mcp-servers/nadlan-mcp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Main updates of Phase 1

Browse Source
This commit is contained in:
Nitzan Pomerantz
2025-10-19 00:58:46 +03:00
parent b844760147
commit 85f52a8108
5 changed files with 1169 additions and 116 deletions
Download Patch File Download Diff File Expand all files Collapse all files
TASKS.md
+242
View File
@@ -0,0 +1,242 @@
# Nadlan-MCP Implementation Tasks
This document tracks the implementation progress of the Nadlan-MCP improvement plan.
## ✅ Completed Tasks
### Phase 1: Code Quality & Reliability
- ✅ Created configuration management system (`config.py`)
- ✅ Added retry logic with exponential backoff
- ✅ Implemented rate limiting protection
- ✅ Standardized error handling (raise exceptions, not return empty lists)
- ✅ Added comprehensive input validation
- ✅ Updated requirements.txt with pinned versions
- ✅ Created requirements-dev.txt for development dependencies
### Documentation
- ✅ Updated USECASES.md with status indicators and roadmap
- ✅ Created ARCHITECTURE.md with system design documentation
- ✅ Marked amenity scoring as future feature with clear roadmap
### Cleanup
- ✅ Deleted redundant mcp_server_concept.py file
## 🚧 In Progress
### Phase 2: Missing Core Functionality
#### 2.1 Property Valuation Data Provision
- [ ] Create `get_comparable_properties()` function with flexible filtering
- [ ] Create `get_deal_statistics()` helper for statistical aggregations
- [ ] Add MCP tool: `get_valuation_comparables`
- [ ] Add MCP tool: `get_deal_statistics`
- [ ] Update documentation with valuation tool usage
#### 2.2 Market Activity & Investment Analysis
- [ ] Create `market_analysis.py` module
- [ ] Implement `calculate_market_activity_score()`
- [ ] Implement `analyze_investment_potential()`
- [ ] Implement `get_market_liquidity()`
- [ ] Add MCP tools for market analysis
- [ ] Update documentation
#### 2.3 Enhanced Deal Filtering & Search
- [ ] Implement `filter_deals_by_criteria()` in govmap.py
- [ ] Add property type filtering
- [ ] Add room count filtering
- [ ] Add price range filtering
- [ ] Add area range filtering
- [ ] Add floor range filtering
- [ ] Update existing functions to support new filters
- [ ] Add tests for filtering logic
## 📋 To-Do (Next Priority)
### Phase 3: Architecture Improvements
#### 3.1 Data Models
- [ ] Create `models.py` with Pydantic models
- [ ] Deal model
- [ ] Address model
- [ ] MarketAnalysis model
- [ ] PropertyValuation model
- [ ] Filter models (DealFilters, etc.)
- [ ] Update functions to use models
- [ ] Add model validation tests
#### 3.2 Separation of Concerns
- [ ] Refactor fastmcp_server.py:
- [ ] Move analysis logic to dedicated modules
- [ ] Keep only MCP tool definitions in fastmcp_server.py
- [ ] Create `api_client.py` for pure API interactions
- [ ] Create `analyzers/` package:
- [ ] `analyzers/market.py` - Market analysis functions
- [ ] `analyzers/filtering.py` - Deal filtering logic
- [ ] `analyzers/valuation.py` - Valuation helpers
- [ ] Update imports and dependencies
- [ ] Update tests
#### 3.3 LLM-Friendly Tool Design
- [ ] Add `summarized_response: bool = False` parameter to all tools
- [ ] Implement summarization logic for each tool
- [ ] Update tool docstrings with parameter descriptions
- [ ] Test both modes (structured and summarized)
- [ ] Update documentation with examples
### Phase 4: Testing & Quality
#### 4.1 Expand Test Coverage
- [ ] Add integration tests (with @pytest.mark.integration)
- [ ] Add edge case tests for all functions
- [ ] Add parametrized tests for address formats
- [ ] Add tests for new valuation tools
- [ ] Add tests for market analysis tools
- [ ] Add tests for enhanced filtering
- [ ] Add tests for `analyze_market_trends`
- [ ] Add tests for `compare_addresses`
- [ ] Add tests for `_is_same_building` logic
#### 4.2 Validation Tests
- [ ] Create `tests/test_validation.py`
- [ ] Test address validation
- [ ] Test coordinate validation
- [ ] Test integer validation
- [ ] Test configuration validation
- [ ] Test model validation (Pydantic)
#### 4.3 Mock External APIs
- [ ] Update `tests/conftest.py` with comprehensive fixtures
- [ ] Add VCR.py for recording/replaying API calls
- [ ] Create fixture for deal responses
- [ ] Create fixture for autocomplete responses
- [ ] Create fixture for error scenarios
### Phase 5: Documentation
#### 5.1 Additional Documentation Files
- [ ] Create `DEPLOYMENT.md` - Deployment guide
- [ ] Create `CONTRIBUTING.md` - Contribution guidelines
- [ ] Create `API_REFERENCE.md` - Detailed API docs
- [ ] Create `CLAUDE.md` - Instructions for AI coding agents
- [ ] Create `docs/` directory for additional docs
#### 5.2 Code Documentation
- [ ] Add module-level docstrings to all Python files
- [ ] Enhance function docstrings with examples
- [ ] Add type hints to remaining functions
- [ ] Add inline comments for complex logic
- [ ] Review and improve existing documentation
#### 5.3 Usage Examples
- [ ] Create `examples/` directory
- [ ] Create `examples/basic_search.py`
- [ ] Create `examples/market_analysis.py`
- [ ] Create `examples/investment_analysis.py`
- [ ] Create `examples/llm_integration.py`
- [ ] Add README in examples/ directory
#### 5.4 README Updates
- [ ] Update README.md with current feature list
- [ ] Add configuration documentation
- [ ] Add troubleshooting section
- [ ] Add API limitations section
- [ ] Add examples from examples/ directory
### Phase 6: Code Quality & Polish
#### 6.2 Code Style & Linting
- [ ] Create `.pre-commit-config.yaml`
- [ ] Setup black formatter
- [ ] Setup isort for imports
- [ ] Setup flake8 linter
- [ ] Setup mypy for type checking
- [ ] Format all code with black
- [ ] Sort all imports with isort
- [ ] Fix all flake8 warnings
- [ ] Fix all mypy errors
- [ ] Add pre-commit hooks to CI
#### 6.3 Remaining Cleanup
- [ ] Remove any remaining unused imports
