lexoid 0.1.10__tar.gz → 0.1.11__tar.gz

{lexoid-0.1.10 → lexoid-0.1.11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: lexoid
-Version: 0.1.10
+Version: 0.1.11
 Summary: 
 Requires-Python: >=3.10,<4.0
 Classifier: Programming Language :: Python :: 3
@@ -110,20 +110,23 @@ print(parsed_md)
 - **kwargs: Additional arguments for the parser.
 
 ## Benchmark
-Initial results (_more updates soon_)
+Results aggregated across 5 iterations each for 5 documents.
 
 _Note:_ Benchmarks are currently done in the zero-shot setting.
 
-| Rank | Model/Framework | Similarity | Time (s) |
-|------|-----------|------------|----------|
-| 1 | gpt-4o | 0.799 | 21.77|
-| 2 | gemini-2.0-flash-exp | 0.797 | 13.47 |
-| 3 | gemini-exp-1121 | 0.779 | 30.88 |
-| 4 | gemini-1.5-pro | 0.742 | 15.77 |
-| 5 | gpt-4o-mini | 0.721 | 14.86 |
-| 6 | gemini-1.5-flash | 0.702 | 4.56 |
-| 7 | Llama-3.2-11B-Vision-Instruct (via HF) | 0.582 | 21.74 |
-| 8 | Llama-3.2-11B-Vision-Instruct-Turbo (via Together AI) | 0.556 | 4.58 |
-| 9 | Llama-3.2-90B-Vision-Instruct-Turbo (via Together AI) | 0.527 | 10.57 |
-| 10 | Llama-Vision-Free (via Together AI) | 0.435 | 8.42 |
+| Rank | Model | Mean Similarity | Std. Dev. | Time (s) |
+|---|---|---|---|---|
+| 1 | gemini-2.0-flash | 0.829 | 0.102 | 7.41 |
+| 2 | gemini-2.0-flash-001 | 0.814 | 0.176 | 6.85 |
+| 3 | gemini-1.5-flash | 0.797 | 0.143 | 9.54 |
+| 4 | gemini-2.0-pro-exp | 0.764 | 0.227 | 11.95 |
+| 5 | gemini-2.0-flash-thinking-exp | 0.746 | 0.266 | 10.46 |
+| 6 | gemini-1.5-pro | 0.732 | 0.265 | 11.44 |
+| 7 | gpt-4o | 0.687 | 0.247 | 10.16 |
+| 8 | gpt-4o-mini | 0.642 | 0.213 | 9.71 |
+| 9 | gemini-1.5-flash-8b | 0.551 | 0.223 | 3.91 |
+| 10 | Llama-Vision-Free (via Together AI) | 0.531 | 0.198 | 6.93 |
+| 11 | Llama-3.2-11B-Vision-Instruct-Turbo (via Together AI) | 0.524 | 0.192 | 3.68 |
+| 12 | Llama-3.2-90B-Vision-Instruct-Turbo (via Together AI) | 0.461 | 0.306 | 19.26 |
+| 13 | Llama-3.2-11B-Vision-Instruct (via Hugging Face) | 0.451 | 0.257 | 4.54 |
 

{lexoid-0.1.10 → lexoid-0.1.11}/README.md

@@ -77,19 +77,22 @@ print(parsed_md)
 - **kwargs: Additional arguments for the parser.
 
 ## Benchmark
-Initial results (_more updates soon_)
+Results aggregated across 5 iterations each for 5 documents.
 
 _Note:_ Benchmarks are currently done in the zero-shot setting.
 
-| Rank | Model/Framework | Similarity | Time (s) |
-|------|-----------|------------|----------|
-| 1 | gpt-4o | 0.799 | 21.77|
-| 2 | gemini-2.0-flash-exp | 0.797 | 13.47 |
-| 3 | gemini-exp-1121 | 0.779 | 30.88 |
-| 4 | gemini-1.5-pro | 0.742 | 15.77 |
-| 5 | gpt-4o-mini | 0.721 | 14.86 |
-| 6 | gemini-1.5-flash | 0.702 | 4.56 |
-| 7 | Llama-3.2-11B-Vision-Instruct (via HF) | 0.582 | 21.74 |
-| 8 | Llama-3.2-11B-Vision-Instruct-Turbo (via Together AI) | 0.556 | 4.58 |
-| 9 | Llama-3.2-90B-Vision-Instruct-Turbo (via Together AI) | 0.527 | 10.57 |
-| 10 | Llama-Vision-Free (via Together AI) | 0.435 | 8.42 |
+| Rank | Model | Mean Similarity | Std. Dev. | Time (s) |
+|---|---|---|---|---|
+| 1 | gemini-2.0-flash | 0.829 | 0.102 | 7.41 |
+| 2 | gemini-2.0-flash-001 | 0.814 | 0.176 | 6.85 |
+| 3 | gemini-1.5-flash | 0.797 | 0.143 | 9.54 |
+| 4 | gemini-2.0-pro-exp | 0.764 | 0.227 | 11.95 |
+| 5 | gemini-2.0-flash-thinking-exp | 0.746 | 0.266 | 10.46 |
+| 6 | gemini-1.5-pro | 0.732 | 0.265 | 11.44 |
+| 7 | gpt-4o | 0.687 | 0.247 | 10.16 |
+| 8 | gpt-4o-mini | 0.642 | 0.213 | 9.71 |
+| 9 | gemini-1.5-flash-8b | 0.551 | 0.223 | 3.91 |
+| 10 | Llama-Vision-Free (via Together AI) | 0.531 | 0.198 | 6.93 |
+| 11 | Llama-3.2-11B-Vision-Instruct-Turbo (via Together AI) | 0.524 | 0.192 | 3.68 |
+| 12 | Llama-3.2-90B-Vision-Instruct-Turbo (via Together AI) | 0.461 | 0.306 | 19.26 |
+| 13 | Llama-3.2-11B-Vision-Instruct (via Hugging Face) | 0.451 | 0.257 | 4.54 |

