kensho-kfinance 1.2.0__py3-none-any.whl → 2.0.0__py3-none-any.whl

{kensho_kfinance-1.2.0.dist-info → kensho_kfinance-2.0.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: kensho-kfinance
-Version: 1.2.0
+Version: 2.0.0
 Summary: Python CLI for kFinance
 Author-email: Luke Brown <luke.brown@kensho.com>, Michelle Keoy <michelle.keoy@kensho.com>, Keith Page <keith.page@kensho.com>, Matthew Rosen <matthew.rosen@kensho.com>, Nick Roshdieh <nick.roshdieh@kensho.com>
 Project-URL: source, https://github.com/kensho-technologies/kfinance
@@ -12,30 +12,34 @@ Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
 License-File: AUTHORS.md
-Requires-Dist: python-dateutil<2.9,>=2.8.2
-Requires-Dist: requests<3,>=2.22.0
-Requires-Dist: urllib3>=1.21.1
-Requires-Dist: pyjwt>=2.8.0
+Requires-Dist: langchain-core>=0.3.15
+Requires-Dist: langchain-google-genai<3,>=2.1.0
 Requires-Dist: numpy>=1.22.4
 Requires-Dist: pandas>=2.0.0
-Requires-Dist: types-requests<3,>=2.22.0
 Requires-Dist: pillow>=10
-Requires-Dist: langchain-core>=0.3.15
+Requires-Dist: pydantic<3,>=2.10.0
+Requires-Dist: pyjwt>=2.8.0
+Requires-Dist: python-dateutil<2.9,>=2.8.2
 Requires-Dist: strenum>=0.4.15
+Requires-Dist: tabulate>=0.9.0
+Requires-Dist: types-requests<3,>=2.22.0
+Requires-Dist: requests<3,>=2.22.0
+Requires-Dist: urllib3>=1.21.1
 Provides-Extra: dev
 Requires-Dist: coverage<8,>=7.6.10; extra == "dev"
 Requires-Dist: mypy<2,>=1.15.0; extra == "dev"
 Requires-Dist: pytest<7,>=6.1.2; extra == "dev"
 Requires-Dist: pytest-cov<7,>=6.0.0; extra == "dev"
+Requires-Dist: requests_mock<2,>=1.12; extra == "dev"
 Requires-Dist: ruff<1,>=0.9.4; extra == "dev"
-Requires-Dist: requests_mock<1.2,>=1.1; extra == "dev"
+Requires-Dist: time_machine<3,>=2.1; extra == "dev"
 Dynamic: license-file
 
 # kFinance
 
 The kFinance Python library provides a simple interface for the LLM-ready API, streamlining API requests and response handling. It can be used on its own, with LLMs, or integrated into applications.
 
-For a complete overview of the functions, usage, and features of the kFinance Python library, please refer to documentation [here](https://kfinance.kensho.com/docs).
+For a complete overview of the functions, usage, and features of the kFinance Python library, please refer to documentation [here](https://kensho-kfinance.readthedocs.io/en/stable/).
 
 Any questions or suggestions can be sent to the [kFinance Maintainers](kfinance-maintainers@kensho.com).
 