- [ ] Consolidate duplicate code
- [ ] Refactor long functions (>100 lines)
- [ ] Improve naming consistency
## 🔮 Future Features (Backlog)
### Phase 7.1: Amenity Scoring
- [ ] Research Google Places API integration
- [ ] Research OpenStreetMap integration
- [ ] Research Ministry of Education data sources
- [ ] Research Ministry of Health data sources
- [ ] Design amenity scoring algorithm
- [ ] Implement `amenities.py` module
- [ ] Add amenity MCP tools
- [ ] Add amenity tests
- [ ] Document amenity scoring methodology
### Phase 7.2: Caching System
- [ ] Design caching strategy
- [ ] Implement in-memory cache with TTL
- [ ] Add cache configuration options
- [ ] Add cache statistics/monitoring
- [ ] Test cache invalidation
- [ ] Document caching behavior
- [ ] (Later) Implement Redis integration
- [ ] (Later) Add cache warming
### Phase 7.3: Performance Optimizations
- [ ] Research async/await patterns
- [ ] Convert to async HTTP with httpx
- [ ] Implement parallel polygon queries
- [ ] Add performance benchmarks
- [ ] Optimize token usage in responses
- [ ] (Later) Database integration design
- [ ] (Later) SQLite implementation
- [ ] (Later) PostgreSQL migration path
### Phase 7.4: Multi-language Support
- [ ] Add English address support
- [ ] Add translation service integration
- [ ] Implement language detection
- [ ] Update documentation for multiple languages
- [ ] Add language selection parameter
### Phase 7.5: Advanced Valuation Helper
- [ ] Design calculation algorithm
- [ ] Implement `calculate_valuation_from_comparables()`
- [ ] Add detailed breakdown in response
- [ ] Test calculation accuracy
- [ ] Document methodology
## 📊 Progress Summary
**Overall Progress:** ~15% complete
### By Phase:
- Phase 1 (Code Quality): ✅ 100% complete
- Phase 2 (Core Features): 🚧 0% complete
- Phase 3 (Architecture): 📋 0% started
- Phase 4 (Testing): 📋 0% started
- Phase 5 (Documentation): 🚧 20% complete (USECASES, ARCHITECTURE done)
- Phase 6 (Polish): 🚧 33% complete (cleanup done, linting pending)
- Phase 7 (Future): 📋 Backlog
### High Priority (MVP) Status:
- ✅ Phase 1: Code Quality & Reliability - COMPLETE
- 🚧 Phase 2: Missing Core Functionality - IN PROGRESS (next focus)
- 📋 Phase 6.1: Cleanup - COMPLETE
## 🎯 Current Sprint Focus
1. **Implement valuation data provision tools** (Phase 2.1)
2. **Implement market analysis tools** (Phase 2.2)
3. **Implement enhanced filtering** (Phase 2.3)
## Notes
- All configuration is now externalized and documented
- Error handling is robust with retry logic
- Documentation is aligned with actual implementation
- Ready to add new features on solid foundation
USECASES.md
+103 -26
View File
@@ -2,54 +2,81 @@
The nadlan-mcp MCP server provides comprehensive Israeli real estate data analysis capabilities. Here's what you can do with it:
## 🏠 **Address & Location Services**
**Legend:**
- ✅ **Implemented**: Feature is fully available
- 🚧 **In Progress**: Feature is currently being developed
- 📋 **Planned**: Feature is planned for future release
## 🏠 **Address & Location Services** ✅
- **Address Autocomplete**: Search and autocomplete Israeli addresses (supports both Hebrew and English)
- **Location-based Deals**: Find real estate deals within a specific radius of any coordinates
## 📊 **Real Estate Deal Analysis**
## 📊 **Real Estate Deal Analysis** ✅
- **Recent Deals by Address**: Find recent real estate transactions for any specific address
- **Street-level Analysis**: Get all recent deals for an entire street/area
- **Neighborhood Analysis**: Analyze deals across entire neighborhoods
- **Deal Filtering**: Filter deals by property type, room count, price range, area, and floor (🚧 enhanced filtering in progress)
## 📈 **Market Intelligence** ✅
## 📈 **Market Intelligence**
- **Market Trends Analysis**: Analyze price patterns and market trends over time for any area
- **Price per Square Meter Trends**: Track how property values change over time
- **Market Activity Levels**: See how active the real estate market is in different areas
- **Market Activity Levels**: See how active the real estate market is in different areas (🚧 enhanced metrics in progress)
## 🔍 **Comparative Analysis** ✅
## 🔍 **Comparative Analysis**
- **Multi-Address Comparison**: Compare real estate markets between multiple addresses side-by-side
- **Investment Analysis**: Compare different areas for investment potential
- **Investment Analysis**: Compare different areas for investment potential (🚧 advanced metrics in progress)
## 🏆 **Amenity Scoring & Quality of Life Analysis**
- **Address Amenity Rating**: Score individual addresses based on proximity to essential amenities
## 🏆 **Amenity Scoring & Quality of Life Analysis** 📋
**Status: Planned for Future Release**
This comprehensive feature will provide amenity-based location scoring using multiple data sources:
- **Address Amenity Rating**: Score individual addresses based on proximity AND quality of essential amenities
- **Street-level Amenity Analysis**: Rate entire streets by their access to schools, parks, healthcare, and other services
- **Neighborhood Quality Scoring**: Comprehensive neighborhood ratings based on amenity density and accessibility
- **Neighborhood Quality Scoring**: Comprehensive neighborhood ratings based on amenity density, accessibility, and quality
- **Lifestyle Compatibility**: Match addresses to lifestyle preferences based on nearby amenities
**Data Sources (Planned):**
- Google Places API / OpenStreetMap - amenity locations
- Ministry of Education - school rankings and quality metrics
- Ministry of Health - healthcare facility ratings
- CBS (Central Bureau of Statistics) - demographic data
- Public transport APIs - station locations and service frequency
**Note:** This feature will combine proximity with quality metrics (not just distance) to provide meaningful amenity scores.
