medicafe 0.250728.9__py3-none-any.whl → 0.250805.2__py3-none-any.whl

MediBot/MediBot_Preprocessor.py

@@ -1,32 +1,53 @@
 #MediBot_Preprocessor.py
 import os, re, sys, argparse
 from collections import OrderedDict # so that the field_mapping stays in order.
-import MediBot_Crosswalk_Library
+import time # Added for timing instrumentation
 
 # Add parent directory of the project to the Python path
 project_dir = os.path.abspath(os.path.join(os.path.dirname(__file__), ".."))
 sys.path.append(project_dir)
 
-try: 
-    from MediLink import MediLink_ConfigLoader
-except ImportError:
-    import MediLink_ConfigLoader
+# Use core utilities for standardized imports
+from MediCafe.core_utils import (
+    import_medibot_module,
+    get_config_loader_with_fallback
+)
 
-try:
-    import MediBot_Preprocessor_lib
-except ImportError:
-    from MediBot import MediBot_Preprocessor_lib
+# Initialize configuration loader with fallback
+MediLink_ConfigLoader = get_config_loader_with_fallback()
+
+# Import MediBot modules using centralized import functions
+MediBot_Crosswalk_Library = import_medibot_module('MediBot_Crosswalk_Library')
+MediBot_Preprocessor_lib = import_medibot_module('MediBot_Preprocessor_lib')
+
+# Initialize API client variables
+get_api_client = None
 
 try:
-    from MediLink.MediLink_API_v3 import APIClient
+    # Try to use the factory for enhanced features (circuit breakers, shared clients, etc.)
+    from MediCafe.api_factory import APIClientFactory
+    factory = APIClientFactory()
+    get_api_client = lambda: factory.get_shared_client()  # Use shared client for token caching benefits
+    MediLink_ConfigLoader.log("MediBot_Preprocessor using API Factory with shared client", level="INFO")
 except ImportError as e:
-    raise ImportError("Failed to import APIClient from MediLink.MediLink_API_v3. "
-                        "Ensure the MediLink_API_v3 module is installed and accessible.") from e
+    # Fallback to basic API client
+    try:
+        from MediCafe.core_utils import get_api_client
+        MediLink_ConfigLoader.log("MediBot_Preprocessor using fallback API client", level="WARNING")
+    except ImportError as e2:
+        # Make API client optional - don't log warning for now
+        get_api_client = None
+
+# Configuration will be loaded when needed
+_config_cache = None
+_crosswalk_cache = None
 
-# Load configuration
-# Should this also take args? Path for ./MediLink needed to be added for this to resolve
-# Use cached configuration to avoid repeated I/O operations
-config, crosswalk = MediBot_Preprocessor_lib.get_cached_configuration()
+def _get_config():
+    """Get configuration, loading it if not already cached."""
+    global _config_cache, _crosswalk_cache
+    if _config_cache is None:
+        _config_cache, _crosswalk_cache = MediBot_Preprocessor_lib.get_cached_configuration()
+    return _config_cache, _crosswalk_cache
 
 # CSV Preprocessor built for Carol
 def preprocess_csv_data(csv_data, crosswalk):
@@ -43,8 +64,22 @@ def preprocess_csv_data(csv_data, crosswalk):
         MediBot_Preprocessor_lib.filter_rows(csv_data)
         
         # Convert 'Surgery Date' from string format to datetime objects for sorting purposes.
+        
+        # TIMING: Start surgery date conversion timing (SLOW OPERATION)
+        date_start_time = time.time()
+        print("Starting surgery date conversion at: {}".format(time.strftime("%H:%M:%S")))
+        MediLink_ConfigLoader.log("Starting surgery date conversion at: {}".format(time.strftime("%H:%M:%S")), level="INFO")
+        
         MediBot_Preprocessor_lib.convert_surgery_date(csv_data)
         
+        # TIMING: End surgery date conversion timing
+        date_end_time = time.time()
+        date_duration = date_end_time - date_start_time
+        print("Surgery date conversion completed at: {} (Duration: {:.2f} seconds)".format(
+            time.strftime("%H:%M:%S"), date_duration))
+        MediLink_ConfigLoader.log("Surgery date conversion completed at: {} (Duration: {:.2f} seconds)".format(
+            time.strftime("%H:%M:%S"), date_duration), level="INFO")
+        
         # Update the CSV data to include only unique patient records.
         # Re-sort the CSV data after deduplication to ensure the correct order.
         # Sort the patients by 'Surgery Date' and then by 'Patient Last' name alphabetically.
@@ -71,8 +106,23 @@ def preprocess_csv_data(csv_data, crosswalk):
         # Update the "Ins1 Insurance ID" column based on the crosswalk and the "Ins1 Payer ID" column for each row.
         # If the Payer ID is not found in the crosswalk, create a placeholder entry in the crosswalk and mark the row for review.
         MediLink_ConfigLoader.log("CSV Pre-processor: Populating 'Ins1 Insurance ID' based on Crosswalk...", level="INFO")
+        config, _ = _get_config()
+        
+        # TIMING: Start insurance ID update timing (SLOWEST OPERATION)
+        insurance_start_time = time.time()
+        print("Starting insurance ID updates at: {}".format(time.strftime("%H:%M:%S")))
+        MediLink_ConfigLoader.log("Starting insurance ID updates at: {}".format(time.strftime("%H:%M:%S")), level="INFO")
+        
         MediBot_Preprocessor_lib.update_insurance_ids(csv_data, config, crosswalk)
         
+        # TIMING: End insurance ID update timing
+        insurance_end_time = time.time()
+        insurance_duration = insurance_end_time - insurance_start_time
+        print("Insurance ID updates completed at: {} (Duration: {:.2f} seconds)".format(
+            time.strftime("%H:%M:%S"), insurance_duration))
+        MediLink_ConfigLoader.log("Insurance ID updates completed at: {} (Duration: {:.2f} seconds)".format(
+            time.strftime("%H:%M:%S"), insurance_duration), level="INFO")
+        
         # Enrich the "Default Diagnosis #1" column based on the parsed docx for each row.
         # This needs to handle the different patient dates correctly so we get the right diagnosis code assigned to the right patient on the right date of service.
         # Currently, we've deleted all the second date entries for patients. As long as they exist in the system, they're just deleted.
@@ -124,6 +174,7 @@ def check_existing_patients(selected_patient_ids, MAPAT_MED_PATH):
 def intake_scan(csv_headers, field_mapping):
     identified_fields = OrderedDict()
     missing_fields_warnings = []
+    config, _ = _get_config()
     required_fields = config["required_fields"]
     
     MediLink_ConfigLoader.log("Intake Scan - Field Mapping: {}".format(field_mapping), level="DEBUG")
@@ -209,7 +260,11 @@ def main():
     
     # Initialize API client only when needed
     if args.update_crosswalk or args.init_crosswalk:
-        client = APIClient()
+        if get_api_client is not None:
+            client = get_api_client()
+        else:
+            print("Error: No API client available")
+            client = None
     
     if args.update_crosswalk:
         print("Updating the crosswalk...")