aiagents4pharma 1.43.0__py3-none-any.whl → 1.45.0__py3-none-any.whl

aiagents4pharma/talk2scholars/tools/paper_download/utils/base_paper_downloader.py

@@ -8,7 +8,7 @@ import logging
 import re
 import tempfile
 from abc import ABC, abstractmethod
-from typing import Any, Dict, List, Optional, Tuple
+from typing import Any
 
 import requests
 
@@ -58,8 +58,8 @@ class BasePaperDownloader(ABC):
 
     @abstractmethod
     def extract_paper_metadata(
-        self, metadata: Any, identifier: str, pdf_result: Optional[Tuple[str, str]]
-    ) -> Dict[str, Any]:
+        self, metadata: Any, identifier: str, pdf_result: tuple[str, str] | None
+    ) -> dict[str, Any]:
         """
         Extract and structure metadata into standardized format.
 
@@ -89,9 +89,7 @@ class BasePaperDownloader(ABC):
         raise NotImplementedError
 
     # Common methods shared by all services
-    def download_pdf_to_temp(
-        self, pdf_url: str, identifier: str
-    ) -> Optional[Tuple[str, str]]:
+    def download_pdf_to_temp(self, pdf_url: str, identifier: str) -> tuple[str, str] | None:
         """
         Download PDF from URL to a temporary file.
 
@@ -103,9 +101,7 @@ class BasePaperDownloader(ABC):
             Tuple of (temp_file_path, filename) or None if failed
         """
         if not pdf_url:
-            logger.info(
-                "No PDF URL available for %s %s", self.get_identifier_name(), identifier
-            )
+            logger.info("No PDF URL available for %s %s", self.get_identifier_name(), identifier)
             return None
 
         try:
@@ -141,14 +137,11 @@ class BasePaperDownloader(ABC):
 
             if "filename=" in content_disposition:
                 try:
-
                     filename_match = re.search(
                         r'filename[*]?=(?:"([^"]+)"|([^;]+))', content_disposition
                     )
                     if filename_match:
-                        extracted_filename = filename_match.group(
-                            1
-                        ) or filename_match.group(2)
+                        extracted_filename = filename_match.group(1) or filename_match.group(2)
                         extracted_filename = extracted_filename.strip().strip('"')
                         if extracted_filename and extracted_filename.endswith(".pdf"):
                             filename = extracted_filename
@@ -189,7 +182,7 @@ class BasePaperDownloader(ABC):
 
         return snippet
 
-    def create_error_entry(self, identifier: str, error_msg: str) -> Dict[str, Any]:
+    def create_error_entry(self, identifier: str, error_msg: str) -> dict[str, Any]:
         """
         Create standardized error entry for failed paper processing.
 
@@ -215,7 +208,7 @@ class BasePaperDownloader(ABC):
             # Service-specific identifier field will be added by subclasses
         }
 
-    def build_summary(self, article_data: Dict[str, Any]) -> str:
+    def build_summary(self, article_data: dict[str, Any]) -> str:
         """
         Build a summary string for up to three papers with snippets.
 
@@ -226,7 +219,7 @@ class BasePaperDownloader(ABC):
             Formatted summary string
         """
         top = list(article_data.values())[:3]
-        lines: List[str] = []
+        lines: list[str] = []
         downloaded_count = sum(
             1
             for paper in article_data.values()
@@ -240,7 +233,7 @@ class BasePaperDownloader(ABC):
             snippet = self.get_snippet(paper.get("Abstract", ""))
 
             # Build paper line with service-specific identifier info
-            line = f"{idx+1}. {title}"
+            line = f"{idx + 1}. {title}"
             line += self._get_paper_identifier_info(paper)
             line += f"\n   Access: {access_type}"
 
@@ -264,7 +257,7 @@ class BasePaperDownloader(ABC):
         )
 
     @abstractmethod
-    def _get_paper_identifier_info(self, paper: Dict[str, Any]) -> str:
+    def _get_paper_identifier_info(self, paper: dict[str, Any]) -> str:
         """
         Get service-specific identifier info for paper summary.
 
