File size: 12,636 Bytes
1b9ef96 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 |
import pandas as pd
import re
import torch
import threading
from transformers import BertTokenizerFast, DistilBertTokenizer, DistilBertForSequenceClassification
from tqdm import tqdm
import tkinter as tk
from tkinter import filedialog, messagebox, scrolledtext, ttk
from tkinter.font import Font
# Check if Metal is available
device = torch.device('mps') if torch.backends.mps.is_available() else torch.device('cpu')
def replace_titles_and_abbreviations(text):
replacements = {
r"Mr\.": "<MR>", r"Ms\.": "<MS>", r"Mrs\.": "<MRS>", r"Dr\.": "<DR>",
r"Prof\.": "<PROF>", r"Rev\.": "<REV>", r"Gen\.": "<GEN>", r"Sen\.": "<SEN>",
r"Rep\.": "<REP>", r"Gov\.": "<GOV>", r"Lt\.": "<LT>", r"Sgt\.": "<SGT>",
r"Capt\.": "<CAPT>", r"Cmdr\.": "<CMDR>", r"Adm\.": "<ADM>", r"Maj\.": "<MAJ>",
r"Col\.": "<COL>", r"St\.": "<ST>", r"Co\.": "<CO>", r"Inc\.": "<INC>",
r"Corp\.": "<CORP>", r"Ltd\.": "<LTD>", r"Jr\.": "<JR>", r"Sr\.": "<SR>",
r"Ph\.D\.": "<PHD>", r"M\.D\.": "<MD>", r"B\.A\.": "<BA>", r"B\.S\.": "<BS>",
r"M\.A\.": "<MA>", r"M\.S\.": "<MS>", r"LL\.B\.": "<LLB>", r"LL\.M\.": "<LLM>",
r"J\.D\.": "<JD>", r"Esq\.": "<ESQ>",
}
for pattern, replacement in replacements.items():
text = re.sub(pattern, replacement, text)
return text
def revert_titles_and_abbreviations(text):
replacements = {
"<MR>": "Mr.", "<MS>": "Ms.", "<MRS>": "Mrs.", "<DR>": "Dr.",
"<PROF>": "Prof.", "<REV>": "Rev.", "<GEN>": "Gen.", "<SEN>": "Sen.",
"<REP>": "Rep.", "<GOV>": "Gov.", "<LT>": "Lt.", "<SGT>": "Sgt.",
"<CAPT>": "Capt.", "<CMDR>": "Cmdr.", "<ADM>": "Adm.", "<MAJ>": "Maj.",
"<COL>": "Col.", "<ST>": "St.", "<CO>": "Co.", "<INC>": "Inc.",
"<CORP>": "Corp.", "<LTD>": "Ltd.", "<JR>": "Jr.", "<SR>": "Sr.",
"<PHD>": "Ph.D.", "<MD>": "M.D.", "<BA>": "B.A.", "<BS>": "B.S.",
"<MA>": "M.A.", "<MS>": "M.S.", "<LLB>": "LL.B.", "<LLM>": "LL.M.",
"<JD>": "J.D.", "<ESQ>": "Esq.",
}
for placeholder, original in replacements.items():
text = re.sub(placeholder, original, text)
return text
def split_text_by_pauses(text):
text = replace_titles_and_abbreviations(text)
pattern = r'[.!,;?:]'
parts = [part.strip() for part in re.split(pattern, text) if part.strip()]
parts_with_punctuation = [
part + text[text.find(part) + len(part)]
if text.find(part) + len(part) < len(text) and text[text.find(part) + len(part)] in '.!,;?'
