Create app.py

app.py

@@ -0,0 +1,192 @@
+import os
+import re
+import json
+import fitz
+from PIL import Image
+import pytesseract
+import spacy
+import gradio as gr
+
+# --- Global Configuration and Initialization ---
+# Load the spaCy model once globally
+nlp = spacy.load("en_core_web_sm")
+
+# On Hugging Face Spaces, Tesseract is usually in the PATH.
+# If you encounter issues, you might need to specify the path, but generally not needed.
+# pytesseract.pytesseract.tesseract_cmd = r'/usr/bin/tesseract' # Example path for Linux
+
+def extract_text_from_pdf(pdf_path):
+    """Extracts text from a PDF file."""
+    text = ""
+    try:
+        with fitz.open(pdf_path) as doc:
+            for page in doc:
+                text += page.get_text()
+    except Exception as e:
+        print(f"Error reading PDF {pdf_path}: {e}")
+    return text
+
+def extract_text_from_image(image_path):
+    """Extracts text from an image file using OCR."""
+    text = ""
+    try:
+        text = pytesseract.image_to_string(Image.open(image_path))
+    except Exception as e:
+        print(f"Error reading image {image_path}: {e}")
+    return text
+
+def parse_sections(text):
+    """Splits the resume text into logical sections."""
+    sections = {
+        'contact_info': '',
+        'experience': '',
+        'education': '',
+        'projects': '',
+        'skills': '',
+        'summary': ''
+    }
+
+    section_keywords = {
+        'experience': [r'\bexperience\b', r'work history', r'professional experience'],
+        'education': [r'\beducation\b'],
+        'projects': [r'\bprojects\b', r'personal projects'],
+        'skills': [r'\bskills\b', r'technical skills'],
+        'summary': [r'\bsummary\b', r'profile', r'objective']
+    }
+
+    lines = text.split('\n')
+    current_section = 'contact_info'
+
+    for line in lines:
+        if not line.strip():
+            continue
+
+        found_section = False
+        for section, keywords in section_keywords.items():
+            for keyword in keywords:
+                if re.search(keyword, line, re.IGNORECASE):
+                    current_section = section
+                    found_section = True
+                    break
+            if found_section:
+                break
+
+        if current_section:
+            sections[current_section] += line + '\n'
+
+    return sections
+
+def extract_accurate_information(text):
+    """Extracts structured information from raw text using a section-based approach."""
+
+    data = {
+        "first_name": None, "middle_name": None, "last_name": None, "email": None,
+        "phone": None, "major": None, "graduation_year": None,
+        "experience_years": None, "experience": [], "project_names": [],
+        "location": None
+    }
+
+    sections = parse_sections(text)
+    contact_section = sections['contact_info']
+
+    # Regex for email and Egyptian phone numbers
+    email_regex = r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b'
+    phone_regex = r'\b(01[0125]\d{8})\b'
+
+    data['email'] = re.search(email_regex, contact_section).group(0) if re.search(email_regex, contact_section) else None
+    data['phone'] = re.search(phone_regex, contact_section).group(0) if re.search(phone_regex, contact_section) else None
+
+    # Extract Name
+    contact_lines = [line.strip() for line in contact_section.split('\n') if line.strip()]
+    if contact_lines:
+        full_name = contact_lines[0]
+        if not data['email'] or data['email'] not in full_name:
+            if not data['phone'] or data['phone'] not in full_name:
+                name_parts = full_name.split()
+                if len(name_parts) > 0:
+                    data['first_name'] = name_parts[0]
+                if len(name_parts) > 2:
+                    data['middle_name'] = " ".join(name_parts[1:-1])
+                    data['last_name'] = name_parts[-1]
+                elif len(name_parts) == 2:
+                    data['last_name'] = name_parts[1]
+
+    # Extract Location using spaCy (globally loaded nlp object)
+    doc = nlp(contact_section)
+    for ent in doc.ents:
+        if ent.label_ == "GPE":
+            data["location"] = ent.text
+            break
+
+    # Education
+    education_section = sections['education']
+    if education_section:
+        years = re.findall(r'\b(20\d{2})\b', education_section)
+        if years:
+            data['graduation_year'] = max([int(y) for y in years])
+
+        for line in education_section.split('\n'):
+            if "bachelor" in line.lower() or "business information system" in line.lower():
+                data['major'] = line.strip()
+                break
+
+    # Experience
+    experience_section = sections['experience']
+    if experience_section:
+        data['experience'] = [
+            line.strip() for line in experience_section.split('\n')
+            if line.strip() and not re.match(r'\bexperience\b', line, re.IGNORECASE)
+        ]
+
+    # Projects
+    projects_section = sections['projects']
+    if projects_section:
+        project_lines = [
+            line.strip() for line in projects_section.split('\n')
+            if line.strip() and not re.match(r'\bprojects\b', line, re.IGNORECASE)
+        ]
+        data['project_names'] = [re.sub(r'^[•\-\*]\s*', '', line).strip('.') for line in project_lines]
+
+    return data
+
+def process_resume(file):
+    """Gradio interface function to process an uploaded resume file."""
+    if file is None:
+        return "Please upload a resume file.", {}
+
+    file_path = file.name # Gradio passes a NamedTemporaryFile object
+    _, file_extension = os.path.splitext(file_path)
+    text = ""
+
+    if file_extension.lower() == ".pdf":
+        text = extract_text_from_pdf(file_path)
+    elif file_extension.lower() in [".png", ".jpg", ".jpeg", ".tiff"]:
+        text = extract_text_from_image(file_path)
+    else:
+        return f"Unsupported file format: {file_extension}. Please upload a PDF or image file.", {}
+
+    if text:
+        extracted_data = extract_accurate_information(text)
+        if extracted_data:
+            return "Resume processed successfully!", json.dumps(extracted_data, indent=4)
+    return "Failed to extract information from the resume. Please check the file format and content.", {}
+
+# --- Gradio Interface ---
+iface = gr.Interface(
+    fn=process_resume,
+    inputs=gr.File(type="filepath", label="Upload Resume (PDF or Image)"),
+    outputs=[
+        gr.Textbox(label="Status"),
+        gr.Json(label="Extracted Data")
+    ],
+    title="Resume Parser",
+    description="Upload a resume (PDF or image) to extract key information.",
+    allow_flagging="never",
+    examples=[
+        # You can add example files here if you have them.
+        # For example: "./examples/sample_resume.pdf"
+    ]
+)
+
+if __name__ == "__main__":
+    iface.launch()