truthound-dashboard 1.0.0__py3-none-any.whl

truthound_dashboard/schemas/schema.py

@@ -0,0 +1,121 @@
+"""Schema-related Pydantic schemas.
+
+This module defines schemas for learned schema API operations.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any
+
+from pydantic import Field
+
+from .base import BaseSchema, IDMixin, TimestampMixin
+
+
+class ColumnSchema(BaseSchema):
+    """Schema for a single column."""
+
+    name: str = Field(..., description="Column name")
+    dtype: str = Field(..., description="Data type")
+    nullable: bool = Field(default=True, description="Whether nulls are allowed")
+    unique: bool = Field(default=False, description="Whether values must be unique")
+    min_value: Any | None = Field(default=None, description="Minimum value constraint")
+    max_value: Any | None = Field(default=None, description="Maximum value constraint")
+    allowed_values: list[Any] | None = Field(
+        default=None,
+        description="List of allowed values",
+    )
+    pattern: str | None = Field(default=None, description="Regex pattern constraint")
+    min_length: int | None = Field(default=None, description="Minimum string length")
+    max_length: int | None = Field(default=None, description="Maximum string length")
+
+    # Statistics
+    null_ratio: float | None = Field(default=None, description="Ratio of null values")
+    unique_ratio: float | None = Field(
+        default=None, description="Ratio of unique values"
+    )
+    mean: float | None = Field(default=None, description="Mean value (numeric)")
+    std: float | None = Field(default=None, description="Standard deviation (numeric)")
+
+
+class SchemaLearnRequest(BaseSchema):
+    """Request to learn schema from source."""
+
+    infer_constraints: bool = Field(
+        default=True,
+        description="Infer constraints from data statistics",
+    )
+
+
+class SchemaUpdateRequest(BaseSchema):
+    """Request to update schema YAML."""
+
+    schema_yaml: str = Field(
+        ...,
+        min_length=1,
+        description="Updated schema in YAML format",
+    )
+
+
+class SchemaResponse(BaseSchema, IDMixin, TimestampMixin):
+    """Full schema response."""
+
+    source_id: str = Field(..., description="Parent source ID")
+    schema_yaml: str = Field(..., description="Schema in YAML format")
+    schema_json: dict[str, Any] | None = Field(
+        default=None,
+        description="Schema as JSON object",
+    )
+    row_count: int | None = Field(
+        default=None,
+        description="Row count when schema was learned",
+    )
+    column_count: int | None = Field(
+        default=None,
+        description="Number of columns in schema",
+    )
+    columns: list[str] = Field(
+        default_factory=list,
+        description="List of column names",
+    )
+    version: str | None = Field(default=None, description="Schema version")
+    is_active: bool = Field(default=True, description="Whether this schema is active")
+
+    @classmethod
+    def from_model(cls, schema: Any) -> SchemaResponse:
+        """Create response from model.
+
+        Args:
+            schema: Schema model instance.
+
+        Returns:
+            SchemaResponse instance.
+        """
+        columns = []
+        if schema.schema_json and "columns" in schema.schema_json:
+            columns = list(schema.schema_json["columns"].keys())
+
+        return cls(
+            id=schema.id,
+            source_id=schema.source_id,
+            schema_yaml=schema.schema_yaml,
+            schema_json=schema.schema_json,
+            row_count=schema.row_count,
+            column_count=schema.column_count,
+            columns=columns,
+            version=schema.version,
+            is_active=schema.is_active,
+            created_at=schema.created_at,
+            updated_at=schema.updated_at,
+        )
+
+
+class SchemaSummary(IDMixin):
+    """Minimal schema summary."""
+
+    source_id: str
+    column_count: int | None = None
+    row_count: int | None = None
+    is_active: bool = True
+    created_at: datetime