{lexoid-0.1.10 → lexoid-0.1.11}/lexoid/api.py

@@ -20,6 +20,7 @@ from lexoid.core.utils import (
     router,
     split_pdf,
     create_sub_pdf,
+    get_webpage_soup,
 )
 
 
@@ -102,7 +103,7 @@ def parse_chunk_list(
 
 def parse(
     path: str,
-    parser_type: Union[str, ParserType] = "LLM_PARSE",
+    parser_type: Union[str, ParserType] = "AUTO",
     pages_per_split: int = 4,
     max_processes: int = 4,
     **kwargs,
@@ -149,6 +150,7 @@ def parse(
             if is_supported_url_file_type(path):
                 path = download_file(path, download_dir)
             elif as_pdf:
+                kwargs["title"] = get_webpage_soup(path).title.string.strip()
                 pdf_filename = kwargs.get("save_filename", f"webpage_{int(time())}.pdf")
                 if not pdf_filename.endswith(".pdf"):
                     pdf_filename += ".pdf"

{lexoid-0.1.10 → lexoid-0.1.11}/lexoid/core/parse_type/llm_parser.py

@@ -50,7 +50,7 @@ def retry_on_http_error(func):
 @retry_on_http_error
 def parse_llm_doc(path: str, **kwargs) -> List[Dict] | str:
     if "model" not in kwargs:
-        kwargs["model"] = "gemini-1.5-flash"
+        kwargs["model"] = "gemini-2.0-flash"
     model = kwargs.get("model")
     if model.startswith("gemini"):
         return parse_with_gemini(path, **kwargs)

{lexoid-0.1.10 → lexoid-0.1.11}/lexoid/core/utils.py

@@ -299,18 +299,7 @@ def html_to_markdown(html: str, title: str, url: str) -> str:
 
     return content
 
-
-def read_html_content(url: str) -> Dict:
-    """
-    Reads the content of an HTML page from the given URL and converts it to markdown or structured content.
-
-    Args:
-        url (str): The URL of the HTML page.
-
-    Returns:
-        Dict: Dictionary containing parsed document data
-    """
-
+def get_webpage_soup(url: str) -> BeautifulSoup:
     try:
         from playwright.async_api import async_playwright
 
@@ -371,6 +360,21 @@ def read_html_content(url: str) -> Dict:
         soup = BeautifulSoup(
             response.content, "html.parser", from_encoding="iso-8859-1"
         )
+    return soup
+
+
+def read_html_content(url: str) -> Dict:
+    """
+    Reads the content of an HTML page from the given URL and converts it to markdown or structured content.
+
+    Args:
+        url (str): The URL of the HTML page.
+
+    Returns:
+        Dict: Dictionary containing parsed document data
+    """
+
+    soup = get_webpage_soup(url)
     title = soup.title.string.strip() if soup.title else "No title"
     url_hash = md5(url.encode("utf-8")).hexdigest()[:8]
     full_title = f"{title} - {url_hash}"
@@ -542,23 +546,37 @@ def has_hyperlink_in_pdf(path: str):
     )
 
 
-def router(path: str):
+def router(path: str, priority: str = "accuracy") -> str:
+    """
+    Routes the file path to the appropriate parser based on the file type.
+
+    Args:
+        path (str): The file path to route.
+        priority (str): The priority for routing: "accuracy" (preference to LLM_PARSE) or "speed" (preference to STATIC_PARSE).
+    """
     file_type = get_file_type(path)
     if file_type.startswith("text/"):
         return "STATIC_PARSE"
-    # Naive routing strategy for now.
-    # Current routing strategy,
-    # 1. If the PDF has hidden hyperlinks (as alias) and no images: STATIC_PARSE
-    # 2. Other scenarios: LLM_PARSE
-    # If you have other needs, do reach out or create an issue.
-    if (
-        file_type == "application/pdf"
-        and not has_image_in_pdf(path)
-        and has_hyperlink_in_pdf(path)
-    ):
-        return "STATIC_PARSE"
-    return "LLM_PARSE"
-
+    
+    if priority == "accuracy":
+        # If the file is a PDF without images but has hyperlinks, use STATIC_PARSE
+        # Otherwise, use LLM_PARSE
+        if (
+            file_type == "application/pdf"
+            and not has_image_in_pdf(path)
+            and has_hyperlink_in_pdf(path)
+        ):
+            return "STATIC_PARSE"
+        return "LLM_PARSE"
+    else:
+        # If the file is a PDF without images, use STATIC_PARSE
+        # Otherwise, use LLM_PARSE
+        if (
+            file_type == "application/pdf"
+            and not has_image_in_pdf(path)
+        ):
+            return "STATIC_PARSE"
+        return "LLM_PARSE"
 
 def convert_doc_to_pdf(input_path: str, temp_dir: str) -> str:
     temp_path = os.path.join(

{lexoid-0.1.10 → lexoid-0.1.11}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "lexoid"
-version = "0.1.10"
+version = "0.1.11"
 description = ""
 authors = []
 readme = "README.md"