tools-only Schema Validation for Construction Data
Overview
Validate data structures against defined schemas for construction data exchange. Ensure API payloads, file imports, and BIM exports conform to expected formats before processing.
Schema Validation Framework
Core Schema Validator
from dataclasses import dataclass, field
from typing import Any, Dict, List, Optional, Union
from enum import Enum
import json
import re
from datetime import datetime
class SchemaType(Enum):
STRING = "string"
NUMBER = "number"
INTEGER = "integer"
BOOLEAN = "boolean"
ARRAY = "array"
OBJECT = "object"
DATE = "date"
DATETIME = "datetime"
CSI_CODE = "csi_code"
CURRENCY = "currency"
GUID = "guid"
@dataclass
class SchemaField:
name: str
type: SchemaType
required: bool = True
nullable: bool = False
min_value: Optional[float] = None
max_value: Optional[float] = None
min_length: Optional[int] = None
max_length: Optional[int] = None
pattern: Optional[str] = None
enum_values: Optional[List[Any]] = None
items_schema: Optional['Schema'] = None # For arrays
properties: Optional[Dict[str, 'SchemaField']] = None # For objects
description: str = ""
@dataclass
class Schema:
name: str
version: str
fields: Dict[str, SchemaField]
description: str = ""
@dataclass
class SchemaValidationError:
path: str
message: str
expected: str
actual: Any
@dataclass
class SchemaValidationResult:
is_valid: bool
errors: List[SchemaValidationError] = field(default_factory=list)
schema_name: str = ""
schema_version: str = ""
def add_error(self, path: str, message: str, expected: str, actual: Any):
self.errors.append(SchemaValidationError(path, message, expected, actual))
self.is_valid = False
def to_report(self) -> str:
lines = [
f"Schema Validation: {self.schema_name} v{self.schema_version}",
"=" * 50,
f"Status: {'✓ VALID' if self.is_valid else '✗ INVALID'}",
f"Errors: {len(self.errors)}",
""
]
for error in self.errors:
lines.append(f"❌ {error.path}")
lines.append(f" {error.message}")
lines.append(f" Expected: {error.expected}")
lines.append(f" Actual: {error.actual}")
lines.append("")
return "\n".join(lines)
class SchemaValidator:
"""Validate data against schemas."""
# Custom type patterns
PATTERNS = {
SchemaType.CSI_CODE: r'^\d{2}\s?\d{2}\s?\d{2}$',
SchemaType.GUID: r'^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$',
SchemaType.CURRENCY: r'^-?\d+(\.\d{2})?$',
SchemaType.DATE: r'^\d{4}-\d{2}-\d{2}$',
SchemaType.DATETIME: r'^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}',
}
def validate(self, data: Any, schema: Schema) -> SchemaValidationResult:
result = SchemaValidationResult(
is_valid=True,
schema_name=schema.name,
schema_version=schema.version
)
self._validate_object(data, schema.fields, "", result)
return result
def _validate_object(self, data: Any, fields: Dict[str, SchemaField], path: str, result: SchemaValidationResult):
if not isinstance(data, dict):
result.add_error(path or "root", "Expected object", "object", type(data).__name__)
return
# Check required fields
for field_name, field_schema in fields.items():
field_path = f"{path}.{field_name}" if path else field_name
if field_name not in data:
if field_schema.required:
result.add_error(field_path, "Required field missing", "present", "missing")
continue
value = data[field_name]
# Check nullable
if value is None:
if not field_schema.nullable:
result.add_error(field_path, "Field cannot be null", "non-null", "null")
continue
# Validate type
self._validate_field(value, field_schema, field_path, result)
# Check for extra fields (warning only)
for key in data.keys():
if key not in fields:
# Could add warning here if needed
pass
def _validate_field(self, value: Any, schema: SchemaField, path: str, result: SchemaValidationResult):
# Type validation
if not self._check_type(value, schema.type):
result.add_error(path, f"Invalid type", schema.type.value, type(value).__name__)
