Add Back document and page content regex functionality

Author: kazmer97

docs/classification.md

@@ -607,7 +607,7 @@ classes:
     x-aws-idp-document-type: Payslip
     type: object
     description: "Employee wage statement showing earnings and deductions"
-    document_name_regex: "(?i).*(payslip|paystub|salary|wage).*"
+    x-aws-idp-document-name-regex: "(?i).*(payslip|paystub|salary|wage).*"
     properties:
       EmployeeName:
         type: string
@@ -640,7 +640,7 @@ classes:
     x-aws-idp-document-type: Invoice
     type: object
     description: "Business invoice document"
-    document_page_content_regex: "(?i)(invoice\\s+number|bill\\s+to|amount\\s+due)"
+    x-aws-idp-document-page-content-regex: "(?i)(invoice\\s+number|bill\\s+to|amount\\s+due)"
     properties:
       InvoiceNumber:
         type: string
@@ -650,7 +650,7 @@ classes:
     x-aws-idp-document-type: Payslip
     type: object
     description: "Employee wage statement"
-    document_page_content_regex: "(?i)(gross\\s+pay|net\\s+pay|employee\\s+id)"
+    x-aws-idp-document-page-content-regex: "(?i)(gross\\s+pay|net\\s+pay|employee\\s+id)"
     properties:
       EmployeeName:
         type: string

lib/idp_common_pkg/idp_common/classification/service.py

@@ -40,7 +40,9 @@
 from idp_common.config.models import IDPConfig
 from idp_common.config.schema_constants import (
     X_AWS_IDP_CLASSIFICATION,
+    X_AWS_IDP_DOCUMENT_NAME_REGEX,
     X_AWS_IDP_DOCUMENT_TYPE,
+    X_AWS_IDP_PAGE_CONTENT_REGEX,
 )
 from idp_common.models import Document, Section, Status
 from idp_common.utils import extract_json_from_text, extract_structured_data_from_text
@@ -168,14 +170,21 @@ def _load_document_types(self) -> List[DocumentType]:
         classes = self.config.classes
         for schema in classes:
             classification_meta = schema.get(X_AWS_IDP_CLASSIFICATION, {})
+
+            # Support both new top-level format and legacy nested format for regex patterns
+            document_name_regex = schema.get(
+                X_AWS_IDP_DOCUMENT_NAME_REGEX
+            ) or classification_meta.get("documentNamePattern")
+            document_page_content_regex = schema.get(
+                X_AWS_IDP_PAGE_CONTENT_REGEX
+            ) or classification_meta.get("pageContentPattern")
+
             doc_types.append(
                 DocumentType(
                     type_name=schema.get(X_AWS_IDP_DOCUMENT_TYPE, ""),
                     description=schema.get("description", ""),
-                    document_name_regex=classification_meta.get("documentNamePattern"),
-                    document_page_content_regex=classification_meta.get(
-                        "pageContentPattern"
-                    ),
+                    document_name_regex=document_name_regex,
+                    document_page_content_regex=document_page_content_regex,
                 )
             )
 

lib/idp_common_pkg/idp_common/config/migration.py

@@ -24,6 +24,8 @@
     X_AWS_IDP_CLASS_PROMPT,
     X_AWS_IDP_ATTRIBUTES_PROMPT,
     X_AWS_IDP_IMAGE_PATH,
+    X_AWS_IDP_DOCUMENT_NAME_REGEX,
+    X_AWS_IDP_PAGE_CONTENT_REGEX,
     VALID_EVALUATION_METHODS,
     MAX_PROMPT_OVERRIDE_LENGTH,
     # Attribute types (for legacy migration only)
@@ -46,6 +48,8 @@
     LEGACY_CLASS_PROMPT,
     LEGACY_ATTRIBUTES_PROMPT,
     LEGACY_IMAGE_PATH,
+    LEGACY_DOCUMENT_NAME_REGEX,
+    LEGACY_DOCUMENT_PAGE_CONTENT_REGEX,
 )
 
