gateway/modules/workflows/processing/shared/promptFactoryPlaceholders.py

"""
Placeholder-based prompt factory for dynamic AI calls.
This module provides prompt templates with placeholders that can be filled dynamically.
"""

import json
import logging
from typing import Dict, Any

logger = logging.getLogger(__name__)
from modules.workflows.processing.shared.promptFactory import (
    getAvailableDocuments, 
    getPreviousRoundContext, 
    getMethodsList, 
    getEnhancedDocumentContext,
    getConnectionReferenceList,
    methods,
    discoverMethods
)


def createTaskPlanningPromptTemplate() -> str:
    """Create task planning prompt template with placeholders."""
    return """# Task Planning

Break down user requests into logical, executable task steps.

## 📋 Context

### User Request
{{KEY:USER_PROMPT}}

### Available Documents
{{KEY:AVAILABLE_DOCUMENTS}}

### Previous Workflow Rounds
{{KEY:WORKFLOW_HISTORY}}

## 📝 Task Planning Rules

### High-Level Focus
- **Create HIGH-LEVEL tasks** - one topic per task, not detailed implementation steps
- **Focus on DELIVERING** what the user asked for, not how to do it
- **Keep tasks simple** and focused on outcomes, not implementation details
- **Each task should produce** usable results for subsequent tasks

### Request Type Handling
- **DATA requests** (numbers, lists, calculations): Plan to deliver the actual data
- **DOCUMENT requests** (Word, PDF, Excel): Plan to create the formatted document
- **ANALYSIS requests**: Plan to analyze and deliver insights

### Retry Handling
- **If retry request**: Analyze previous rounds to understand what failed
- **Learn from mistakes**: Improve the plan based on previous failures

## 📊 Required JSON Structure

```json
{
    "overview": "Brief description of the overall plan",
    "languageUserDetected": "en",
    "userMessage": "User-friendly message explaining the task plan",
    "tasks": [
        {
            "id": "task_1",
            "objective": "Clear business objective focusing on what to deliver",
            "dependencies": ["task_0"],
            "success_criteria": ["measurable criteria 1", "measurable criteria 2"],
            "estimated_complexity": "low|medium|high",
            "userMessage": "What this task will accomplish"
        }
    ]
}
```

## 🎯 Task Structure Guidelines

### Task ID Format
- Use sequential numbering: `task_1`, `task_2`, `task_3`
- Keep IDs simple and clear

### Objective Writing
- **Focus on business value** - what will be delivered
- **Be specific** about the expected outcome
- **Avoid technical jargon** - use business language

### Dependencies
- **List prerequisite tasks** that must complete first
- **Use task IDs** from the same plan
- **Keep dependencies minimal** - avoid complex chains

### Success Criteria
- **Make them measurable** - specific, quantifiable outcomes
- **Focus on deliverables** - what the user will receive
- **Keep criteria realistic** - achievable within the task scope

### Complexity Estimation
- **Low**: Simple, straightforward tasks (1-2 actions)
- **Medium**: Moderate complexity (3-5 actions)
- **High**: Complex tasks requiring multiple steps (6+ actions)

## 🚀 Response Format
Return ONLY the JSON object."""


def createActionDefinitionPromptTemplate() -> str:
    """Create action definition prompt template with placeholders."""
    return """# Action Definition

Generate the next action to advance toward completing the task objective.

## 📋 Context

### Task Objective
{{KEY:USER_PROMPT}}

### Available Documents
{{KEY:AVAILABLE_DOCUMENTS}}

### Workflow History
{{KEY:WORKFLOW_HISTORY}}

### Available Methods
{{KEY:AVAILABLE_METHODS}}

### Available Connections
{{KEY:AVAILABLE_CONNECTIONS}}

### User Language
{{KEY:USER_LANGUAGE}}

## ⚠️ RULES

### Action Names
- **Use EXACT compound action names** from AVAILABLE_METHODS (e.g., "ai.process", "document.extract", "web.search")
- **DO NOT create** new action names - only use those listed in AVAILABLE_METHODS
- **DO NOT separate** method and action names - use the full compound name

### Parameter Guidelines
- **Use exact document references** from AVAILABLE_DOCUMENTS
- **Use exact connection references** from AVAILABLE_CONNECTIONS
- **Include user language** if relevant
- **Avoid unnecessary fields** - host applies defaults

