Spaces:
Running
Running
| # app_logic.py | |
| from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM | |
| import fitz, docx, pptx, openpyxl, re, nltk, tempfile, os, easyocr, hashlib, datetime | |
| from nltk.tokenize import sent_tokenize | |
| from fpdf import FPDF | |
| from gtts import gTTS | |
| nltk.download('punkt', quiet=True) | |
| # Load once | |
| MODEL_NAME = "facebook/bart-large-cnn" | |
| tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME) | |
| model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME) | |
| summarizer = pipeline("summarization", model=model, tokenizer=tokenizer, device=-1, batch_size=4) | |
| reader = easyocr.Reader(['en'], gpu=False) | |
| summary_cache = {} | |
| def clean_text(text): | |
| text = re.sub(r'\s+', ' ', text) | |
| text = re.sub(r'\u2022\s*|\d\.\s+', '', text) | |
| text = re.sub(r'\[.*?\]|\(.*?\)', '', text) | |
| text = re.sub(r'\bPage\s*\d+\b', '', text, flags=re.IGNORECASE) | |
| return text.strip() | |
| def extract_text(file_path, file_extension): | |
| try: | |
| if file_extension in ["pdf"]: | |
| with fitz.open(file_path) as doc: | |
| text = "\n".join(page.get_text("text") for page in doc) | |
| if len(text.strip()) < 50: | |
| images = [page.get_pixmap() for page in doc] | |
| temp_img = tempfile.NamedTemporaryFile(suffix=".png", delete=False) | |
| images[0].save(temp_img.name) | |
| ocr_result = reader.readtext(temp_img.name, detail=0) | |
| os.unlink(temp_img.name) | |
| text = "\n".join(ocr_result) if ocr_result else text | |
| elif file_extension in ["docx"]: | |
| doc = docx.Document(file_path) | |
| text = "\n".join(p.text for p in doc.paragraphs) | |
| elif file_extension in ["pptx"]: | |
| prs = pptx.Presentation(file_path) | |
| text = "\n".join(shape.text for slide in prs.slides for shape in slide.shapes if hasattr(shape, "text")) | |
| elif file_extension in ["xlsx"]: | |
| wb = openpyxl.load_workbook(file_path, read_only=True) | |
| text = "\n".join([" ".join(str(cell) for cell in row if cell) for sheet in wb.sheetnames for row in wb[sheet].iter_rows(values_only=True)]) | |
| else: | |
| return "", "Unsupported file type" | |
| return clean_text(text), "" | |
| except Exception as e: | |
| return "", f"Extraction error: {e}" | |
| def chunk_text(text, max_tokens=950): | |
| sentences = sent_tokenize(text) | |
| chunks, current_chunk = [], "" | |
| for sentence in sentences: | |
| if len(tokenizer.encode(current_chunk + " " + sentence)) <= max_tokens: | |
| current_chunk += " " + sentence | |
| else: | |
| chunks.append(current_chunk.strip()) | |
| current_chunk = sentence | |
| if current_chunk: | |
| chunks.append(current_chunk.strip()) | |
| return chunks | |
| def generate_summary(text, length="medium"): | |
| cache_key = hashlib.md5((text + length).encode()).hexdigest() | |
| if cache_key in summary_cache: | |
| return summary_cache[cache_key] | |
| params = {"short": (30, 80), "medium": (80, 200), "long": (210, 300)}[length] | |
| min_len, max_len = params | |
| chunks = chunk_text(text) | |
| summaries = summarizer(chunks, max_length=max_len, min_length=min_len, do_sample=False) | |
| final_summary = " ".join(s['summary_text'] for s in summaries) | |
| summary_cache[cache_key] = final_summary | |
| return final_summary | |
| def text_to_speech(text): | |
| try: | |
| tts = gTTS(text) | |
| temp_audio = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") | |
| tts.save(temp_audio.name) | |
| return temp_audio.name | |
| except: | |
| return "" | |
| def create_pdf(summary, original_filename): | |
| try: | |
| pdf = FPDF() | |
| pdf.add_page() | |
| pdf.set_font("Arial", 'B', 16) | |
| pdf.cell(200, 10, "Summary", ln=True, align='C') | |
| pdf.set_font("Arial", size=12) | |
| pdf.multi_cell(0, 10, summary) | |
| temp_pdf = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") | |
| pdf.output(temp_pdf.name) | |
| return temp_pdf.name | |
| except: | |
| return "" | |