LinkedinMonitor / run_agentic_pipeline.py
GuglielmoTor's picture
Update run_agentic_pipeline.py
d40cdca verified
raw
history blame
5.43 kB
import gradio as gr
import pandas as pd
import logging
from typing import Dict, Any, List, Optional
# Import the reconstruction function that now expects a cache dictionary
from services.report_data_handler import fetch_and_reconstruct_data_from_bubble
# UI formatting functions
try:
from ui.insights_ui_generator import (
format_report_for_display,
extract_key_results_for_selection,
format_single_okr_for_display
)
AGENTIC_MODULES_LOADED = True
except ImportError as e:
logging.error(f"Could not import agentic pipeline display modules: {e}. Tabs 3 and 4 will be disabled.")
AGENTIC_MODULES_LOADED = False
def format_report_for_display(report_data): return "Agentic modules not loaded."
def extract_key_results_for_selection(okrs_dict): return []
def format_single_okr_for_display(okr_data, **kwargs): return "Agentic modules not loaded."
logger = logging.getLogger(__name__)
def load_and_display_agentic_results(token_state: dict, session_cache: dict):
"""
Loads agentic results from state, populates the report library, and displays
the LATEST report and its fully reconstructed OKRs by default, using a session-specific cache.
Args:
token_state: The main state dictionary with Bubble data.
session_cache: The session-specific cache for reconstructed data.
Returns:
A tuple of Gradio updates, including the updated cache.
"""
initial_updates = (
"No agentic analysis data found in Bubble.",
gr.update(choices=[], value=None, interactive=False),
gr.update(choices=[], value=[], interactive=False),
"No OKRs to display.",
None, [], [], "Status: No agentic analysis data found.",
session_cache # Return the cache unchanged
)
if not AGENTIC_MODULES_LOADED:
error_updates = list(initial_updates)
error_updates[7] = "Status: Critical module import error."
return tuple(error_updates)
agentic_df = token_state.get("bubble_agentic_analysis_data")
if agentic_df is None or agentic_df.empty:
logger.warning("Agentic analysis DataFrame is missing or empty.")
return initial_updates
try:
if 'Created Date' not in agentic_df.columns or '_id' not in agentic_df.columns:
raise KeyError("Required columns ('Created Date', '_id') not found.")
agentic_df['Created Date'] = pd.to_datetime(agentic_df['Created Date'])
agentic_df = agentic_df.sort_values(by='Created Date', ascending=False).reset_index(drop=True)
report_choices = [(f"{row.get('report_type', 'Report')} - {row['Created Date'].strftime('%Y-%m-%d %H:%M')}", row['_id'])
for _, row in agentic_df.iterrows()]
if not report_choices:
return initial_updates
quarterly_reports_df = agentic_df[agentic_df['report_type'] == 'Quarter'].copy()
latest_report_series = quarterly_reports_df.iloc[0]
latest_report_id = latest_report_series['_id']
report_display_md = format_report_for_display(latest_report_series)
report_selector_update = gr.update(choices=report_choices, value=latest_report_id, interactive=True)
# --- MODIFIED: Use the session cache for data reconstruction ---
reconstructed_data, updated_cache = fetch_and_reconstruct_data_from_bubble(latest_report_series, session_cache)
raw_results_state, okr_details_md = None, "No OKRs found in the latest report."
key_results_cbg_update = gr.update(choices=[], value=[], interactive=False)
all_krs_state = []
if reconstructed_data:
raw_results_state = reconstructed_data
actionable_okrs_dict = raw_results_state.get("actionable_okrs", {})
if actionable_okrs_dict:
all_krs_state = extract_key_results_for_selection(actionable_okrs_dict)
if all_krs_state:
kr_choices = [(kr['kr_description'], kr['unique_kr_id']) for kr in all_krs_state]
key_results_cbg_update = gr.update(choices=kr_choices, value=[], interactive=True)
okrs_list = actionable_okrs_dict.get("okrs", [])
output_md_parts = [format_single_okr_for_display(okr, okr_main_index=i) for i, okr in enumerate(okrs_list)]
okr_details_md = "\n\n---\n\n".join(output_md_parts) if output_md_parts else okr_details_md
else:
logger.error(f"Failed to reconstruct data for latest report ID {latest_report_id}")
okr_details_md = "Error: Could not reconstruct OKR data for this report."
status_update = f"Status: Loaded {len(agentic_df)} reports. Displaying latest from {latest_report_series['Created Date'].strftime('%Y-%m-%d')}."
return (
report_display_md, report_selector_update, key_results_cbg_update,
okr_details_md, raw_results_state, [], all_krs_state, status_update,
updated_cache # Return the potentially updated cache
)
except Exception as e:
logger.error(f"Failed to process and display agentic results: {e}", exc_info=True)
error_updates = list(initial_updates)
error_updates[0] = f"An error occurred while loading reports: {e}"
error_updates[7] = f"Status: Error - {e}"
return tuple(error_updates)