feat: rename pdf to markdown in agent

src/deepengineer/deepsearch/analyse_markdown_agent.py

@@ -1,5 +1,5 @@
 from smolagents import CodeAgent, tool, Tool, LiteLLMModel
-from deepengineer.webcrawler.pdf_utils import get_markdown_by_page_numbers, get_table_of_contents_per_page_pdf, find_in_pdf, convert_ocr_response_to_markdown
+from deepengineer.webcrawler.pdf_utils import get_markdown_by_page_numbers, get_table_of_contents_per_page_markdown, find_in_markdown, convert_ocr_response_to_markdown
 from mistralai import OCRResponse
 from enum import Enum
 from pathlib import Path
@@ -8,7 +8,7 @@ class ToolNames(Enum):
     GET_TABLE_OF_CONTENTS = "get_table_of_contents"
     GET_MARKDOWN = "get_markdown"
     GET_PAGES_CONTENT = "get_pages_content"
-    FIND_IN_PDF = "find_in_pdf"
+    FIND_IN_MARKDOWN = "find_in_markdown"
 
 class GetTableOfContentsTool(Tool):
     name = ToolNames.GET_TABLE_OF_CONTENTS.value
@@ -19,7 +19,7 @@ class GetTableOfContentsTool(Tool):
     def __init__(self, markdown: OCRResponse):
         super().__init__()
         self.markdown: OCRResponse = markdown
-        self.table_of_contents: str = get_table_of_contents_per_page_pdf(self.markdown)
+        self.table_of_contents: str = get_table_of_contents_per_page_markdown(self.markdown)
         
     def forward(self) -> str:
         return self.table_of_contents
@@ -57,8 +57,8 @@ class GetPagesContentTool(Tool):
     def forward(self, page_numbers: list[int]) -> str:
         return get_markdown_by_page_numbers(self.markdown, page_numbers)
     
-class FindInPdfTool(Tool):
-    name = ToolNames.FIND_IN_PDF.value
+class FindInMarkdownTool(Tool):
+    name = ToolNames.FIND_IN_MARKDOWN.value
     description = f"Finds the page numbers of the document that contain the search queries. If you are looking for a specific information, you can use this tool to find the page numbers of the document that contain the information and then use {ToolNames.GET_PAGES_CONTENT.value} to get the content of the pages."
     inputs = {
         "search_queries": {
@@ -73,7 +73,7 @@ class FindInPdfTool(Tool):
         self.markdown: OCRResponse = markdown
     
     def forward(self, search_queries: list[str]) -> list[int]:
-        return find_in_pdf(self.markdown, search_queries)
+        return find_in_markdown(self.markdown, search_queries)
 
 
         
@@ -81,21 +81,21 @@ def create_agent(markdown: OCRResponse, model_id="deepseek/deepseek-chat"):
 
     model = LiteLLMModel(model_id=model_id)
 
-    PDFS_TOOLS = [
+    MARKDOWN_TOOLS = [
         GetTableOfContentsTool(markdown),
         GetMarkdownTool(markdown),
         GetPagesContentTool(markdown),
-        FindInPdfTool(markdown),
+        FindInMarkdownTool(markdown),
     ]
-    pdf_agent = CodeAgent(
+    markdown_agent = CodeAgent(
         model=model,
-        tools=PDFS_TOOLS,
+        tools=MARKDOWN_TOOLS,
         max_steps=20,
         verbosity_level=2,
         planning_interval=4,
-        name="pdf_agent",
-        description="""A team member that will search the internet to answer your question.""",
+        name="markdown_agent",
+        description="""A team member that can analyse a markdown.""",
     )
-    pdf_agent.prompt_templates["managed_agent"]["task"] += """You can navigate to .txt online files."""
+    markdown_agent.prompt_templates["managed_agent"]["task"] += """You can navigate to .txt online files."""
 
-    return pdf_agent
+    return markdown_agent

src/deepengineer/webcrawler/pdf_utils.py

@@ -62,7 +62,7 @@ def get_markdown_by_page_numbers(markdown: OCRResponse, page_numbers: list[int],
         markdowns.append(f"*Page {page_number}*\n{markdown.pages[page_number].markdown}")
     return "\n\n".join(markdowns)
 
-def find_in_pdf(markdown: OCRResponse, search_queries: list[str]) -> list[int]:
+def find_in_markdown(markdown: OCRResponse, search_queries: list[str]) -> list[int]:
     """
     Find the page numbers of the pdf that contain the search query.
 
@@ -80,7 +80,7 @@ def find_in_pdf(markdown: OCRResponse, search_queries: list[str]) -> list[int]:
                 page_numbers.append(page_number)
     return page_numbers
 
-def get_table_of_contents_per_page_pdf(markdown: OCRResponse) -> str:
+def get_table_of_contents_per_page_markdown(markdown: OCRResponse) -> str:
     """
     Get the table of contents of the pdf.
     

src/deepengineer/webcrawler/tools.py

@@ -1,6 +1,6 @@
 from deepengineer.webcrawler.async_search import linkup_search_async, tavily_search_async, arxiv_search_async, pubmed_search_async, sciencedirect_search_async, scientific_search_async
 from deepengineer.webcrawler.async_crawl import crawl4ai_extract_markdown_of_url_async, arxiv_download_pdf_async, download_pdf_async
-from deepengineer.webcrawler.pdf_utils import get_table_of_contents_per_page_pdf
+from deepengineer.webcrawler.pdf_utils import get_table_of_contents_per_page_markdown
 from typing import Callable
 from smolagents.tools import get_json_schema
 
@@ -14,4 +14,4 @@ def print_function_signature_smolagents(tool_function: Callable):
     print("inputs: ", tool_json_schema["parameters"]["properties"])
     print("output_type: ", tool_json_schema["return"]["type"])
 
-print_function_signature_smolagents(get_table_of_contents_per_page_pdf)
+print_function_signature_smolagents(get_table_of_contents_per_page_markdown)

tests/webcrawler/test_pdfs_utils.py

@@ -1,4 +1,4 @@
-from deepengineer.webcrawler.pdf_utils import convert_pdf_to_markdown_async, convert_ocr_response_to_markdown, find_in_pdf, get_table_of_contents_per_page_pdf, get_markdown_by_page_numbers
+from deepengineer.webcrawler.pdf_utils import convert_pdf_to_markdown_async, convert_ocr_response_to_markdown, find_in_markdown, get_table_of_contents_per_page_markdown, get_markdown_by_page_numbers
 from mistralai import OCRResponse
 from deepengineer.common_path import DATA_DIR
 import pytest
@@ -22,12 +22,12 @@ async def test_convert_pdf_to_markdown_async():
 
 def test_table_of_contents_per_page_pdf():
     ocr_response = load_mock_ocr_response()
-    table_of_contents = get_table_of_contents_per_page_pdf(ocr_response)
+    table_of_contents = get_table_of_contents_per_page_markdown(ocr_response)
     assert "References - Page 15" in table_of_contents
 
 def test_find_in_pdf():
     ocr_response = load_mock_ocr_response()
-    page_numbers = find_in_pdf(ocr_response, "where each cylinder represent")
+    page_numbers = find_in_markdown(ocr_response, "where each cylinder represent")
     assert page_numbers == [7]
 
 def test_get_markdown_by_page_numbers():