kensho_kfinance-2.0.0.dist-info/RECORD

@@ -0,0 +1,40 @@
+kensho_kfinance-2.0.0.dist-info/licenses/AUTHORS.md,sha256=0h9ClbI0pu1oKj1M28ROUsaxrbZg-6ukQGl6X4y9noI,68
+kensho_kfinance-2.0.0.dist-info/licenses/LICENSE,sha256=bsY4blvSgq6o0FMQ3RXa2NCgco--nHCCchLXzxr6kms,83
+kfinance/CHANGELOG.md,sha256=Gjau3iwqbJoOw-ReXpn9H3mkRhsWNTttBljc7D2-i4Y,641
+kfinance/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+kfinance/batch_request_handling.py,sha256=s0uXRY4CC-GnShI0i8PLPgLddRdqLyQ8jnqDKdmaZzs,5531
+kfinance/constants.py,sha256=HrTh2QuhbA70xxcJ6N8WfTOJs6YiM30P59GbSDu_3AY,48312
+kfinance/fetch.py,sha256=ygvp8q1o8mqPYY4ctQwch37dmprO0UMDteDumk8A-Gw,20172
+kfinance/kfinance.py,sha256=nw2SPkixFht3PkcH4tZ1FnLhVxmyZm1_gbOqnmC6PBQ,49219
+kfinance/meta_classes.py,sha256=sNezG7uHAQOWj2Xdjao6qc3z_gHppgJat7O9J2k_yqI,15730
+kfinance/prompt.py,sha256=PtVB8c_FcSlVdyGgByAnIFGzuUuBaEjciCqnBJl1hSQ,25133
+kfinance/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+kfinance/server_thread.py,sha256=jUnt1YGoYDkqqz1MbCwd44zJs1T_Z2BCgvj75bdtLgA,2574
+kfinance/version.py,sha256=2thmcF9DS_Zp1zHI3N0kjBeMAuCP9mdDGnL0clqQpS8,511
+kfinance/tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+kfinance/tests/test_batch_requests.py,sha256=NsldoZqjjQpv0QKM2th80F1nru2tIF4LJeXU-RamQvA,9836
+kfinance/tests/test_fetch.py,sha256=iJLOPBOb8xdAGF6Jel5jPhWb4iW6yPpt9f5tW4uZEyo,12189
+kfinance/tests/test_objects.py,sha256=egXkhfoK2MepZdXtrBxzWxWni7f-zVCefUbnyDnWDOE,22554
+kfinance/tests/test_tools.py,sha256=YG5UVWnR4EyHGsL4bBwUp2auSd55lcvUEEncXcMXw1Q,16941
+kfinance/tool_calling/README.md,sha256=omJq7Us6r4U45QB7hRpLjRJ5BMalCkZkh4uXBjTbJXc,2022
+kfinance/tool_calling/__init__.py,sha256=GDEeK2zM2tOfYam6ejwYACXecoRdlkRve_ZC5VTKugo,2038
+kfinance/tool_calling/get_business_relationship_from_identifier.py,sha256=S_-nCsppi9tvwC81GMKvJd-eZSwRejT96PZhlOf5Wv4,1381
+kfinance/tool_calling/get_capitalization_from_identifier.py,sha256=LmH5yEWSRJfAP4kRYHREdbMtrmxcZfgfbNB4dGQwyb4,1435
+kfinance/tool_calling/get_company_id_from_identifier.py,sha256=18XAmWxTXRw1qaqw4WuA-yQ4d7kI2HnbMi32d-xNGzE,485
+kfinance/tool_calling/get_cusip_from_ticker.py,sha256=uzVMbtcV4eKtWO4bYn-cicCo8tfAEGk-v95ZOON_cbA,533
+kfinance/tool_calling/get_earnings_call_datetimes_from_identifier.py,sha256=3v73ZZz9n-Rdp5vp1RPmFD41nJ0Z3EgTjFSbnm30xdA,665
+kfinance/tool_calling/get_financial_line_item_from_identifier.py,sha256=zJ-kcr_B9N3ZIXK1W-2TEUlF6tj4rhzmLSsW8EUhJn4,2064
+kfinance/tool_calling/get_financial_statement_from_identifier.py,sha256=LyGmh81-_yljOhMVpsXa1M4u6hNJNP69Ms1M1NsxwPM,1817
+kfinance/tool_calling/get_history_metadata_from_identifier.py,sha256=0h7xdhrIJg3pJJrrCgcsiZd3ft1GpConhNZe2pcOREA,666
+kfinance/tool_calling/get_info_from_identifier.py,sha256=bv4lRhXWG9N7U8XEli9v44QEXNgf7-ZTuTDmE6XIeEg,659
+kfinance/tool_calling/get_isin_from_ticker.py,sha256=yz7SOCFgwL03Zd8zG3pIfa3McB_tdRU-_oXYfSfqWoo,527
+kfinance/tool_calling/get_latest.py,sha256=U_5SuGlD0G-u7YreiNkwEtx2zkUhZzUvwUwdL_tmqFk,724
+kfinance/tool_calling/get_n_quarters_ago.py,sha256=VVKKyUNbmw7fS7e-vi4YXSdd_OO2bo8k50Grt5F_Mvo,651
+kfinance/tool_calling/get_prices_from_identifier.py,sha256=b-cK9dWXbVBjppTORlC4HK9zd_qe8V0n98s823v0dCc,1796
+kfinance/tool_calling/get_security_id_from_identifier.py,sha256=vPFYuFydUGxQ0b23CWkU3YxWVeoW-nzsrPbB9ZH8U_w,489
+kfinance/tool_calling/get_trading_item_id_from_identifier.py,sha256=fB8QBFzhn-kgUL8dSF6w_rtO3zG5yLbm5s29Xz6U1AY,504
+kfinance/tool_calling/shared_models.py,sha256=6y74yJCORHDVBxhhBJhTUOWCF9D11Lgg9eYQeDIT3aY,2047
+kensho_kfinance-2.0.0.dist-info/METADATA,sha256=g5E_LG0HZWOwX6V8Bd7OE4Eeu7HoD69scUh5QqH3hSs,3051
+kensho_kfinance-2.0.0.dist-info/WHEEL,sha256=SmOxYU7pzNKBqASvQJ7DjX3XGUF92lrGhMb3R6_iiqI,91
+kensho_kfinance-2.0.0.dist-info/top_level.txt,sha256=kT_kNwVhfQoOAecY8W7uYah5xaHMoHoAdBIvXh6DaKM,9
+kensho_kfinance-2.0.0.dist-info/RECORD,,

