- Organise retrievers

- Tuning moved to Retriever iso in the configuration, as this is an attribute that should be available for all types of Retrievers

CHANGELOG.md

@@ -25,6 +25,39 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Security
 - In case of vulnerabilities.
 
+## [1.0.13-alfa]
+
+### Added
+- Finished Catalog introduction
+- Reinitialization of WordPress site for syncing
+
+### Changed
+- Modification of WordPress Sync Component
+- Cleanup of attributes in Tenant
+
+### Fixed
+- Overall bugfixes as result from the Catalog introduction
+
+## [1.0.12-alfa]
+
+### Added
+- Added Catalog functionality
+
+### Changed
+- For changes in existing functionality.
+
+### Deprecated
+- For soon-to-be removed features.
+
+### Removed
+- For now removed features.
+
+### Fixed
+- Set default language when registering Documents or URLs.
+
+### Security
+- In case of vulnerabilities.
+
 ## [1.0.11-alfa]
 
 ### Added
@@ -37,7 +70,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - repopack can now split for different components
 
 ### Fixed
-- Various fixes as consequece of changing file_location / file_name ==> bucket_name / object_name
+- Various fixes as consequence of changing file_location / file_name ==> bucket_name / object_name
 - Celery Routing / Queuing updated
 
 ## [1.0.10-alfa]

common/langchain/retrievers/eveai_default_rag_retriever.py

@@ -11,16 +11,22 @@ from common.utils.datetime_utils import get_date_in_timezone
 from common.utils.model_utils import ModelVariables
 
 
-class EveAIRetriever(BaseRetriever, BaseModel):
+class EveAIDefaultRagRetriever(BaseRetriever, BaseModel):
+    _catalog_id: int = PrivateAttr()
     _model_variables: ModelVariables = PrivateAttr()
     _tenant_info: Dict[str, Any] = PrivateAttr()
 
-    def __init__(self, model_variables: ModelVariables, tenant_info: Dict[str, Any]):
+    def __init__(self, catalog_id: int, model_variables: ModelVariables, tenant_info: Dict[str, Any]):
         super().__init__()
         current_app.logger.debug(f'Model variables type: {type(model_variables)}')
+        self._catalog_id = catalog_id
         self._model_variables = model_variables
         self._tenant_info = tenant_info
 
+    @property
+    def catalog_id(self) -> int:
+        return self._catalog_id
+
     @property
     def model_variables(self) -> ModelVariables:
         return self._model_variables
@@ -38,7 +44,7 @@ class EveAIRetriever(BaseRetriever, BaseModel):
         similarity_threshold = self.model_variables['similarity_threshold']
         k = self.model_variables['k']
 
-        if self.tenant_info['rag_tuning']:
+        if self.model_variables['rag_tuning']:
             try:
                 current_date = get_date_in_timezone(self.tenant_info['timezone'])
                 current_app.rag_tuning_logger.debug(f'Current date: {current_date}\n')
@@ -73,7 +79,7 @@ class EveAIRetriever(BaseRetriever, BaseModel):
                 current_app.logger.error(f'Error generating overview: {e}')
                 db.session.rollback()
 
-        if self.tenant_info['rag_tuning']:
+        if self.model_variables['rag_tuning']:
             current_app.rag_tuning_logger.debug(f'Parameters for Retrieval of documents: \n')
             current_app.rag_tuning_logger.debug(f'Similarity Threshold: {similarity_threshold}\n')
             current_app.rag_tuning_logger.debug(f'K: {k}\n')
@@ -100,20 +106,21 @@ class EveAIRetriever(BaseRetriever, BaseModel):
                 .filter(
                     or_(Document.valid_from.is_(None), func.date(Document.valid_from) <= current_date),
                     or_(Document.valid_to.is_(None), func.date(Document.valid_to) >= current_date),
-                    (1 - db_class.embedding.cosine_distance(query_embedding)) > similarity_threshold
+                    (1 - db_class.embedding.cosine_distance(query_embedding)) > similarity_threshold,
+                    Document.catalog_id == self._catalog_id
                 )
                 .order_by(desc('similarity'))
                 .limit(k)
             )
 
-            if self.tenant_info['rag_tuning']:
+            if self.model_variables['rag_tuning']:
                 current_app.rag_tuning_logger.debug(f'Query executed for Retrieval of documents: \n')
                 current_app.rag_tuning_logger.debug(f'{query_obj.statement}\n')
                 current_app.rag_tuning_logger.debug(f'---------------------------------------\n')
 
             res = query_obj.all()
 