truthound_dashboard/schemas/source.py

@@ -0,0 +1,138 @@
+"""Source-related Pydantic schemas.
+
+This module defines schemas for data source API operations.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Literal
+
+from pydantic import Field, field_validator
+
+from .base import BaseSchema, IDMixin, ListResponseWrapper, TimestampMixin
+
+# Supported source types
+SourceType = Literal["file", "postgresql", "mysql", "snowflake", "bigquery"]
+
+
+class SourceBase(BaseSchema):
+    """Base source schema with common fields."""
+
+    name: str = Field(
+        ...,
+        min_length=1,
+        max_length=255,
+        description="Human-readable source name",
+        examples=["Sales Data", "User Events"],
+    )
+    type: SourceType = Field(
+        ...,
+        description="Data source type",
+        examples=["file", "postgresql"],
+    )
+    config: dict[str, Any] = Field(
+        ...,
+        description="Source-specific configuration",
+        examples=[{"path": "/data/sales.csv"}],
+    )
+    description: str | None = Field(
+        default=None,
+        max_length=1000,
+        description="Optional source description",
+    )
+
+
+class SourceCreate(SourceBase):
+    """Schema for creating a new source."""
+
+    @field_validator("config")
+    @classmethod
+    def validate_config(cls, v: dict[str, Any]) -> dict[str, Any]:
+        """Validate configuration has required fields."""
+        if not v:
+            raise ValueError("Config cannot be empty")
+        return v
+
+
+class SourceUpdate(BaseSchema):
+    """Schema for updating an existing source."""
+
+    name: str | None = Field(
+        default=None,
+        min_length=1,
+        max_length=255,
+        description="New source name",
+    )
+    config: dict[str, Any] | None = Field(
+        default=None,
+        description="New source configuration",
+    )
+    description: str | None = Field(
+        default=None,
+        max_length=1000,
+        description="New description",
+    )
+    is_active: bool | None = Field(
+        default=None,
+        description="Whether source is active",
+    )
+
+
+class SourceResponse(SourceBase, IDMixin, TimestampMixin):
+    """Schema for source responses."""
+
+    is_active: bool = Field(default=True, description="Whether source is active")
+    last_validated_at: datetime | None = Field(
+        default=None,
+        description="Last validation timestamp",
+    )
+
+    # Computed properties from relationships
+    has_schema: bool = Field(default=False, description="Whether schema exists")
+    latest_validation_status: str | None = Field(
+        default=None,
+        description="Status of latest validation",
+    )
+
+    @classmethod
+    def from_model(cls, source: Any) -> SourceResponse:
+        """Create response from model with computed fields.
+
+        Args:
+            source: Source model instance.
+
+        Returns:
+            SourceResponse with computed fields.
+        """
+        return cls(
+            id=source.id,
+            name=source.name,
+            type=source.type,
+            config=source.config,
+            description=source.description,
+            is_active=source.is_active,
+            created_at=source.created_at,
+            updated_at=source.updated_at,
+            last_validated_at=source.last_validated_at,
+            has_schema=source.latest_schema is not None,
+            latest_validation_status=(
+                source.latest_validation.status if source.latest_validation else None
+            ),
+        )
+
+
+class SourceListResponse(ListResponseWrapper[SourceResponse]):
+    """Paginated source list response."""
+
+    pass
+
+
+class SourceSummary(BaseSchema):
+    """Minimal source summary for lists."""
+
+    id: str
+    name: str
+    type: SourceType
+    is_active: bool
+    last_validated_at: datetime | None = None