@@ -276,7 +269,7 @@ class BasePaperDownloader(ABC):
         """
         raise NotImplementedError
 
-    def process_identifiers(self, identifiers: List[str]) -> Dict[str, Any]:
+    def process_identifiers(self, identifiers: list[str]) -> dict[str, Any]:
         """
         Main processing loop for downloading papers.
 
@@ -293,7 +286,7 @@ class BasePaperDownloader(ABC):
             identifiers,
         )
 
-        article_data: Dict[str, Any] = {}
+        article_data: dict[str, Any] = {}
 
         for identifier in identifiers:
             logger.info("Processing %s: %s", self.get_identifier_name(), identifier)
@@ -332,7 +325,7 @@ class BasePaperDownloader(ABC):
         return article_data
 
     @abstractmethod
-    def _add_service_identifier(self, entry: Dict[str, Any], identifier: str) -> None:
+    def _add_service_identifier(self, entry: dict[str, Any], identifier: str) -> None:
         """
         Add service-specific identifier field to entry.
 

aiagents4pharma/talk2scholars/tools/paper_download/utils/biorxiv_downloader.py

@@ -6,7 +6,7 @@ BioRxiv paper downloader implementation.
 import logging
 import re
 import tempfile
-from typing import Any, Dict, Optional, Tuple
+from typing import Any
 
 import cloudscraper
 import requests
@@ -43,9 +43,7 @@ class BiorxivDownloader(BasePaperDownloader):
         # CloudScraper specific settings
         self.cf_clearance_timeout = getattr(config, "cf_clearance_timeout", 30)
         self.session_reuse = getattr(config, "session_reuse", True)
-        self.browser_config_type = getattr(config, "browser_config", {}).get(
-            "type", "custom"
-        )
+        self.browser_config_type = getattr(config, "browser_config", {}).get("type", "custom")
 
         # Initialize shared CloudScraper session if enabled
         self._scraper = None
@@ -55,7 +53,7 @@ class BiorxivDownloader(BasePaperDownloader):
                 delay=self.cf_clearance_timeout,
             )
 
-    def fetch_metadata(self, identifier: str) -> Dict[str, Any]:
+    def fetch_metadata(self, identifier: str) -> dict[str, Any]:
         """
         Fetch paper metadata from bioRxiv API.
 
@@ -88,7 +86,7 @@ class BiorxivDownloader(BasePaperDownloader):
 
         return paper_data
 
-    def construct_pdf_url(self, metadata: Dict[str, Any], identifier: str) -> str:
+    def construct_pdf_url(self, metadata: dict[str, Any], identifier: str) -> str:
         """
         Construct PDF URL from bioRxiv metadata and DOI.
 
@@ -111,9 +109,7 @@ class BiorxivDownloader(BasePaperDownloader):
 
         return pdf_url
 
-    def download_pdf_to_temp(
-        self, pdf_url: str, identifier: str
-    ) -> Optional[Tuple[str, str]]:
+    def download_pdf_to_temp(self, pdf_url: str, identifier: str) -> tuple[str, str] | None:
         """
         Override base method to use CloudScraper for bioRxiv PDF downloads.
         Includes landing page visit to handle CloudFlare protection.
@@ -188,9 +184,7 @@ class BiorxivDownloader(BasePaperDownloader):
                     r'filename[*]?=(?:"([^"]+)"|([^;]+))', content_disposition
                 )
                 if filename_match:
-                    extracted_filename = filename_match.group(
-                        1
-                    ) or filename_match.group(2)
+                    extracted_filename = filename_match.group(1) or filename_match.group(2)
                     extracted_filename = extracted_filename.strip().strip('"')
                     if extracted_filename and extracted_filename.endswith(".pdf"):
                         filename = extracted_filename
@@ -202,10 +196,10 @@ class BiorxivDownloader(BasePaperDownloader):
 
     def extract_paper_metadata(
         self,
-        metadata: Dict[str, Any],
+        metadata: dict[str, Any],
         identifier: str,
-        pdf_result: Optional[Tuple[str, str]],
-    ) -> Dict[str, Any]:
+        pdf_result: tuple[str, str] | None,
+    ) -> dict[str, Any]:
         """
         Extract structured metadata from bioRxiv API response.
 