## 📊 Required JSON Structure

```json
{
    "actions": [
        {
            "action": "method.action_name",
            "parameters": {},
            "resultLabel": "round{current_round}_task{current_task}_action{action_number}_{descriptive_label}",
            "description": "What this action accomplishes",
            "userMessage": "User-friendly message in {{KEY:USER_LANGUAGE}}"
        }
    ]
}
```

## ✅ Correct Example

```json
{
    "actions": [
        {
            "action": "document.extract",
            "parameters": {"documentList": ["docList:msg_123:results"]},
            "resultLabel": "round1_task1_action1_extract_results",
            "description": "Extract data from documents",
            "userMessage": "Extracting data from documents"
        }
    ]
}
```


## 🎯 Action Planning Guidelines

### Method Selection
- **Choose appropriate method** based on task requirements
- **Consider available resources** (documents, connections)
- **Match method capabilities** to task objectives

### Parameter Design
- **Use ACTION SIGNATURE** to understand required parameters
- **Convert objective** into appropriate parameter values
- **Include all required parameters** for the action

### Result Labeling
- **Use descriptive labels** that explain what the action produces
- **Follow naming convention**: `round{round}_task{task}_action{action}_{label}`
- **Make labels meaningful** for future reference

### User Messages
- **Write in user language** ({{KEY:USER_LANGUAGE}})
- **Explain what's happening** in user-friendly terms
- **Keep messages concise** but informative

## 🚀 Response Format
Return ONLY the JSON object."""


def createActionSelectionPromptTemplate() -> str:
    """Create action selection prompt template with placeholders."""
    return """# Action Selection

Select exactly one action to advance the task.

## 📋 Context

### Objective
{{KEY:USER_PROMPT}}

### Available Documents
{{KEY:AVAILABLE_DOCUMENTS}}

### User Language
{{KEY:USER_LANGUAGE}}

### Available Methods
{{KEY:AVAILABLE_METHODS}}

## ⚠️ CRITICAL RULES

### Selection Requirements
- **Return ONLY the compound action name**
- **Do NOT include parameters or prompts**
- **Use EXACT compound action names** from AVAILABLE_METHODS above
- **DO NOT create** new action names

### Action Format
- **Compound action names**: Use exact names from AVAILABLE_METHODS (e.g., "ai.process", "document.extract", "web.search")
- **Single field format**: Use the full compound action name as a single string

## 📝 Required JSON Format

```json
{"action":"method.action_name"}
```

## ✅ Correct Examples

```json
{"action":"ai.process"}
{"action":"document.extract"}
{"action":"web.search"}
```


## 🎯 Selection Guidelines

### Choose Appropriate Action
- **Match action to objective** - select the most relevant action
- **Consider available resources** - ensure required documents/connections are available
- **Think about the next step** - what action will advance the task

### Method Selection
- **AI methods**: For processing, analysis, or generation tasks
- **Document methods**: For document operations (extract, generate, etc.)
- **Web methods**: For web searches or external data retrieval
- **Other methods**: Based on specific requirements

## 🚀 Response Format
Return ONLY the JSON object."""


def createActionParameterPromptTemplate() -> str:
    """Create action parameter prompt template with placeholders."""
    return """# Action Parameter Generation

You are an AI assistant tasked with generating parameters for a selected action.

## 🎯 Your Goal
Provide the EXACT parameters required by the ACTION SIGNATURE, using information from the OBJECTIVE, AVAILABLE DOCUMENTS, and AVAILABLE CONNECTIONS.

