#!/usr/bin/env python # coding: utf-8 # In[1]: import validators, re from fake_useragent import UserAgent from bs4 import BeautifulSoup import streamlit as st from transformers import pipeline import time import base64 import requests import docx2txt from io import StringIO from PyPDF2 import PdfFileReader import warnings from nltk import sent_tokenize warnings.filterwarnings("ignore") # In[2]: time_str = time.strftime("%d%m%Y-%H%M%S") #Functions def article_text_extractor(url: str): '''Extract text from url and divide text into chunks if length of text is more than 500 words''' ua = UserAgent() headers = {'User-Agent':str(ua.chrome)} r = requests.get(url,headers=headers) soup = BeautifulSoup(r.text, "html.parser") title_text = soup.find_all(["h1"]) para_text = soup.find_all(["p"]) article_text = [result.text for result in para_text] article_header = [result.text for result in title_text][0] article = " ".join(article_text) article = article.replace(".", ".") article = article.replace("!", "!") article = article.replace("?", "?") sentences = article.split("") current_chunk = 0 chunks = [] for sentence in sentences: if len(chunks) == current_chunk + 1: if len(chunks[current_chunk]) + len(sentence.split(" ")) <= 500: chunks[current_chunk].extend(sentence.split(" ")) else: current_chunk += 1 chunks.append(sentence.split(" ")) else: print(current_chunk) chunks.append(sentence.split(" ")) for chunk_id in range(len(chunks)): chunks[chunk_id] = " ".join(chunks[chunk_id]) return article_header, chunks def chunk_clean_text(text): sentences = sent_tokenize(text) current_chunk = 0 chunks = [] for sentence in sentences: if len(chunks) == current_chunk + 1: if len(chunks[current_chunk]) + len(sentence.split(" ")) <= 500: chunks[current_chunk].extend(sentence.split(" ")) else: current_chunk += 1 chunks.append(sentence.split(" ")) else: print(current_chunk) chunks.append(sentence.split(" ")) for chunk_id in range(len(chunks)): chunks[chunk_id] = " ".join(chunks[chunk_id]) return chunks def preprocess_plain_text(x): x = x.encode("ascii", "ignore").decode() # unicode x = re.sub(r"https*\S+", " ", x) # url x = re.sub(r"@\S+", " ", x) # mentions x = re.sub(r"#\S+", " ", x) # hastags x = re.sub(r"\s{2,}", " ", x) # over spaces x = re.sub("[^.,!?A-Za-z0-9]+", " ", x) # special charachters except .,!? return x def extract_pdf(file): '''Extract text from PDF file''' pdfReader = PdfFileReader(file) count = pdfReader.numPages all_text = "" for i in range(count): page = pdfReader.getPage(i) all_text += page.extractText() return all_text def extract_text_from_file(file): '''Extract text from uploaded file''' # read text file if file.type == "text/plain": # To convert to a string based IO: stringio = StringIO(file.getvalue().decode("utf-8")) # To read file as string: file_text = stringio.read() # read pdf file elif file.type == "application/pdf": file_text = extract_pdf(file) # read docx file elif ( file.type == "application/vnd.openxmlformats-officedocument.wordprocessingml.document" ): file_text = docx2txt.process(file) return file_text def summary_downloader(raw_text): b64 = base64.b64encode(raw_text.encode()).decode() new_filename = "new_text_file_{}_.txt".format(time_str) st.markdown("#### Download Summary as a File ###") href = f'Click to Download!!' st.markdown(href,unsafe_allow_html=True) @st.cache(allow_output_mutation=True) def facebook_model(): summarizer = pipeline('summarization',model='facebook/bart-large-cnn') return summarizer @st.cache(allow_output_mutation=True) def schleifer_model(): summarizer = pipeline('summarization',model='sshleifer/distilbart-cnn-12-6') return summarizer #Streamlit App st.title("Article Text and Link Extractive Summarizer 📝") model_type = st.sidebar.selectbox( "Model type", options=["Facebook-Bart", "Sshleifer-DistilBart"] ) max_len= st.sidebar.slider("Maximum length of the summarized text",min_value=100,max_value=500) min_len= st.sidebar.slider("Minimum length of the summarized text",min_value=30) st.markdown( "Model Source: [Facebook-Bart-large-CNN](https://huggingface.co/facebook/bart-large-cnn) and [Sshleifer-distilbart-cnn-12-6](https://huggingface.co/sshleifer/distilbart-cnn-12-6)" ) st.markdown( """The app supports extractive summarization which aims to identify the salient information that is then extracted and grouped together to form a concise summary. For documents or text that is more than 500 words long, the app will divide the text into chunks and summarize each chunk. Please note when using the sidebar slider, those values represent the min/max text length per chunk of text to be summarized. If your article to be summarized is 1000 words, it will be divided into two chunks of 500 words first then the default max length of 100 words is applied per chunk, resulting in a summarized text with 200 words maximum. There are two models available to choose from:""") st.markdown(""" - Facebook-Bart, trained on large [CNN and Daily Mail](https://huggingface.co/datasets/cnn_dailymail) news articles. - Sshleifer-Distilbart, which is a distilled (smaller) version of the large Bart model.""" ) st.markdown("""Please do note that the model will take longer to generate summaries for documents that are too long.""") st.markdown( "The app only ingests the below formats for summarization task:" ) st.markdown( """- Raw text entered in text box. - URL of an article to be summarized. - Documents with .txt, .pdf or .docx file formats.""" ) st.markdown("---") url_text = st.text_input("Please Enter a url here") st.markdown( "

OR

", unsafe_allow_html=True, ) plain_text = st.text_input("Please Paste/Enter plain text here") st.markdown( "

OR

", unsafe_allow_html=True, ) upload_doc = st.file_uploader( "Upload a .txt, .pdf, .docx file for summarization" ) is_url = validators.url(url_text) if is_url: # complete text, chunks to summarize (list of sentences for long docs) article_title,chunks = article_text_extractor(url=url_text) elif upload_doc: clean_text = chunk_clean_text(preprocess_plain_text(extract_text_from_file(upload_doc))) else: clean_text = chunk_clean_text(preprocess_plain_text(plain_text)) summarize = st.button("Summarize") # called on toggle button [summarize] if summarize: if model_type == "Facebook-Bart": if is_url: text_to_summarize = chunks else: text_to_summarize = clean_text with st.spinner( text="Loading Facebook-Bart Model and Extracting summary. This might take a few seconds depending on the length of your text..." ): summarizer_model = facebook_model() summarized_text = summarizer_model(text_to_summarize, max_length=max_len, min_length=min_len) summarized_text = ' '.join([summ['summary_text'] for summ in summarized_text]) elif model_type == "Sshleifer-DistilBart": if is_url: text_to_summarize = chunks else: text_to_summarize = clean_text with st.spinner( text="Loading Sshleifer-DistilBart Model and Extracting summary. This might take a few seconds depending on the length of your text..." ): summarizer_model = schleifer_model() summarized_text = summarizer_model(text_to_summarize, max_length=max_len, min_length=min_len) summarized_text = ' '.join([summ['summary_text'] for summ in summarized_text]) # final summarized output st.subheader("Summarized text") if is_url: # view summarized text (expander) st.markdown(f"Article title: {article_title}") st.write(summarized_text) summary_downloader(summarized_text) # In[ ]: