add resume key value extract API

any_parser/any_parser.py

@@ -43,6 +43,7 @@ def __init__(self, api_key: str, base_url: str = PUBLIC_SHARED_BASE_URL) -> None
         """
         self._sync_extract_url = f"{base_url}/extract"
         self._sync_json_url = f"{base_url}/json/extract"
+        self._sync_resume_url = f"{base_url}/resume/extract"
         self._sync_refined_url = f"{base_url}/refined_parse"
         self._async_upload_url = f"{base_url}/async/upload"
         self._async_fetch_url = f"{base_url}/async/fetch"
@@ -188,6 +189,66 @@ def extract_key_value(
         else:
             return f"Error: {response.status_code} {response.text}", None
 
+    def resume_extract(
+        self,
+        file_path: str,
+    ) -> Tuple[str, str]:
+        """Extract resume in real-time.
+
+        Args:
+            file_path (str): The path to the file to be parsed.
+        Returns:
+            tuple(str, str): The extracted data and the time taken.
+                extracted data includes:
+                    - "education": Education
+                    - "work_experience": Work Experience
+                    - "personal_info": Personal Information
+                    - "skills": Skills
+                    - "certifications": Certifications
+                    - "projects": Projects
+                    - "pii": Personally Identifiable Information - includes only name, email, and phone
+        """
+        file_extension = Path(file_path).suffix.lower().lstrip(".")
+
+        # Check if the file exists and file_type
+        error = check_file_type_and_path(file_path, file_extension)
+        if error:
+            return error, None
+
+        # Encode the file content in base64
+        with open(file_path, "rb") as file:
+            encoded_file = base64.b64encode(file.read()).decode("utf-8")
+
+        # Create the JSON payload
+        payload = {
+            "file_content": encoded_file,
+            "file_type": file_extension,
+        }
+
+        # Send the POST request
+        start_time = time.time()
+        response = requests.post(
+            self._sync_resume_url,
+            headers=self._headers,
+            data=json.dumps(payload),
+            timeout=TIMEOUT,
+        )
+        end_time = time.time()
+
+        # Check if the request was successful
+        if response.status_code == 200:
+            try:
+                response_data = response.json()
+                result = response_data["extraction_result"]
+                return (
+                    result,
+                    f"Time Elapsed: {end_time - start_time:.2f} seconds",
+                )
+            except json.JSONDecodeError:
+                return f"Error: Invalid JSON response: {response.text}", None
+        else:
+            return f"Error: {response.status_code} {response.text}", None
+
     def async_extract(
         self,
         file_path: str,
@@ -289,6 +350,44 @@ def async_extract_key_value(
         # If response successful, upload the file
         return upload_file_to_presigned_url(file_path, response)
 
+    def async_extract_resume_key_value(
+        self,
+        file_path: str,
+    ) -> str:
+        """Extract key-value pairs from a file asynchronously.
+
+        Args:
+            file_path (str): The path to the file to be parsed.
+        Returns:
+            str: The file id of the uploaded file.
+        """
+        file_extension = Path(file_path).suffix.lower().lstrip(".")
