OASIS expects the field name 'username' (no underscore)
|
|
"name": self.name,
|
|
"bio": self.bio,
|
|
"persona": self.persona,
|
|
"friend_count": self.friend_count,
|
|
"follower_count": self.follower_count,
|
|
"statuses_count": self.statuses_count,
|
|
"created_at": self.created_at,
|
|
}
|
|
|
|
# Add extra persona information
|
|
if self.age:
|
|
profile["age"] = self.age
|
|
if self.gender:
|
|
profile["gender"] = self.gender
|
|
if self.mbti:
|
|
profile["mbti"] = self.mbti
|
|
if self.country:
|
|
profile["country"] = self.country
|
|
if self.profession:
|
|
profile["profession"] = self.profession
|
|
if self.interested_topics:
|
|
profile["interested_topics"] = self.interested_topics
|
|
|
|
return profile
|
|
|
|
def to_dict(self) -> Dict[str, Any]:
|
|
"""Convert to the full dict format"""
|
|
return {
|
|
"user_id": self.user_id,
|
|
"user_name": self.user_name,
|
|
"name": self.name,
|
|
"bio": self.bio,
|
|
"persona": self.persona,
|
|
"karma": self.karma,
|
|
"friend_count": self.friend_count,
|
|
"follower_count": self.follower_count,
|
|
"statuses_count": self.statuses_count,
|
|
"age": self.age,
|
|
"gender": self.gender,
|
|
"mbti": self.mbti,
|
|
"country": self.country,
|
|
"profession": self.profession,
|
|
"interested_topics": self.interested_topics,
|
|
"source_entity_uuid": self.source_entity_uuid,
|
|
"source_entity_type": self.source_entity_type,
|
|
"created_at": self.created_at,
|
|
}
|
|
|
|
|
|
class OasisProfileGenerator:
|
|
"""
|
|
OASIS Profile generator
|
|
|
|
Convert entities from the Zep graph into the Agent Profile format required by OASIS simulation.
|
|
|
|
Optimisation features:
|
|
1. Use the Zep graph retrieval function to fetch richer context.
|
|
2. Generate very detailed personas (including basic info, career history, personality traits, social-media behaviour, etc.).
|
|
3. Distinguish between individual and abstract-group entities.
|
|
"""
|
|
|
|
# MBTI type list
|
|
MBTI_TYPES = [
|
|
"INTJ", "INTP", "ENTJ", "ENTP",
|
|
"INFJ", "INFP", "ENFJ", "ENFP",
|
|
"ISTJ", "ISFJ", "ESTJ", "ESFJ",
|
|
"ISTP", "ISFP", "ESTP", "ESFP"
|
|
]
|
|
|
|
# Common country list
|
|
COUNTRIES = [
|
|
"China", "US", "UK", "Japan", "Germany", "France",
|
|
"Canada", "Australia", "Brazil", "India", "South Korea"
|
|
]
|
|
|
|
# Individual-type entities (require a concrete persona)
|
|
INDIVIDUAL_ENTITY_TYPES = [
|
|
"student", "alumni", "professor", "person", "publicfigure",
|
|
"expert", "faculty", "official", "journalist", "activist"
|
|
]
|
|
|
|
# Group/institutional-type entities (require a representative persona)
|
|
GROUP_ENTITY_TYPES = [
|
|
"university", "governmentagency", "organization", "ngo",
|
|
"mediaoutlet", "company", "institution", "group", "community"
|
|
]
|
|
|
|
def __init__(
|
|
self,
|
|
api_key: Optional[str] = None,
|
|
base_url: Optional[str] = None,
|
|
model_name: Optional[str] = None,
|
|
zep_api_key: Optional[str] = None,
|
|
graph_id: Optional[str] = None
|
|
):
|
|
self.api_key = api_key or Config.LLM_API_KEY
|
|
self.base_url = base_url or Config.LLM_BASE_URL
|
|
self.model_name = model_name or Config.LLM_MODEL_NAME
|
|
|
|
if not self.api_key:
|
|
raise ValueError("LLM_API_KEY is not configured")
|
|
|
|
self.client = OpenAI(
|
|
api_key=self.api_key,
|
|
base_url=self.base_url
|
|
)
|
|
|
|
# Graphiti adapter (replaces Zep) for retrieval-enrichment
|
|
self.zep_api_key = zep_api_key # kept for signature compat
|
|
from .graphiti_service import get_graphiti_adapter
|
|
self.zep_client = get_graphiti_adapter()
|
|
self.graph_id = graph_id
|
|
|
|
def generate_profile_from_entity(
|
|
self,
|
|
entity: EntityNode,
|
|
user_id: int,
|
|
use_llm: bool = True
|
|
) -> OasisAgentProfile:
|
|
"""
|
|
Generate an OASIS Agent Profile from a Zep entity.
|
|
|
|
Args:
|
|
entity: Zep entity node.
|
|
user_id: User ID (used by OASIS).
|
|
use_llm: Whether to use LLM to generate a detailed persona.