## 💡 **Practical Use Cases**
### Property Valuation
### Property Valuation ✅
- Research recent sales to estimate property values
- Understand pricing trends in specific neighborhoods
- Compare similar properties in the area
- **Note:** The MCP provides comprehensive data; the LLM performs the valuation analysis
### Investment Research
### Investment Research ✅
- Identify trending neighborhoods and price patterns
- Analyze market activity levels across different areas
- Analyze market activity levels across different areas (🚧 enhanced metrics in progress)
- Track price per square meter trends over time
### Market Analysis
### Market Analysis ✅
- Understand local real estate dynamics
- Monitor market trends and patterns
- Identify emerging or declining areas
### Due Diligence
### Due Diligence ✅
- Research an area before buying/selling property
- Understand recent transaction history
- Compare multiple potential locations
### Amenity-Based Location Selection
- Score addresses based on proximity to schools, parks, healthcare facilities
### Amenity-Based Location Selection 📋
**Planned for Future Release**
- Score addresses based on proximity AND quality of schools, parks, healthcare facilities
- Find family-friendly neighborhoods with high amenity scores
- Identify areas with the best access to public transportation and services
- Compare lifestyle compatibility across different locations
@@ -63,13 +90,15 @@ You can ask questions like:
- "Find all recent sales on [street name]"
- "Analyze market trends around [specific address]"
- "What's the average price per square meter in [area]?"
- "Rate this address based on nearby amenities like schools and parks"
- "Which street has better access to healthcare and education facilities?"
- "Score this neighborhood for family-friendliness based on amenities"
- "Compare amenity scores between these two addresses"
- "Rate this address based on nearby amenities like schools and parks" *(📋 Planned)*
- "Which street has better access to healthcare and education facilities?" *(📋 Planned)*
- "Score this neighborhood for family-friendliness based on amenities" *(📋 Planned)*
- "Compare amenity scores between these two addresses" *(📋 Planned)*
## 📋 **Available Functions**
### ✅ Implemented
- `autocomplete_address` - Search and autocomplete Israeli addresses
- `get_deals_by_radius` - Get deals within a radius of coordinates
- `get_street_deals` - Get deals for a specific street polygon
@@ -78,16 +107,64 @@ You can ask questions like:
- `analyze_market_trends` - Analyze market trends and price patterns
- `compare_addresses` - Compare real estate markets between multiple addresses
### 🚧 In Progress
- `get_valuation_comparables` - Get comparable properties for valuation analysis
- `get_deal_statistics` - Calculate statistical aggregations on deal data
- `get_market_activity_metrics` - Detailed market activity and velocity metrics
- Enhanced filtering for all deal functions (property type, rooms, price range, area, floor)
### 📋 Planned (Future Release)
- `get_address_amenity_rating` - Comprehensive amenity analysis with quality metrics
- `compare_addresses_by_amenities` - Side-by-side amenity comparison
- `find_amenities_near_address` - Raw amenity list with quality data
## 🎯 **Data Coverage**
The data covers recent Israeli real estate transactions and can help with:
- Property research and valuation
- Investment analysis and decision-making
- Market understanding and trends
- Comparative analysis across locations
- Due diligence for real estate decisions
- Amenity-based location scoring and quality of life analysis
- Property research and valuation ✅
- Investment analysis and decision-making ✅ (🚧 advanced metrics in progress)
- Market understanding and trends ✅
- Comparative analysis across locations ✅
- Due diligence for real estate decisions ✅
- Amenity-based location scoring and quality of life analysis 📋 (planned for future release)
## 🚀 **Getting Started**
Simply connect to the nadlan-mcp server through your MCP client (like Cursor) and start asking questions about Israeli real estate data. The server supports both Hebrew and English addresses and provides comprehensive market analysis capabilities.
## 🔄 **Roadmap & Timeline**
### Phase 1: Core Reliability & Quality (Current)
- ✅ Configuration management system
- ✅ Retry logic with exponential backoff
- ✅ Rate limiting protection
- ✅ Input validation and error handling
- 🚧 Enhanced deal filtering
- 🚧 Valuation data provision tools
- 🚧 Market activity metrics
### Phase 2: Architecture Improvements (Next)
- Data models with Pydantic
- Separation of concerns (API client, analyzers, tools)
- LLM-friendly tool design with summarized_response parameter
- Comprehensive testing suite
### Phase 3: Documentation & Developer Experience
- Architecture documentation
- API reference guide
- Usage examples
- Contributing guidelines
### Phase 4: Future Features
- Amenity scoring with quality metrics
- In-memory caching
- Async/parallel processing
- Production-ready caching (Redis)
- Multi-language support
- Database integration
## 📞 **Support & Contributions**
For issues, questions, or contributions, please create an issue in the repository. The project follows semantic versioning and maintains backward compatibility.
nadlan_mcp/config.py
+131
View File
@@ -0,0 +1,131 @@
"""
Configuration management for Nadlan MCP.
This module provides centralized configuration for API clients, timeouts,
rate limiting, and other settings. Configuration can be set via environment
variables or code.
"""
import os
from typing import Optional
from dataclasses import dataclass, field
@dataclass
class GovmapConfig:
"""Configuration for Govmap API client."""