{kensho_kfinance-1.2.0.dist-info → kensho_kfinance-2.0.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (78.1.0)
+Generator: setuptools (79.0.1)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

kfinance/CHANGELOG.md

@@ -1,4 +1,14 @@
 # Changelog
+
+## v2.0.0
+- Refactor llm tools to use langchain `BaseTool`.
+
+## v1.2.2
+- Add tabulate and pydantic as dependencies
+
+## v1.2.1
+- Add fetch for industry_code, industry_classification and gics_code.
+
 ## v1.2.0
 - Add batch requests for iterable classes
 

kfinance/constants.py

@@ -1,4 +1,5 @@
 from datetime import date
+from enum import Enum
 from itertools import chain
 from typing import TypedDict
 
@@ -26,7 +27,45 @@ class IdentificationTriple(TypedDict):
     company_id: int
 
 
+class Capitalization(Enum):
+    """The capitalization type"""
+
+    market_cap = "market_cap"
+    tev = "tev"
+    shares_outstanding = "shares_outstanding"
+
+
+class PeriodType(Enum):
+    """The period type"""
+
+    annual = "annual"
+    quarterly = "quarterly"
+    ltm = "ltm"
+    ytd = "ytd"
+
+
+class Periodicity(Enum):
+    """The frequency or interval at which the historical data points are sampled or aggregated. Periodicity is not the same as the date range. The date range specifies the time span over which the data is retrieved, while periodicity determines how the data within that date range is aggregated."""
+
+    day = "day"
+    week = "week"
+    month = "month"
+    year = "year"
+
+
+class StatementType(Enum):
+    """The type of financial statement"""
+
+    balance_sheet = "balance_sheet"
+    bs = "balance_sheet"
+    income_statement = "income_statement"
+    cashflow = "cashflow"
+    cf = "cashflow"
+
+
 class BusinessRelationshipType(StrEnum):
+    """The type of business relationship"""
+
     supplier = "supplier"
     customer = "customer"
     distributor = "distributor"
@@ -76,6 +115,17 @@ class LatestPeriods(TypedDict):
     now: CurrentPeriod
 
 
+class IndustryClassification(StrEnum):
+    sic = "sic"
+    naics = "naics"
+    nace = "nace"
+    anzsic = "anzsic"
+    spcapiqetf = "spcapiqetf"
+    spratings = "spratings"
+    gics = "gics"
+    simple = "simple"
+
+
 # all of these values must be lower case keys
 LINE_ITEMS: list[LineItemType] = [
     {
@@ -1661,18 +1711,11 @@ LINE_ITEMS: list[LineItemType] = [
     },
 ]
 
-LINE_ITEMS_TO_DATA_ITEM_ID = {
-    line_item["name"]: line_item["dataitemid"] for line_item in LINE_ITEMS
-}
 
-LINE_ITEM_NAMES_AND_ALIASES = list(
+LINE_ITEM_NAMES_AND_ALIASES: list[str] = list(
     chain(*[[line_item["name"]] + list(line_item["aliases"]) for line_item in LINE_ITEMS])
 )
 
-LINE_ITEM_SYNONYMS = {
-    alias: line_item["dataitemid"] for line_item in LINE_ITEMS for alias in line_item["aliases"]
-}
-
 