|
|
|
|
Returns:
|
|
OasisAgentProfile
|
|
"""
|
|
entity_type = entity.get_entity_type() or "Entity"
|
|
|
|
# Basic info
|
|
name = entity.name
|
|
user_name = self._generate_username(name)
|
|
|
|
# Build context information
|
|
context = self._build_entity_context(entity)
|
|
|
|
if use_llm:
|
|
# Use LLM to generate a detailed persona
|
|
profile_data = self._generate_profile_with_llm(
|
|
entity_name=name,
|
|
entity_type=entity_type,
|
|
entity_summary=entity.summary,
|
|
entity_attributes=entity.attributes,
|
|
context=context
|
|
)
|
|
else:
|
|
# Use rule-based generation for a basic persona
|
|
profile_data = self._generate_profile_rule_based(
|
|
entity_name=name,
|
|
entity_type=entity_type,
|
|
entity_summary=entity.summary,
|
|
entity_attributes=entity.attributes
|
|
)
|
|
|
|
return OasisAgentProfile(
|
|
user_id=user_id,
|
|
user_name=user_name,
|
|
name=name,
|
|
bio=profile_data.get("bio", f"{entity_type}: {name}"),
|
|
persona=profile_data.get("persona", entity.summary or f"A {entity_type} named {name}."),
|
|
karma=profile_data.get("karma", random.randint(500, 5000)),
|
|
friend_count=profile_data.get("friend_count", random.randint(50, 500)),
|
|
follower_count=profile_data.get("follower_count", random.randint(100, 1000)),
|
|
statuses_count=profile_data.get("statuses_count", random.randint(100, 2000)),
|
|
age=profile_data.get("age"),
|
|
gender=profile_data.get("gender"),
|
|
mbti=profile_data.get("mbti"),
|
|
country=profile_data.get("country"),
|
|
profession=profile_data.get("profession"),
|
|
interested_topics=profile_data.get("interested_topics", []),
|
|
source_entity_uuid=entity.uuid,
|
|
source_entity_type=entity_type,
|
|
)
|
|
|
|
def _generate_username(self, name: str) -> str:
|
|
"""Generate a username"""
|
|
# Strip special characters and lower-case
|
|
username = name.lower().replace(" ", "_")
|
|
username = ''.join(c for c in username if c.isalnum() or c == '_')
|
|
|
|
# Add a random suffix to avoid duplicates
|
|
suffix = random.randint(100, 999)
|
|
return f"{username}_{suffix}"
|
|
|
|
def _search_zep_for_entity(self, entity: EntityNode) -> Dict[str, Any]:
|
|
"""
|
|
Use Zep's graph hybrid search to fetch rich information related to the entity.
|
|
|
|
Zep has no built-in hybrid-search endpoint, so we search edges and nodes separately
|
|
and merge the results. The two requests run in parallel for efficiency.
|
|
|
|
Args:
|
|
entity: Entity node object.
|
|
|
|
Returns:
|
|
A dictionary containing facts, node_summaries, and context.
|
|
"""
|
|
import concurrent.futures
|
|
|
|
if not self.zep_client:
|
|
return {"facts": [], "node_summaries": [], "context": ""}
|
|
|
|
entity_name = entity.name
|
|
|
|
results = {
|
|
"facts": [],
|
|
"node_summaries": [],
|
|
"context": ""
|
|
}
|
|
|
|
# A graph_id is required to perform a search
|
|
if not self.graph_id:
|
|
logger.debug(f"Skipping Zep retrieval: graph_id not set")
|
|
return results
|
|
|
|
comprehensive_query = t('progress.zepSearchQuery', name=entity_name)
|
|
|
|
def search_edges():
|
|
"""Search edges (facts/relationships) — with retry mechanism."""
|
|
max_retries = 3
|
|
last_exception = None
|
|
delay = 2.0
|
|
|
|
for attempt in range(max_retries):
|
|
try:
|
|
return self.zep_client.search(
|
|
graph_id=self.graph_id,
|
|
query=comprehensive_query,
|
|
limit=30,
|
|
scope="edges",
|
|
)
|
|
except Exception as e:
|
|
last_exception = e
|
|
if attempt < max_retries - 1:
|
|
logger.debug(f"Zep edge search attempt {attempt + 1} failed: {str(e)[:80]}, retrying...")
|
|
time.sleep(delay)
|
|
delay *= 2
|
|
else:
|
|
logger.debug(f"Zep edge search failed after {max_retries} attempts: {e}")
|
|
return None
|
|
|
|
def search_nodes():
|
|
"""Search nodes (entity summaries) — with retry mechanism."""
|
|
max_retries = 3
|
|
last_exception = None
|
|
delay = 2.0
|
|
|
|
for attempt in range(max_retries):
|
|
try:
|
|
return self.zep_client.search(
|
|
graph_id=self.graph_id or "",
|
|
query=comprehensive_query,
|
|
limit=20,
|
|
scope="nodes",
|
|
)
|
|
except Exception as e:
|
|
last_exception = e
|
|
if attempt < max_retries - 1:
|
|
logger.debug(f"Zep node search attempt {attempt + 1} failed: {str(e)[:80]}, retrying...")
|
|
time.sleep(delay)
|
|
delay *= 2
|
|
else:
|
|
logger.debug(f"Zep node search failed after {max_retries} attempts: {e}")
|
|
return None
|
|
|
|
try:
|
|
# Run edges and nodes searches in parallel
|
|
with concurrent.futures.ThreadPoolExecutor(max_workers=2) as executor:
|
|
edge_future = executor.submit(search_edges)
|
|
node_future = executor.submit(search_nodes)
|
|
|
|
# Collect results
|
|
edge_result = edge_future.result(timeout=30)
|
|
node_result = node_future.result(timeout=30)
|
|
|
|
# Process edge-search results
|
|
all_facts = set()
|
|
if edge_result and hasattr(edge_result, 'edges') and edge_result.edges:
|
|
for edge in edge_result.edges:
|
|
if hasattr(edge, 'fact') and edge.fact:
|
|
all_facts.add(edge.fact)
|
|
results["facts"] = list(all_facts)
|
|
|
|
# Process node-search results
|
|
all_summaries = set()
|
|
if node_result and hasattr(node_result, 'nodes') and node_result.nodes:
|
|
for node in node_result.nodes:
|
|
if hasattr(node, 'summary') and node.summary:
|
|
all_summaries.add(node.summary)
|
|
if hasattr(node, 'name') and node.name and node.name != entity_name:
|
|
all_summaries.add(f"Related entity: {node.name}")
|
|
results["node_summaries"] = list(all_summaries)
|
|
|
|
# Build combined context
|
|
context_parts = []
|
|
if results["facts"]:
|
|
context_parts.append("Facts:\n" + "\n".join(f"- {f}" for f in results["facts"][:20]))
|
|
if results["node_summaries"]:
|
|
context_parts.append("Related entities:\n" + "\n".join(f"- {s}" for s in results["node_summaries"][:10]))
|
|
results["context"] = "\n\n".join(context_parts)
|
|
|
|
logger.info(f"Zep hybrid retrieval done: {entity_name}, fetched {len(results['facts'])} facts and {len(results['node_summaries'])} related nodes")
|
|
|
|
except concurrent.futures.TimeoutError:
|
|
logger.warning(f"Zep retrieval timed out ({entity_name})")
|
|
except Exception as e:
|
|
logger.warning(f"Zep retrieval failed ({entity_name}): {e}")
|
|
|
|
return results
|
|
|
|
def _build_entity_context(self, entity: EntityNode) -> str:
|
|
"""
|
|
Build the full context for an entity.