return
# String validations
if schema.type == SchemaType.STRING:
if schema.min_length and len(value) < schema.min_length:
result.add_error(path, f"String too short", f"min {schema.min_length}", len(value))
if schema.max_length and len(value) > schema.max_length:
result.add_error(path, f"String too long", f"max {schema.max_length}", len(value))
if schema.pattern and not re.match(schema.pattern, value):
result.add_error(path, "Pattern mismatch", schema.pattern, value)
# Numeric validations
if schema.type in (SchemaType.NUMBER, SchemaType.INTEGER):
if schema.min_value is not None and value < schema.min_value:
result.add_error(path, "Value below minimum", f">= {schema.min_value}", value)
if schema.max_value is not None and value > schema.max_value:
result.add_error(path, "Value above maximum", f"<= {schema.max_value}", value)
# Enum validation
if schema.enum_values and value not in schema.enum_values:
result.add_error(path, "Invalid enum value", str(schema.enum_values), value)
# Array validation
if schema.type == SchemaType.ARRAY and schema.items_schema:
for i, item in enumerate(value):
item_path = f"{path}[{i}]"
if schema.items_schema.fields:
self._validate_object(item, schema.items_schema.fields, item_path, result)
# Nested object validation
if schema.type == SchemaType.OBJECT and schema.properties:
self._validate_object(value, schema.properties, path, result)
# Custom type validation
if schema.type in self.PATTERNS:
pattern = self.PATTERNS[schema.type]
if not re.match(pattern, str(value)):
result.add_error(path, f"Invalid {schema.type.value} format", pattern, value)
def _check_type(self, value: Any, expected: SchemaType) -> bool:
type_checks = {
SchemaType.STRING: lambda v: isinstance(v, str),
SchemaType.NUMBER: lambda v: isinstance(v, (int, float)),
SchemaType.INTEGER: lambda v: isinstance(v, int) and not isinstance(v, bool),
SchemaType.BOOLEAN: lambda v: isinstance(v, bool),
SchemaType.ARRAY: lambda v: isinstance(v, list),
SchemaType.OBJECT: lambda v: isinstance(v, dict),
SchemaType.DATE: lambda v: isinstance(v, str),
SchemaType.DATETIME: lambda v: isinstance(v, str),
SchemaType.CSI_CODE: lambda v: isinstance(v, str),
SchemaType.CURRENCY: lambda v: isinstance(v, (int, float, str)),
SchemaType.GUID: lambda v: isinstance(v, str),
}
return type_checks.get(expected, lambda v: True)(value)
Construction Data Schemas
Cost Estimate Schema
# Define schema for cost estimate data
COST_ESTIMATE_SCHEMA = Schema(
name="CostEstimate",
version="1.0",
description="Schema for construction cost estimates",
fields={
"project_id": SchemaField(
name="project_id",
type=SchemaType.STRING,
required=True,
description="Unique project identifier"
),
"project_name": SchemaField(
name="project_name",
type=SchemaType.STRING,
required=True,
max_length=200
),
"estimate_type": SchemaField(
name="estimate_type",
type=SchemaType.STRING,
required=True,
enum_values=["conceptual", "schematic", "design_development", "construction_documents", "bid"]
),
"estimate_date": SchemaField(
name="estimate_date",
type=SchemaType.DATE,
required=True
),
"currency": SchemaField(
name="currency",
type=SchemaType.STRING,
required=False,
enum_values=["USD", "EUR", "GBP", "CAD"],
nullable=True
),
"gross_area": SchemaField(
name="gross_area",
type=SchemaType.NUMBER,
required=True,
min_value=0,
description="Gross floor area in SF or SM"
),
"line_items": SchemaField(
name="line_items",
type=SchemaType.ARRAY,
required=True,
items_schema=Schema(
name="LineItem",
version="1.0",
fields={
"id": SchemaField(name="id", type=SchemaType.STRING, required=True),
"csi_code": SchemaField(name="csi_code", type=SchemaType.CSI_CODE, required=False, nullable=True),
"description": SchemaField(name="description", type=SchemaType.STRING, required=True, max_length=500),
"quantity": SchemaField(name="quantity", type=SchemaType.NUMBER, required=True, min_value=0),
