add connection to another inference

.gitignore

@@ -0,0 +1 @@
+.env.local

handler.py

@@ -10,18 +10,28 @@ from skillNer.general_params import SKILL_DB
 from skillNer.skill_extractor_class import SkillExtractor
 import torch
 from transformers import LongformerTokenizer
+import requests
+import os
+from dotenv import load_dotenv
 
 import torch
 import torch.nn.functional as F
 from transformers import LongformerTokenizer
 import re
 from datetime import datetime
+import time
+
+# Load environment variables from .env.local
+load_dotenv('.env.local')
 
 Resume_num_labels = None
 class EndpointHandler():
     def __init__(self, path=""):
         # Label mapping as provided
         # Resume Label Mapping
+        self.hf_token = os.getenv('HUGGINGFACE_TOKEN')
+        if not self.hf_token:
+            print("Warning: HUGGINGFACE_TOKEN environment variable not set")
         self.Resume_label_map = {
             "RT": 0,    # Resume Title
             "SST": 1,   # Summary Section Title
@@ -276,24 +286,81 @@ class EndpointHandler():
         print("Defaulted to: ", possible_dates[0][1])
         return possible_dates[0][0]  # Return chosen date
     
+    def extract_dates_from_context(self, context):
+        """Extract dates from context using the date extraction endpoint."""
+        max_retries = 5  # Increased retries for startup
+        retry_delay = 5  # Increased delay for startup
+        startup_delay = 10  # Longer delay for startup state
+        
+        for attempt in range(max_retries):
+            try:
+                headers = {
+                    "Authorization": f"Bearer {self.hf_token}"
+                }
+                response = requests.post(
+                    "https://wsk6b4yr3gxrlm4v.us-east-1.aws.endpoints.huggingface.cloud",
+                    json={"inputs": context},
+                    headers=headers,
+                    timeout=30
+                )
+                
+                if response.status_code == 200:
+                    return response.json()
+                elif response.status_code == 503:
+                    if attempt < max_retries - 1:
+                        if attempt == 0:
+                            print(f"Service temporarily unavailable (503). Waiting 20 seconds... (Attempt {attempt + 1}/{max_retries})")
+                            time.sleep(20)
+                        else:
+                            print(f"Service temporarily unavailable (503). Waiting 2 seconds... (Attempt {attempt + 1}/{max_retries})")
+                            time.sleep(2)
+                        continue
+                    else:
+                        print("Service unavailable after maximum retries")
+                        return {"start_date": None, "end_date": None}
+                elif response.status_code == 404:
+                    print("Endpoint not found. Please check if the endpoint URL is correct.")
+                    return {"start_date": None, "end_date": None}
+                elif response.status_code == 401:
+                    print("Authentication failed. Please check your Hugging Face token.")
+                    return {"start_date": None, "end_date": None}
+                else:
+                    print(f"Error calling date extraction endpoint: {response.status_code}")
+                    print(f"Response: {response.text}")
+                    return {"start_date": None, "end_date": None}
+                    
+            except requests.exceptions.Timeout:
+                print(f"Request timed out. Attempt {attempt + 1}/{max_retries}")
+                if attempt < max_retries - 1:
+                    time.sleep(retry_delay)
+                    continue
+                return {"start_date": None, "end_date": None}
+            except Exception as e:
+                print(f"Exception while calling date extraction endpoint: {str(e)}")
+                if attempt < max_retries - 1:
+                    time.sleep(retry_delay)
+                    continue
+                return {"start_date": None, "end_date": None}
+        
+        return {"start_date": None, "end_date": None}
+
     def label_resume(self, text):
         results = self.extract_resume_roles(text)
         for item in results:
             # Extracting dates
             context = (" ".join(item["title"]))
-            date_started = "2020-01-01"  # Random start date
-            date_ended = "2023-12-31"    # Random end date
+            dates = self.extract_dates_from_context(context)
+            date_started = dates.get("start_date")
+            date_ended = dates.get("end_date")
 
             # Try parsing the dates; default to 0 for role_length if parsing fails.
             try:
-                date_started_formatted = self.parse_date(date_started)
+                date_started_formatted = self.parse_date(date_started) if date_started else None
             except ValueError:
                 date_started_formatted = None
 
-            # date_started_formatted = parse_date(date_started)
-            # date_ended_formatted = parse_date(date_ended)
             try:
-                date_ended_formatted = self.parse_date(date_ended)
+                date_ended_formatted = self.parse_date(date_ended) if date_ended else None
             except ValueError:
                 date_ended_formatted = None
 

requirements.txt

@@ -1,4 +1,5 @@
 skillNer>=1.0.0
 spacy>=3.7.2
 en-core-web-lg @ https://github.com/explosion/spacy-models/releases/download/en_core_web_lg-3.7.0/en_core_web_lg-3.7.0-py3-none-any.whl
-ipython>=8.12.0
+ipython>=8.12.0
+python-dotenv>=1.0.0