|
|
|
|
Includes:
|
|
1. The entity's own edges (facts).
|
|
2. Detailed information about related nodes.
|
|
3. Rich information from Zep hybrid retrieval.
|
|
"""
|
|
context_parts = []
|
|
|
|
# 1. Add entity attribute information
|
|
if entity.attributes:
|
|
attrs = []
|
|
for key, value in entity.attributes.items():
|
|
if value and str(value).strip():
|
|
attrs.append(f"- {key}: {value}")
|
|
if attrs:
|
|
context_parts.append("### Entity attributes\n" + "\n".join(attrs))
|
|
|
|
# 2. Add related edge information (facts/relationships)
|
|
existing_facts = set()
|
|
if entity.related_edges:
|
|
relationships = []
|
|
for edge in entity.related_edges: # no count limit
|
|
fact = edge.get("fact", "")
|
|
edge_name = edge.get("edge_name", "")
|
|
direction = edge.get("direction", "")
|
|
|
|
if fact:
|
|
relationships.append(f"- {fact}")
|
|
existing_facts.add(fact)
|
|
elif edge_name:
|
|
if direction == "outgoing":
|
|
relationships.append(f"- {entity.name} --[{edge_name}]--> (related entity)")
|
|
else:
|
|
relationships.append(f"- (related entity) --[{edge_name}]--> {entity.name}")
|
|
|
|
if relationships:
|
|
context_parts.append("### Related facts and relationships\n" + "\n".join(relationships))
|
|
|
|
# 3. Add detailed information about related nodes
|
|
if entity.related_nodes:
|
|
related_info = []
|
|
for node in entity.related_nodes: # no count limit
|
|
node_name = node.get("name", "")
|
|
node_labels = node.get("labels", [])
|
|
node_summary = node.get("summary", "")
|
|
|
|
# Filter out default labels
|
|
custom_labels = [l for l in node_labels if l not in ["Entity", "Node"]]
|
|
label_str = f" ({', '.join(custom_labels)})" if custom_labels else ""
|
|
|
|
if node_summary:
|
|
related_info.append(f"- **{node_name}**{label_str}: {node_summary}")
|
|
else:
|
|
related_info.append(f"- **{node_name}**{label_str}")
|
|
|
|
if related_info:
|
|
context_parts.append("### Related entity information\n" + "\n".join(related_info))
|
|
|
|
# 4. Use Zep hybrid retrieval to fetch richer information
|
|
zep_results = self._search_zep_for_entity(entity)
|
|
|
|
if zep_results.get("facts"):
|
|
# Deduplicate: exclude facts already present
|
|
new_facts = [f for f in zep_results["facts"] if f not in existing_facts]
|
|
if new_facts:
|
|
context_parts.append("### Facts retrieved from Zep\n" + "\n".join(f"- {f}" for f in new_facts[:15]))
|
|
|
|
if zep_results.get("node_summaries"):
|
|
context_parts.append("### Related nodes retrieved from Zep\n" + "\n".join(f"- {s}" for s in zep_results["node_summaries"][:10]))
|
|
|
|
return "\n\n".join(context_parts)
|
|
|
|
def _is_individual_entity(self, entity_type: str) -> bool:
|
|
"""Determine whether the entity is an individual-type entity."""
|
|
return entity_type.lower() in self.INDIVIDUAL_ENTITY_TYPES
|
|
|
|
def _is_group_entity(self, entity_type: str) -> bool:
|
|
"""Determine whether the entity is a group/institutional-type entity."""
|
|
return entity_type.lower() in self.GROUP_ENTITY_TYPES
|
|
|
|
def _generate_profile_with_llm(
|
|
self,
|
|
entity_name: str,
|
|
entity_type: str,
|
|
entity_summary: str,
|
|
entity_attributes: Dict[str, Any],
|
|
context: str
|
|
) -> Dict[str, Any]:
|
|
"""
|
|
Use LLM to generate a very detailed persona.
|
|
|
|
Differs by entity type:
|
|
- Individual entity: produce a concrete character profile.
|
|
- Group/institutional entity: produce a representative account profile.
|
|
"""
|
|
|
|
is_individual = self._is_individual_entity(entity_type)
|
|
|
|
if is_individual:
|
|
prompt = self._build_individual_persona_prompt(
|
|
entity_name, entity_type, entity_summary, entity_attributes, context
|
|
)
|
|
else:
|
|
prompt = self._build_group_persona_prompt(
|
|
entity_name, entity_type, entity_summary, entity_attributes, context
|
|
)
|
|
|
|
# Try multiple times until success or max retries reached
|
|
max_attempts = 3
|
|
last_error = None
|
|
|
|
for attempt in range(max_attempts):
|
|
try:
|
|
response = self.client.chat.completions.create(
|
|
model=self.model_name,
|
|
messages=[
|
|
{"role": "system", "content": self._get_system_prompt(is_individual)},
|
|
{"role": "user", "content": prompt}
|
|
],
|
|
response_format={"type": "json_object"},
|
|
temperature=0.7 - (attempt * 0.1) # Lower temperature on each retry
|
|
# Do not set max_tokens; let the LLM decide.
|
|
)
|
|
|
|
content = response.choices[0].message.content
|
|
|
|
# Check whether the output was truncated (finish_reason is not 'stop')
|
|
finish_reason = response.choices[0].finish_reason
|
|
if finish_reason == 'length':
|
|
logger.warning(f"LLM output was truncated (attempt {attempt+1}), attempting to repair...")