"unit": SchemaField(name="unit", type=SchemaType.STRING, required=True),
"unit_cost": SchemaField(name="unit_cost", type=SchemaType.NUMBER, required=True, min_value=0),
"amount": SchemaField(name="amount", type=SchemaType.NUMBER, required=True, min_value=0),
}
)
),
"subtotal": SchemaField(
name="subtotal",
type=SchemaType.NUMBER,
required=True,
min_value=0
),
"contingency_percent": SchemaField(
name="contingency_percent",
type=SchemaType.NUMBER,
required=False,
min_value=0,
max_value=50
),
"total": SchemaField(
name="total",
type=SchemaType.NUMBER,
required=True,
min_value=0
)
}
)
Schedule Data Schema
SCHEDULE_SCHEMA = Schema(
name="ProjectSchedule",
version="1.0",
description="Schema for project schedule data",
fields={
"project_id": SchemaField(name="project_id", type=SchemaType.STRING, required=True),
"schedule_name": SchemaField(name="schedule_name", type=SchemaType.STRING, required=True),
"data_date": SchemaField(name="data_date", type=SchemaType.DATE, required=True),
"start_date": SchemaField(name="start_date", type=SchemaType.DATE, required=True),
"finish_date": SchemaField(name="finish_date", type=SchemaType.DATE, required=True),
"calendar": SchemaField(
name="calendar",
type=SchemaType.STRING,
required=False,
enum_values=["5-day", "6-day", "7-day"],
nullable=True
),
"tasks": SchemaField(
name="tasks",
type=SchemaType.ARRAY,
required=True,
items_schema=Schema(
name="Task",
version="1.0",
fields={
"id": SchemaField(name="id", type=SchemaType.STRING, required=True),
"wbs": SchemaField(name="wbs", type=SchemaType.STRING, required=False, nullable=True),
"name": SchemaField(name="name", type=SchemaType.STRING, required=True, max_length=300),
"start_date": SchemaField(name="start_date", type=SchemaType.DATE, required=True),
"finish_date": SchemaField(name="finish_date", type=SchemaType.DATE, required=True),
"duration": SchemaField(name="duration", type=SchemaType.INTEGER, required=True, min_value=0),
"percent_complete": SchemaField(name="percent_complete", type=SchemaType.NUMBER, required=False, min_value=0, max_value=100),
"predecessors": SchemaField(name="predecessors", type=SchemaType.ARRAY, required=False, nullable=True),
"resources": SchemaField(name="resources", type=SchemaType.ARRAY, required=False, nullable=True),
}
)
)
}
)
BIM Element Schema
BIM_ELEMENT_SCHEMA = Schema(
name="BIMElement",
version="1.0",
description="Schema for BIM element data",
fields={
"guid": SchemaField(name="guid", type=SchemaType.GUID, required=True),
"ifc_class": SchemaField(
name="ifc_class",
type=SchemaType.STRING,
required=True,
pattern=r'^Ifc[A-Z][a-zA-Z]+$'
),
"name": SchemaField(name="name", type=SchemaType.STRING, required=False, nullable=True),
"description": SchemaField(name="description", type=SchemaType.STRING, required=False, nullable=True),
"level": SchemaField(name="level", type=SchemaType.STRING, required=False, nullable=True),
"classification": SchemaField(
name="classification",
type=SchemaType.OBJECT,
required=False,
nullable=True,
properties={
"system": SchemaField(name="system", type=SchemaType.STRING, required=True),
"code": SchemaField(name="code", type=SchemaType.STRING, required=True),
"name": SchemaField(name="name", type=SchemaType.STRING, required=False, nullable=True),
}
),
"quantities": SchemaField(
name="quantities",
type=SchemaType.OBJECT,
required=False,
nullable=True,
properties={
"area": SchemaField(name="area", type=SchemaType.NUMBER, required=False, min_value=0, nullable=True),
"volume": SchemaField(name="volume", type=SchemaType.NUMBER, required=False, min_value=0, nullable=True),
"length": SchemaField(name="length", type=SchemaType.NUMBER, required=False, min_value=0, nullable=True),
"count": SchemaField(name="count", type=SchemaType.INTEGER, required=False, min_value=0, nullable=True),
}
),
"properties": SchemaField(name="properties", type=SchemaType.OBJECT, required=False, nullable=True)
