feat: removed semchunk and used tikton

Author: VinciGit00

pyproject.toml

@@ -29,7 +29,6 @@ dependencies = [
     "playwright>=1.43.0",
     "undetected-playwright>=0.3.0",
     "google>=3.0.0",
-    "semchunk>=1.0.1",
     "langchain-ollama>=0.1.3",
 ]
 

requirements-dev.lock

@@ -71,7 +71,6 @@ cycler==0.12.1
 dataclasses-json==0.6.7
     # via langchain-community
 dill==0.3.8
-    # via multiprocess
     # via pylint
 distro==1.9.0
     # via openai
@@ -236,13 +235,9 @@ mdurl==0.1.2
     # via markdown-it-py
 minify-html==0.15.0
     # via scrapegraphai
-mpire==2.10.2
-    # via semchunk
 multidict==6.0.5
     # via aiohttp
     # via yarl
-multiprocess==0.70.16
-    # via mpire
 mypy-extensions==1.0.0
     # via typing-inspect
 narwhals==1.3.0
@@ -325,7 +320,6 @@ pyee==11.1.0
     # via playwright
 pygments==2.18.0
     # via furo
-    # via mpire
     # via rich
     # via sphinx
 pylint==3.2.6
@@ -373,8 +367,6 @@ rsa==4.9
     # via google-auth
 s3transfer==0.10.2
     # via boto3
-semchunk==2.2.0
-    # via scrapegraphai
 sf-hamilton==1.73.1
     # via burr
 six==1.16.0
@@ -436,10 +428,8 @@ tornado==6.4.1
 tqdm==4.66.5
     # via google-generativeai
     # via huggingface-hub
-    # via mpire
     # via openai
     # via scrapegraphai
-    # via semchunk
 typing-extensions==4.12.2
     # via altair
     # via anyio

requirements.lock

@@ -41,8 +41,6 @@ charset-normalizer==3.3.2
     # via requests
 dataclasses-json==0.6.7
     # via langchain-community
-dill==0.3.8
-    # via multiprocess
 distro==1.9.0
     # via openai
 exceptiongroup==1.2.2
@@ -155,13 +153,9 @@ marshmallow==3.21.3
     # via dataclasses-json
 minify-html==0.15.0
     # via scrapegraphai
-mpire==2.10.2
-    # via semchunk
 multidict==6.0.5
     # via aiohttp
     # via yarl
-multiprocess==0.70.16
-    # via mpire
 mypy-extensions==1.0.0
     # via typing-inspect
 numpy==1.26.4
@@ -211,8 +205,6 @@ pydantic-core==2.20.1
     # via pydantic
 pyee==11.1.0
     # via playwright
-pygments==2.18.0
-    # via mpire
 pyparsing==3.1.2
     # via httplib2
 python-dateutil==2.9.0.post0
@@ -241,8 +233,6 @@ rsa==4.9
     # via google-auth
 s3transfer==0.10.2
     # via boto3
-semchunk==2.2.0
-    # via scrapegraphai
 six==1.16.0
     # via python-dateutil
 sniffio==1.3.1
@@ -266,10 +256,8 @@ tokenizers==0.19.1
 tqdm==4.66.4
     # via google-generativeai
     # via huggingface-hub
-    # via mpire
     # via openai
     # via scrapegraphai
-    # via semchunk
 typing-extensions==4.12.2
     # via anyio
     # via google-generativeai

scrapegraphai/nodes/parse_node.py

@@ -2,11 +2,10 @@
 ParseNode Module
 """
 from typing import List, Optional
-from semchunk import chunk
 from langchain_community.document_transformers import Html2TextTransformer
 from langchain_core.documents import Document
 from .base_node import BaseNode
-from tokenizer import num_tokens_calculus
+from ..utils.split_text_into_chunks import split_text_into_chunks
 
 class ParseNode(BaseNode):
     """
@@ -69,26 +68,20 @@ def execute(self, state: dict) -> dict:
             docs_transformed = Html2TextTransformer(ignore_links=False).transform_documents(input_data[0])
             docs_transformed = docs_transformed[0]
 
-            chunks = chunk(text=docs_transformed.page_content,
-                            chunk_size=self.node_config.get("chunk_size", 4096)-250,
-                            token_counter=lambda text: len(text.split()),
-                            memoize=False)
+            chunks = split_text_into_chunks(text=docs_transformed.page_content,
+                                            chunk_size=self.node_config.get("chunk_size", 4096)-250)
         else:
             docs_transformed = docs_transformed[0]
 
             chunk_size = self.node_config.get("chunk_size", 4096)
             chunk_size = min(chunk_size - 500, int(chunk_size * 0.9))
 
             if isinstance(docs_transformed, Document):
-                chunks = chunk(text=docs_transformed.page_content,
-                            chunk_size=chunk_size,
-                            token_counter=lambda text: len(text.split()),
-                            memoize=False)
+                chunks = split_text_into_chunks(text=docs_transformed.page_content,
+                                                chunk_size=chunk_size)
             else:
-                chunks = chunk(text=docs_transformed,
-                                chunk_size=chunk_size,
-                                token_counter=lambda text: len(text.split()),
-                                memoize=False)
+                chunks = split_text_into_chunks(text=docs_transformed,
+                                                chunk_size=chunk_size)
 
         state.update({self.output[0]: chunks})
 

scrapegraphai/utils/__init__.py

@@ -1,7 +1,6 @@
 """
     __init__.py file for utils folder
 """
-
 from .convert_to_csv import convert_to_csv
 from .convert_to_json import convert_to_json
 from .prettify_exec_info import prettify_exec_info
@@ -17,3 +16,4 @@
                                                          crop_image)
 from .screenshot_scraping.text_detection import detect_text
 from .tokenizer import num_tokens_calculus
+from .split_text_into_chunks import split_text_into_chunks

scrapegraphai/utils/split_text_into_chunks.py

@@ -0,0 +1,40 @@
+"""
+split_text_into_chunks module
+"""
+from typing import List
+from .tokenizer import num_tokens_calculus  # Import the new tokenizing function
+
+def split_text_into_chunks(text: str, chunk_size: int) -> List[str]:
+    """
+    Splits the text into chunks based on the number of tokens.
+
+    Args:
+        text (str): The text to split.
+        chunk_size (int): The maximum number of tokens per chunk.
+
+    Returns:
+        List[str]: A list of text chunks.
+    """
+    tokens = num_tokens_calculus(text)
+    if tokens <= chunk_size:
+        return [text]
+
+    chunks = []
+    current_chunk = []
+    current_length = 0
+
+    words = text.split()
+    for word in words:
+        word_tokens = num_tokens_calculus(word)
+        if current_length + word_tokens > chunk_size:
+            chunks.append(' '.join(current_chunk))
+            current_chunk = [word]
+            current_length = word_tokens
+        else:
+            current_chunk.append(word)
+            current_length += word_tokens
+
+    if current_chunk:
+        chunks.append(' '.join(current_chunk))
+
+    return chunks