arionxiv 1.0.32__py3-none-any.whl

arionxiv/prompts/prompts.py

@@ -0,0 +1,247 @@
+"""
+Centralized Prompt Templates for ArionXiv
+
+Prompts are stored in MongoDB and fetched on-demand with TTL caching.
+Admin can edit prompts directly in the database (prompts collection).
+
+Fallback prompts are used when database is unavailable.
+"""
+
+from typing import Dict, Any, Optional
+import asyncio
+import logging
+
+logger = logging.getLogger(__name__)
+
+# ============================================================================
+# FALLBACK PROMPTS - Used when database is unavailable
+# Master prompts are stored in MongoDB (prompts collection)
+# ============================================================================
+
+DEFAULT_PROMPTS: Dict[str, str] = {
+    
+    "comprehensive_paper_analysis": """You are an expert research analyst. Analyze this research paper thoroughly.
+
+Paper Content:
+{content}
+
+IMPORTANT: You MUST respond with ONLY valid JSON. No explanations, no markdown, no text before or after the JSON.
+
+Return EXACTLY this JSON structure (replace placeholders with your analysis):
+{{
+  "summary": "A comprehensive 3-4 sentence summary of the paper's main contribution and findings",
+  "key_findings": ["Finding 1", "Finding 2", "Finding 3", "Finding 4"],
+  "methodology": "Description of the research methodology, algorithms, or experimental approach used",
+  "strengths": ["Strength 1", "Strength 2", "Strength 3"],
+  "limitations": ["Limitation 1", "Limitation 2"],
+  "technical_details": "Detailed technical explanation of the core innovations, architectures, or implementations",
+  "broader_impact": "Analysis of the paper's broader implications, applications, and future research directions",
+  "confidence_score": 0.85,
+  "relevance_tags": ["tag1", "tag2", "tag3"],
+  "technical_level": "intermediate",
+  "novelty_score": 0.8,
+  "impact_potential": "Description of the potential real-world impact"
+}}
+
+Remember: Output ONLY the JSON object. No other text.""",
+
+    "enhanced_paper_analysis": """You are an elite research analyst. Provide an exceptionally thorough analysis of this research paper.
+
+Paper Content:
+{content}
+
+Provide a detailed analysis in JSON format with: summary, key_findings, methodology, strengths, limitations, technical_details, broader_impact, confidence_score, relevance_tags, technical_level.""",
+
+    "paper_analysis": """Analyze this research paper comprehensively.
+
+Title: {title}
+Abstract: {abstract}
+Content: {content}
+
+Provide JSON with: summary, key_findings, methodology, strengths, limitations.""",
+
+    "quick_analysis": """Provide a quick analysis of this paper.
+
+Title: {title}
+Abstract: {abstract}
+
+Return JSON with: summary, main_contribution, relevance_score.""",
+
+    "research_analysis": """Analyze this research paper in depth.
+
+Title: {title}
+Content: {content}
+
+Provide JSON with: detailed_summary, methodology, findings, implications.""",
+
+    "summary_analysis": """Provide a summary analysis of this paper.
+
+Content: {content}
+
+Focus on key points and main contributions.""",
+
+    "detailed_analysis": """Provide a detailed analysis of this paper.
+
+Content: {content}
+
+Include methodology, findings, strengths, and limitations.""",
+
+    "technical_analysis": """Provide a technical analysis of this paper.
+
+Content: {content}
+
+Focus on technical contributions, algorithms, and implementation details.""",
+
+    "insights_analysis": """Extract key insights from this paper.
+
+Content: {content}
+
+Provide actionable insights and implications for the field.""",
+
+    "daily_dose_analysis": """Analyze this research paper and provide a thorough analysis.
+
+Title: {title}
+
+Authors: {authors}
+
+Categories: {categories}
+
+Abstract: {abstract}
+
+Provide a structured analysis with the following sections:
+1. SUMMARY: A concise 2-3 sentence summary of the paper's main contribution.
+2. KEY FINDINGS: List the 3-4 most important findings or contributions (as numbered items).
+3. METHODOLOGY: Brief description of the approach or methods used.
+4. SIGNIFICANCE: Why this paper matters and its potential impact.
+5. LIMITATIONS: Any noted limitations or areas for future work.
+6. RELEVANCE SCORE: Rate from 1-10 how impactful this paper is for practitioners.
+
+IMPORTANT: Do NOT use any markdown formatting (no asterisks, no bold, no headers). Write in plain text only.
+Be concise but thorough. Focus on actionable insights.""",
+
+    "trend_analysis": """You are a research trend analyst. Based on the following research papers, generate insights about current trends and directions.
+
+Papers:
+{papers_data}
+