+
+        # Check if the file exists and file_type
+        error = check_file_type_and_path(file_path, file_extension)
+
+        if error:
+            return error, None
+
+        file_name = Path(file_path).name
+
+        # Create the JSON payload
+        payload = {
+            "file_name": file_name,
+            "process_type": "resume_extract",
+        }
+
+        # Send the POST request
+        response = requests.post(
+            self._async_upload_url,
+            headers=self._headers,
+            data=json.dumps(payload),
+            timeout=TIMEOUT,
+        )
+
+        # If response successful, upload the file
+        return upload_file_to_presigned_url(file_path, response)
+
     def async_fetch(
         self,
         file_id: str,
@@ -340,6 +439,8 @@ def async_fetch(
             result = response.json()
             if "json" in result:
                 return result["json"]
+            elif "resume_extraction" in result:
+                return result["resume_extraction"]
             elif "markdown" in result:
                 markdown_list = result["markdown"]
                 return "\n".join(markdown_list)

examples/async_extract_resume_key_value.ipynb

@@ -0,0 +1,261 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Install the libraries (ipython is used for displaying markdown in this demo)\n",
+    "# !pip3 install --upgrade ipython\n",
+    "# !pip3 install --upgrade any-parser"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from IPython.display import display\n",
+    "from any_parser import AnyParser"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "ap = AnyParser(api_key=\"...\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "file_path = \"./sample_data/resume_1.pdf\"\n",
+    "file_id = ap.async_extract_resume_key_value(file_path)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Waiting for response...\n",
+      "Waiting for response...\n",
+      "Waiting for response...\n",
+      "Waiting for response...\n",
+      "Waiting for response...\n",
+      "Waiting for response...\n"
+     ]
+    }
+   ],
+   "source": [
+    "json_result = ap.async_fetch(file_id=file_id)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "{'pii': {'full_name': 'Gary Jiang',\n",
+       "  'email': '[email protected]',\n",
+       "  'phone': '+1 (213) 725-7637'},\n",
+       " 'education': [{'organization': 'Shenyang University of Technology',\n",
+       "   'degree': \"Bachelor's Degree\",\n",
+       "   'major': 'Computer Science',\n",
+       "   'start_date': '2008-01-01',\n",
+       "   'end_date': '2012-12-31',\n",
+       "   'courses': None,\n",
+       "   'achievements': None}],\n",
+       " 'work_experience': [{'job_title': 'Full Stack Developer',\n",
+       "   'company_name': 'VIMMERSE',\n",
+       "   'location': None,\n",
+       "   'start_date': '2023-06-01',\n",
+       "   'end_date': 'present',\n",
+       "   'job_type': None,\n",
+       "   'summary': 'Developed an AI-powered editor web application that brings photos to life by animating images in 3D.',\n",
+       "   'bullet_points': ['Developed robust back-end services utilizing Python (Flask/FastAPI) and Node.js (AWS Lambda) for efficient and scalable web applications.',\n",
+       "    'Built user-friendly and interactive websites using Next.js, ensuring a seamless user experience.',\n",
+       "    'Deployed and managed AWS infrastructure, including EC2 instances, S3 storage buckets, DynamoDB for NoSQL data management, and Cognito user pools for secure user authentication.',\n",
+       "    'Experienced Agile and Scrum methodologies within a fast-paced startup environment to ensure efficient project delivery and continuous improvement.',\n",
+       "    'Collaborated effectively with cross-functional teams (design, product management) to deliver projects on time, fostering a positive and collaborative work environment.']},\n",
+       "  {'job_title': 'Full Stack Developer',\n",
+       "   'company_name': 'VIKING SASQUATCH',\n",
+       "   'location': None,\n",
+       "   'start_date': '2023-01-01',\n",
+       "   'end_date': '2023-06-01',\n",
+       "   'job_type': None,\n",
+       "   'summary': 'Developed APIs and Integrations for all of the parties that work on Real Estate Transactions.',\n",