else part for part in parts
]
parts_with_punctuation = [revert_titles_and_abbreviations(part) for part in parts_with_punctuation]
return parts_with_punctuation
def Process_txt_into_BERT_quotes_input_dataframe(filepath):
tokenizer = BertTokenizerFast.from_pretrained('bert-base-uncased')
with open(filepath, 'r', encoding='utf-8') as file:
text = file.read()
sentences = split_text_by_pauses(text)
data = {
'Text': [],
'Context': [],
'Text start char': [],
'Text end char': [],
'Context start char': [],
'Context end char': [],
'Is Quote': [],
'Speaker': []
}
tokenized_text = tokenizer.tokenize(text)
encoded_text = tokenizer.encode_plus(text, add_special_tokens=False, return_offsets_mapping=True)
offsets = encoded_text['offset_mapping']
for sentence in sentences:
start_idx, end_idx = text.find(sentence), text.find(sentence) + len(sentence)
start_token_idx = next((i for i, offset in enumerate(offsets) if offset[0] == start_idx), None)
end_token_idx = next((i for i, offset in enumerate(offsets) if offset[1] == end_idx), None)
if start_token_idx is not None and end_token_idx is not None:
context_start_token_idx = max(0, start_token_idx - 200)
context_end_token_idx = min(len(tokenized_text), end_token_idx + 200)
context_start_char = offsets[context_start_token_idx][0]
context_end_char = offsets[min(context_end_token_idx, len(offsets) - 1)][1]
context = text[context_start_char:context_end_char]
data['Text'].append(sentence)
data['Context'].append(context)
data['Text start char'].append(start_idx)
data['Text end char'].append(end_idx)
data['Context start char'].append(context_start_char)
data['Context end char'].append(context_end_char)
data['Is Quote'].append('')
data['Speaker'].append('')
df = pd.DataFrame(data)
return df
def predict_quote(context, text, model_checkpoint_path="./quotation_identifer_model/checkpoint-1000"):
formatted_input = f"{context} : Is Sentence Quote : {text}"
model = DistilBertForSequenceClassification.from_pretrained(model_checkpoint_path).to(device)
tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased')
tokenized_input = tokenizer(formatted_input, padding="max_length", truncation=True, max_length=512, return_tensors="pt").to(device)
with torch.no_grad():
outputs = model(**tokenized_input)
predicted_label = torch.argmax(outputs.logits).item()
label_encoder = {0: "Not a Quote", 1: "Quote"}
return label_encoder[predicted_label] == "Quote"
def fill_is_quote_column(df, model_checkpoint_path="./quotation_identifer_model/checkpoint-1000"):
if 'Is Quote' not in df.columns:
df['Is Quote'] = None
tqdm.pandas(desc="Processing rows", unit="row")
for index, row in tqdm(df.iterrows(), total=len(df)):
context = row['Context']
text = row['Text']
df.at[index, 'Is Quote'] = predict_quote(context, text, model_checkpoint_path)
return df
def transfer_quotes(complete_df, incomplete_df):
for index, row in complete_df.iterrows():
is_quote = row['Is Quote']
if pd.notna(is_quote):
incomplete_df.at[index, 'Is Quote'] = is_quote
return incomplete_df
def visualize_quotes(df, is_dark_mode=False):
root = tk.Toplevel()
root.title("Text Visualization")
root.geometry("800x600")
style = ttk.Style(root)
style.theme_use('clam')
main_frame = ttk.Frame(root, padding="20")
main_frame.pack(fill=tk.BOTH, expand=True)
title_font = Font(family="Helvetica", size=24, weight="bold")
title_label = ttk.Label(main_frame, text="Quote Visualization (Identified quotes are highlighted in blue)", font=title_font)
title_label.pack(pady=(0, 20))
text_box = scrolledtext.ScrolledText(main_frame, width=80, height=30, wrap=tk.WORD, font=("Helvetica", 12))
text_box.pack(fill=tk.BOTH, expand=True)
def set_color_scheme(is_dark):
if is_dark:
style.configure("TFrame", background="#2c2c2c")
style.configure("TLabel", background="#2c2c2c", foreground="white")
text_box.config(bg="#2c2c2c", fg="white", insertbackground="white")
text_box.tag_configure('quote', background='#4a86e8', foreground='white')
root.configure(bg="#2c2c2c")
else:
style.configure("TFrame", background="#f0f0f0")
style.configure("TLabel", background="#f0f0f0", foreground="black")
text_box.config(bg="white", fg="black", insertbackground="black")
text_box.tag_configure('quote', background='#4a86e8', foreground='black')