}
)
RFI Schema
RFI_SCHEMA = Schema(
name="RFI",
version="1.0",
description="Schema for Request for Information",
fields={
"rfi_number": SchemaField(name="rfi_number", type=SchemaType.STRING, required=True, pattern=r'^RFI-\d+$'),
"project_id": SchemaField(name="project_id", type=SchemaType.STRING, required=True),
"subject": SchemaField(name="subject", type=SchemaType.STRING, required=True, max_length=500),
"status": SchemaField(
name="status",
type=SchemaType.STRING,
required=True,
enum_values=["draft", "submitted", "in_review", "answered", "closed"]
),
"priority": SchemaField(
name="priority",
type=SchemaType.STRING,
required=False,
enum_values=["low", "medium", "high", "critical"],
nullable=True
),
"date_submitted": SchemaField(name="date_submitted", type=SchemaType.DATE, required=True),
"date_required": SchemaField(name="date_required", type=SchemaType.DATE, required=True),
"from_company": SchemaField(name="from_company", type=SchemaType.STRING, required=True),
"to_company": SchemaField(name="to_company", type=SchemaType.STRING, required=True),
"spec_section": SchemaField(name="spec_section", type=SchemaType.CSI_CODE, required=False, nullable=True),
"drawing_reference": SchemaField(name="drawing_reference", type=SchemaType.STRING, required=False, nullable=True),
"question": SchemaField(name="question", type=SchemaType.STRING, required=True),
"response": SchemaField(name="response", type=SchemaType.STRING, required=False, nullable=True),
"date_responded": SchemaField(name="date_responded", type=SchemaType.DATE, required=False, nullable=True),
"attachments": SchemaField(name="attachments", type=SchemaType.ARRAY, required=False, nullable=True)
}
)
Schema Registry
class ConstructionSchemaRegistry:
"""Registry of construction data schemas."""
def __init__(self):
self.schemas: Dict[str, Schema] = {}
self._register_defaults()
def _register_defaults(self):
self.register(COST_ESTIMATE_SCHEMA)
self.register(SCHEDULE_SCHEMA)
self.register(BIM_ELEMENT_SCHEMA)
self.register(RFI_SCHEMA)
def register(self, schema: Schema):
key = f"{schema.name}:{schema.version}"
self.schemas[key] = schema
# Also register without version for latest
self.schemas[schema.name] = schema
def get(self, name: str, version: str = None) -> Optional[Schema]:
if version:
return self.schemas.get(f"{name}:{version}")
return self.schemas.get(name)
def validate(self, data: Any, schema_name: str, version: str = None) -> SchemaValidationResult:
schema = self.get(schema_name, version)
if not schema:
result = SchemaValidationResult(is_valid=False)
result.add_error("schema", f"Schema '{schema_name}' not found", "valid schema", "not found")
return result
validator = SchemaValidator()
return validator.validate(data, schema)
def list_schemas(self) -> List[str]:
return [k for k in self.schemas.keys() if ':' in k]
Usage Examples
# Initialize registry
registry = ConstructionSchemaRegistry()
# Validate cost estimate
estimate_data = {
"project_id": "PROJ-001",
"project_name": "Downtown Office Tower",
"estimate_type": "schematic",
"estimate_date": "2026-01-15",
"gross_area": 50000,
"line_items": [
{
"id": "1",
"csi_code": "03 30 00",
"description": "Cast-in-place concrete",
"quantity": 5000,
"unit": "CY",
"unit_cost": 150.00,
"amount": 750000.00
}
],
"subtotal": 750000.00,
"contingency_percent": 10,
"total": 825000.00
}
result = registry.validate(estimate_data, "CostEstimate")
print(result.to_report())
# Validate RFI
rfi_data = {
"rfi_number": "RFI-042",
"project_id": "PROJ-001",
"subject": "Concrete mix design clarification",
"status": "submitted",
"priority": "high",
"date_submitted": "2026-01-20",
"date_required": "2026-01-27",
"from_company": "ABC Concrete",
"to_company": "XYZ Architects",
"spec_section": "03 30 00",
"question": "Please clarify the required PSI for the foundation pour."