@@ -234,9 +228,7 @@ class BiorxivDownloader(BasePaperDownloader):
             **pdf_metadata,
         }
 
-    def _extract_basic_metadata(
-        self, paper: Dict[str, Any], identifier: str
-    ) -> Dict[str, Any]:
+    def _extract_basic_metadata(self, paper: dict[str, Any], identifier: str) -> dict[str, Any]:
         """Extract basic metadata from paper data."""
         # Extract basic fields
         title = paper.get("title", "N/A").strip()
@@ -267,8 +259,8 @@ class BiorxivDownloader(BasePaperDownloader):
         return [author.strip() for author in authors_str.split(";") if author.strip()]
 
     def _extract_pdf_metadata(
-        self, pdf_result: Optional[Tuple[str, str]], identifier: str
-    ) -> Dict[str, Any]:
+        self, pdf_result: tuple[str, str] | None, identifier: str
+    ) -> dict[str, Any]:
         """Extract PDF-related metadata."""
         if pdf_result:
             temp_file_path, filename = pdf_result
@@ -301,7 +293,7 @@ class BiorxivDownloader(BasePaperDownloader):
         # Sanitize DOI for filename use
         return f"{identifier.replace('/', '_').replace('.', '_')}.pdf"
 
-    def _get_paper_identifier_info(self, paper: Dict[str, Any]) -> str:
+    def _get_paper_identifier_info(self, paper: dict[str, Any]) -> str:
         """Get bioRxiv-specific identifier info for paper summary."""
         doi = paper.get("DOI", "N/A")
         pub_date = paper.get("Publication Date", "N/A")
@@ -313,7 +305,7 @@ class BiorxivDownloader(BasePaperDownloader):
 
         return info
 
-    def _add_service_identifier(self, entry: Dict[str, Any], identifier: str) -> None:
+    def _add_service_identifier(self, entry: dict[str, Any], identifier: str) -> None:
         """Add DOI and bioRxiv-specific fields to entry."""
         entry["DOI"] = identifier
         entry["Category"] = "N/A"

aiagents4pharma/talk2scholars/tools/paper_download/utils/medrxiv_downloader.py

@@ -4,7 +4,7 @@ MedRxiv paper downloader implementation.
 """
 
 import logging
-from typing import Any, Dict, Optional, Tuple
+from typing import Any
 
 import requests
 
@@ -27,7 +27,7 @@ class MedrxivDownloader(BasePaperDownloader):
         )
         self.default_version = getattr(config, "default_version", "1")
 
-    def fetch_metadata(self, identifier: str) -> Dict[str, Any]:
+    def fetch_metadata(self, identifier: str) -> dict[str, Any]:
         """
         Fetch paper metadata from medRxiv API.
 
@@ -54,7 +54,7 @@ class MedrxivDownloader(BasePaperDownloader):
 
         return paper_data
 
-    def construct_pdf_url(self, metadata: Dict[str, Any], identifier: str) -> str:
+    def construct_pdf_url(self, metadata: dict[str, Any], identifier: str) -> str:
         """
         Construct PDF URL from medRxiv metadata and DOI.
 
@@ -79,10 +79,10 @@ class MedrxivDownloader(BasePaperDownloader):
 
     def extract_paper_metadata(
         self,
-        metadata: Dict[str, Any],
+        metadata: dict[str, Any],
         identifier: str,
-        pdf_result: Optional[Tuple[str, str]],
-    ) -> Dict[str, Any]:
+        pdf_result: tuple[str, str] | None,
+    ) -> dict[str, Any]:
         """
         Extract structured metadata from medRxiv API response.
 