+Analyze: emerging trends, common methodologies, novel approaches, cross-paper themes, future opportunities, field evolution.""",
+
+    "enhanced_trend_analysis": """You are an expert research trend analyst. Analyze the following papers for emerging trends and patterns.
+
+Papers:
+{papers_data}
+
+Provide JSON with: emerging_trends, common_themes, methodology_patterns, future_directions, key_insights.""",
+
+    "paper_summary": """Summarize the following collection of research papers concisely.
+
+Papers:
+{papers_data}
+
+Provide a brief, coherent summary of the main themes and contributions.""",
+
+    "personalized_recommendations": """Based on the user's research interests and history, recommend relevant research directions.
+
+User Profile:
+{user_profile}
+
+Recent Activity:
+{recent_activity}
+
+Provide JSON with: recommended_topics, suggested_papers, emerging_areas, learning_path.""",
+
+    "rag_chat": """You are an AI research assistant helping users understand research papers. Your role is to provide accurate, helpful answers based on the paper content.
+
+PAPER METADATA:
+Title: {paper_title}
+Authors: {paper_authors}
+Published on arXiv: {paper_published}
+
+RELEVANT SECTIONS FROM THE PAPER:
+{context}
+
+CONVERSATION HISTORY:
+{history}
+
+USER QUESTION: {message}
+
+Instructions:
+- You are discussing the paper "{paper_title}" - always remember and reference this title when asked about which paper you're discussing
+- Provide comprehensive, detailed answers using specific details from the paper
+- Quote relevant passages when appropriate
+- If the answer is not in the provided context, say so clearly
+- Be conversational but maintain technical accuracy
+- Structure longer answers with clear sections""",
+
+}
+
+
+# ============================================================================
+# PROMPT SERVICE INTEGRATION
+# ============================================================================
+
+try:
+    from ..services.unified_prompt_service import unified_prompt_service
+    PROMPT_SERVICE_AVAILABLE = True
+except ImportError:
+    PROMPT_SERVICE_AVAILABLE = False
+    # Debug level - this is expected during module initialization
+    logger.debug("UnifiedPromptService not available - using fallback prompts")
+
+def format_prompt(prompt_name: str, **kwargs) -> str:
+    """
+    Format a prompt template with provided variables.
+    Fetches from MongoDB with caching, falls back to local prompts if unavailable.
+    """
+    if PROMPT_SERVICE_AVAILABLE:
+        try:
+            try:
+                loop = asyncio.get_running_loop()
+                raise RuntimeError("In async context")
+            except RuntimeError:
+                try:
+                    loop = asyncio.get_event_loop()
+                    if loop.is_closed():
+                        loop = asyncio.new_event_loop()
+                        asyncio.set_event_loop(loop)
+                except RuntimeError:
+                    loop = asyncio.new_event_loop()
+                    asyncio.set_event_loop(loop)
+                
+                if not loop.is_running():
+                    formatted = loop.run_until_complete(
+                        unified_prompt_service.format_prompt(prompt_name, **kwargs)
+                    )
+                    return formatted
+            
+        except Exception as e:
+            logger.debug(f"Using fallback prompt for {prompt_name}: {str(e)}")
+    
+    # Fallback to local prompts
+    if prompt_name in DEFAULT_PROMPTS:
+        return DEFAULT_PROMPTS[prompt_name].format(**kwargs)
+    else:
+        logger.warning(f"Unknown prompt '{prompt_name}', using generic template")
+        return f"Process the following with context: {kwargs}"
+
+async def format_prompt_async(prompt_name: str, **kwargs) -> str:
+    """
+    Async version of format_prompt for use in async contexts.
+    """
+    if PROMPT_SERVICE_AVAILABLE:
+        try:
+            return await unified_prompt_service.format_prompt(prompt_name, **kwargs)
+        except Exception as e:
+            logger.debug(f"Using fallback prompt for {prompt_name}: {str(e)}")
+    
+    # Fallback to local prompts
+    if prompt_name in DEFAULT_PROMPTS:
+        return DEFAULT_PROMPTS[prompt_name].format(**kwargs)
+    else:
+        logger.warning(f"Unknown prompt '{prompt_name}', using generic template")
+        return f"Process the following with context: {kwargs}"
+
+
+def get_all_prompts() -> Dict[str, str]:
+    """Get all available prompt names"""
+    return DEFAULT_PROMPTS

arionxiv/rag_techniques/__init__.py

@@ -0,0 +1,8 @@
+"""
+RAG Techniques Module for ArionXiv
+Provides various Retrieval-Augmented Generation implementations
+"""
+
+from .basic_rag import BasicRAG
+
+__all__ = ['BasicRAG']