 FINANCIAL_STATEMENTS: dict[str, int] = {
     "income_statement": 1,

kfinance/fetch.py

@@ -7,7 +7,13 @@ from uuid import uuid4
 import jwt
 import requests
 
-from .constants import BusinessRelationshipType, IdentificationTriple
+from .constants import (
+    BusinessRelationshipType,
+    IdentificationTriple,
+    IndustryClassification,
+    Periodicity,
+    PeriodType,
+)
 
 
 # version.py gets autogenerated by setuptools-scm and is not available
@@ -228,14 +234,14 @@ class KFinanceApiClient:
         is_adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
-        periodicity: Optional[str] = None,
+        periodicity: Optional[Periodicity] = None,
     ) -> dict:
         """Get the pricing history."""
         url = (
             f"{self.url_base}pricing/{trading_item_id}/"
             f"{start_date if start_date is not None else 'none'}/"
             f"{end_date if end_date is not None else 'none'}/"
-            f"{periodicity if periodicity is not None else 'none'}/"
+            f"{periodicity.value if periodicity else 'none'}/"
             f"{'adjusted' if is_adjusted else 'unadjusted'}"
         )
         return self.fetch(url)
@@ -265,14 +271,14 @@ class KFinanceApiClient:
         is_adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
-        periodicity: Optional[str] = "",
+        periodicity: Optional[Periodicity] = None,
     ) -> bytes:
         """Get the price chart."""
         url = (
             f"{self.url_base}price_chart/{trading_item_id}/"
             f"{start_date if start_date is not None else 'none'}/"
             f"{end_date if end_date is not None else 'none'}/"
-            f"{periodicity if periodicity is not None else 'none'}/"
+            f"{periodicity.value if periodicity else 'none'}/"
             f"{'adjusted' if is_adjusted else 'unadjusted'}"
         )
 
@@ -291,7 +297,7 @@ class KFinanceApiClient:
         self,
         company_id: int,
         statement_type: str,
-        period_type: Optional[str] = None,
+        period_type: Optional[PeriodType] = None,
         start_year: Optional[int] = None,
         end_year: Optional[int] = None,
         start_quarter: Optional[int] = None,
@@ -300,7 +306,7 @@ class KFinanceApiClient:
         """Get a specified financial statement for a specified duration."""
         url = (
             f"{self.url_base}statements/{company_id}/{statement_type}/"
-            f"{period_type if period_type is not None else 'none'}/"
+            f"{period_type.value if period_type else 'none'}/"
             f"{start_year if start_year is not None else 'none'}/"
             f"{end_year if end_year is not None else 'none'}/"
             f"{start_quarter if start_quarter is not None else 'none'}/"
@@ -312,7 +318,7 @@ class KFinanceApiClient:
         self,
         company_id: int,
         line_item: str,
-        period_type: Optional[str] = None,
+        period_type: Optional[PeriodType] = None,
         start_year: Optional[int] = None,
         end_year: Optional[int] = None,
         start_quarter: Optional[int] = None,
@@ -321,7 +327,7 @@ class KFinanceApiClient:
         """Get a specified financial line item for a specified duration."""
         url = (
             f"{self.url_base}line_item/{company_id}/{line_item}/"
-            f"{period_type if period_type is not None else 'none'}/"
+            f"{period_type.value if period_type else 'none'}/"
             f"{start_year if start_year is not None else 'none'}/"
             f"{end_year if end_year is not None else 'none'}/"
             f"{start_quarter if start_quarter is not None else 'none'}/"
@@ -368,26 +374,6 @@ class KFinanceApiClient:
             country_iso_code=country_iso_code, state_iso_code=state_iso_code, fetch_ticker=False
         )
 