## ⚠️ CRITICAL RULES
- **MUST respond with a JSON object**
- **All parameters MUST be wrapped in a "parameters" object**
- **ONLY include parameters listed in the ACTION SIGNATURE**
- **Do NOT use code blocks or markdown in your response**
- **Return ONLY the JSON object**

## 📋 Document & Connection References
- **Document references**: Copy the EXACT reference string from AVAILABLE DOCUMENTS (e.g., `docList:msg_UUID:label`)
- **Connection references**: Copy the EXACT reference string from AVAILABLE CONNECTIONS (e.g., `connection:msft:user@domain.com:uuid [status:active, token:valid]`)
- **Do NOT invent, shorten, or modify any references**
- **If unsure**: Use "UNCLEAR_REFERENCE" or "UNCLEAR_OBJECTIVE" and explain in a comment

## 📝 Input Context

### Selected Action
{{KEY:SELECTED_ACTION}}

### Objective
{{KEY:USER_PROMPT}}

### Available Documents
{{KEY:AVAILABLE_DOCUMENTS}}

### Available Connections
{{KEY:AVAILABLE_CONNECTIONS}}

### User Language
{{KEY:USER_LANGUAGE}}

### Action Requirements
{{KEY:ACTION_SIGNATURE}}

## 📚 Reference Types

### Document References
- **docItem**: Reference to a single document (e.g., "docItem:uuid:filename.pdf")
- **docList**: Reference to a group of documents (e.g., "docList:msg_123:AnalysisResults")
- **Use EXACT reference strings** shown in AVAILABLE_DOCUMENTS

### Connection References
- **Use exact connection references** from AVAILABLE CONNECTIONS
- **Examples**: "connection:msft:user@domain.com:uuid [status:active, token:valid]", "connection:sp:user@domain.com:uuid [status:active, token:valid]"

## 💡 Basic Examples

```json
{"parameters":{"aiPrompt": "Summarize the document"}}
{"parameters":{"documentList": ["docList:msg_UUID:label"]}}
{"parameters":{"connectionReference": "connection:msft:user@domain.com:uuid [status:active, token:valid]"}}
```

## ❌ Wrong Format (DO NOT USE)

```json
{"aiPrompt": "Your prompt here"}
```

```json
{"parameters":{"aiPrompt": "Your prompt here"}}
```

## 🎯 Parameter Guidelines

### Required Parameters
- **Use ACTION SIGNATURE** to understand what parameters are required
- **Convert objective** into appropriate parameter values
- **Include user language** if relevant
- **Avoid unnecessary fields** - host applies defaults

### Document Reference Rules
- **ONLY use exact document reference strings** from AVAILABLE_DOCUMENTS
- **DO NOT add file paths** or individual filenames to document references
- **For documentList parameters**: Use the EXACT reference strings shown in AVAILABLE_DOCUMENTS

### Connection Reference Rules
- **ONLY use exact connection references** from AVAILABLE CONNECTIONS
- **For connectionReference parameters**: Use the exact connection reference from AVAILABLE CONNECTIONS

## 🚀 Response Format
Return your JSON response immediately after this prompt."""


def createRefinementPromptTemplate() -> str:
    """Create refinement prompt template with placeholders."""
    return """# Workflow Refinement Decision

Decide the next step based on the observation.

## 📋 Context

### Objective
{{KEY:USER_PROMPT}}

### Observation
{{KEY:REVIEW_CONTENT}}

## ⚠️ CRITICAL RULES

### Data Requirements
- **If user wants DATA** (numbers, lists, calculations): Ensure AI delivers the actual data, not code
- **If user wants DOCUMENTS** (Word, PDF, Excel): Ensure appropriate method is used to create the document
- **If user wants ANALYSIS**: Ensure AI analyzes and delivers insights
- **NEVER accept code when user wants data** - demand the actual data
- **NEVER accept algorithms when user wants results** - demand the actual results

## 🤔 Decision Rules

### Continue Conditions
- The objective is **NOT fulfilled** (user didn't get what they asked for)
- More data or processing is needed
- The current result is incomplete

### Stop Conditions
- The objective is **fulfilled** (user got what they asked for)
- All required data has been delivered
- The task is complete

### Focus
- Focus on what the user actually wants, not what was delivered
- Consider the user's original request carefully

## 📝 Response Format

```json
{"decision":"continue","reason":"Need more data"}
```

### Decision Options
- `"continue"` - Keep working on the objective
- `"stop"` - Objective has been fulfilled

### Reason Examples
- `"Need more data"`
- `"Objective fulfilled"`
- `"User got the requested document"`
- `"Analysis complete"`

## 🎯 Decision Guidelines

### When to Continue
- **Incomplete results** - User didn't get what they asked for
- **Missing data** - Need to gather more information
- **Partial success** - Some but not all requirements met
- **Technical issues** - Action failed and needs retry

### When to Stop
- **Complete success** - User got exactly what they asked for
- **All criteria met** - Success criteria have been achieved
- **Document created** - Required document has been generated
- **Data delivered** - All requested data has been provided

### Quality Assessment
- **Check completeness** - Is the result complete?
- **Verify accuracy** - Is the data correct?
- **Assess usefulness** - Does it meet the user's needs?
- **Consider format** - Is it in the requested format?

