"""Pydantic models for tool input parameters."""
from __future__ import annotations
from pydantic import BaseModel, Field
from law_scrapper_mcp.models.enums import (
DetailLevel,
MetadataCategory,
Publisher,
RelationshipType,
)
class SearchRequest(BaseModel):
"""Parameters for searching legal acts."""
publisher: Publisher = Field(
default=Publisher.DU, description="Publisher code (DU or MP)"
)
year: int | None = Field(default=None, description="Year of publication")
keywords: list[str] | None = Field(
default=None,
description="Keywords to search for (AND logic - all must be present)",
)
date_from: str | None = Field(
default=None, description="Start date for date range filter (YYYY-MM-DD)"
)
date_to: str | None = Field(
default=None, description="End date for date range filter (YYYY-MM-DD)"
)
title: str | None = Field(default=None, description="Title search phrase")
act_type: str | None = Field(default=None, description="Type of legal act")
pub_date_from: str | None = Field(
default=None, description="Publication date from (YYYY-MM-DD)"
)
pub_date_to: str | None = Field(
default=None, description="Publication date to (YYYY-MM-DD)"
)
in_force: bool | None = Field(
default=None, description="Filter by acts currently in force"
)
limit: int | None = Field(default=None, description="Maximum number of results")
offset: int | None = Field(default=None, description="Number of results to skip")
detail_level: DetailLevel = Field(
default=DetailLevel.STANDARD, description="Level of detail in results"
)
class BrowseRequest(BaseModel):
"""Parameters for browsing acts by publisher and year."""
publisher: Publisher = Field(description="Publisher code (DU or MP)")
year: int = Field(description="Year to browse")
detail_level: DetailLevel = Field(
default=DetailLevel.STANDARD, description="Level of detail in results"
)
class ActDetailsRequest(BaseModel):
"""Parameters for retrieving act details."""
eli: str = Field(
description="ELI identifier of the act (e.g., 'DU/2024/1' or full URL)"
)
load_content: bool = Field(
default=False,
description="Whether to load full content (HTML/PDF) into document store",
)
class ReadContentRequest(BaseModel):
"""Parameters for reading act content."""
eli: str = Field(description="ELI identifier of the act")
section: str | None = Field(
default=None,
description="Section ID to read (if None, returns table of contents)",
)
class SearchInActRequest(BaseModel):
"""Parameters for searching within an act."""
eli: str = Field(description="ELI identifier of the act")
query: str = Field(description="Search query text")
context_chars: int = Field(
default=500, description="Number of characters of context around matches"
)
class MetadataRequest(BaseModel):
"""Parameters for retrieving metadata."""
category: MetadataCategory = Field(
default=MetadataCategory.ALL, description="Category of metadata to retrieve"
)
class RelationshipsRequest(BaseModel):
"""Parameters for retrieving act relationships."""
eli: str = Field(description="ELI identifier of the act")
relationship_type: RelationshipType | None = Field(
default=None, description="Specific relationship type to retrieve"
)
class TrackChangesRequest(BaseModel):
"""Parameters for tracking changes to legal acts."""
publisher: Publisher = Field(
default=Publisher.DU, description="Publisher code (DU or MP)"
)
date_from: str = Field(description="Start date for tracking (YYYY-MM-DD)")
date_to: str | None = Field(
default=None, description="End date for tracking (YYYY-MM-DD, default: today)"
)
keywords: list[str] | None = Field(
default=None, description="Filter by keywords"
)
class DateCalculationRequest(BaseModel):
"""Parameters for date calculation."""
days: int = Field(default=0, description="Number of days to add (positive=future, negative=past)")
months: int = Field(default=0, description="Number of months to add (positive=future, negative=past)")
years: int = Field(default=0, description="Number of years to add (positive=future, negative=past)")
base_date: str | None = Field(
default=None, description="Base date for calculation (YYYY-MM-DD, default: today)"
)
def parse_eli(eli: str) -> tuple[str, int, int]:
"""Parse ELI string into publisher, year, and position.
Args:
eli: ELI identifier (e.g., "DU/2024/1" or full URL)
Returns:
Tuple of (publisher, year, pos)
Raises:
ValueError: If ELI format is invalid
"""
# Strip URL prefix if present
if "api.sejm.gov.pl/eli/" in eli:
eli = eli.split("api.sejm.gov.pl/eli/")[1]
elif eli.startswith("http"):
raise ValueError(f"Invalid ELI URL format: {eli}")
# Remove trailing slashes and split
eli = eli.rstrip("/")
parts = eli.split("/")
if len(parts) != 3:
raise ValueError(
f"Invalid ELI format: {eli}. Expected format: PUBLISHER/YEAR/POS"
)
publisher, year_str, pos_str = parts
try:
year = int(year_str)
pos = int(pos_str)
except ValueError as e:
raise ValueError(f"Invalid year or position in ELI: {eli}") from e
return publisher, year, pos