-    def fetch_simple_industry_groups(
-        self, simple_industry: str, fetch_ticker: bool = True
-    ) -> dict[str, list]:
-        """Fetch simple industry groups"""
-        url = f"{self.url_base}{'ticker_groups' if fetch_ticker else 'company_groups'}/industry/simple/{simple_industry}"
-        return self.fetch(url)
-
-    def fetch_ticker_simple_industry_groups(
-        self, simple_industry: str
-    ) -> dict[str, list[IdentificationTriple]]:
-        """Fetch ticker simple industry groups"""
-        return self.fetch_simple_industry_groups(simple_industry=simple_industry, fetch_ticker=True)
-
-    def fetch_company_simple_industry_groups(self, simple_industry: str) -> dict[str, list[int]]:
-        """Fetch company simple industry groups"""
-        return self.fetch_simple_industry_groups(
-            simple_industry=simple_industry,
-            fetch_ticker=False,
-        )
-
     def fetch_exchange_groups(
         self, exchange_code: str, fetch_ticker: bool = True
     ) -> dict[str, list]:
@@ -456,3 +442,56 @@ class KFinanceApiClient:
         """
         url = f"{self.url_base}relationship/{company_id}/{relationship_type}"
         return self.fetch(url)
+
+    def fetch_from_industry_code(
+        self,
+        industry_code: str,
+        industry_classification: IndustryClassification,
+        fetch_ticker: bool = True,
+    ) -> dict[str, list]:
+        """Fetches a list of companies or identification triples that are classified in the given industry_code and industry_classification."""
+
+        url = f"{self.url_base}{'ticker_groups' if fetch_ticker else 'company_groups'}/industry/{industry_classification}/{industry_code}"
+        return self.fetch(url)
+
+    def fetch_ticker_from_industry_code(
+        self,
+        industry_code: str,
+        industry_classification: IndustryClassification,
+    ) -> dict[str, list[IdentificationTriple]]:
+        """Fetches a list of identification triples that are classified in the given industry_code and industry_classification.
+
+        Returns a dictionary of shape {"tickers": List[{“company_id”: <company_id>, “security_id”: <security_id>, “trading_item_id”: <trading_item_id>}]}.
+        :param industry_code: The industry_code to filter on. The industry_code is a string corresponding to the Industry classifications ontology.
+        :type industry_code: str
+        :param industry_classification: The type of industry_classification to filter on.
+        :type industry_classification: IndustryClassification
+        :return: A dictionary containing the list of identification triple [company_id, security_id, trading_item_id] that are classified in the given industry_code and industry_classification.
+        :rtype: dict[str, list[IdentificationTriple]]
+        """
+        return self.fetch_from_industry_code(
+            industry_code=industry_code,
+            industry_classification=industry_classification,
+            fetch_ticker=True,
+        )
+
+    def fetch_company_from_industry_code(
+        self,
+        industry_code: str,
+        industry_classification: IndustryClassification,
+    ) -> dict[str, list[int]]:
+        """Fetches a list of companies that are classified in the given industry_code and industry_classification.
+
+        Returns a dictionary of shape {"companies": List[<company_id>]}.
+        :param industry_code: The industry_code to filter on. The industry_code is a string corresponding to the Industry classifications ontology.
+        :type industry_code: str
+        :param industry_classification: The type of industry_classification to filter on.
+        :type industry_classification: IndustryClassification
+        :return: A dictionary containing the list of companies that are classified in the given industry_code and industry_classification.
+        :rtype: dict[str, list[int]]
+        """
+        return self.fetch_from_industry_code(
+            industry_code=industry_code,
+            industry_classification=industry_classification,
+            fetch_ticker=False,
+        )

kfinance/kfinance.py

@@ -7,16 +7,25 @@ from io import BytesIO
 import logging
 import re
 from sys import stdout
-from typing import Iterable, NamedTuple, Optional
+from typing import TYPE_CHECKING, Any, Callable, Iterable, NamedTuple, Optional
 from urllib.parse import urljoin
 import webbrowser
 
+import google.ai.generativelanguage_v1beta.types as gapic
+from langchain_core.utils.function_calling import convert_to_openai_tool
+from langchain_google_genai._function_utils import convert_to_genai_function_declarations
 import numpy as np
 import pandas as pd
 from PIL.Image import Image, open as image_open
 
 from .batch_request_handling import add_methods_of_singular_class_to_iterable_class
-from .constants import HistoryMetadata, IdentificationTriple, LatestPeriods, YearAndQuarter
+from .constants import (
+    HistoryMetadata,
+    IdentificationTriple,
+    LatestPeriods,
+    Periodicity,
+    YearAndQuarter,
+)
 from .fetch import (
     DEFAULT_API_HOST,
     DEFAULT_API_VERSION,
@@ -24,15 +33,6 @@ from .fetch import (
     DEFAULT_OKTA_HOST,
     KFinanceApiClient,
 )
-from .llm_tools import (
-    _llm_tools,
-    anthropic_tool_descriptions,
-    gemini_tool_descriptions,
-    get_latest,
-    get_n_quarters_ago,
-    langchain_tools,
-    openai_tool_descriptions,
-)
 from .meta_classes import (
     CompanyFunctionsMetaClass,
     DelegatedCompanyFunctionsMetaClass,
@@ -41,6 +41,9 @@ from .prompt import PROMPT
 from .server_thread import ServerThread
 