 
@@ -153,6 +157,17 @@ def migrate_legacy_to_schema(
         if LEGACY_EXAMPLES in class_config:
             migrated_class[X_AWS_IDP_EXAMPLES] = class_config[LEGACY_EXAMPLES]
 
+        # Migrate regex patterns if present
+        if LEGACY_DOCUMENT_NAME_REGEX in class_config:
+            migrated_class[X_AWS_IDP_DOCUMENT_NAME_REGEX] = class_config[
+                LEGACY_DOCUMENT_NAME_REGEX
+            ]
+
+        if LEGACY_DOCUMENT_PAGE_CONTENT_REGEX in class_config:
+            migrated_class[X_AWS_IDP_PAGE_CONTENT_REGEX] = class_config[
+                LEGACY_DOCUMENT_PAGE_CONTENT_REGEX
+            ]
+
         legacy_attributes = class_config.get(LEGACY_ATTRIBUTES, [])
 
         for attr in legacy_attributes:
@@ -456,6 +471,17 @@ def _convert_classes_to_json_schema(
         ):
             schema[X_AWS_IDP_EXAMPLES] = doc_type_class[X_AWS_IDP_EXAMPLES]
 
+        # Add regex patterns if present
+        if X_AWS_IDP_DOCUMENT_NAME_REGEX in doc_type_class:
+            schema[X_AWS_IDP_DOCUMENT_NAME_REGEX] = doc_type_class[
+                X_AWS_IDP_DOCUMENT_NAME_REGEX
+            ]
+
+        if X_AWS_IDP_PAGE_CONTENT_REGEX in doc_type_class:
+            schema[X_AWS_IDP_PAGE_CONTENT_REGEX] = doc_type_class[
+                X_AWS_IDP_PAGE_CONTENT_REGEX
+            ]
+
         if defs:
             schema[DEFS_FIELD] = defs
 

lib/idp_common_pkg/idp_common/config/schema_constants.py

@@ -25,6 +25,10 @@
 # Classification metadata for document type
 X_AWS_IDP_CLASSIFICATION = "x-aws-idp-classification"
 
+# Regex patterns for classification optimization
+X_AWS_IDP_DOCUMENT_NAME_REGEX = "x-aws-idp-document-name-regex"
+X_AWS_IDP_PAGE_CONTENT_REGEX = "x-aws-idp-document-page-content-regex"
+
 # ============================================================================
 # Legacy Attribute Type Values (for migration only)
 # ============================================================================
@@ -52,8 +56,7 @@
 X_AWS_IDP_EVALUATION_METHOD = "x-aws-idp-evaluation-method"
 
 
-
-X_AWS_IDP_EXAMPLES= "x-aws-idp-examples"
+X_AWS_IDP_EXAMPLES = "x-aws-idp-examples"
 
 # Valid evaluation methods
 EVALUATION_METHOD_EXACT = "EXACT"
@@ -114,6 +117,10 @@
 LEGACY_ATTRIBUTES_PROMPT = "attributesPrompt"
 LEGACY_IMAGE_PATH = "imagePath"
 
+# Legacy regex fields (same name in both legacy and new format)
+LEGACY_DOCUMENT_NAME_REGEX = "document_name_regex"
+LEGACY_DOCUMENT_PAGE_CONTENT_REGEX = "document_page_content_regex"
+
 # ============================================================================
 # JSON Schema Standard Property Names
 # ============================================================================