@@ -111,9 +111,7 @@ class MedrxivDownloader(BasePaperDownloader):
             **pdf_metadata,
         }
 
-    def _extract_basic_metadata(
-        self, paper: Dict[str, Any], identifier: str
-    ) -> Dict[str, Any]:
+    def _extract_basic_metadata(self, paper: dict[str, Any], identifier: str) -> dict[str, Any]:
         """Extract basic metadata from paper data."""
         # Extract basic fields
         title = paper.get("title", "N/A").strip()
@@ -144,8 +142,8 @@ class MedrxivDownloader(BasePaperDownloader):
         return [author.strip() for author in authors_str.split(";") if author.strip()]
 
     def _extract_pdf_metadata(
-        self, pdf_result: Optional[Tuple[str, str]], identifier: str
-    ) -> Dict[str, Any]:
+        self, pdf_result: tuple[str, str] | None, identifier: str
+    ) -> dict[str, Any]:
         """Extract PDF-related metadata."""
         if pdf_result:
             temp_file_path, filename = pdf_result
@@ -178,7 +176,7 @@ class MedrxivDownloader(BasePaperDownloader):
         # Sanitize DOI for filename use
         return f"{identifier.replace('/', '_').replace('.', '_')}.pdf"
 
-    def _get_paper_identifier_info(self, paper: Dict[str, Any]) -> str:
+    def _get_paper_identifier_info(self, paper: dict[str, Any]) -> str:
         """Get medRxiv-specific identifier info for paper summary."""
         doi = paper.get("DOI", "N/A")
         pub_date = paper.get("Publication Date", "N/A")
@@ -190,7 +188,7 @@ class MedrxivDownloader(BasePaperDownloader):
 
         return info
 
-    def _add_service_identifier(self, entry: Dict[str, Any], identifier: str) -> None:
+    def _add_service_identifier(self, entry: dict[str, Any], identifier: str) -> None:
         """Add DOI and medRxiv-specific fields to entry."""
         entry["DOI"] = identifier
         entry["Category"] = "N/A"

aiagents4pharma/talk2scholars/tools/paper_download/utils/pubmed_downloader.py

@@ -5,7 +5,7 @@ PubMed paper downloader implementation.
 
 import logging
 import xml.etree.ElementTree as ET
-from typing import Any, Dict, Optional, Tuple, cast
+from typing import Any, cast
 
 import requests
 from bs4 import BeautifulSoup, Tag
@@ -37,7 +37,7 @@ class PubmedDownloader(BasePaperDownloader):
         self.pdf_meta_name = getattr(config, "pdf_meta_name", "citation_pdf_url")
         self.default_error_code = getattr(config, "default_error_code", "unknown")
 
-    def fetch_metadata(self, identifier: str) -> Dict[str, Any]:
+    def fetch_metadata(self, identifier: str) -> dict[str, Any]:
         """
         Fetch paper metadata from PubMed ID Converter API.
 
@@ -52,9 +52,7 @@ class PubmedDownloader(BasePaperDownloader):
             RuntimeError: If no records found in response
         """
         query_url = f"{self.id_converter_url}?ids={identifier}&format={self.id_converter_format}"
-        logger.info(
-            "Fetching metadata from ID converter for PMID %s: %s", identifier, query_url
-        )
+        logger.info("Fetching metadata from ID converter for PMID %s: %s", identifier, query_url)
 
         response = requests.get(query_url, timeout=self.request_timeout)
         response.raise_for_status()
@@ -67,7 +65,7 @@ class PubmedDownloader(BasePaperDownloader):
 
         return result
 
-    def construct_pdf_url(self, metadata: Dict[str, Any], identifier: str) -> str:
+    def construct_pdf_url(self, metadata: dict[str, Any], identifier: str) -> str:
         """
         Construct PDF URL using multiple fallback strategies.
 