+       "   'bullet_points': ['Connecting Mortgage, Title, and Real Estate to solve pain points and improve automation and create efficiencies.',\n",
+       "    'Implemented a user-friendly front-end interface using Nuxt.js, ensuring a seamless user experience.',\n",
+       "    'Built backend APIs utilizing Node.js serverless functions for optimal performance.',\n",
+       "    'Managed data storage and security by implementing a MySQL database.',\n",
+       "    'Collaborated effectively within a team using Agile methodologies like sprint planning, daily standups, retrospectives to ensure project delivery and continuous improvement.']},\n",
+       "  {'job_title': 'Full Stack Developer',\n",
+       "   'company_name': 'ROX PAY SRL',\n",
+       "   'location': None,\n",
+       "   'start_date': '2021-12-01',\n",
+       "   'end_date': '2022-12-31',\n",
+       "   'job_type': None,\n",
+       "   'summary': 'Built Fintech Software House that aims to optimize B2B payments by offering a systemic solution that gives value-added services in collection of payments, financial information and corporate liquidity.',\n",
+       "   'bullet_points': ['Developed front-end by using React.js and Redux, Javascript/Typescript.',\n",
+       "    'Contributed developing backend utilizing Django/Python.']},\n",
+       "  {'job_title': 'Freelancer',\n",
+       "   'company_name': 'FREELANCE',\n",
+       "   'location': None,\n",
+       "   'start_date': '2017-09-01',\n",
+       "   'end_date': '2021-10-31',\n",
+       "   'job_type': None,\n",
+       "   'summary': 'Developed and managed many web and mobile applications while working as freelancer at Internet Dzyns LLC company.',\n",
+       "   'bullet_points': ['Developed multiple web applications, participating in the whole process of their development: product design and estimation, code design and development, DevOps, UI/UX design, product launch and maintenance.',\n",
+       "    'Developed cross-platform mobile application using Flutter and Ionic/Angular.',\n",
+       "    'Developed NFT marketplace websites and wrote smart contracts.']},\n",
+       "  {'job_title': 'Server Administrator, Java Developer',\n",
+       "   'company_name': 'NEUSOFT',\n",
+       "   'location': None,\n",
+       "   'start_date': '2014-06-01',\n",
+       "   'end_date': '2017-08-31',\n",
+       "   'job_type': None,\n",
+       "   'summary': 'Worked as intern and software developer after graduated university.',\n",
+       "   'bullet_points': ['Correct analytical and reasoning skills to troubleshoot and repair server issues.',\n",
+       "    'Operating Systems & Security Software.',\n",
+       "    'Java / Spring Boot / Hibernate.']}],\n",
+       " 'personal_info': {'name': 'GARY JIANG',\n",
+       "  'phone_number': '+1-213-725-7637',\n",
+       "  'address': None,\n",
+       "  'email_address': '[email protected]',\n",
+       "  'linkedin_url': 'linkedin.com/in/gary-jiang',\n",
+       "  'github_url': None,\n",
+       "  'summary': None},\n",
+       " 'skills': {'Programming Languages': ['Python',\n",
+       "   'PHP',\n",
+       "   'Javascript',\n",
+       "   'Typescript',\n",
+       "   'HTML',\n",
+       "   'CSS'],\n",
+       "  'Tools': ['Flask',\n",
+       "   'Django',\n",
+       "   'FastAPI',\n",
+       "   'Laravel',\n",
+       "   'Node.js',\n",
+       "   'SQL databases',\n",
+       "   'Next.js',\n",
+       "   'React',\n",
+       "   'Redux',\n",
+       "   'Nuxt.js',\n",
+       "   'Vue',\n",
+       "   'AWS Lambda',\n",
+       "   'Cognito',\n",
+       "   'EC2',\n",
+       "   'S3',\n",
+       "   'DynamoDB',\n",
+       "   'API Gateway',\n",
+       "   'Flutter',\n",
+       "   'Ionic',\n",
+       "   'Angular',\n",
+       "   'Git',\n",
+       "   'Version Control'],\n",
+       "  'Other': ['Startup Experience',\n",
+       "   'Adaptable',\n",
+       "   'Resourceful',\n",
+       "   'Prioritization',\n",
+       "   'Hybrid Mobile App Development',\n",
+       "   'AGILE',\n",
+       "   'SCRUM',\n",
+       "   'DevOps',\n",
+       "   'CI/CD']},\n",
+       " 'certifications': [],\n",
+       " 'projects': [{'organization': 'VIMMERSE',\n",
+       "   'project_name': 'AI-powered Editor Web Application',\n",
+       "   'location': None,\n",
+       "   'start_date': '2023-06-01',\n",