# API settings
base_url: str = field(
default_factory=lambda: os.getenv(
"GOVMAP_BASE_URL",
"https://www.govmap.gov.il/api/"
)
)
# Timeout settings (in seconds)
connect_timeout: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_CONNECT_TIMEOUT", "10"))
)
read_timeout: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_READ_TIMEOUT", "30"))
)
# Retry settings
max_retries: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_MAX_RETRIES", "3"))
)
retry_min_wait: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_RETRY_MIN_WAIT", "1"))
)
retry_max_wait: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_RETRY_MAX_WAIT", "10"))
)
# Rate limiting
requests_per_second: float = field(
default_factory=lambda: float(os.getenv("GOVMAP_REQUESTS_PER_SECOND", "5.0"))
)
# Default search parameters
default_radius_meters: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_DEFAULT_RADIUS", "50"))
)
default_years_back: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_DEFAULT_YEARS_BACK", "2"))
)
default_deal_limit: int = field(
default_factory=lambda: int(os.getenv("GOVMAP_DEFAULT_DEAL_LIMIT", "100"))
)
# User agent
user_agent: str = field(
default_factory=lambda: os.getenv(
"GOVMAP_USER_AGENT",
"NadlanMCP/1.0.0"
)
)
def __post_init__(self):
"""Validate configuration after initialization."""
self._validate()
def _validate(self):
"""Validate configuration values."""
if self.connect_timeout <= 0:
raise ValueError("connect_timeout must be positive")
if self.read_timeout <= 0:
raise ValueError("read_timeout must be positive")
if self.max_retries < 0:
raise ValueError("max_retries must be non-negative")
if self.retry_min_wait <= 0:
raise ValueError("retry_min_wait must be positive")
if self.retry_max_wait < self.retry_min_wait:
raise ValueError("retry_max_wait must be >= retry_min_wait")
if self.requests_per_second <= 0:
raise ValueError("requests_per_second must be positive")
if self.default_radius_meters <= 0:
raise ValueError("default_radius_meters must be positive")
if self.default_years_back <= 0:
raise ValueError("default_years_back must be positive")
if self.default_deal_limit <= 0:
raise ValueError("default_deal_limit must be positive")
if not self.base_url:
raise ValueError("base_url cannot be empty")
if not self.user_agent:
raise ValueError("user_agent cannot be empty")
# Global configuration instance
_config: Optional[GovmapConfig] = None
def get_config() -> GovmapConfig:
"""
Get the global configuration instance.
Returns:
GovmapConfig: The global configuration object
"""
global _config
if _config is None:
_config = GovmapConfig()
return _config
def set_config(config: GovmapConfig):
"""
Set the global configuration instance.
Args:
config: The new configuration object
"""
global _config
_config = config
def reset_config():
"""Reset the global configuration to default values."""
global _config
_config = None
nadlan_mcp/fastmcp_server.py
+147 -1
View File
@@ -8,7 +8,7 @@ using the FastMCP library with simplified, working functions.
import json
import logging
from typing import List, Dict
from typing import List, Dict, Optional
from mcp.server.fastmcp import FastMCP
from nadlan_mcp.govmap import GovmapClient
@@ -532,6 +532,152 @@ def compare_addresses(addresses: List[str]) -> str:
logger.error(f"Error in compare_addresses: {e}")
return f"Error comparing addresses: {str(e)}"
@mcp.tool()
def get_valuation_comparables(
address: str,
years_back: int = 2,
property_type: Optional[str] = None,
min_rooms: Optional[float] = None,
max_rooms: Optional[float] = None,
min_price: Optional[float] = None,
max_price: Optional[float] = None,
min_area: Optional[float] = None,
max_area: Optional[float] = None,
min_floor: Optional[int] = None,
max_floor: Optional[int] = None
) -> str:
"""Get comparable properties for valuation analysis.
This tool provides detailed comparable deals filtered by your criteria.
The LLM can then analyze these comparables and estimate property values.
Args:
address: The address to find comparables for (in Hebrew or English)
years_back: How many years back to search (default: 2)
property_type: Filter by property type (e.g., "דירה", "בית", "פנטהאוז")
min_rooms: Minimum number of rooms
max_rooms: Maximum number of rooms
min_price: Minimum deal amount (NIS)
max_price: Maximum deal amount (NIS)
min_area: Minimum asset area (square meters)
max_area: Maximum asset area (square meters)
min_floor: Minimum floor number
max_floor: Maximum floor number
Returns:
JSON string containing filtered comparable deals with full details
"""
try:
# Get all deals for the address
deals = client.find_recent_deals_for_address(address, years_back)
if not deals:
return json.dumps({
"address": address,
"years_back": years_back,
"comparables": [],
"message": "No deals found for this address"
}, ensure_ascii=False, indent=2)
# Apply filters
filtered_deals = client.filter_deals_by_criteria(
deals,
property_type=property_type,
min_rooms=min_rooms,
max_rooms=max_rooms,
min_price=min_price,
max_price=max_price,
min_area=min_area,
max_area=max_area,
min_floor=min_floor,
max_floor=max_floor
)
# Calculate statistics on filtered comparables
stats = client.calculate_deal_statistics(filtered_deals)
return json.dumps({
"address": address,
"years_back": years_back,
"filters_applied": {
"property_type": property_type,
"rooms": f"{min_rooms}-{max_rooms}" if min_rooms or max_rooms else None,
"price": f"{min_price}-{max_price}" if min_price or max_price else None,
"area": f"{min_area}-{max_area}" if min_area or max_area else None,
"floor": f"{min_floor}-{max_floor}" if min_floor or max_floor else None,
},
"total_comparables": len(filtered_deals),
"statistics": stats,
"comparables": filtered_deals
}, ensure_ascii=False, indent=2)
except Exception as e:
logger.error(f"Error in get_valuation_comparables: {e}")
return f"Error getting valuation comparables: {str(e)}"
@mcp.tool()
def get_deal_statistics(
address: str,
years_back: int = 2,
property_type: Optional[str] = None,
min_rooms: Optional[float] = None,
max_rooms: Optional[float] = None
) -> str:
"""Calculate statistical aggregations on deal data for an address.
This tool provides quick statistical summaries without returning all raw deals.
Useful when LLM needs calculations on large datasets without full details.
Args:
address: The address to analyze (in Hebrew or English)
years_back: How many years back to analyze (default: 2)
property_type: Filter by property type (e.g., "דירה", "בית")
min_rooms: Minimum number of rooms
max_rooms: Maximum number of rooms
Returns:
JSON string containing statistical metrics (mean, median, percentiles, etc.)