|
|
content = self._fix_truncated_json(content)
|
|
|
|
# Try to parse JSON
|
|
try:
|
|
result = json.loads(content)
|
|
|
|
# Validate required fields
|
|
if "bio" not in result or not result["bio"]:
|
|
result["bio"] = entity_summary[:200] if entity_summary else f"{entity_type}: {entity_name}"
|
|
if "persona" not in result or not result["persona"]:
|
|
result["persona"] = entity_summary or f"{entity_name} is a {entity_type}."
|
|
|
|
return result
|
|
|
|
except json.JSONDecodeError as je:
|
|
logger.warning(f"JSON parse failed (attempt {attempt+1}): {str(je)[:80]}")
|
|
|
|
# Try to repair the JSON
|
|
result = self._try_fix_json(content, entity_name, entity_type, entity_summary)
|
|
if result.get("_fixed"):
|
|
del result["_fixed"]
|
|
return result
|
|
|
|
last_error = je
|
|
|
|
except Exception as e:
|
|
logger.warning(f"LLM call failed (attempt {attempt+1}): {str(e)[:80]}")
|
|
last_error = e
|
|
import time
|
|
time.sleep(1 * (attempt + 1)) # Exponential backoff
|
|
|
|
logger.warning(f"LLM persona generation failed (after {max_attempts} attempts): {last_error}, falling back to rule-based")
|
|
return self._generate_profile_rule_based(
|
|
entity_name, entity_type, entity_summary, entity_attributes
|
|
)
|
|
|
|
def _fix_truncated_json(self, content: str) -> str:
|
|
"""Repair truncated JSON (output cut off by max_tokens limit)."""
|
|
import re
|
|
|
|
# If the JSON was truncated, try to close it
|
|
content = content.strip()
|
|
|
|
# Count unclosed brackets
|
|
open_braces = content.count('{') - content.count('}')
|
|
open_brackets = content.count('[') - content.count(']')
|
|
|
|
# Check whether a string is unclosed
|
|
# Simple check: if the last character is not a quote, comma, or closing bracket, the string is likely truncated
|
|
if content and content[-1] not in '",}]':
|
|
# Try to close the string
|
|
content += '"'
|
|
|
|
# Close brackets
|
|
content += ']' * open_brackets
|
|
content += '}' * open_braces
|
|
|
|
return content
|
|
|
|
def _try_fix_json(self, content: str, entity_name: str, entity_type: str, entity_summary: str = "") -> Dict[str, Any]:
|
|
"""Try to repair broken JSON."""
|
|
import re
|
|
|
|
# 1. First try to handle the truncated case
|
|
content = self._fix_truncated_json(content)
|
|
|
|
# 2. Try to extract the JSON portion
|
|
json_match = re.search(r'\{[\s\S]*\}', content)
|
|
if json_match:
|
|
json_str = json_match.group()
|
|
|
|
# 3. Handle embedded newlines in strings
|
|
# Find all string values and replace embedded newlines
|
|
def fix_string_newlines(match):
|
|
s = match.group(0)
|
|
# Replace actual newlines inside strings with spaces
|
|
s = s.replace('\n', ' ').replace('\r', ' ')
|
|
# Collapse extra whitespace
|
|
s = re.sub(r'\s+', ' ', s)
|
|
return s
|
|
|
|
# Match JSON string values
|
|
json_str = re.sub(r'"[^"\\]*(?:\\.[^"\\]*)*"', fix_string_newlines, json_str)
|
|
|
|
# 4. Try to parse
|
|
try:
|
|
result = json.loads(json_str)
|
|
result["_fixed"] = True
|
|
return result
|
|
except json.JSONDecodeError as e:
|
|
# 5. If it still fails, try more aggressive repair
|
|
try:
|
|
# Remove all control characters
|
|
json_str = re.sub(r'[\x00-\x1f\x7f-\x9f]', ' ', json_str)
|
|
# Collapse consecutive whitespace
|
|
json_str = re.sub(r'\s+', ' ', json_str)
|
|
result = json.loads(json_str)
|
|
result["_fixed"] = True
|
|
return result
|
|
except:
|
|
pass
|
|
|
|
# 6. Try to extract partial information from the content
|
|
bio_match = re.search(r'"bio"\s*:\s*"([^"]*)"', content)
|
|
persona_match = re.search(r'"persona"\s*:\s*"([^"]*)', content) # possibly truncated
|
|
|
|
bio = bio_match.group(1) if bio_match else (entity_summary[:200] if entity_summary else f"{entity_type}: {entity_name}")
|
|
persona = persona_match.group(1) if persona_match else (entity_summary or f"{entity_name} is a {entity_type}.")
|
|
|
|
# If we extracted any meaningful content, mark as repaired
|
|
if bio_match or persona_match:
|
|
logger.info(f"Extracted partial information from broken JSON")
|
|
return {
|
|
"bio": bio,
|
|
"persona": persona,
|
|
"_fixed": True
|
|
}
|
|
|
|
# 7. Total failure: return a basic structure
|
|
logger.warning(f"JSON repair failed, returning basic structure")
|
|
return {
|
|
"bio": entity_summary[:200] if entity_summary else f"{entity_type}: {entity_name}",
|
|
"persona": entity_summary or f"{entity_name} is a {entity_type}."
|
|
}
|
|
|
|
def _get_system_prompt(self, is_individual: bool) -> str:
|
|
"""Get the system prompt."""