 
+if TYPE_CHECKING:
+    from kfinance.tool_calling.shared_models import KfinanceTool
+
 logger = logging.getLogger(__name__)
 
 
@@ -124,23 +127,20 @@ class TradingItem:
 
     def history(
         self,
-        periodicity: str = "day",
+        periodicity: Periodicity = Periodicity.day,
         adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
     ) -> pd.DataFrame:
         """Retrieves the historical price data for a given asset over a specified date range.
 
-        :param str periodicity: Determines the frequency of the historical data returned. Options are "day", "week", "month" and "year". This default to "day"
+        :param periodicity: Determines the frequency of the historical data returned. Defaults to Periodicity.day.
         :param Optional[bool] adjusted: Whether to retrieve adjusted prices that account for corporate actions such as dividends and splits, it defaults True
         :param Optional[str] start_date: The start date for historical price retrieval in format "YYYY-MM-DD", default to None
         :param Optional[str] end_date: The end date for historical price retrieval in format "YYYY-MM-DD", default to None
-        :return: A pd.DataFrame containing historical price data with columns corresponding to the specified periodicity, with Date as the index, and columns "open", "high", "low", "close", "volume" in type decimal. The Date index is a string that depends on the periodicity. If periodicity="day", the Date index is the day in format "YYYY-MM-DD", eg "2024-05-13" If periodicity="week", the Date index is the week number of the year in format "YYYY Week ##", eg "2024 Week 2" If periodicity="month", the Date index is the month name of the year in format "<Month> YYYY", eg "January 2024". If periodicity="year", the Date index is the year in format "YYYY", eg "2024".
+        :return: A pd.DataFrame containing historical price data with columns corresponding to the specified periodicity, with Date as the index, and columns "open", "high", "low", "close", "volume" in type decimal. The Date index is a string that depends on the periodicity. If Periodicity.day, the Date index is the day in format "YYYY-MM-DD", eg "2024-05-13" If Periodicity.week, the Date index is the week number of the year in format "YYYY Week ##", eg "2024 Week 2" If Periodicity.month, the Date index is the month name of the year in format "<Month> YYYY", eg "January 2024". If Periodicity.year, the Date index is the year in format "YYYY", eg "2024".
         :rtype: pd.DataFrame
         """
-        if periodicity not in {"day", "week", "month", "year"}:
-            raise RuntimeError(f"Periodicity type {periodicity} is not valid.")
-
         if start_date and end_date:
             if (
                 datetime.strptime(start_date, "%Y-%m-%d").date()
@@ -165,14 +165,14 @@ class TradingItem:
 
     def price_chart(
         self,
-        periodicity: str = "day",
+        periodicity: Periodicity = Periodicity.day,
         adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
     ) -> Image:
         """Get the price chart.
 
-        :param str periodicity: Determines the frequency of the historical data returned. Options are "day", "week", "month" and "year". This default to "day"
+        :param str periodicity: Determines the frequency of the historical data returned. Defaults to Periodicity.day.
         :param Optional[bool] adjusted: Whether to retrieve adjusted prices that account for corporate actions such as dividends and splits, it defaults True
         :param Optional[str] start_date: The start date for historical price retrieval in format "YYYY-MM-DD", default to None
         :param Optional[str] end_date: The end date for historical price retrieval in format "YYYY-MM-DD", default to None
@@ -180,9 +180,6 @@ class TradingItem:
         :rtype: Image
         """
 
-        if periodicity not in {"day", "week", "month", "year"}:
-            raise RuntimeError(f"Periodicity type {periodicity} is not valid.")
-
         if start_date and end_date:
             if (
                 datetime.strptime(start_date, "%Y-%m-%d").date()
@@ -855,22 +852,22 @@ class Ticker(DelegatedCompanyFunctionsMetaClass):
 
     def history(
         self,
-        periodicity: str = "day",
+        periodicity: Periodicity = Periodicity.day,
         adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
     ) -> pd.DataFrame:
         """Retrieves the historical price data for a given asset over a specified date range.
 