## 🚀 Response Format
Return your JSON response immediately after this prompt."""


def createResultReviewPromptTemplate() -> str:
    """Create result review prompt template with placeholders."""
    return """# Result Review & Validation

Review task execution outcomes and determine success, retry needs, or failure.

## 📋 Context

### Task Objective
{{KEY:USER_PROMPT}}

### Execution Results
{{KEY:REVIEW_CONTENT}}

## 🔍 Validation Criteria

### Action Assessment
- **Review each action's success/failure status**
- **Check if required documents were produced**
- **Validate document quality and completeness**
- **Assess if success criteria were met**
- **Identify any missing or incomplete outputs**

### Decision Making
- **Determine if retry would help** or if task should be marked as failed
- **Consider business value** and user satisfaction
- **Evaluate technical execution** and results quality

## 📊 Required JSON Structure

```json
{
    "status": "success|retry|failed",
    "reason": "Detailed explanation of the validation decision",
    "improvements": ["specific improvement 1", "specific improvement 2"],
    "quality_score": 8,
    "met_criteria": ["criteria1", "criteria2"],
    "unmet_criteria": ["criteria3", "criteria4"],
    "confidence": 0.85,
    "userMessage": "User-friendly message explaining the validation result"
}
```

## 🎯 Validation Principles

### Assessment Approach
- **Be thorough but fair** in assessment
- **Focus on business value** and outcomes
- **Consider both technical execution** and business results
- **Provide specific, actionable** improvement suggestions

### Quality Scoring
- **Use quality scores** to track progress across retries
- **Scale 1-10**: 1 = Poor, 5 = Average, 10 = Excellent
- **Consider completeness, accuracy, and usefulness**

### Criteria Evaluation
- **Clearly identify** which success criteria were met vs. unmet
- **List specific criteria** that were achieved
- **Note missing requirements** that need attention

### Confidence Levels
- **Set appropriate confidence levels** based on evidence quality
- **Scale 0.0-1.0**: 0.0 = No confidence, 1.0 = Complete confidence
- **Consider data quality** and result reliability

## 📝 Status Definitions

### Success
- **All objectives met** - User got what they asked for
- **Quality standards met** - Results are complete and accurate
- **No retry needed** - Task is fully complete

### Retry
- **Partial success** - Some but not all objectives met
- **Improvement possible** - Retry could lead to better results
- **Technical issues** - Action failures that can be resolved

### Failed
- **No progress made** - Objectives not achieved
- **Technical limitations** - Cannot be resolved with retry
- **Resource constraints** - Missing required inputs

## 💡 Improvement Suggestions

### Actionable Improvements
- **Be specific** - Don't just say "improve quality"
- **Focus on process** - How to do better next time
- **Consider resources** - What additional inputs might help
- **Technical fixes** - Address specific technical issues

### Examples
- "Use more specific document references from AVAILABLE_DOCUMENTS"
- "Include user language parameter for better localization"
- "Break down complex objective into smaller, focused actions"
- "Verify document references before processing"

## 🚀 Response Format
Return ONLY the JSON object. Do not include any explanatory text."""


# Helper functions to extract content for placeholders

def extractUserPrompt(context) -> str:
    """Extract user prompt from context."""
    if hasattr(context, 'task_step') and context.task_step:
        return context.task_step.objective or 'No request specified'
    return 'No request specified'


def extractAvailableDocuments(context) -> str:
    """Extract available documents from context."""
    if hasattr(context, 'available_documents') and context.available_documents:
        return context.available_documents
    return "No documents available"


def extractWorkflowHistory(service, context) -> str:
    """Extract workflow history from context."""
    if hasattr(context, 'workflow') and context.workflow:
        return getPreviousRoundContext(service, context.workflow) or "No previous workflow rounds - this is the first round."
    return "No previous workflow rounds - this is the first round."