|
|
base_prompt = "You are an expert at generating social media user personas. Generate detailed, realistic personas for public opinion simulation, faithfully reflecting any existing real-world context. You must return a valid JSON format, and all string values must not contain unescaped newlines."
|
|
return f"{base_prompt}\n\n{get_language_instruction()}"
|
|
|
|
def _build_individual_persona_prompt(
|
|
self,
|
|
entity_name: str,
|
|
entity_type: str,
|
|
entity_summary: str,
|
|
entity_attributes: Dict[str, Any],
|
|
context: str
|
|
) -> str:
|
|
"""Build the detailed persona prompt for an individual entity."""
|
|
|
|
attrs_str = json.dumps(entity_attributes, ensure_ascii=False) if entity_attributes else "None"
|
|
context_str = context[:3000] if context else "No additional context"
|
|
|
|
return f"""Generate a detailed social media user persona for the entity, faithfully reflecting any existing real-world context.
|
|
|
|
Entity Name: {entity_name}
|
|
Entity Type: {entity_type}
|
|
Entity Summary: {entity_summary}
|
|
Entity Attributes: {attrs_str}
|
|
|
|
Context Information:
|
|
{context_str}
|
|
|
|
Please generate JSON containing the following fields:
|
|
|
|
1. bio: Social media bio, 200 words
|
|
2. persona: Detailed persona description (2000 words of plain text), must include:
|
|
- Basic information (age, profession, education background, location)
|
|
- Background (important experiences, connection to events, social relationships)
|
|
- Personality traits (MBTI type, core personality, emotional expression style)
|
|
- Social media behavior (posting frequency, content preferences, interaction style, language characteristics)
|
|
- Stance and views (attitude toward topics, content that may anger/move them)
|
|
- Unique features (catchphrases, special experiences, personal hobbies)
|
|
- Personal memory (an important part of the persona; describe the individual's connection to the event and any prior actions and reactions in the event)
|
|
3. age: Age number (must be an integer)
|
|
4. gender: Gender, must be in English: "male" or "female"
|
|
5. mbti: MBTI type (e.g., INTJ, ENFP, etc.)
|
|
6. country: Country (use the local-language country name, e.g., "China")
|
|
7. profession: Profession
|
|
8. interested_topics: Array of topics of interest
|
|
|
|
Important:
|
|
- All field values must be strings or numbers; do not use newline characters
|
|
- persona must be a coherent paragraph of text
|
|
- {get_language_instruction()} (the gender field must use English male/female)
|
|
- Content must be consistent with entity information
|
|
- age must be a valid integer, gender must be "male" or "female"
|
|
"""
|
|
|
|
def _build_group_persona_prompt(
|
|
self,
|
|
entity_name: str,
|
|
entity_type: str,
|
|
entity_summary: str,
|
|
entity_attributes: Dict[str, Any],
|
|
context: str
|
|
) -> str:
|
|
"""Build the detailed persona prompt for a group/institutional entity."""
|
|
|
|
attrs_str = json.dumps(entity_attributes, ensure_ascii=False) if entity_attributes else "None"
|
|
context_str = context[:3000] if context else "No additional context"
|
|
|
|
return f"""Generate a detailed social media account profile for the institution/group entity, faithfully reflecting any existing real-world context.
|
|
|
|
Entity Name: {entity_name}
|
|
Entity Type: {entity_type}
|
|
Entity Summary: {entity_summary}
|
|
Entity Attributes: {attrs_str}
|
|
|
|
Context Information:
|
|
{context_str}
|
|
|
|
Please generate JSON containing the following fields:
|
|
|
|
1. bio: Official account bio, 200 words, professional and appropriate
|
|
2. persona: Detailed account profile description (2000 words of plain text), must include:
|
|
- Institution basic information (official name, institution nature, founding background, main functions)
|
|
- Account positioning (account type, target audience, core functions)
|
|
- Speaking style (language characteristics, common expressions, taboo topics)
|
|
- Content publishing characteristics (content type, publishing frequency, active time periods)
|
|
- Stance and attitude (official stance on core topics, approach to handling controversy)
|
|
- Special notes (representative group profile, operating habits)
|
|
- Institutional memory (an important part of the institutional persona; describe the institution's connection to the event and any prior actions and reactions in the event)
|
|
3. age: Always 30 (virtual age of the institutional account)
|
|
4. gender: Always "other" (institutional accounts use "other" to indicate non-individual)
|
|
5. mbti: MBTI type, used to describe the account style, e.g., ISTJ for rigorous and conservative
|
|
6. country: Country (use the local-language country name, e.g., "China")
|
|
7. profession: Description of institutional function
|
|
8. interested_topics: Array of focus areas
|
|
|
|
Important:
|
|
- All field values must be strings or numbers; null values are not allowed
|
|
- persona must be a coherent paragraph of text; do not use newline characters
|
|
- {get_language_instruction()} (the gender field must use English "other")
|
|
- age must be the integer 30, gender must be the string "other"
|
|
- Institutional account statements must be consistent with its identity positioning"""
|
|
|
|
def _generate_profile_rule_based(
|
|
self,
|
|
entity_name: str,
|
|
entity_type: str,
|
|
entity_summary: str,
|
|
entity_attributes: Dict[str, Any]
|
|
) -> Dict[str, Any]:
|
|
"""Generate a basic persona using rules."""