@@ -145,18 +143,14 @@ class PubmedDownloader(BasePaperDownloader):
             if error_elem is not None:
                 error_code = error_elem.get("code", self.default_error_code)
                 error_text = error_elem.text or "unknown error"
-                logger.info(
-                    "OA API error for PMCID %s: %s - %s", pmcid, error_code, error_text
-                )
+                logger.info("OA API error for PMCID %s: %s - %s", pmcid, error_code, error_text)
                 return ""
 
             # Look for PDF link
             pdf_link = root.find(".//link[@format='pdf']")
             if pdf_link is not None:
                 pdf_url = pdf_link.get("href", "")
-                logger.info(
-                    "Found PDF URL from OA API for PMCID %s: %s", pmcid, pdf_url
-                )
+                logger.info("Found PDF URL from OA API for PMCID %s: %s", pmcid, pdf_url)
 
                 # Convert FTP links to HTTPS for download compatibility
                 if pdf_url.startswith(self.ftp_base_url):
@@ -188,15 +182,11 @@ class PubmedDownloader(BasePaperDownloader):
     def _try_pmc_page_scraping(self, pmcid: str) -> str:
         """Try scraping PMC page for PDF meta tag."""
         pmc_page_url = f"{self.pmc_page_base_url}/{pmcid}/"
-        logger.info(
-            "Scraping PMC page for PDF meta tag for %s: %s", pmcid, pmc_page_url
-        )
+        logger.info("Scraping PMC page for PDF meta tag for %s: %s", pmcid, pmc_page_url)
 
         try:
             headers = {"User-Agent": self.user_agent}
-            response = requests.get(
-                pmc_page_url, headers=headers, timeout=self.request_timeout
-            )
+            response = requests.get(pmc_page_url, headers=headers, timeout=self.request_timeout)
             response.raise_for_status()
 
             soup = BeautifulSoup(response.content, "html.parser")
@@ -238,10 +228,10 @@ class PubmedDownloader(BasePaperDownloader):
 
     def extract_paper_metadata(
         self,
-        metadata: Dict[str, Any],
+        metadata: dict[str, Any],
         identifier: str,
-        pdf_result: Optional[Tuple[str, str]],
-    ) -> Dict[str, Any]:
+        pdf_result: tuple[str, str] | None,
+    ) -> dict[str, Any]:
         """
         Extract structured metadata from PubMed ID converter response.
 
@@ -310,15 +300,11 @@ class PubmedDownloader(BasePaperDownloader):
 
     def get_snippet(self, abstract: str) -> str:
         """Override to handle PubMed-specific abstract placeholder."""
-        if (
-            not abstract
-            or abstract == "N/A"
-            or abstract == "Abstract available in PubMed"
-        ):
+        if not abstract or abstract == "N/A" or abstract == "Abstract available in PubMed":
             return ""
         return super().get_snippet(abstract)
 
-    def _get_paper_identifier_info(self, paper: Dict[str, Any]) -> str:
+    def _get_paper_identifier_info(self, paper: dict[str, Any]) -> str:
         """Get PubMed-specific identifier info for paper summary."""
         pmid = paper.get("PMID", "N/A")
         pmcid = paper.get("PMCID", "N/A")
@@ -329,7 +315,7 @@ class PubmedDownloader(BasePaperDownloader):
 
         return info
 
-    def _add_service_identifier(self, entry: Dict[str, Any], identifier: str) -> None:
+    def _add_service_identifier(self, entry: dict[str, Any], identifier: str) -> None:
         """Add PMID and PubMed-specific fields to entry."""
         entry["PMID"] = identifier
         entry["PMCID"] = "N/A"

aiagents4pharma/talk2scholars/tools/pdf/question_and_answer.py

@@ -25,11 +25,11 @@ from langgraph.prebuilt import InjectedState
 from langgraph.types import Command
 from pydantic import BaseModel, Field
 