-        :param periodicity: Determines the frequency of the historical data returned. Options are "day", "week", "month" and "year". This default to "day"
-        :type periodicity: str
+        :param periodicity: Determines the frequency of the historical data returned. Defaults to Periodicity.day.
+        :type periodicity: Periodicity
         :param adjusted: Whether to retrieve adjusted prices that account for corporate actions such as dividends and splits, it defaults True
         :type adjusted: bool, optional
         :param start_date: The start date for historical price retrieval in format "YYYY-MM-DD", default to None
         :type start_date: str, optional
         :param end_date: The end date for historical price retrieval in format "YYYY-MM-DD", default to None
         :type end_date: str, optional
-        :return: A pd.DataFrame containing historical price data with columns corresponding to the specified periodicity, with Date as the index, and columns "open", "high", "low", "close", "volume" in type decimal. The Date index is a string that depends on the periodicity. If periodicity="day", the Date index is the day in format "YYYY-MM-DD", eg "2024-05-13" If periodicity="week", the Date index is the week number of the year in format "YYYY Week ##", eg "2024 Week 2" If periodicity="month", the Date index is the month name of the year in format "<Month> YYYY", eg "January 2024". If periodicity="year", the Date index is the year in format "YYYY", eg "2024".
+        :return: A pd.DataFrame containing historical price data with columns corresponding to the specified periodicity, with Date as the index, and columns "open", "high", "low", "close", "volume" in type decimal. The Date index is a string that depends on the periodicity. If Periodicity.day, the Date index is the day in format "YYYY-MM-DD", eg "2024-05-13" If Periodicity.week, the Date index is the week number of the year in format "YYYY Week ##", eg "2024 Week 2" If Periodicity.month, the Date index is the month name of the year in format "<Month> YYYY", eg "January 2024". If Periodicity.year, the Date index is the year in format "YYYY", eg "2024".
         :rtype: pd.DataFrame
         """
         return self.primary_trading_item.history(
@@ -882,15 +879,15 @@ class Ticker(DelegatedCompanyFunctionsMetaClass):
 
     def price_chart(
         self,
-        periodicity: str = "day",
+        periodicity: Periodicity = Periodicity.day,
         adjusted: bool = True,
         start_date: Optional[str] = None,
         end_date: Optional[str] = None,
     ) -> Image:
         """Get the price chart.
 
-        :param periodicity: Determines the frequency of the historical data returned. Options are "day", "week", "month" and "year". This default to "day"
-        :type periodicity: str
+        :param str periodicity: Determines the frequency of the historical data returned. Defaults to Periodicity.day.
+        :type periodicity: Periodicity
         :param adjusted: Whether to retrieve adjusted prices that account for corporate actions such as dividends and splits, it defaults True
         :type adjusted: bool, optional
         :param start_date: The start date for historical price retrieval in format "YYYY-MM-DD", default to None
@@ -1121,11 +1118,63 @@ class Client:
             )
             stdout.write("Login credentials received.\n")
 
-        self.tools = _llm_tools(self)
-        self.langchain_tools = langchain_tools(self.tools)
-        self.anthropic_tool_descriptions = anthropic_tool_descriptions
-        self.gemini_tool_descriptions = gemini_tool_descriptions
-        self.openai_tool_descriptions = openai_tool_descriptions
+        self._tools: list[KfinanceTool] | None = None
+
+    @property
+    def langchain_tools(self) -> list["KfinanceTool"]:
+        """Return a list of all Kfinance tools for tool calling."""
+        if self._tools is None:
+            from kfinance.tool_calling import ALL_TOOLS
+
+            self._tools = [t(kfinance_client=self) for t in ALL_TOOLS]  # type: ignore[call-arg]
+        return self._tools
+
+    @property
+    def tools(self) -> dict[str, Callable]:
+        """Return a mapping of tool calling function names to the corresponding functions.
+
+        `tools` is intended for running without langchain. When running with langchain,
+        use `langchain_tools`.
+        """
+        return {t.name: t.run_without_langchain for t in self.langchain_tools}
+
+    @property
+    def anthropic_tool_descriptions(self) -> list[dict[str, Any]]:
+        """Return tool descriptions for anthropic"""
+
+        anthropic_tool_descriptions = []
+
+        for tool in self.langchain_tools:
+            # Copied from https://python.langchain.com/api_reference/_modules/langchain_anthropic/chat_models.html#convert_to_anthropic_tool
+            # to avoid adding a langchain-anthropic dependency.
+            oai_formatted = convert_to_openai_tool(tool)["function"]
+            anthropic_tool_descriptions.append(
+                dict(
+                    name=oai_formatted["name"],
+                    description=oai_formatted["description"],
+                    input_schema=oai_formatted["parameters"],
+                )
+            )
+
+        return anthropic_tool_descriptions
+
+    @property
+    def gemini_tool_descriptions(self) -> gapic.Tool:
+        """Return tool descriptions for gemini.
+
+        The conversion from BaseTool -> openai tool description -> google tool mirrors the
+        langchain implementation.
+        """
+        openai_tool_descriptions = [
+            convert_to_openai_tool(t)["function"] for t in self.langchain_tools
+        ]
+        return convert_to_genai_function_declarations(openai_tool_descriptions)
+
+    @property
+    def openai_tool_descriptions(self) -> list[dict[str, Any]]:
+        """Return tool descriptions for gemini"""
+        openai_tool_descriptions = [convert_to_openai_tool(t) for t in self.langchain_tools]
+        return openai_tool_descriptions
 