|
|
|
|
# Produce different personas depending on the entity type
|
|
entity_type_lower = entity_type.lower()
|
|
|
|
if entity_type_lower in ["student", "alumni"]:
|
|
return {
|
|
"bio": f"{entity_type} with interests in academics and social issues.",
|
|
"persona": f"{entity_name} is a {entity_type.lower()} who is actively engaged in academic and social discussions. They enjoy sharing perspectives and connecting with peers.",
|
|
"age": random.randint(18, 30),
|
|
"gender": random.choice(["male", "female"]),
|
|
"mbti": random.choice(self.MBTI_TYPES),
|
|
"country": random.choice(self.COUNTRIES),
|
|
"profession": "Student",
|
|
"interested_topics": ["Education", "Social Issues", "Technology"],
|
|
}
|
|
|
|
elif entity_type_lower in ["publicfigure", "expert", "faculty"]:
|
|
return {
|
|
"bio": f"Expert and thought leader in their field.",
|
|
"persona": f"{entity_name} is a recognized {entity_type.lower()} who shares insights and opinions on important matters. They are known for their expertise and influence in public discourse.",
|
|
"age": random.randint(35, 60),
|
|
"gender": random.choice(["male", "female"]),
|
|
"mbti": random.choice(["ENTJ", "INTJ", "ENTP", "INTP"]),
|
|
"country": random.choice(self.COUNTRIES),
|
|
"profession": entity_attributes.get("occupation", "Expert"),
|
|
"interested_topics": ["Politics", "Economics", "Culture & Society"],
|
|
}
|
|
|
|
elif entity_type_lower in ["mediaoutlet", "socialmediaplatform"]:
|
|
return {
|
|
"bio": f"Official account for {entity_name}. News and updates.",
|
|
"persona": f"{entity_name} is a media entity that reports news and facilitates public discourse. The account shares timely updates and engages with the audience on current events.",
|
|
"age": 30, # Virtual age for institutional accounts
|
|
"gender": "other", # Institutions use 'other'
|
|
"mbti": "ISTJ", # Institutional style: rigorous and conservative
|
|
"country": "China",
|
|
"profession": "Media",
|
|
"interested_topics": ["General News", "Current Events", "Public Affairs"],
|
|
}
|
|
|
|
elif entity_type_lower in ["university", "governmentagency", "ngo", "organization"]:
|
|
return {
|
|
"bio": f"Official account of {entity_name}.",
|
|
"persona": f"{entity_name} is an institutional entity that communicates official positions, announcements, and engages with stakeholders on relevant matters.",
|
|
"age": 30, # Virtual age for institutional accounts
|
|
"gender": "other", # Institutions use 'other'
|
|
"mbti": "ISTJ", # Institutional style: rigorous and conservative
|
|
"country": "China",
|
|
"profession": entity_type,
|
|
"interested_topics": ["Public Policy", "Community", "Official Announcements"],
|
|
}
|
|
|
|
else:
|
|
# Default persona
|
|
return {
|
|
"bio": entity_summary[:150] if entity_summary else f"{entity_type}: {entity_name}",
|
|
"persona": entity_summary or f"{entity_name} is a {entity_type.lower()} participating in social discussions.",
|
|
"age": random.randint(25, 50),
|
|
"gender": random.choice(["male", "female"]),
|
|
"mbti": random.choice(self.MBTI_TYPES),
|
|
"country": random.choice(self.COUNTRIES),
|
|
"profession": entity_type,
|
|
"interested_topics": ["General", "Social Issues"],
|
|
}
|
|
|
|
def set_graph_id(self, graph_id: str):
|
|
"""Set the graph ID used for Zep retrieval."""
|
|
self.graph_id = graph_id
|
|
|
|
def generate_profiles_from_entities(
|
|
self,
|
|
entities: List[EntityNode],
|
|
use_llm: bool = True,
|
|
progress_callback: Optional[callable] = None,
|
|
graph_id: Optional[str] = None,
|
|
parallel_count: int = 5,
|
|
realtime_output_path: Optional[str] = None,
|
|
output_platform: str = "reddit"
|
|
) -> List[OasisAgentProfile]:
|
|
"""
|
|
Batch-generate Agent Profiles from entities (supports parallel generation).
|
|
|
|
Args:
|
|
entities: List of entities.
|
|
use_llm: Whether to use LLM to generate detailed personas.
|
|
progress_callback: Progress callback function (current, total, message).
|
|
graph_id: Graph ID, used by Zep retrieval to fetch richer context.
|
|
parallel_count: Number of parallel generations (default 5).
|
|
realtime_output_path: Real-time write file path. If provided, every newly generated
|
|
profile is written immediately.
|
|
output_platform: Output platform format ("reddit" or "twitter").
|
|
|
|
Returns:
|
|
List of Agent Profiles
|
|
"""
|
|
import concurrent.futures
|
|
from threading import Lock
|
|
|
|
# Set graph_id for Zep retrieval
|
|
if graph_id:
|
|
self.graph_id = graph_id
|
|
|
|
total = len(entities)
|
|
profiles = [None] * total # pre-allocate list to preserve order
|
|
completed_count = [0] # use a list so the closure can mutate it
|
|
lock = Lock()
|
|
|
|
# Helper: write generated profiles to disk in real time
|
|
def save_profiles_realtime():
|
|
"""Save the profiles generated so far to disk in real time."""
|
|
if not realtime_output_path:
|
|
return
|
|
|
|
with lock:
|
|
# Filter out profiles that are already generated
|
|
existing_profiles = [p for p in profiles if p is not None]
|
|
if not existing_profiles:
|
|
return
|
|
|
|
try:
|
|
if output_platform == "reddit":
|
|
# Reddit JSON format
|
|
profiles_data = [p.to_reddit_format() for p in existing_profiles]
|
|
with open(realtime_output_path, 'w', encoding='utf-8') as f:
|
|
json.dump(profiles_data, f, ensure_ascii=False, indent=2)
|
|
else:
|
|
# Twitter CSV format
|
|
import csv
|
|
profiles_data = [p.to_twitter_format() for p in existing_profiles]
|
|
if profiles_data:
|
|
fieldnames = list(profiles_data[0].keys())
|
|
with open(realtime_output_path, 'w', encoding='utf-8', newline='') as f:
|
|
writer = csv.DictWriter(f, fieldnames=fieldnames)
|
|
writer.writeheader()
|
|
writer.writerows(profiles_data)
|
|
except Exception as e:
|
|
logger.warning(f"Real-time save of profiles failed: {e}")
|
|
|
|
# Capture locale before spawning thread pool workers
|
|
current_locale = get_locale()
|
|
|
|
def generate_single_profile(idx: int, entity: EntityNode) -> tuple:
|
|
"""Worker function that generates a single profile."""