+from .utils.answer_formatter import format_answer
 from .utils.generate_answer import load_hydra_config
-from .utils.tool_helper import QAToolHelper
 from .utils.paper_loader import load_all_papers
 from .utils.rag_pipeline import retrieve_and_rerank_chunks
-from .utils.answer_formatter import format_answer
+from .utils.tool_helper import QAToolHelper
 
 # Helper for managing state, vectorstore, reranking, and formatting
 helper = QAToolHelper()
@@ -56,9 +56,7 @@ class QuestionAndAnswerInput(BaseModel):
         - llm_model: chat/LLM instance for answer generation.
     """
 
-    question: str = Field(
-        description="User question for generating a PDF-based answer."
-    )
+    question: str = Field(description="User question for generating a PDF-based answer.")
     tool_call_id: Annotated[str, InjectedToolCallId]
     state: Annotated[dict, InjectedState]
 
@@ -133,9 +131,7 @@ def question_and_answer(
     )
 
     # Retrieve and rerank chunks in one step
-    reranked_chunks = retrieve_and_rerank_chunks(
-        vs, question, config, call_id, helper.has_gpu
-    )
+    reranked_chunks = retrieve_and_rerank_chunks(vs, question, config, call_id, helper.has_gpu)
 
     if not reranked_chunks:
         msg = f"No relevant chunks found for question: '{question}'"

aiagents4pharma/talk2scholars/tools/pdf/utils/__init__.py

@@ -2,20 +2,22 @@
 Utility modules for the PDF question_and_answer tool.
 """
 
-from . import answer_formatter
-from . import batch_processor
-from . import collection_manager
-from . import generate_answer
-from . import get_vectorstore
-from . import gpu_detection
-from . import nvidia_nim_reranker
-from . import paper_loader
-from . import rag_pipeline
-from . import retrieve_chunks
-from . import singleton_manager
-from . import tool_helper
-from . import vector_normalization
-from . import vector_store
+from . import (
+    answer_formatter,
+    batch_processor,
+    collection_manager,
+    generate_answer,
+    get_vectorstore,
+    gpu_detection,
+    nvidia_nim_reranker,
+    paper_loader,
+    rag_pipeline,
+    retrieve_chunks,
+    singleton_manager,
+    tool_helper,
+    vector_normalization,
+    vector_store,
+)
 
 __all__ = [
     "answer_formatter",

aiagents4pharma/talk2scholars/tools/pdf/utils/answer_formatter.py

@@ -3,7 +3,7 @@ Format the final answer text with source attributions and hardware info.
 """
 
 import logging
-from typing import Any, Dict, List
+from typing import Any
 
 from .generate_answer import generate_answer
 
@@ -12,9 +12,9 @@ logger = logging.getLogger(__name__)
 
 def format_answer(
     question: str,
-    chunks: List[Any],
+    chunks: list[Any],
     llm: Any,
-    articles: Dict[str, Any],
+    articles: dict[str, Any],
     config: Any,
     **kwargs: Any,
 ) -> str:
@@ -27,7 +27,7 @@ def format_answer(
     answer = result.get("output_text", "No answer generated.")
 
     # Get unique paper titles for source attribution
-    titles: Dict[str, str] = {}
+    titles: dict[str, str] = {}
     for pid in result.get("papers_used", []):
         if pid in articles:
             titles[pid] = articles[pid].get("Title", "Unknown paper")

aiagents4pharma/talk2scholars/tools/pdf/utils/batch_processor.py

@@ -5,7 +5,7 @@ Batch processing utilities for adding multiple papers to vector store.
 import concurrent.futures
 import logging
 import time
-from typing import Any, Dict, List, Set, Tuple
+from typing import Any
 
 from langchain_core.documents import Document
 
@@ -15,11 +15,11 @@ logger = logging.getLogger(__name__)
 
 
 def add_papers_batch(
-    papers_to_add: List[Tuple[str, str, Dict[str, Any]]],
+    papers_to_add: list[tuple[str, str, dict[str, Any]]],
     vector_store: Any,
-    loaded_papers: Set[str],
-    paper_metadata: Dict[str, Dict[str, Any]],
-    documents: Dict[str, Document],
+    loaded_papers: set[str],
+    paper_metadata: dict[str, dict[str, Any]],
+    documents: dict[str, Document],
     **kwargs: Any,
 ) -> None:
     """
@@ -43,9 +43,7 @@ def add_papers_batch(
         logger.info("No papers to add")
         return
 
