Demo.py

import streamlit as st
import sparknlp
import os
import pandas as pd

from sparknlp.base import *
from sparknlp.annotator import *
from pyspark.ml import Pipeline
from sparknlp.pretrained import PretrainedPipeline

# Page configuration
st.set_page_config(
    layout="wide",  
    initial_sidebar_state="auto"
)

# CSS for styling
st.markdown("""
    <style>
        .main-title {
            font-size: 36px;
            color: #4A90E2;
            font-weight: bold;
            text-align: center;
        }
        .section p, .section ul {
            color: #666666;
        }
    </style>
""", unsafe_allow_html=True)

@st.cache_resource
def init_spark():
    return sparknlp.start()

@st.cache_resource
def create_pipeline(model):
  document_assembler = DocumentAssembler()\
      .setInputCol("text")\
      .setOutputCol("document")

  sentence_detector = SentenceDetector() \
      .setInputCols(["document"]) \
      .setOutputCol("sentence")

  tokenizer = Tokenizer() \
      .setInputCols(["sentence"]) \
      .setOutputCol("token")

  word_embeddings = WordEmbeddingsModel()\
      .pretrained('urduvec_140M_300d', 'ur')\
      .setInputCols(["sentence",'token'])\
      .setOutputCol("word_embeddings")

  sentence_embeddings = SentenceEmbeddings() \
      .setInputCols(["sentence", "word_embeddings"]) \
      .setOutputCol("sentence_embeddings") \
      .setPoolingStrategy("AVERAGE")

  classifier = SentimentDLModel.pretrained('sentimentdl_urduvec_imdb', 'ur' )\
      .setInputCols(['sentence_embeddings'])\
      .setOutputCol('sentiment')

  nlpPipeline = Pipeline(
      stages=[ 
          document_assembler,
          sentence_detector, 
          tokenizer,
          word_embeddings,
          sentence_embeddings,
          classifier ])
  
  return nlpPipeline

def fit_data(pipeline, data):
    empty_df = spark.createDataFrame([['']]).toDF('text')
    pipeline_model = pipeline.fit(empty_df)
    model = LightPipeline(pipeline_model)
    results = model.fullAnnotate(data)[0]

    return results['sentiment'][0].result

# Set up the page layout
st.markdown('<div class="main-title">State-of-the-Art Urdu Sentiment Detection with Spark NLP</div>', unsafe_allow_html=True)

# Sidebar content
model = st.sidebar.selectbox(
    "Choose the pretrained model",
    ["sentimentdl_urduvec_imdb"],
    help="For more info about the models visit: https://sparknlp.org/models"
)

# Reference notebook link in sidebar
link = """
<a href="https://colab.research.google.com/github/JohnSnowLabs/spark-nlp-workshop/blob/master/tutorials/streamlit_notebooks/public/SENTIMENT_UR.ipynb">
    <img src="https://colab.research.google.com/assets/colab-badge.svg" style="zoom: 1.3" alt="Open In Colab"/>
</a>
"""
st.sidebar.markdown('Reference notebook:')
st.sidebar.markdown(link, unsafe_allow_html=True)

# Load examples
folder_path = f"inputs/{model}"
examples = [
    lines[1].strip() 
    for filename in os.listdir(folder_path) 
    if filename.endswith('.txt') 
    for lines in [open(os.path.join(folder_path, filename), 'r', encoding='utf-8').readlines()] 
    if len(lines) >= 2
]

selected_text = st.selectbox("Select a sample", examples)
custom_input = st.text_input("Try it for yourself!")

if custom_input:
    selected_text = custom_input
elif selected_text:
    selected_text = selected_text

st.subheader('Selected Text')
st.write(selected_text)

# Initialize Spark and create pipeline
spark = init_spark()
pipeline = create_pipeline(model)
output = fit_data(pipeline, selected_text)

# Display output sentence
if output.lower() in ['pos', 'positive']:
  st.markdown("""<h3>This seems like a <span style="color: green">{}</span> text. <span style="font-size:35px;">&#128515;</span></h3>""".format('positive'), unsafe_allow_html=True)
elif output.lower() in ['neg', 'negative']:
  st.markdown("""<h3>This seems like a <span style="color: red">{}</span> text. <span style="font-size:35px;">&#128544;</span?</h3>""".format('negative'), unsafe_allow_html=True)