notebooks/examples/step2_classification_with_regex.ipynb

@@ -124,7 +124,7 @@
      "        'x-aws-idp-document-type': 'BankStatement',\n",
      "        'type': 'object',\n",
      "        'description': 'Employee wage statement',\n",
-     "        'document_name_regex': r'(?i).*(statement).*',\n",
+     "        'x-aws-idp-document-name-regex': r'(?i).*(statement).*',\n",
      "        'properties': {\n",
      "            'Name': {\n",
      "                'type': 'string',\n",
@@ -192,7 +192,7 @@
      "        'x-aws-idp-document-type': 'Payslip',\n",
      "        'type': 'object',\n",
      "        'description': 'Employee wage statement',\n",
-     "        'document_page_content_regex': r'(?i)(gross\\s+pay|net\\s+pay|employee\\s+id)',\n",
+     "        'x-aws-idp-document-page-content-regex': r'(?i)(gross\\s+pay|net\\s+pay|employee\\s+id)',\n",
      "        'properties': {\n",
      "            'EmployeeName': {\n",
      "                'type': 'string',\n",
@@ -206,7 +206,7 @@
      "        'x-aws-idp-document-type': 'Invoice',\n",
      "        'type': 'object',\n",
      "        'description': 'Business invoice',\n",
-     "        'document_page_content_regex': r'(?i)(invoice\\s+number|bill\\s+to|amount\\s+due)',\n",
+     "        'x-aws-idp-document-page-content-regex': r'(?i)(invoice\\s+number|bill\\s+to|amount\\s+due)',\n",
      "        'properties': {\n",
      "            'InvoiceNumber': {\n",
      "                'type': 'string',\n",

src/ui/src/components/json-schema-builder/SchemaInspector.jsx

@@ -15,6 +15,8 @@ import {
   X_AWS_IDP_EVALUATION_METHOD,
   X_AWS_IDP_CONFIDENCE_THRESHOLD,
   X_AWS_IDP_EXAMPLES,
+  X_AWS_IDP_DOCUMENT_NAME_REGEX,
+  X_AWS_IDP_PAGE_CONTENT_REGEX,
 } from '../../constants/schemaConstants';
 
 const SchemaInspector = ({
@@ -99,10 +101,34 @@ const SchemaInspector = ({
           </FormField>
 
           {selectedClass[X_AWS_IDP_DOCUMENT_TYPE] && (
-            <ExamplesEditor
-              examples={selectedClass[X_AWS_IDP_EXAMPLES] || []}
-              onChange={(examples) => onUpdateClass({ [X_AWS_IDP_EXAMPLES]: examples })}
-            />
+            <>
+              <ExamplesEditor
+                examples={selectedClass[X_AWS_IDP_EXAMPLES] || []}
+                onChange={(examples) => onUpdateClass({ [X_AWS_IDP_EXAMPLES]: examples })}
+              />
+              
+              <FormField
+                label="Document Name Regex (Optional)"
+                description="Pattern to match document ID/name. When matched, instantly classifies all pages as this type (single-class configs only). Use case-insensitive patterns like (?i).*(invoice|bill).*"
+              >
+                <Input
+                  value={selectedClass[X_AWS_IDP_DOCUMENT_NAME_REGEX] || ''}
+                  onChange={({ detail }) => onUpdateClass({ [X_AWS_IDP_DOCUMENT_NAME_REGEX]: detail.value || undefined })}
+                  placeholder="e.g., (?i).*(invoice|bill).*"
+                />
+              </FormField>
+
+              <FormField
+                label="Page Content Regex (Optional)"
+                description="Pattern to match page text content. When matched during page-level classification, classifies the page as this type. Use case-insensitive patterns like (?i)(invoice\\s+number|amount\\s+due)"
+              >
+                <Input
+                  value={selectedClass[X_AWS_IDP_PAGE_CONTENT_REGEX] || ''}
+                  onChange={({ detail }) => onUpdateClass({ [X_AWS_IDP_PAGE_CONTENT_REGEX]: detail.value || undefined })}
+                  placeholder="e.g., (?i)(invoice\\s+number|bill\\s+to)"
+                />
+              </FormField>
+            </>
           )}
 
           {usedIn.length > 0 && (

src/ui/src/constants/schemaConstants.js

@@ -67,6 +67,10 @@ export const X_AWS_IDP_DOCUMENT_TYPE = 'x-aws-idp-document-type';
 /** Classification metadata for document type */
 export const X_AWS_IDP_CLASSIFICATION = 'x-aws-idp-classification';
 
+/** Regex patterns for classification optimization */
+export const X_AWS_IDP_DOCUMENT_NAME_REGEX = 'x-aws-idp-document-name-regex';
+export const X_AWS_IDP_PAGE_CONTENT_REGEX = 'x-aws-idp-document-page-content-regex';
+
 // ============================================================================
 // AWS IDP List-Specific Extensions
 // ============================================================================