"""
try:
# Get all deals for the address
deals = client.find_recent_deals_for_address(address, years_back)
if not deals:
return json.dumps({
"address": address,
"years_back": years_back,
"statistics": {
"count": 0,
"message": "No deals found for this address"
}
}, ensure_ascii=False, indent=2)
# Apply filters if provided
if property_type or min_rooms or max_rooms:
deals = client.filter_deals_by_criteria(
deals,
property_type=property_type,
min_rooms=min_rooms,
max_rooms=max_rooms
)
# Calculate statistics
stats = client.calculate_deal_statistics(deals)
return json.dumps({
"address": address,
"years_back": years_back,
"filters_applied": {
"property_type": property_type,
"rooms": f"{min_rooms}-{max_rooms}" if min_rooms or max_rooms else None,
},
"statistics": stats
}, ensure_ascii=False, indent=2)
except Exception as e:
logger.error(f"Error in get_deal_statistics: {e}")
return f"Error calculating deal statistics: {str(e)}"
# Run the server
if __name__ == "__main__":
mcp.run()
nadlan_mcp/govmap.py
+546 -89
View File
@@ -1,10 +1,12 @@
import requests
import logging
import json
import time
from datetime import datetime, timedelta
from typing import Any, Dict, List, Tuple, Optional
from nadlan_mcp.config import get_config, GovmapConfig
logger = logging.getLogger(__name__)
@@ -13,22 +15,114 @@ class GovmapClient:
A client for interacting with the Israeli government's Govmap API.
This class provides methods to search for properties, find block/parcel information,
and retrieve real estate deal data.
and retrieve real estate deal data with automatic retries and rate limiting.
Attributes:
config: Configuration object with API settings
session: Requests session for connection pooling
last_request_time: Timestamp of last API request for rate limiting
"""
def __init__(self, base_url: str = "https://www.govmap.gov.il/api/"):
def __init__(self, config: Optional[GovmapConfig] = None):
"""
Initialize the GovmapClient.
Args:
base_url: The base URL for the Govmap API
config: Optional configuration object. If None, uses global config.
"""
self.base_url = base_url.rstrip('/')
self.config = config or get_config()
self.base_url = self.config.base_url.rstrip('/')
self.session = requests.Session()
self.session.headers.update({
'Content-Type': 'application/json',
'User-Agent': 'NadlanMCP/1.0.0'
'User-Agent': self.config.user_agent
})
self.last_request_time = 0.0
def _rate_limit(self):
"""
Enforce rate limiting by sleeping if necessary.
Ensures requests don't exceed the configured requests_per_second.
"""
min_interval = 1.0 / self.config.requests_per_second
elapsed = time.time() - self.last_request_time
if elapsed < min_interval:
time.sleep(min_interval - elapsed)
self.last_request_time = time.time()
def _validate_address(self, address: str) -> str:
"""
Validate and sanitize address input.
Args:
address: Address string to validate
Returns:
Sanitized address string
Raises:
ValueError: If address is invalid
"""
if not address or not isinstance(address, str):
raise ValueError("Address must be a non-empty string")
address = address.strip()
if not address:
raise ValueError("Address cannot be empty or whitespace only")
if len(address) > 500:
raise ValueError("Address is too long (max 500 characters)")
return address
def _validate_coordinates(self, point: Tuple[float, float]) -> Tuple[float, float]:
"""
Validate coordinate input.
Args:
point: Tuple of (longitude, latitude)
Returns:
Validated coordinate tuple
Raises:
ValueError: If coordinates are invalid
"""
if not isinstance(point, (tuple, list)) or len(point) != 2:
raise ValueError("Point must be a tuple of (longitude, latitude)")
try:
lon, lat = float(point[0]), float(point[1])
except (TypeError, ValueError):
raise ValueError("Coordinates must be numeric values")
# Basic validation for Israeli coordinates (ITM projection)
if not (0 < lon < 400000): # Rough bounds for Israeli ITM longitude
logger.warning(f"Longitude {lon} may be outside Israeli bounds")
if not (0 < lat < 1400000): # Rough bounds for Israeli ITM latitude
logger.warning(f"Latitude {lat} may be outside Israeli bounds")
return (lon, lat)
def _validate_positive_int(self, value: int, name: str, max_value: Optional[int] = None) -> int:
"""
Validate positive integer input.
Args:
value: Value to validate
name: Name of the parameter (for error messages)
max_value: Optional maximum allowed value
Returns:
Validated integer
Raises:
ValueError: If value is invalid
"""
if not isinstance(value, int):
raise ValueError(f"{name} must be an integer")
if value <= 0:
raise ValueError(f"{name} must be positive")
if max_value and value > max_value:
raise ValueError(f"{name} must be <= {max_value}")
return value
def autocomplete_address(self, search_text: str) -> Dict[str, Any]:
"""
@@ -41,9 +135,10 @@ class GovmapClient:
Dict containing the JSON response from the API with coordinates
Raises:
requests.RequestException: If the API request fails
ValueError: If the response is invalid
requests.RequestException: If the API request fails after retries
ValueError: If the response is invalid or input is invalid
"""
search_text = self._validate_address(search_text)
url = f"{self.base_url}/search-service/autocomplete"
payload = {
@@ -53,23 +148,35 @@ class GovmapClient:
"maxResults": 10
}
try:
logger.info(f"Searching for address: {search_text}")
response = self.session.post(url, json=payload)
response.raise_for_status()
# Retry logic with exponential backoff
for attempt in range(self.config.max_retries + 1):
try:
self._rate_limit()
logger.info(f"Searching for address: {search_text} (attempt {attempt + 1}/{self.config.max_retries + 1})")
timeout = (self.config.connect_timeout, self.config.read_timeout)
response = self.session.post(url, json=payload, timeout=timeout)
response.raise_for_status()
data = response.json()
if not data or 'results' not in data:
raise ValueError("Invalid response format from autocomplete API")
data = response.json()
if not data or 'results' not in data:
raise ValueError("Invalid response format from autocomplete API")
return data
except requests.RequestException as e:
logger.error(f"Error calling autocomplete API: {e}")
raise
except json.JSONDecodeError as e:
logger.error(f"Error parsing JSON response: {e}")
raise ValueError("Invalid JSON response from API")
return data
except (requests.RequestException, requests.Timeout) as e:
if attempt < self.config.max_retries:
wait_time = min(
self.config.retry_min_wait * (2 ** attempt),
self.config.retry_max_wait
)
logger.warning(f"Request failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
time.sleep(wait_time)
else:
logger.error(f"Request failed after {self.config.max_retries + 1} attempts: {e}")
raise
# This line should never be reached but satisfies type checker
raise RuntimeError("Unexpected error: retry loop exited without return or raise")
def get_gush_helka(self, point: Tuple[float, float]) -> Dict[str, Any]:
"""
@@ -82,9 +189,10 @@ class GovmapClient:
Dict containing the JSON response with block and parcel data
Raises:
requests.RequestException: If the API request fails
ValueError: If the response is invalid
requests.RequestException: If the API request fails after retries
ValueError: If the response or input is invalid
"""
point = self._validate_coordinates(point)
url = f"{self.base_url}/layers-catalog/entitiesByPoint"
payload = {
@@ -93,20 +201,32 @@ class GovmapClient:
"tolerance": 0
}
try:
logger.info(f"Getting Gush/Helka for point: {point}")
response = self.session.post(url, json=payload)
response.raise_for_status()
# Retry logic with exponential backoff
for attempt in range(self.config.max_retries + 1):
try:
self._rate_limit()
logger.info(f"Getting Gush/Helka for point: {point} (attempt {attempt + 1}/{self.config.max_retries + 1})")
timeout = (self.config.connect_timeout, self.config.read_timeout)
response = self.session.post(url, json=payload, timeout=timeout)
response.raise_for_status()
data = response.json()
return data
except requests.RequestException as e:
logger.error(f"Error calling entitiesByPoint API: {e}")
raise
except json.JSONDecodeError as e:
logger.error(f"Error parsing JSON response: {e}")
raise ValueError("Invalid JSON response from API")
data = response.json()
return data
except (requests.RequestException, requests.Timeout) as e:
if attempt < self.config.max_retries:
wait_time = min(
self.config.retry_min_wait * (2 ** attempt),
self.config.retry_max_wait
)
logger.warning(f"Request failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
time.sleep(wait_time)
else:
logger.error(f"Request failed after {self.config.max_retries + 1} attempts: {e}")
raise
# This line should never be reached but satisfies type checker
raise RuntimeError("Unexpected error: retry loop exited without return or raise")
def get_deals_by_radius(self, point: Tuple[float, float], radius: int = 50) -> List[Dict[str, Any]]:
"""
@@ -120,24 +240,41 @@ class GovmapClient:
List of deals found within the radius
Raises:
requests.RequestException: If the API request fails
requests.RequestException: If the API request fails after retries
ValueError: If the response or input is invalid
"""
point = self._validate_coordinates(point)
radius = self._validate_positive_int(radius, "radius", max_value=5000)
url = f"{self.base_url}/real-estate/deals/{point[0]},{point[1]}/{radius}"
try:
logger.info(f"Getting deals by radius for point: {point}, radius: {radius}m")
response = self.session.get(url)
response.raise_for_status()
# Retry logic with exponential backoff
for attempt in range(self.config.max_retries + 1):
try:
self._rate_limit()
logger.info(f"Getting deals by radius for point: {point}, radius: {radius}m (attempt {attempt + 1}/{self.config.max_retries + 1})")
timeout = (self.config.connect_timeout, self.config.read_timeout)
response = self.session.get(url, timeout=timeout)
response.raise_for_status()
data = response.json()
return data if isinstance(data, list) else []
except requests.RequestException as e:
logger.error(f"Error calling deals by radius API: {e}")
raise
except json.JSONDecodeError as e:
logger.error(f"Error parsing JSON response: {e}")
return []
data = response.json()
if not isinstance(data, list):
raise ValueError(f"Expected list response, got {type(data).__name__}")
return data
except (requests.RequestException, requests.Timeout) as e:
if attempt < self.config.max_retries:
wait_time = min(
self.config.retry_min_wait * (2 ** attempt),
self.config.retry_max_wait
)
logger.warning(f"Request failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
time.sleep(wait_time)
else:
logger.error(f"Request failed after {self.config.max_retries + 1} attempts: {e}")
raise
# This line should never be reached but satisfies type checker
raise RuntimeError("Unexpected error: retry loop exited without return or raise")
def get_street_deals(self, polygon_id: str, limit: int = 10,
start_date: Optional[str] = None, end_date: Optional[str] = None,
@@ -156,8 +293,19 @@ class GovmapClient:
List of detailed deal information for the street
Raises:
requests.RequestException: If the API request fails
requests.RequestException: If the API request fails after retries
ValueError: If the response or input is invalid
"""
if not polygon_id or not isinstance(polygon_id, str):
raise ValueError("polygon_id must be a non-empty string")
polygon_id = polygon_id.strip()
if not polygon_id:
raise ValueError("polygon_id cannot be empty or whitespace only")
limit = self._validate_positive_int(limit, "limit", max_value=1000)
if deal_type not in (1, 2):
raise ValueError("deal_type must be 1 (first hand/new) or 2 (second hand/used)")
url = f"{self.base_url}/real-estate/street-deals/{polygon_id}"
params: Dict[str, Any] = {"limit": limit, "dealType": deal_type}
@@ -166,23 +314,40 @@ class GovmapClient:
if end_date:
params["endDate"] = end_date
try:
logger.info(f"Getting street deals for polygon: {polygon_id}, dealType: {deal_type}")
response = self.session.get(url, params=params)
response.raise_for_status()
# Retry logic with exponential backoff
for attempt in range(self.config.max_retries + 1):
try:
self._rate_limit()
logger.info(f"Getting street deals for polygon: {polygon_id}, dealType: {deal_type} (attempt {attempt + 1}/{self.config.max_retries + 1})")