}
result = registry.validate(rfi_data, "RFI")
if result.is_valid:
print("RFI data is valid")
else:
print(result.to_report())
JSON Schema Export
def export_to_json_schema(schema: Schema) -> dict:
"""Export DDC schema to JSON Schema format."""
def field_to_json_schema(field: SchemaField) -> dict:
type_map = {
SchemaType.STRING: "string",
SchemaType.NUMBER: "number",
SchemaType.INTEGER: "integer",
SchemaType.BOOLEAN: "boolean",
SchemaType.ARRAY: "array",
SchemaType.OBJECT: "object",
SchemaType.DATE: "string",
SchemaType.DATETIME: "string",
SchemaType.CSI_CODE: "string",
SchemaType.CURRENCY: "number",
SchemaType.GUID: "string",
}
js = {"type": type_map.get(field.type, "string")}
if field.description:
js["description"] = field.description
if field.min_value is not None:
js["minimum"] = field.min_value
if field.max_value is not None:
js["maximum"] = field.max_value
if field.min_length is not None:
js["minLength"] = field.min_length
if field.max_length is not None:
js["maxLength"] = field.max_length
if field.pattern:
js["pattern"] = field.pattern
if field.enum_values:
js["enum"] = field.enum_values
if field.type == SchemaType.DATE:
js["format"] = "date"
if field.type == SchemaType.DATETIME:
js["format"] = "date-time"
if field.type == SchemaType.GUID:
js["format"] = "uuid"
return js
properties = {}
required = []
for name, field in schema.fields.items():
properties[name] = field_to_json_schema(field)
if field.required:
required.append(name)
return {
"$schema": "http://json-schema.org/draft-07/schema#",
"title": schema.name,
"description": schema.description,
"type": "object",
"properties": properties,
"required": required
}
# Export to JSON Schema
json_schema = export_to_json_schema(COST_ESTIMATE_SCHEMA)
print(json.dumps(json_schema, indent=2))
Integration with DDC Pipeline
# Validate API request before processing
def validate_api_request(endpoint: str, payload: dict) -> SchemaValidationResult:
schema_map = {
'/api/estimates': 'CostEstimate',
'/api/schedules': 'ProjectSchedule',
'/api/rfis': 'RFI',
'/api/bim/elements': 'BIMElement'
}
schema_name = schema_map.get(endpoint)
if not schema_name:
result = SchemaValidationResult(is_valid=True)
return result
registry = ConstructionSchemaRegistry()
return registry.validate(payload, schema_name)
# Use in API handler
@app.post('/api/estimates')
def create_estimate(payload: dict):
validation = validate_api_request('/api/estimates', payload)
if not validation.is_valid:
return {'error': 'Validation failed', 'details': [e.__dict__ for e in validation.errors]}, 400
# Process valid data
return process_estimate(payload)
Resources
- JSON Schema: https://json-schema.org/
- CSI MasterFormat: Standard classification codes
- IFC Schema: https://standards.buildingsmart.org/IFC/