import openai import gradio as gr import PyPDF2 import docx from docx import Document from gradio import CSVLogger import os from reportlab.lib.pagesizes import letter from reportlab.pdfgen import canvas from reportlab.lib.styles import getSampleStyleSheet from reportlab.platypus import Paragraph from reportlab.lib.units import inch from reportlab.lib.pagesizes import letter from reportlab.pdfgen import canvas from docx import Document from docx.shared import Pt from docx.enum.text import WD_UNDERLINE from docx.enum.text import WD_ALIGN_PARAGRAPH from docx.shared import RGBColor from docx.enum.text import WD_PARAGRAPH_ALIGNMENT from docx.oxml import parse_xml from docx.oxml.ns import nsdecls openai.api_key = "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" messages = [] latest_resume_text = "" desired_position = "" default_text_1 = "Hello, I will share a candidate's resume with you. Please provide a summary for me" default_text_2 = "Hello, I will share a candidate's resume with you. Please desensitize the resume by removing or redacting any personal contact information like the phone number, home address, and email address. Also remove or mask any private details like ID numbers, exact dates for education and work history, and compensation details. Double check that no other private information is included. The goal is to remove personally identifying details while still allowing this candidate's professional qualifications to be visible to potential employers." default_text_3 = "请帮我阅读候选人简历并总结出以下几个部分。第一个板块是“个人信息”，请总结出此后选人的姓名，性别，工作经验（多少年），最高学历，毕业院校，专业，和毕业时间。第二个部分是此候选人的本人评价，放进一个自然段里。第三个部分是具体得工作经历。第四个部分是做过的项目经验。请根据你对于这位候选人简历的最细致的阅读排出以上几个部分。每个部分由自己的标题：五个标题为 个人信息，本人评价，工作经历，和项目经验。请把每一个小项写得细致一点，并且用数字排序！" def set_text_1(): return default_text_1 def set_text_2(): return default_text_2 def set_text_3(): return default_text_3 # # extracting text from pdf # def extract_text_from_pdf(file): # reader = PyPDF2.PdfReader(file) # text = "" # for page in reader.pages: # text += page.extract_text() # return text #write a function that extract text from a pdf that could contain multiple pages def extract_text_from_pdf(file): reader = PyPDF2.PdfReader(file) text = "" for page in reader.pages: text += page.extract_text() return text # extracting text from doc def extract_text_from_docx(file): doc = docx.Document(file) all_text = [] for para in doc.paragraphs: # Extract the text from the current paragraph paragraph_text = para.text all_text.append(paragraph_text) combined_text = " ".join(all_text) return combined_text def handle_file_upload(uploaded_files): combined_text = "" if uploaded_files is None: return "" for uploaded_file in uploaded_files: file_type = uploaded_file.name.split('.')[-1].lower() if file_type == 'pdf': combined_text += extract_text_from_pdf(uploaded_file) elif file_type in ['docx', 'doc']: combined_text += extract_text_from_docx(uploaded_file) else: combined_text += "Unsupported file format, please upload a PDF or Word document.\n" return combined_text def clear_inputs(): messages = [] return "", None def log_conversation(input, reply): with open("log.csv", "a") as log_file: log_file.write(f"user_input: {input},\n\n Chatgpt: {reply}\n\n") import re from docx import Document from docx.enum.text import WD_ALIGN_PARAGRAPH from docx.shared import RGBColor from docx.enum.text import WD_PARAGRAPH_ALIGNMENT def is_chinese_char(char): """Check if a given character is a Chinese character.""" return '\u4e00' <= char <= '\u9fff' def count_chinese_chars(text): """Count the number of Chinese characters in a string.""" return sum(is_chinese_char(char) for char in text) def generate_resume_document(latest_resume_text): # Define keywords that indicate the start of a new section section_keywords = ["个人信息", "本人评价", "工作经历", "项目经验"] # Create a new Document doc = Document() # Add a title to the document heading = doc.add_heading('候选人简历', level=0) run = heading.add_run() run.underline = WD_UNDERLINE.SINGLE # Center-align the heading paragraph_format = heading.paragraph_format paragraph_format.alignment = WD_ALIGN_PARAGRAPH.CENTER # Check if the text is empty if not latest_resume_text.strip(): doc.add_paragraph("No resume text found") return None # Remove non-word characters from the text except for spaces and new lines # Remove non-word characters from the text except for "-", ",", "." and numbers cleaned_text = re.sub(r'[^\w\s\-\,\.\d]', '', latest_resume_text) # Split the resume text into lines lines = cleaned_text.split('\n') # Variables to store the current section title and its content current_section = None current_content = [] # Function to add a section to the document with bullet points def add_section_to_doc(section, content): global name if section and content: sec = doc.add_heading(section, level=0) for run in sec.runs: run.font.size = Pt(14) if section == "个人信息": # Combine all content lines into a single line separated by a semicolon combined_content = '；'.join(content) p = doc.add_paragraph() # Create a new paragraph for combined content last_index = 0 # Keep track of the last index processed for keyword in ["姓名", "性别", "工作经验", "最高学历", "毕业院校", "专业", "毕业时间"]: if keyword in combined_content: start_index = combined_content.index(keyword, last_index) # Add text before the keyword as a normal run p.add_run(combined_content[last_index:start_index]) # Add the keyword and the colon as a bold run bold_run = p.add_run(keyword + '：') bold_run.bold = True # Update the last index processed last_index = start_index + len(keyword) if keyword == "姓名": name_start_index = last_index try: # Try to find the start index of the next keyword "性别" name_end_index = combined_content.index("性别", name_start_index) # Store the content into name variable name = combined_content[name_start_index:name_end_index] # Remove any non-Chinese characters name = re.sub(r'[^\u4e00-\u9fff]', '', name) except ValueError: # Handle the case where "性别" is not found name = "需手动填写" # Add any remaining text after the last keyword p.add_run(combined_content[last_index:]) else: for line in content: # # Count Chinese characters in the line # chinese_char_count = count_chinese_chars(line) # if chinese_char_count < 5: # doc.add_heading(line, level=0) # p = doc.add_paragraph() # p.alignment = WD_PARAGRAPH_ALIGNMENT.CENTER # for run in sec.runs: # run.font.size = Pt(14) # else: doc.add_paragraph(line) for line in lines: # Check for section titles (considering case-insensitivity) if any(line.strip().upper() == keyword for keyword in section_keywords): # Add the previous section to the document before starting a new one add_section_to_doc(current_section, current_content) current_section = line.strip().title() # Title Case for headings current_content = [] else: # Clean line and add to current content clean_line = line.strip() if clean_line: # Ignore empty lines # Append clean_line to current_content, removing any additional internal line breaks current_content.append(re.sub(r'\s+', ' ', clean_line)) # Add the last section to the document add_section_to_doc(current_section, current_content) # Save the document word_filename = "博网科技-" + name + "-" + desired_position + ".docx" doc.save(word_filename) return word_filename def CustomChatGPT(user_input, uploaded_file): global messages, latest_resume_text, desired_position, name # Declare latest_resume_text as global if it's used globally resume_text = "" resume_text = handle_file_upload(uploaded_file) print("Resume text from the file:", resume_text) if resume_text: messages.append({"role": "system", "content": resume_text}) # if the resume text contains "期望职位", extract the desired position and store it in desired_position if "期望职位" in resume_text: match = re.search(r'期望职位：(.+?)\s', resume_text) if match: desired_position = match.group(1) # Remove non-Chinese characters desired_position = re.sub(r'[^\u4e00-\u9fff]', '', desired_position) else: desired_position = "需手动填写" messages.append({"role": "user", "content": user_input}) response = openai.ChatCompletion.create( model="gpt-3.5-turbo", messages=messages ) ChatGPT_reply = response.choices[0].message.content messages.append({"role": "assistant", "content": ChatGPT_reply}) combined_input = user_input + " " + resume_text log_conversation(combined_input, ChatGPT_reply) latest_resume_text = ChatGPT_reply if "请帮我阅读候选人简历" in user_input: word_filename = generate_resume_document(latest_resume_text) ChatGPT_reply = "Resume generated successfully. Please download the file using the link below." messages = [] else: doc = Document() doc.add_paragraph(ChatGPT_reply) word_filename = "generated_response.docx" doc.save(word_filename) # reset the messages to start a new conversation messages = [] return ChatGPT_reply, word_filename def get_log_file(): return "log.csv" def read_log_file(): if os.path.exists("log.csv"): with open("log.csv", "r") as file: return file.read() else: return "Log file is empty or doesn't exist." def clear_log_file(): if not os.path.exists("log.csv"): return "file not exist" else: with open("log.csv", "w") as log_file: log_file.write("") return "Log cleared" def main(): custom_css = """ """ with gr.Blocks(css=custom_css) as demo: gr.Markdown("# Resume Analysis Tool", elem_classes=["centered-title"]) with gr.Row(): with gr.Column(): text_input = gr.Textbox() file_input = gr.File(file_count="multiple", label="Upload Resume") with gr.Row(): btn1 = gr.Button("Summary", elem_classes=["custom-button"]) btn2 = gr.Button("Desensitization", elem_classes=["custom-button"]) btn3 = gr.Button("简历生成", elem_classes=["custom-button"]) btn1.click(fn=set_text_1, inputs=[], outputs=text_input) btn2.click(fn=set_text_2, inputs=[], outputs=text_input) btn3.click(fn=set_text_3, inputs=[], outputs=text_input) with gr.Row(): submit_btn = gr.Button("Submit", elem_classes=["custom-button"]) clear_btn = gr.Button("Clear", elem_classes=["custom-button"]) clear_btn.click(fn=clear_inputs, inputs=[], outputs=[]) with gr.Column(): with gr.Row(): output_text = gr.Textbox(label="ChatGPT Reply", interactive=True, lines=1) with gr.Row(): output_word = gr.File(label="Download Word File") # output_pdf = gr.File(label="Download PDF File") submit_btn.click(fn=CustomChatGPT, inputs=[text_input, file_input], outputs=[output_text, output_word]) with gr.Row(): log_text = gr.Textbox(label="Log Content", interactive=True, lines=1) with gr.Row(): view_log_button = gr.Button("View Log", elem_classes=["custom-button"]) download_log_button = gr.Button("Download Log File", elem_classes=["custom-button"]) clear_log_button = gr.Button("Clear Log File", elem_classes=["custom-button"]) view_log_button.click(fn=read_log_file, inputs=[], outputs=log_text) download_log_button.click(fn=get_log_file, inputs=[], outputs=[]) clear_log_button.click(fn=clear_log_file, inputs=[], outputs=[]) demo.launch(share=True) if __name__ == "__main__": main()