timeout = (self.config.connect_timeout, self.config.read_timeout)
response = self.session.get(url, params=params, timeout=timeout)
response.raise_for_status()
data = response.json()
# API returns {data: [...], totalCount: ..., limit: ..., offset: ...}
if isinstance(data, dict) and 'data' in data:
return data['data'] if isinstance(data['data'], list) else []
return data if isinstance(data, list) else []
except requests.RequestException as e:
logger.error(f"Error calling street deals API: {e}")
raise
except json.JSONDecodeError as e:
logger.error(f"Error parsing JSON response: {e}")
return []
data = response.json()
# API returns {data: [...], totalCount: ..., limit: ..., offset: ...}
if isinstance(data, dict) and 'data' in data:
if not isinstance(data['data'], list):
raise ValueError(f"Expected list in 'data' field, got {type(data['data']).__name__}")
return data['data']
elif isinstance(data, list):
return data
else:
raise ValueError(f"Unexpected response format: {type(data).__name__}")
except (requests.RequestException, requests.Timeout) as e:
if attempt < self.config.max_retries:
wait_time = min(
self.config.retry_min_wait * (2 ** attempt),
self.config.retry_max_wait
)
logger.warning(f"Request failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
time.sleep(wait_time)
else:
logger.error(f"Request failed after {self.config.max_retries + 1} attempts: {e}")
raise
# This line should never be reached but satisfies type checker
raise RuntimeError("Unexpected error: retry loop exited without return or raise")
def get_neighborhood_deals(self, polygon_id: str, limit: int = 10,
start_date: Optional[str] = None, end_date: Optional[str] = None,
@@ -201,8 +366,19 @@ class GovmapClient:
List of deals in the neighborhood
Raises:
requests.RequestException: If the API request fails
requests.RequestException: If the API request fails after retries
ValueError: If the response or input is invalid
"""
if not polygon_id or not isinstance(polygon_id, str):
raise ValueError("polygon_id must be a non-empty string")
polygon_id = polygon_id.strip()
if not polygon_id:
raise ValueError("polygon_id cannot be empty or whitespace only")
limit = self._validate_positive_int(limit, "limit", max_value=1000)
if deal_type not in (1, 2):
raise ValueError("deal_type must be 1 (first hand/new) or 2 (second hand/used)")
url = f"{self.base_url}/real-estate/neighborhood-deals/{polygon_id}"
params: Dict[str, Any] = {"limit": limit, "dealType": deal_type}
@@ -211,23 +387,40 @@ class GovmapClient:
if end_date:
params["endDate"] = end_date
try:
logger.info(f"Getting neighborhood deals for polygon: {polygon_id}, dealType: {deal_type}")
response = self.session.get(url, params=params)
response.raise_for_status()
# Retry logic with exponential backoff
for attempt in range(self.config.max_retries + 1):
try:
self._rate_limit()
logger.info(f"Getting neighborhood deals for polygon: {polygon_id}, dealType: {deal_type} (attempt {attempt + 1}/{self.config.max_retries + 1})")
timeout = (self.config.connect_timeout, self.config.read_timeout)
response = self.session.get(url, params=params, timeout=timeout)
response.raise_for_status()
data = response.json()
# API returns {data: [...], totalCount: ..., limit: ..., offset: ...}
if isinstance(data, dict) and 'data' in data:
return data['data'] if isinstance(data['data'], list) else []
return data if isinstance(data, list) else []
except requests.RequestException as e:
logger.error(f"Error calling neighborhood deals API: {e}")
raise
except json.JSONDecodeError as e:
logger.error(f"Error parsing JSON response: {e}")
return []
data = response.json()
# API returns {data: [...], totalCount: ..., limit: ..., offset: ...}
if isinstance(data, dict) and 'data' in data:
if not isinstance(data['data'], list):
raise ValueError(f"Expected list in 'data' field, got {type(data['data']).__name__}")
return data['data']
elif isinstance(data, list):
return data
else:
raise ValueError(f"Unexpected response format: {type(data).__name__}")
except (requests.RequestException, requests.Timeout) as e:
if attempt < self.config.max_retries:
wait_time = min(
self.config.retry_min_wait * (2 ** attempt),
self.config.retry_max_wait
)
logger.warning(f"Request failed (attempt {attempt + 1}), retrying in {wait_time}s: {e}")
time.sleep(wait_time)
else:
logger.error(f"Request failed after {self.config.max_retries + 1} attempts: {e}")
raise
# This line should never be reached but satisfies type checker
raise RuntimeError("Unexpected error: retry loop exited without return or raise")
def find_recent_deals_for_address(self, address: str, years_back: int = 2,
radius: int = 30, max_deals: int = 50,
@@ -251,9 +444,17 @@ class GovmapClient:
then street deals, then neighborhood deals
Raises:
ValueError: If address cannot be found or processed
requests.RequestException: If API requests fail
ValueError: If address cannot be found or processed, or input is invalid
requests.RequestException: If API requests fail after retries
"""
# Validate inputs
address = self._validate_address(address)
years_back = self._validate_positive_int(years_back, "years_back", max_value=50)
radius = self._validate_positive_int(radius, "radius", max_value=5000)
max_deals = self._validate_positive_int(max_deals, "max_deals", max_value=10000)
if deal_type not in (1, 2):
raise ValueError("deal_type must be 1 (first hand/new) or 2 (second hand/used)")
try:
# Step 1: Get coordinates for the address
logger.info(f"Starting search for address: {address}, dealType: {deal_type}")
@@ -438,4 +639,260 @@ class GovmapClient:
if search_address in deal_address or deal_address in search_address:
return True
return False
return False
def filter_deals_by_criteria(
self,
deals: List[Dict[str, Any]],
property_type: Optional[str] = None,
min_rooms: Optional[float] = None,
max_rooms: Optional[float] = None,
min_price: Optional[float] = None,
max_price: Optional[float] = None,
min_area: Optional[float] = None,
max_area: Optional[float] = None,
min_floor: Optional[int] = None,
max_floor: Optional[int] = None,
) -> List[Dict[str, Any]]:
"""
Filter deals by various criteria.