     @property
     def access_token(self) -> str:
@@ -1154,7 +1203,7 @@ class Client:
         :rtype: Ticker
         """
         if function_called:
-            self.kfinance_api_client.user_agent_source = "function_calling"
+            self.kfinance_api_client.user_agent_source = "tool_calling"
         return Ticker(self.kfinance_api_client, str(identifier), exchange_code)
 
     def tickers(
@@ -1222,14 +1271,42 @@ class Client:
         )
 
     @staticmethod
-    def get_latest() -> LatestPeriods:
+    def get_latest(use_local_timezone: bool = True) -> LatestPeriods:
         """Get the latest annual reporting year, latest quarterly reporting quarter and year, and current date.
 
+        :param use_local_timezone: whether to use the local timezone of the user
+        :type use_local_timezone: bool
         :return: A dict in the form of {"annual": {"latest_year": int}, "quarterly": {"latest_quarter": int, "latest_year": int}, "now": {"current_year": int, "current_quarter": int, "current_month": int, "current_date": str of Y-m-d}}
         :rtype: Latest
         """
 
-        return get_latest()
+        datetime_now = datetime.now() if use_local_timezone else datetime.now(timezone.utc)
+        current_year = datetime_now.year
+        current_qtr = (datetime_now.month - 1) // 3 + 1
+
+        # Quarterly data. Get most recent year and quarter
+        if current_qtr == 1:
+            most_recent_year_qtrly = current_year - 1
+            most_recent_qtr = 4
+        else:
+            most_recent_year_qtrly = current_year
+            most_recent_qtr = current_qtr - 1
+
+        # Annual data. Get most recent year
+        most_recent_year_annual = current_year - 1
+
+        current_month = datetime_now.month
+        latest: LatestPeriods = {
+            "annual": {"latest_year": most_recent_year_annual},
+            "quarterly": {"latest_quarter": most_recent_qtr, "latest_year": most_recent_year_qtrly},
+            "now": {
+                "current_year": current_year,
+                "current_quarter": current_qtr,
+                "current_month": current_month,
+                "current_date": datetime_now.date().isoformat(),
+            },
+        }
+        return latest
 
     @staticmethod
     def get_n_quarters_ago(n: int) -> YearAndQuarter:
@@ -1240,4 +1317,17 @@ class Client:
         :rtype: YearAndQuarter
         """
 
-        return get_n_quarters_ago(n)
+        datetime_now = datetime.now()
+        current_qtr = (datetime_now.month - 1) // 3 + 1
+        total_quarters_completed = datetime_now.year * 4 + current_qtr - 1
+        total_quarters_completed_n_quarters_ago = total_quarters_completed - n
+
+        year_n_quarters_ago = total_quarters_completed_n_quarters_ago // 4
+        quarter_n_quarters_ago = total_quarters_completed_n_quarters_ago % 4 + 1
+
+        year_quarter_n_quarters_ago: YearAndQuarter = {
+            "year": year_n_quarters_ago,
+            "quarter": quarter_n_quarters_ago,
+        }
+
+        return year_quarter_n_quarters_ago