root.configure(bg="#f0f0f0")
def highlight_text():
text_box.delete('1.0', tk.END)
for _, row in df.iterrows():
text = row['Text']
is_quote = row['Is Quote']
if is_quote:
text_box.insert(tk.END, text + "\n", 'quote')
else:
text_box.insert(tk.END, text + "\n")
set_color_scheme(is_dark_mode)
highlight_text()
root.mainloop()
class QuoteIdentifierApp:
def __init__(self, master):
self.master = master
self.master.title("Quote Identifier")
self.master.geometry("600x450")
self.master.resizable(False, False)
self.style = ttk.Style()
self.style.theme_use('clam')
self.is_dark_mode = False
self.create_widgets()
self.set_light_mode()
def create_widgets(self):
self.main_frame = ttk.Frame(self.master, padding="20")
self.main_frame.pack(fill=tk.BOTH, expand=True)
title_font = Font(family="Helvetica", size=24, weight="bold")
title_label = ttk.Label(self.main_frame, text="Quote Identifier", font=title_font)
title_label.pack(pady=(0, 20))
btn_frame = ttk.Frame(self.main_frame)
btn_frame.pack(fill=tk.X, pady=10)
self.open_file_btn = ttk.Button(btn_frame, text="Open Text File", command=self.open_file, style="AccentButton.TButton")
self.open_file_btn.pack(side=tk.LEFT, padx=(0, 10))
self.identify_quotes_btn = ttk.Button(btn_frame, text="Run Identify Quotes", command=self.identify_quotes, style="AccentButton.TButton")
self.identify_quotes_btn.pack(side=tk.LEFT)
self.dark_mode_btn = ttk.Button(self.main_frame, text="Toggle Dark Mode", command=self.toggle_dark_mode, style="TButton")
self.dark_mode_btn.pack(pady=10)
self.status_label = ttk.Label(self.main_frame, text="Ready", font=("Helvetica", 12))
self.status_label.pack(pady=10)
self.progress_bar = ttk.Progressbar(self.main_frame, orient=tk.HORIZONTAL, length=300, mode='determinate')
self.progress_bar.pack(pady=10)
def set_light_mode(self):
self.style.configure("TFrame", background="#f0f0f0")
self.style.configure("TButton", background="#e0e0e0", foreground="black")
self.style.configure("AccentButton.TButton", background="#4a86e8", foreground="white")
self.style.configure("TLabel", background="#f0f0f0", foreground="black")
self.master.configure(bg="#f0f0f0")
self.is_dark_mode = False
def set_dark_mode(self):
self.style.configure("TFrame", background="#2c2c2c")
self.style.configure("TButton", background="#3c3c3c", foreground="white")
self.style.configure("AccentButton.TButton", background="#4a86e8", foreground="white")
self.style.configure("TLabel", background="#2c2c2c", foreground="white")
self.master.configure(bg="#2c2c2c")
self.is_dark_mode = True
def toggle_dark_mode(self):
if self.is_dark_mode:
self.set_light_mode()
else:
self.set_dark_mode()
def open_file(self):
filepath = filedialog.askopenfilename(filetypes=[("Text files", "*.txt")])
if filepath:
self.status_label.config(text=f"File selected: {filepath}")
self.filepath = filepath
else:
self.status_label.config(text="No file selected")
def identify_quotes(self):
if hasattr(self, 'filepath'):
self.status_label.config(text="Processing... Please wait.")
self.progress_bar['value'] = 0
self.master.update()
def process_quotes():
df = Process_txt_into_BERT_quotes_input_dataframe(self.filepath)
df = self.fill_is_quote_column_with_progress(df)
self.master.after(0, lambda: self.finish_processing(df))
threading.Thread(target=process_quotes, daemon=True).start()
else:
messagebox.showwarning("No File Selected", "Please select a text file first.")
def fill_is_quote_column_with_progress(self, df):
if 'Is Quote' not in df.columns:
df['Is Quote'] = None
total_rows = len(df)
for index, row in enumerate(tqdm(df.iterrows(), total=total_rows, desc="Processing rows", unit="row")):
context = row[1]['Context']
text = row[1]['Text']
df.at[index, 'Is Quote'] = predict_quote(context, text)
progress = (index + 1) / total_rows * 100
self.master.after(0, lambda p=progress: self.update_progress(p))
return df
def update_progress(self, value):
self.progress_bar['value'] = value
self.master.update_idletasks()
def finish_processing(self, df):
self.progress_bar['value'] = 100
self.status_label.config(text="Quote identification complete!")
visualize_quotes(df, self.is_dark_mode)
def create_gui():
root = tk.Tk()
app = QuoteIdentifierApp(root)
root.mainloop()
if __name__ == "__main__":
create_gui() |