awacke1/Wikipedia.Chat.Multiplayer

1#In streamlit and python edit this example and add tracking of the option selections by adding buttons for the three choice sets for options.  Also save the values to text file and show full history after an option is recorded.  import streamlit as st2import spacy3import wikipediaapi4import wikipedia5from wikipedia.exceptions import DisambiguationError6from transformers import TFAutoModel, AutoTokenizer7import numpy as np8import pandas as pd9import faiss10import datetime11import time12 13 14st.markdown("""15Scene 1: The Enchanted Castle16You arrive at the enchanted castle, surrounded by a forest of thorns. You have heard stories of a beautiful princess asleep within, waiting for someone to awaken her.17Option 1: Try to make your way through the thorns.18Option 2: Look for another way in.19Sentiment: Feels like harsher trials after passive sleep.20---21Scene 2: The Castle's Secrets22If you made it past the thorns, you discover that the castle is full of hidden chambers, each containing a different trial. 23These trials are designed to test your limits, reveal your inner most desires, and help you understand the suffering of humankind.24Option 1: Enter the first chamber.25Option 2: Continue exploring the castle.26Sentiment: Comedy ending in marriage.27---28Scene 3: The Princess's Awakening29After navigating the castle's trials, you finally reach the chamber where the princess lies sleeping. 30You are faced with the decision of how to awaken her, knowing that your actions will determine the nature of your relationship with her.31Option 1: Awaken her with a gentle kiss.32Option 2: Awaken her through a more assertive act like lifting her up.33Sentiment: Heart forged awakening with different implications depending on context.34""")35 36try:37    nlp = spacy.load("en_core_web_sm")38except:39    spacy.cli.download("en_core_web_sm")40    nlp = spacy.load("en_core_web_sm")41 42wh_words = ['what', 'who', 'how', 'when', 'which']43 44def get_concepts(text):45    text = text.lower()46    doc = nlp(text)47    concepts = []48    for chunk in doc.noun_chunks:49        if chunk.text not in wh_words:50            concepts.append(chunk.text)51    return concepts52 53def get_passages(text, k=100):54    doc = nlp(text)55    passages = []56    passage_len = 057    passage = ""58    sents = list(doc.sents)59    for i in range(len(sents)):60        sen = sents[i]61        passage_len += len(sen)62        if passage_len >= k:63            passages.append(passage)64            passage = sen.text65            passage_len = len(sen)66            continue67        elif i == (len(sents) - 1):68            passage += " " + sen.text69            passages.append(passage)70            passage = ""71            passage_len = 072            continue73        passage += " " + sen.text74    return passages75 76def get_dicts_for_dpr(concepts, n_results=20, k=100):77    dicts = []78    for concept in concepts:79        wikis = wikipedia.search(concept, results=n_results)80        st.write(f"{concept} No of Wikis: {len(wikis)}")81        for wiki in wikis:82            try:83                html_page = wikipedia.page(title=wiki, auto_suggest=False)84            except DisambiguationError:85                continue86            htmlResults = html_page.content87            passages = get_passages(htmlResults, k=k)88            for passage in passages:89                i_dicts = {}90                i_dicts['text'] = passage91                i_dicts['title'] = wiki92                dicts.append(i_dicts)93    return dicts94 95passage_encoder = TFAutoModel.from_pretrained("nlpconnect/dpr-ctx_encoder_bert_uncased_L-2_H-128_A-2")96query_encoder = TFAutoModel.from_pretrained("nlpconnect/dpr-question_encoder_bert_uncased_L-2_H-128_A-2")97p_tokenizer = AutoTokenizer.from_pretrained("nlpconnect/dpr-ctx_encoder_bert_uncased_L-2_H-128_A-2")98q_tokenizer = AutoTokenizer.from_pretrained("nlpconnect/dpr-question_encoder_bert_uncased_L-2_H-128_A-2")99 100def get_title_text_combined(passage_dicts):101    res = []102    for p in passage_dicts:103        res.append(tuple((p['title'], p['text'])))104    return res105 106def extracted_passage_embeddings(processed_passages, max_length=156):107    passage_inputs = p_tokenizer.batch_encode_plus(108                    processed_passages,109                    add_special_tokens=True,110                    truncation=True,111                    padding="max_length",112                    max_length=max_length,113                    return_token_type_ids=True114                )115    passage_embeddings = passage_encoder.predict([np.array(passage_inputs['input_ids']), np.array(passage_inputs['attention_mask']), 116                                            np.array(passage_inputs['token_type_ids'])], 117                                            batch_size=64, 118                                            verbose=1)119    return passage_embeddings120 121def extracted_query_embeddings(queries, max_length=64):122    query_inputs = q_tokenizer.batch_encode_plus(123        queries,124        add_special_tokens=True,125        truncation=True,126        padding="max_length",127        max_length=max_length,128        return_token_type_ids=True129    )130    131    query_embeddings = query_encoder.predict([np.array(query_inputs['input_ids']),132        np.array(query_inputs['attention_mask']),133        np.array(query_inputs['token_type_ids'])],134        batch_size=1,135        verbose=1)136    return query_embeddings137 138def get_pagetext(page):139    s = str(page).replace("/t","")140    return s141 142def get_wiki_summary(search):143    wiki_wiki = wikipediaapi.Wikipedia('en')144    page = wiki_wiki.page(search)                                   145 146 147def get_wiki_summaryDF(search):148    wiki_wiki = wikipediaapi.Wikipedia('en')149    page = wiki_wiki.page(search)150 151    isExist = page.exists()152    if not isExist:153        return isExist, "Not found", "Not found", "Not found", "Not found"154 155    pageurl = page.fullurl156    pagetitle = page.title157    pagesummary = page.summary[0:60]158    pagetext = get_pagetext(page.text)159 160    backlinks = page.backlinks161    linklist = ""162    for link in backlinks.items():163      pui = link[0]164      linklist += pui + " ,  "165      a=1 166      167    categories = page.categories168    categorylist = ""169    for category in categories.items():170      pui = category[0]171      categorylist += pui + " ,  "172      a=1     173    174    links = page.links175    linklist2 = ""176    for link in links.items():177      pui = link[0]178      linklist2 += pui + " ,  "179      a=1 180      181    sections = page.sections182    183    ex_dic = {184      'Entity' : ["URL","Title","Summary", "Text", "Backlinks", "Links", "Categories"],185      'Value': [pageurl, pagetitle, pagesummary, pagetext, linklist,linklist2, categorylist ]186    }187 188    df = pd.DataFrame(ex_dic)189    190    return df191 192 193def save_message(name, message):194    now = datetime.datetime.now()195    timestamp = now.strftime("%Y-%m-%d %H:%M:%S")196    with open("chat.txt", "a") as f:197        f.write(f"{timestamp} - {name}: {message}\n")198 199def press_release():200    st.markdown("""🎉🎊 Breaking News! 📢📣201Introducing StreamlitWikipediaChat - the ultimate way to chat with Wikipedia and the whole world at the same time! 🌎📚👋202Are you tired of reading boring articles on Wikipedia? Do you want to have some fun while learning new things? Then StreamlitWikipediaChat is just the thing for you! 😃💻203With StreamlitWikipediaChat, you can ask Wikipedia anything you want and get instant responses! Whether you want to know the capital of Madagascar or how to make a delicious chocolate cake, Wikipedia has got you covered. 🍰🌍204But that's not all! You can also chat with other people from around the world who are using StreamlitWikipediaChat at the same time. It's like a virtual classroom where you can learn from and teach others. 🌐👨‍🏫👩‍🏫205And the best part? StreamlitWikipediaChat is super easy to use! All you have to do is type in your question and hit send. That's it! 🤯🙌206So, what are you waiting for? Join the fun and start chatting with Wikipedia and the world today! 😎🎉207StreamlitWikipediaChat - where learning meets fun! 🤓🎈""")208 209 210def main():211    st.title("Streamlit Chat")212 213    name = st.text_input("Enter your name")214    message = st.text_input("Enter a topic to share from Wikipedia")215    if st.button("Submit"):216        217        # wiki218        df = get_wiki_summaryDF(message)219        220        save_message(name, message)221        save_message(name, df)222        223        st.text("Message sent!")224 225    226    st.text("Chat history:")227    with open("chat.txt", "a+") as f:228        f.seek(0)229        chat_history = f.read()230    #st.text(chat_history)231    st.markdown(chat_history)232 233    countdown = st.empty()234    t = 60235    while t:236        mins, secs = divmod(t, 60)237        countdown.text(f"Time remaining: {mins:02d}:{secs:02d}")238        time.sleep(1)239        t -= 1240        if t == 0:241            countdown.text("Time's up!")242            with open("chat.txt", "a+") as f:243                f.seek(0)244                chat_history = f.read()245            #st.text(chat_history)246            st.markdown(chat_history)247 248            press_release()249            250            t = 60251 252if __name__ == "__main__":253    main()254