+       "   'end_date': 'present',\n",
+       "   'descriptions': ['Developed an AI-powered editor web application that brings photos to life by animating images in 3D.',\n",
+       "    'Developed robust back-end services utilizing Python (Flask/FastAPI) and Node.js (AWS Lambda) for efficient and scalable web applications.',\n",
+       "    'Built user-friendly and interactive websites using Next.js, ensuring a seamless user experience.',\n",
+       "    'Deployed and managed AWS infrastructure, including EC2 instances, S3 storage buckets, DynamoDB for NoSQL data management, and Cognito user pools for secure user authentication.',\n",
+       "    'Experienced Agile and Scrum methodologies within a fast-paced startup environment to ensure efficient project delivery and continuous improvement.',\n",
+       "    'Collaborated effectively with cross-functional teams (design, product management) to deliver projects on time, fostering a positive and collaborative work environment.']},\n",
+       "  {'organization': 'VIKING SASQUATCH',\n",
+       "   'project_name': 'Real Estate Transactions API and Integrations',\n",
+       "   'location': None,\n",
+       "   'start_date': '2023-01-01',\n",
+       "   'end_date': '2023-06-30',\n",
+       "   'descriptions': ['Developed APIs and Integrations for all of the parties that work on Real Estate Transactions.',\n",
+       "    'Connecting Mortgage, Title, and Real Estate to solve pain points and improve automation and create efficiencies.',\n",
+       "    'Implemented a user-friendly front-end interface using Nuxt.js, ensuring a seamless user experience.',\n",
+       "    'Built backend APIs utilizing Node.js serverless functions for optimal performance.',\n",
+       "    'Managed data storage and security by implementing a MySQL database.',\n",
+       "    'Collaborated effectively within a team using Agile methodologies like sprint planning, daily standups, retrospectives to ensure project delivery and continuous improvement.']},\n",
+       "  {'organization': 'ROX PAY SRL',\n",
+       "   'project_name': 'Fintech Software House',\n",
+       "   'location': None,\n",
+       "   'start_date': '2021-12-01',\n",
+       "   'end_date': '2022-12-31',\n",
+       "   'descriptions': ['Built Fintech Software House that aims to optimize B2B payments by offering a systemic solution that gives value-added services in collection of payments, financial information and corporate liquidity by essentially creating a Commission Free, Open Loop, Payment Gateway system.',\n",
+       "    'Developed front-end by using React.js and Redux, Javascript/Typescript.',\n",
+       "    'Contributed developing backend utilizing Django/Python.']},\n",
+       "  {'organization': 'FREELANCE',\n",
+       "   'project_name': 'Various Web and Mobile Applications',\n",
+       "   'location': None,\n",
+       "   'start_date': '2017-09-01',\n",
+       "   'end_date': '2021-10-31',\n",
+       "   'descriptions': ['Developed and managed many web and mobile applications while working as freelancer at Internet Dzyns LLC company.',\n",
+       "    'Developed multiple web applications, participating in the whole process of their development: product design and estimation, code design and development, DevOps, UI/UX design, product launch and maintenance.',\n",
+       "    'Developed cross-platform mobile application using Flutter and Ionic/Angular.',\n",
+       "    'Developed NFT marketplace websites and wrote smart contracts.']},\n",
+       "  {'organization': 'NEUSOFT',\n",
+       "   'project_name': 'Server Administration and Java Development',\n",
+       "   'location': None,\n",
+       "   'start_date': '2014-06-01',\n",
+       "   'end_date': '2017-08-31',\n",
+       "   'descriptions': ['Worked as intern and software developer after graduated university.',\n",
+       "    'Correct analytical and reasoning skills to troubleshoot and repair server issues.',\n",
+       "    'Operating Systems & Security Software.',\n",
+       "    'Java / Spring Boot / Hibernate.']}]}"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "display(json_result)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "any",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.10"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}