|
|
set_locale(current_locale)
|
|
entity_type = entity.get_entity_type() or "Entity"
|
|
|
|
try:
|
|
profile = self.generate_profile_from_entity(
|
|
entity=entity,
|
|
user_id=idx,
|
|
use_llm=use_llm
|
|
)
|
|
|
|
# Stream the generated persona to console and log in real time
|
|
self._print_generated_profile(entity.name, entity_type, profile)
|
|
|
|
return idx, profile, None
|
|
|
|
except Exception as e:
|
|
logger.error(f"Failed to generate persona for entity {entity.name}: {str(e)}")
|
|
# Create a basic profile as a fallback
|
|
fallback_profile = OasisAgentProfile(
|
|
user_id=idx,
|
|
user_name=self._generate_username(entity.name),
|
|
name=entity.name,
|
|
bio=f"{entity_type}: {entity.name}",
|
|
persona=entity.summary or f"A participant in social discussions.",
|
|
source_entity_uuid=entity.uuid,
|
|
source_entity_type=entity_type,
|
|
)
|
|
return idx, fallback_profile, str(e)
|
|
|
|
logger.info(f"Starting parallel generation of {total} Agent personas (parallel count: {parallel_count})...")
|
|
print(f"\n{'='*60}")
|
|
print(f"Starting Agent persona generation - {total} entities total, parallel count: {parallel_count}")
|
|
print(f"{'='*60}\n")
|
|
|
|
# Run with a thread pool
|
|
with concurrent.futures.ThreadPoolExecutor(max_workers=parallel_count) as executor:
|
|
# Submit all tasks
|
|
future_to_entity = {
|
|
executor.submit(generate_single_profile, idx, entity): (idx, entity)
|
|
for idx, entity in enumerate(entities)
|
|
}
|
|
|
|
# Collect results
|
|
for future in concurrent.futures.as_completed(future_to_entity):
|
|
idx, entity = future_to_entity[future]
|
|
entity_type = entity.get_entity_type() or "Entity"
|
|
|
|
try:
|
|
result_idx, profile, error = future.result()
|
|
profiles[result_idx] = profile
|
|
|
|
with lock:
|
|
completed_count[0] += 1
|
|
current = completed_count[0]
|
|
|
|
# Write to file in real time
|
|
save_profiles_realtime()
|
|
|
|
if progress_callback:
|
|
progress_callback(
|
|
current,
|
|
total,
|
|
f"Done {current}/{total}: {entity.name} ({entity_type})"
|
|
)
|
|
|
|
if error:
|
|
logger.warning(f"[{current}/{total}] {entity.name} used fallback persona: {error}")
|
|
else:
|
|
logger.info(f"[{current}/{total}] Successfully generated persona: {entity.name} ({entity_type})")
|
|
|
|
except Exception as e:
|
|
logger.error(f"Exception while processing entity {entity.name}: {str(e)}")
|
|
with lock:
|
|
completed_count[0] += 1
|
|
profiles[idx] = OasisAgentProfile(
|
|
user_id=idx,
|
|
user_name=self._generate_username(entity.name),
|
|
name=entity.name,
|
|
bio=f"{entity_type}: {entity.name}",
|
|
persona=entity.summary or "A participant in social discussions.",
|
|
source_entity_uuid=entity.uuid,
|
|
source_entity_type=entity_type,
|
|
)
|
|
# Write to file in real time (even for the fallback persona)
|
|
save_profiles_realtime()
|
|
|
|
print(f"\n{'='*60}")
|
|
print(f"Persona generation complete! Generated {len([p for p in profiles if p])} Agents in total")
|
|
print(f"{'='*60}\n")
|
|
|
|
return profiles
|
|
|
|
def _print_generated_profile(self, entity_name: str, entity_type: str, profile: OasisAgentProfile):
|
|
"""Stream the generated persona to the console in real time (full content, no truncation)."""
|
|
separator = "-" * 70
|
|
|
|
# Build full output content (no truncation)
|
|
topics_str = ', '.join(profile.interested_topics) if profile.interested_topics else 'None'
|
|
|
|
output_lines = [
|
|
f"\n{separator}",
|
|
t('progress.profileGenerated', name=entity_name, type=entity_type),
|
|
f"{separator}",
|
|
f"Username: {profile.user_name}",
|
|
f"",
|
|
f"[Bio]",
|
|
f"{profile.bio}",
|
|
f"",
|
|
f"[Detailed Persona]",
|
|
f"{profile.persona}",
|
|
f"",
|
|
f"[Basic Attributes]",
|
|
f"Age: {profile.age} | Gender: {profile.gender} | MBTI: {profile.mbti}",
|
|
f"Profession: {profile.profession} | Country: {profile.country}",
|
|
f"Interested topics: {topics_str}",
|
|
separator
|
|
]
|
|
|
|
output = "\n".join(output_lines)
|
|
|
|
# Only output to console (avoid duplication; the logger no longer prints the full content)
|
|
print(output)
|
|
|
|
def save_profiles(
|
|
self,
|
|
profiles: List[OasisAgentProfile],
|
|
file_path: str,
|
|
platform: str = "reddit"
|
|
):
|
|
"""
|
|
Save profiles to a file (uses the correct format for the platform).
|
|
|
|
OASIS platform format requirements:
|
|
- Twitter: CSV format
|
|
- Reddit: JSON format
|
|
|
|
Args:
|
|
profiles: List of profiles.
|
|
file_path: File path.
|
|
platform: Platform type ("reddit" or "twitter").