truthound_dashboard/schemas/validation.py

@@ -0,0 +1,192 @@
+"""Validation-related Pydantic schemas.
+
+This module defines schemas for validation API operations.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any, Literal
+
+from pydantic import Field
+
+from .base import BaseSchema, IDMixin, ListResponseWrapper
+
+# Validation status types
+ValidationStatus = Literal["pending", "running", "success", "failed", "error"]
+
+# Issue severity types
+IssueSeverity = Literal["critical", "high", "medium", "low"]
+
+
+class ValidationIssue(BaseSchema):
+    """Single validation issue.
+
+    Represents one issue found during validation, mapping to
+    truthound's ValidationIssue dataclass.
+    """
+
+    column: str = Field(..., description="Column where issue was found")
+    issue_type: str = Field(
+        ..., description="Type of issue", examples=["null_values", "type_mismatch"]
+    )
+    count: int = Field(..., ge=0, description="Number of occurrences")
+    severity: IssueSeverity = Field(..., description="Issue severity level")
+    details: str | None = Field(default=None, description="Detailed description")
+    expected: Any | None = Field(default=None, description="Expected value/type")
+    actual: Any | None = Field(default=None, description="Actual value/type found")
+    sample_values: list[Any] | None = Field(
+        default=None,
+        description="Sample problematic values",
+    )
+
+
+class ValidationRunRequest(BaseSchema):
+    """Request to run validation on a source."""
+
+    validators: list[str] | None = Field(
+        default=None,
+        description="Specific validators to run. If None, all validators are used.",
+        examples=[["null", "duplicate", "schema"]],
+    )
+    schema_path: str | None = Field(
+        default=None,
+        description="Path to schema YAML file for schema validation",
+    )
+    auto_schema: bool = Field(
+        default=False,
+        description="Auto-learn and cache schema for validation",
+    )
+
+
+class ValidationSummary(BaseSchema):
+    """Summary statistics for a validation run."""
+
+    passed: bool = Field(..., description="Whether validation passed")
+    has_critical: bool = Field(default=False, description="Has critical issues")
+    has_high: bool = Field(default=False, description="Has high severity issues")
+    total_issues: int = Field(default=0, ge=0, description="Total issue count")
+    critical_issues: int = Field(default=0, ge=0, description="Critical issue count")
+    high_issues: int = Field(default=0, ge=0, description="High severity issue count")
+    medium_issues: int = Field(
+        default=0, ge=0, description="Medium severity issue count"
+    )
+    low_issues: int = Field(default=0, ge=0, description="Low severity issue count")
+
+
+class ValidationResponse(IDMixin, ValidationSummary):
+    """Full validation response with all details."""
+
+    source_id: str = Field(..., description="Source that was validated")
+    status: ValidationStatus = Field(..., description="Current validation status")
+
+    # Data statistics
+    row_count: int | None = Field(default=None, description="Number of rows validated")
+    column_count: int | None = Field(default=None, description="Number of columns")
+
+    # Issues list (full details)
+    issues: list[ValidationIssue] = Field(
+        default_factory=list,
+        description="List of validation issues",
+    )
+
+    # Error info (if status is 'error')
+    error_message: str | None = Field(
+        default=None,
+        description="Error message if validation failed",
+    )
+
+    # Timing
+    duration_ms: int | None = Field(
+        default=None,
+        ge=0,
+        description="Validation duration in milliseconds",
+    )
+    started_at: datetime | None = Field(default=None, description="Start timestamp")
+    completed_at: datetime | None = Field(
+        default=None, description="Completion timestamp"
+    )
+    created_at: datetime = Field(..., description="Record creation timestamp")
+
+    @classmethod
+    def from_model(cls, validation: Any) -> ValidationResponse:
+        """Create response from model.
+
+        Args:
+            validation: Validation model instance.
+
+        Returns:
+            ValidationResponse instance.
+        """
+        issues = []
+        if validation.result_json and "issues" in validation.result_json:
+            issues = [
+                ValidationIssue(**issue) for issue in validation.result_json["issues"]
+            ]
+
+        return cls(
+            id=validation.id,
+            source_id=validation.source_id,
+            status=validation.status,
+            passed=validation.passed or False,
+            has_critical=validation.has_critical or False,
+            has_high=validation.has_high or False,
+            total_issues=validation.total_issues or 0,
+            critical_issues=validation.critical_issues or 0,
+            high_issues=validation.high_issues or 0,
+            medium_issues=validation.medium_issues or 0,
+            low_issues=validation.low_issues or 0,
+            row_count=validation.row_count,
+            column_count=validation.column_count,
+            issues=issues,
+            error_message=validation.error_message,
+            duration_ms=validation.duration_ms,
+            started_at=validation.started_at,
+            completed_at=validation.completed_at,
+            created_at=validation.created_at,
+        )
+
+
+class ValidationListItem(IDMixin, ValidationSummary):
+    """Validation list item (without full issues)."""
+
+    source_id: str
+    status: ValidationStatus
+    row_count: int | None = None
+    column_count: int | None = None
+    duration_ms: int | None = None
+    created_at: datetime
+
+    @classmethod
+    def from_model(cls, validation: Any) -> ValidationListItem:
+        """Create list item from model.
+
+        Args:
+            validation: Validation model instance.
+
+        Returns:
+            ValidationListItem instance.
+        """
+        return cls(
+            id=validation.id,
+            source_id=validation.source_id,
+            status=validation.status,
+            passed=validation.passed or False,
+            has_critical=validation.has_critical or False,
+            has_high=validation.has_high or False,
+            total_issues=validation.total_issues or 0,
+            critical_issues=validation.critical_issues or 0,
+            high_issues=validation.high_issues or 0,
+            medium_issues=validation.medium_issues or 0,
+            low_issues=validation.low_issues or 0,
+            row_count=validation.row_count,
+            column_count=validation.column_count,
+            duration_ms=validation.duration_ms,
+            created_at=validation.created_at,
+        )
+
+
+class ValidationListResponse(ListResponseWrapper[ValidationListItem]):
+    """Paginated validation list response."""
+
+    pass