-    to_process = [
-        (pid, url, md) for pid, url, md in papers_to_add if pid not in loaded_papers
-    ]
+    to_process = [(pid, url, md) for pid, url, md in papers_to_add if pid not in loaded_papers]
     if not to_process:
         logger.info("Skipping %d already-loaded papers", len(papers_to_add))
         logger.info("All %d papers are already loaded", len(papers_to_add))
@@ -91,16 +89,16 @@ def add_papers_batch(
 
 
 def _parallel_load_and_split(
-    papers: List[Tuple[str, str, Dict[str, Any]]],
+    papers: list[tuple[str, str, dict[str, Any]]],
     config: Any,
-    metadata_fields: List[str],
-    documents: Dict[str, Document],
+    metadata_fields: list[str],
+    documents: dict[str, Document],
     max_workers: int,
-) -> Tuple[List[Document], List[str], List[str]]:
+) -> tuple[list[Document], list[str], list[str]]:
     """Load & split PDFs in parallel, preserving original logic."""
-    all_chunks: List[Document] = []
-    all_ids: List[str] = []
-    success: List[str] = []
+    all_chunks: list[Document] = []
+    all_ids: list[str] = []
+    success: list[str] = []
 
     with concurrent.futures.ThreadPoolExecutor(max_workers=max_workers) as executor:
         futures = {
@@ -138,8 +136,8 @@ def _parallel_load_and_split(
 
 
 def _batch_embed(
-    chunks: List[Document],
-    ids: List[str],
+    chunks: list[Document],
+    ids: list[str],
     store: Any,
     batch_size: int,
     has_gpu: bool,

aiagents4pharma/talk2scholars/tools/pdf/utils/collection_manager.py

@@ -5,7 +5,7 @@ Collection Manager for Milvus
 import logging
 import os
 import threading
-from typing import Any, Dict
+from typing import Any
 
 from pymilvus import (
     Collection,
@@ -28,7 +28,7 @@ _cache_lock = threading.Lock()
 
 
 def ensure_collection_exists(
-    collection_name: str, config: Any, index_params: Dict[str, Any], has_gpu: bool
+    collection_name: str, config: Any, index_params: dict[str, Any], has_gpu: bool
 ) -> Collection:
     """Ensure the Milvus collection exists before trying to sync or add documents."""
 

aiagents4pharma/talk2scholars/tools/pdf/utils/document_processor.py

@@ -3,7 +3,7 @@ Document processing utilities for loading and splitting PDFs.
 """
 
 import logging
-from typing import Any, Dict, List
+from typing import Any
 
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.document_loaders import PyPDFLoader
@@ -15,10 +15,10 @@ logger = logging.getLogger(__name__)
 def load_and_split_pdf(
     paper_id: str,
     pdf_url: str,
-    paper_metadata: Dict[str, Any],
+    paper_metadata: dict[str, Any],
     config: Any,
     **kwargs: Any,
-) -> List[Document]:
+) -> list[Document]:
     """
     Load a PDF and split it into chunks.
 
@@ -35,8 +35,8 @@ def load_and_split_pdf(
     Returns:
         A list of Document chunks, each with updated metadata.
     """
-    metadata_fields: List[str] = kwargs["metadata_fields"]
-    documents_dict: Dict[str, Document] = kwargs["documents_dict"]
+    metadata_fields: list[str] = kwargs["metadata_fields"]
+    documents_dict: dict[str, Document] = kwargs["documents_dict"]
 
     logger.info("Loading PDF for paper %s from %s", paper_id, pdf_url)