|
|
"""
|
|
if platform == "twitter":
|
|
self._save_twitter_csv(profiles, file_path)
|
|
else:
|
|
self._save_reddit_json(profiles, file_path)
|
|
|
|
def _save_twitter_csv(self, profiles: List[OasisAgentProfile], file_path: str):
|
|
"""
|
|
Save Twitter profiles as CSV (per OASIS official requirements).
|
|
|
|
CSV fields required by OASIS Twitter:
|
|
- user_id: User ID (sequential from 0 based on CSV order).
|
|
- name: User's real name.
|
|
- username: Username in the system.
|
|
- user_char: Detailed persona description (injected into the LLM system prompt to
|
|
guide the agent's behaviour).
|
|
- description: Short public bio (shown on the user profile page).
|
|
|
|
user_char vs description:
|
|
- user_char: Internal — used in the LLM system prompt, governs how the agent thinks
|
|
and acts.
|
|
- description: External — the public bio visible to other users.
|
|
"""
|
|
import csv
|
|
|
|
# Ensure the file extension is .csv
|
|
if not file_path.endswith('.csv'):
|
|
file_path = file_path.replace('.json', '.csv')
|
|
|
|
with open(file_path, 'w', newline='', encoding='utf-8') as f:
|
|
writer = csv.writer(f)
|
|
|
|
# Write the OASIS-required header
|
|
headers = ['user_id', 'name', 'username', 'user_char', 'description']
|
|
writer.writerow(headers)
|
|
|
|
# Write data rows
|
|
for idx, profile in enumerate(profiles):
|
|
# user_char: full persona (bio + persona), used in the LLM system prompt
|
|
user_char = profile.bio
|
|
if profile.persona and profile.persona != profile.bio:
|
|
user_char = f"{profile.bio} {profile.persona}"
|
|
# Process newlines (replace with spaces in CSV)
|
|
user_char = user_char.replace('\n', ' ').replace('\r', ' ')
|
|
|
|
# description: short bio for external display
|
|
description = profile.bio.replace('\n', ' ').replace('\r', ' ')
|
|
|
|
row = [
|
|
idx, # user_id: sequential ID starting from 0
|
|
profile.name, # name: real name
|
|
profile.user_name, # username: username
|
|
user_char, # user_char: full persona (internal LLM use)
|
|
description # description: short bio (external display)
|
|
]
|
|
writer.writerow(row)
|
|
|
|
logger.info(f"Saved {len(profiles)} Twitter profiles to {file_path} (OASIS CSV format)")
|
|
|
|
def _normalize_gender(self, gender: Optional[str]) -> str:
|
|
"""
|
|
Normalize the gender field to the English format OASIS requires.
|
|
|
|
OASIS requires: male, female, other.
|
|
"""
|
|
if not gender:
|
|
return "other"
|
|
|
|
gender_lower = gender.lower().strip()
|
|
|
|
# Legacy Chinese-value mapping translated to English keys for backward
|
|
# compatibility with previously generated data; current generation is English-only.
|
|
# Original mapping: "male"→male, "female"→female, "institution"→other, "other"→other.
|
|
gender_map = {
|
|
"male": "male",
|
|
"female": "female",
|
|
"other": "other",
|
|
"institution": "other",
|
|
}
|
|
|
|
return gender_map.get(gender_lower, "other")
|
|
|
|
def _save_reddit_json(self, profiles: List[OasisAgentProfile], file_path: str):
|
|
"""
|
|
Save Reddit profiles as JSON.
|
|
|
|
Uses the same format as to_reddit_format() so OASIS can read it correctly.
|
|
The user_id field is required — it's the key that OASIS agent_graph.get_agent()
|
|
uses to match.
|
|
|
|
Required fields:
|
|
- user_id: User ID (integer, used to match poster_agent_id in initial_posts).
|
|
- username: Username.
|
|
- name: Display name.
|
|
- bio: Bio.
|
|
- persona: Detailed persona.
|
|
- age: Age (integer).
|
|
- gender: "male", "female", or "other".
|
|
- mbti: MBTI type.
|
|
- country: Country.
|
|
"""
|
|
data = []
|
|
for idx, profile in enumerate(profiles):
|
|
# Use the same format as to_reddit_format()
|
|
item = {
|
|
"user_id": profile.user_id if profile.user_id is not None else idx, # Key: must include user_id
|
|
"username": profile.user_name,
|
|
"name": profile.name,
|
|
"bio": profile.bio[:150] if profile.bio else f"{profile.name}",
|
|
"persona": profile.persona or f"{profile.name} is a participant in social discussions.",
|
|
"karma": profile.karma if profile.karma else 1000,
|
|
"created_at": profile.created_at,
|
|
# OASIS-required fields — ensure each has a default
|
|
"age": profile.age if profile.age else 30,
|
|
"gender": self._normalize_gender(profile.gender),
|
|
"mbti": profile.mbti if profile.mbti else "ISTJ",
|
|
"country": profile.country if profile.country else "China",
|
|
}
|
|
|
|
# Optional fields
|
|
if profile.profession:
|
|
item["profession"] = profile.profession
|
|
if profile.interested_topics:
|
|
item["interested_topics"] = profile.interested_topics
|
|
|
|
data.append(item)
|
|
|
|
with open(file_path, 'w', encoding='utf-8') as f:
|
|
json.dump(data, f, ensure_ascii=False, indent=2)
|
|
|
|
logger.info(f"Saved {len(profiles)} Reddit profiles to {file_path} (JSON format, with user_id)")
|
|
|
|
# Keep the old method name as an alias for backward compatibility
|
|
def save_profiles_to_json(
|
|
self,
|
|
profiles: List[OasisAgentProfile],
|
|
file_path: str,
|
|
platform: str = "reddit"
|
|
):
|
|
"""[Deprecated] Please use the save_profiles() method instead."""
|
|
logger.warning("save_profiles_to_json is deprecated; please use save_profiles()")
|
|
self.save_profiles(profiles, file_path, platform)
|
|
|