Args:
deals: List of deal dictionaries to filter
property_type: Property type to filter by (Hebrew description)
min_rooms: Minimum number of rooms
max_rooms: Maximum number of rooms
min_price: Minimum deal amount
max_price: Maximum deal amount
min_area: Minimum asset area (square meters)
max_area: Maximum asset area (square meters)
min_floor: Minimum floor number
max_floor: Maximum floor number
Returns:
Filtered list of deals
Raises:
ValueError: If filter criteria are invalid
"""
if not isinstance(deals, list):
raise ValueError("deals must be a list")
# Validate numeric ranges
if min_rooms is not None and max_rooms is not None and min_rooms > max_rooms:
raise ValueError("min_rooms cannot be greater than max_rooms")
if min_price is not None and max_price is not None and min_price > max_price:
raise ValueError("min_price cannot be greater than max_price")
if min_area is not None and max_area is not None and min_area > max_area:
raise ValueError("min_area cannot be greater than max_area")
if min_floor is not None and max_floor is not None and min_floor > max_floor:
raise ValueError("min_floor cannot be greater than max_floor")
filtered_deals = []
for deal in deals:
# Property type filter
if property_type is not None:
deal_type = deal.get('propertyTypeDescription', deal.get('assetTypeHeb', ''))
if property_type.lower() not in deal_type.lower():
continue
# Room count filter
rooms = deal.get('assetRoomNum')
if rooms is not None:
try:
rooms = float(rooms)
if min_rooms is not None and rooms < min_rooms:
continue
if max_rooms is not None and rooms > max_rooms:
continue
except (TypeError, ValueError):
pass # Skip deals with invalid room data
# Price filter
price = deal.get('dealAmount')
if price is not None:
try:
price = float(price)
if min_price is not None and price < min_price:
continue
if max_price is not None and price > max_price:
continue
except (TypeError, ValueError):
pass # Skip deals with invalid price data
# Area filter
area = deal.get('assetArea')
if area is not None:
try:
area = float(area)
if min_area is not None and area < min_area:
continue
if max_area is not None and area > max_area:
continue
except (TypeError, ValueError):
pass # Skip deals with invalid area data
# Floor filter
floor_str = deal.get('floorNo', '')
if floor_str and isinstance(floor_str, str):
# Try to extract floor number (handles Hebrew floor descriptions)
floor_num = self._extract_floor_number(floor_str)
if floor_num is not None:
if min_floor is not None and floor_num < min_floor:
continue
if max_floor is not None and floor_num > max_floor:
continue
filtered_deals.append(deal)
return filtered_deals
def _extract_floor_number(self, floor_str: str) -> Optional[int]:
"""
Extract numeric floor number from Hebrew floor description.
Args:
floor_str: Floor description string (e.g., "שלישית", "קומה 3", "3")
Returns:
Floor number or None if cannot be extracted
"""
if not floor_str:
return None
# Hebrew ordinal floor names to numbers
hebrew_floors = {
'קרקע': 0, 'מרתף': -1,
'ראשונה': 1, 'שניה': 2, 'שלישית': 3, 'רביעית': 4,
'חמישית': 5, 'שישית': 6, 'שביעית': 7, 'שמינית': 8,
'תשיעית': 9, 'עשירית': 10
}
floor_lower = floor_str.lower().strip()
# Check for direct match with Hebrew names
for heb, num in hebrew_floors.items():
if heb in floor_lower:
return num
# Try to extract number from string
import re
numbers = re.findall(r'\d+', floor_str)
if numbers:
try:
return int(numbers[0])
except ValueError:
pass
return None
def calculate_deal_statistics(
self,
deals: List[Dict[str, Any]]
) -> Dict[str, Any]:
"""
Calculate statistical aggregations on deal data.
Args:
deals: List of deal dictionaries
Returns:
Dictionary with statistical metrics
Raises:
ValueError: If deals is not a valid list
"""
if not isinstance(deals, list):
raise ValueError("deals must be a list")
if not deals:
return {
"count": 0,
"price_stats": {},
"area_stats": {},
"price_per_sqm_stats": {},
"room_distribution": {}
}
# Extract numeric values
prices = []
areas = []
price_per_sqm_values = []
rooms = []
for deal in deals:
price = deal.get('dealAmount')
if isinstance(price, (int, float)) and price > 0:
prices.append(price)
area = deal.get('assetArea')
if isinstance(area, (int, float)) and area > 0:
areas.append(area)
pps = deal.get('price_per_sqm')
if pps is None and price and area and area > 0:
pps = price / area
if isinstance(pps, (int, float)) and pps > 0:
price_per_sqm_values.append(pps)
room_count = deal.get('assetRoomNum')
if isinstance(room_count, (int, float)):
rooms.append(room_count)
# Calculate statistics
stats: Dict[str, Any] = {"count": len(deals)}
# Price statistics
if prices:
sorted_prices = sorted(prices)
stats["price_stats"] = {
"mean": round(sum(prices) / len(prices), 2),
"median": sorted_prices[len(sorted_prices) // 2],
"min": min(prices),
"max": max(prices),
"p25": sorted_prices[len(sorted_prices) // 4],
"p75": sorted_prices[(3 * len(sorted_prices)) // 4],
"std_dev": round(self._calculate_std_dev(prices), 2) if len(prices) > 1 else 0,
"total": sum(prices)
}
# Area statistics
if areas:
sorted_areas = sorted(areas)
stats["area_stats"] = {
"mean": round(sum(areas) / len(areas), 2),
"median": sorted_areas[len(sorted_areas) // 2],
"min": min(areas),
"max": max(areas),
"p25": sorted_areas[len(sorted_areas) // 4],
"p75": sorted_areas[(3 * len(sorted_areas)) // 4]
}
# Price per sqm statistics
if price_per_sqm_values:
sorted_pps = sorted(price_per_sqm_values)
stats["price_per_sqm_stats"] = {
"mean": round(sum(price_per_sqm_values) / len(price_per_sqm_values), 2),
"median": round(sorted_pps[len(sorted_pps) // 2], 2),
"min": round(min(price_per_sqm_values), 2),
"max": round(max(price_per_sqm_values), 2),
"p25": round(sorted_pps[len(sorted_pps) // 4], 2),
"p75": round(sorted_pps[(3 * len(sorted_pps)) // 4], 2)
}
# Room distribution
if rooms:
from collections import Counter
room_counts = Counter(rooms)
stats["room_distribution"] = dict(sorted(room_counts.items()))
return stats
def _calculate_std_dev(self, values: List[float]) -> float:
"""Calculate standard deviation of a list of values."""
if len(values) < 2:
return 0.0
mean = sum(values) / len(values)
variance = sum((x - mean) ** 2 for x in values) / (len(values) - 1)
return variance ** 0.5