utils.py

import re
import functools

import requests
import pandas as pd
import plotly.express as px

import torch
import gradio as gr
from transformers import pipeline, Wav2Vec2ProcessorWithLM
from pyannote.audio import Pipeline
from librosa import load, resample
import whisperx

import re
alphabets= "([A-Za-z])"
prefixes = "(Mr|St|Mrs|Ms|Dr)[.]"
suffixes = "(Inc|Ltd|Jr|Sr|Co)"
starters = "(Mr|Mrs|Ms|Dr|He\s|She\s|It\s|They\s|Their\s|Our\s|We\s|But\s|However\s|That\s|This\s|Wherever)"
acronyms = "([A-Z][.][A-Z][.](?:[A-Z][.])?)"
websites = "[.](com|net|org|io|gov)"

def split_into_sentences(text):
    text = " " + text + "  "
    text = text.replace("\n"," ")
    text = re.sub(prefixes,"\\1<prd>",text)
    text = re.sub(websites,"<prd>\\1",text)
    if "Ph.D" in text: text = text.replace("Ph.D.","Ph<prd>D<prd>")
    text = re.sub("\s" + alphabets + "[.] "," \\1<prd> ",text)
    text = re.sub(acronyms+" "+starters,"\\1<stop> \\2",text)
    text = re.sub(alphabets + "[.]" + alphabets + "[.]" + alphabets + "[.]","\\1<prd>\\2<prd>\\3<prd>",text)
    text = re.sub(alphabets + "[.]" + alphabets + "[.]","\\1<prd>\\2<prd>",text)
    text = re.sub(" "+suffixes+"[.] "+starters," \\1<stop> \\2",text)
    text = re.sub(" "+suffixes+"[.]"," \\1<prd>",text)
    text = re.sub(" " + alphabets + "[.]"," \\1<prd>",text)
    if "”" in text: text = text.replace(".”","”.")
    if "\"" in text: text = text.replace(".\"","\".")
    if "!" in text: text = text.replace("!\"","\"!")
    if "?" in text: text = text.replace("?\"","\"?")
    text = text.replace(".",".<stop>")
    text = text.replace("?","?<stop>")
    text = text.replace("!","!<stop>")
    text = text.replace("<prd>",".")
    sentences = text.split("<stop>")
    sentences = sentences[:-1]
    sentences = [s.strip() for s in sentences]
    return sentences


def summarize(diarized, check, summarization_pipeline):
    """
    diarized: a list of tuples. Each tuple has a string to be displayed and a label for highlighting.
        The start/end times are not highlighted [(speaker text, speaker id), (start time/end time, None)]
    check is a list of speaker ids whose speech will get summarized
    """

    if not check:
        return ""

    # Combine text based on the speaker id
    text_lines = [f"{d[1]}: {d[0]}" if len(check) == 2 and d[1] is not None else d[0] for d in diarized if d[1] in check]
    text = "\n".join(text_lines)

    # Cache the inner function because the outer function cannot be cached
    @functools.lru_cache(maxsize=128)
    def call_summarize_api(text):
        return summarization_pipeline(text)[0]["summary_text"]

    return call_summarize_api(text)


# display if the sentiment value is above these thresholds
thresholds = {
    "joy": 0.99,
    "anger": 0.95,
    "surprise": 0.95,
    "sadness": 0.98,
    "fear": 0.95,
    "love": 0.99,
}

color_map = {
    "joy": "green",
    "anger": "red",
    "surprise": "yellow",
    "sadness": "blue",
    "fear": "orange",
    "love": "purple",
}


def sentiment(diarized, emotion_pipeline):
    def split_into_intervals(speaker_speech, start_time, end_time):
        sentences = split_into_sentences(speaker_speech)
        interval_size = (end_time - start_time) / len(sentences)
        return sentences, interval_size

    def process_customer_emotion(outputs, sentences, start_time, interval_size):
        sentiments = []
        for idx, (o, t) in enumerate(zip(outputs, sentences)):
            sent = "neutral"
            if o["score"] > thresholds[o["label"]]:
                sentiments.append((t + f"({round(idx*interval_size+start_time,1)} s)", o["label"]))
                if o["label"] in {"joy", "love", "surprise"}:
                    sent = "positive"
                elif o["label"] in {"sadness", "anger", "fear"}:
                    sent = "negative"
            if sent != "neutral":
                to_plot.append((start_time + idx * interval_size, sent))
                plot_sentences.append(t)
        return sentiments

    x_min = 100
    x_max = 0

    customer_sentiments, to_plot, plot_sentences = [], [], []

    for i in range(0, len(diarized), 2):
        speaker_speech, speaker_id = diarized[i]
        times, _ = diarized[i + 1]
        start_time, end_time = map(float, times[5:].split("-"))
        x_min, x_max = min(x_min, start_time), max(x_max, end_time)

        if "Customer" in speaker_id:
            sentences, interval_size = split_into_intervals(speaker_speech, start_time, end_time)
            outputs = emotion_pipeline(sentences)
            customer_sentiments.extend(process_customer_emotion(outputs, sentences, start_time, interval_size))

    plot_df = pd.DataFrame(data={"x": [x for x, _ in to_plot], "y": [y for _, y in to_plot], "sentence": plot_sentences})
    fig = px.line(plot_df, x="x", y="y", hover_data={"sentence": True, "x": True, "y": False}, labels={"x": "time (seconds)", "y": "sentiment"}, title=f"Customer sentiment over time", markers=True)
    fig.update_yaxes(categoryorder="category ascending")
    fig.update_layout(font=dict(size=18), xaxis_range=[x_min - 5, x_max + 5])

    return customer_sentiments, fig

def speech_to_text(speech_file, speaker_segmentation, whisper, alignment_model, metadata, whisper_device):
    
    def process_chunks(turn, chunks):
        diarized = ""
        i = 0
        while i < len(chunks) and chunks[i]["end"] <= turn.end:
            diarized += chunks[i]["text"] + " "
            i += 1
        return diarized, i

    speaker_output = speaker_segmentation(speech_file)
    result = whisper.transcribe(speech_file)
    chunks = whisperx.align(result["segments"], alignment_model, metadata, speech_file, whisper_device)["word_segments"]

    diarized_output = []
    i = 0
    speaker_counter = 0

    for turn, _, _ in speaker_output.itertracks(yield_label=True):
        speaker = "Customer" if speaker_counter % 2 == 0 else "Support"
        diarized, i = process_chunks(turn, chunks[i:])
        if diarized:
            diarized_output.extend([(diarized, speaker), (f"from {turn.start:.2f}-{turn.end:.2f}", None)])
            speaker_counter += 1

    return diarized_output