def extractAvailableMethods(service) -> str:
    """Extract available methods for action planning using compound action names."""
    try:
        # Get the methods dictionary directly from the global methods variable
        if not methods:
            discoverMethods(service)
        
        # Create a flat JSON format with compound action names for better AI parsing
        available_actions_json = {}
        for methodName, methodInfo in methods.items():
            # Convert MethodAi -> ai, MethodDocument -> document, etc.
            shortName = methodName.replace('Method', '').lower()
            
            for actionName, actionInfo in methodInfo['actions'].items():
                # Create compound action name: method.action
                compoundActionName = f"{shortName}.{actionName}"
                # Get the action description
                action_description = actionInfo.get('description', f"Execute {actionName} action")
                available_actions_json[compoundActionName] = action_description
        
        return json.dumps(available_actions_json, indent=2, ensure_ascii=False)
    except Exception as e:
        logger.error(f"Error extracting available methods: {str(e)}")
        return json.dumps({}, indent=2, ensure_ascii=False)


def extractUserLanguage(service) -> str:
    """Extract user language from service."""
    return service.user.language if service and service.user else 'en'


def extractReviewContent(context) -> str:
    """Extract review content from context with full document metadata."""
    if hasattr(context, 'action_results') and context.action_results:
        # Build result summary
        result_summary = ""
        for i, result in enumerate(context.action_results):
            result_summary += f"\nRESULT {i+1}:\n"
            result_summary += f"  Success: {result.success}\n"
            if result.error:
                result_summary += f"  Error: {result.error}\n"
            
            if result.documents:
                result_summary += f"  Documents: {len(result.documents)} document(s)\n"
                for doc in result.documents:
                    # Extract all available metadata without content
                    doc_metadata = {
                        "name": getattr(doc, 'documentName', 'Unknown'),
                        "mimeType": getattr(doc, 'mimeType', 'Unknown'),
                        "size": getattr(doc, 'size', 'Unknown'),
                        "created": getattr(doc, 'created', 'Unknown'),
                        "modified": getattr(doc, 'modified', 'Unknown'),
                        "typeGroup": getattr(doc, 'typeGroup', 'Unknown'),
                        "documentId": getattr(doc, 'documentId', 'Unknown'),
                        "reference": getattr(doc, 'reference', 'Unknown')
                    }
                    # Remove 'Unknown' values to keep it clean
                    doc_metadata = {k: v for k, v in doc_metadata.items() if v != 'Unknown'}
                    result_summary += f"    - {json.dumps(doc_metadata, indent=6, ensure_ascii=False)}\n"
            else:
                result_summary += f"  Documents: None\n"
        
        return result_summary
    elif hasattr(context, 'observation') and context.observation:
        # For observation data, show full content but handle documents specially
        if isinstance(context.observation, dict):
            # Create a copy to modify
            obs_copy = context.observation.copy()
            
            # If there are previews with documents, show only metadata
            if 'previews' in obs_copy and isinstance(obs_copy['previews'], list):
                for preview in obs_copy['previews']:
                    if isinstance(preview, dict) and 'snippet' in preview:
                        # Replace snippet with metadata indicator
                        preview['snippet'] = f"[Content: {len(preview.get('snippet', ''))} characters]"
            
            return json.dumps(obs_copy, indent=2, ensure_ascii=False)
        else:
            return json.dumps(context.observation, ensure_ascii=False)
    elif hasattr(context, 'step_result') and context.step_result and 'observation' in context.step_result:
        # For observation data in step_result, show full content but handle documents specially
        observation = context.step_result['observation']
        if isinstance(observation, dict):
            # Create a copy to modify
            obs_copy = observation.copy()
            
            # If there are previews with documents, show only metadata
            if 'previews' in obs_copy and isinstance(obs_copy['previews'], list):
                for preview in obs_copy['previews']:
                    if isinstance(preview, dict) and 'snippet' in preview:
                        # Replace snippet with metadata indicator
                        preview['snippet'] = f"[Content: {len(preview.get('snippet', ''))} characters]"
            
            return json.dumps(obs_copy, indent=2, ensure_ascii=False)
        else:
            return json.dumps(observation, ensure_ascii=False)
    else:
        return "No review content available"