-            if self.tenant_info['rag_tuning']:
+            if self.model_variables['rag_tuning']:
                 current_app.rag_tuning_logger.debug(f'Retrieved {len(res)} relevant documents \n')
                 current_app.rag_tuning_logger.debug(f'Data retrieved: \n')
                 current_app.rag_tuning_logger.debug(f'{res}\n')
@@ -121,7 +128,7 @@ class EveAIRetriever(BaseRetriever, BaseModel):
 
             result = []
             for doc in res:
-                if self.tenant_info['rag_tuning']:
+                if self.model_variables['rag_tuning']:
                     current_app.rag_tuning_logger.debug(f'Document ID: {doc[0].id} - Distance: {doc[1]}\n')
                     current_app.rag_tuning_logger.debug(f'Chunk: \n {doc[0].chunk}\n\n')
                 result.append(f'SOURCE: {doc[0].id}\n\n{doc[0].chunk}\n\n')

common/langchain/retrievers/eveai_dossier_retriever.py

@@ -0,0 +1,154 @@
+from langchain_core.retrievers import BaseRetriever
+from sqlalchemy import func, and_, or_, desc, cast, JSON
+from sqlalchemy.exc import SQLAlchemyError
+from pydantic import BaseModel, Field, PrivateAttr
+from typing import Any, Dict, List, Optional
+from flask import current_app
+from contextlib import contextmanager
+
+from common.extensions import db
+from common.models.document import Document, DocumentVersion, Catalog
+from common.utils.datetime_utils import get_date_in_timezone
+from common.utils.model_utils import ModelVariables
+
+
+class EveAIDossierRetriever(BaseRetriever, BaseModel):
+    _catalog_id: int = PrivateAttr()
+    _model_variables: ModelVariables = PrivateAttr()
+    _tenant_info: Dict[str, Any] = PrivateAttr()
+    _active_filters: Optional[Dict[str, Any]] = PrivateAttr()
+
+    def __init__(self, catalog_id: int, model_variables: ModelVariables, tenant_info: Dict[str, Any]):
+        super().__init__()
+        self._catalog_id = catalog_id
+        self._model_variables = model_variables
+        self._tenant_info = tenant_info
+        self._active_filters = None
+
+    @contextmanager
+    def filtering(self, metadata_filters: Dict[str, Any]):
+        """Context manager for temporarily setting metadata filters"""
+        previous_filters = self._active_filters
+        self._active_filters = metadata_filters
+        try:
+            yield self
+        finally:
+            self._active_filters = previous_filters
+
+    def _build_metadata_filter_conditions(self, query):
+        """Build SQL conditions for metadata filtering"""
+        if not self._active_filters:
+            return query
+
+        conditions = []
+        for field, value in self._active_filters.items():
+            if value is None:
+                continue
+
+            # Handle both single values and lists of values
+            if isinstance(value, (list, tuple)):
+                # Multiple values - create OR condition
+                or_conditions = []
+                for val in value:
+                    or_conditions.append(
+                        cast(DocumentVersion.user_metadata[field].astext, JSON) == str(val)
+                    )
+                if or_conditions:
+                    conditions.append(or_(*or_conditions))
+            else:
+                # Single value - direct comparison
+                conditions.append(
+                    cast(DocumentVersion.user_metadata[field].astext, JSON) == str(value)
+                )
+
+        if conditions:
+            query = query.filter(and_(*conditions))
+
+        return query
+
+    def _get_relevant_documents(self, query: str):
+        current_app.logger.debug(f'Retrieving relevant documents for dossier query: {query}')
+        if self._active_filters:
+            current_app.logger.debug(f'Using metadata filters: {self._active_filters}')
+
+        query_embedding = self._get_query_embedding(query)
+        db_class = self.model_variables['embedding_db_model']
+        similarity_threshold = self.model_variables['similarity_threshold']
+        k = self.model_variables['k']
+
+        try:
+            current_date = get_date_in_timezone(self.tenant_info['timezone'])
+
+            # Subquery to find the latest version of each document
+            subquery = (
+                db.session.query(
+                    DocumentVersion.doc_id,
+                    func.max(DocumentVersion.id).label('latest_version_id')
+                )
+                .group_by(DocumentVersion.doc_id)
+                .subquery()
+            )
+
+            # Build base query
+            # Build base query
+            query_obj = (
+                db.session.query(db_class,
+                                 (1 - db_class.embedding.cosine_distance(query_embedding)).label('similarity'))
+                .join(DocumentVersion, db_class.doc_vers_id == DocumentVersion.id)
+                .join(Document, DocumentVersion.doc_id == Document.id)
+                .join(subquery, DocumentVersion.id == subquery.c.latest_version_id)
+                .filter(
+                    or_(Document.valid_from.is_(None), func.date(Document.valid_from) <= current_date),
+                    or_(Document.valid_to.is_(None), func.date(Document.valid_to) >= current_date),
+                    (1 - db_class.embedding.cosine_distance(query_embedding)) > similarity_threshold,
+                    Document.catalog_id == self._catalog_id
+                )
+            )
+
+            # Apply metadata filters
+            query_obj = self._build_metadata_filter_conditions(query_obj)
+
+            # Order and limit results
+            query_obj = query_obj.order_by(desc('similarity')).limit(k)
+
+            # Debug logging for RAG tuning if enabled
+            if self.model_variables['rag_tuning']:
+                self._log_rag_tuning(query_obj, query_embedding)
+
+            res = query_obj.all()
+
+            result = []
+            for doc in res:
+                if self.model_variables['rag_tuning']:
+                    current_app.logger.debug(f'Document ID: {doc[0].id} - Distance: {doc[1]}\n')
+                    current_app.logger.debug(f'Chunk: \n {doc[0].chunk}\n\n')
+                result.append(f'SOURCE: {doc[0].id}\n\n{doc[0].chunk}\n\n')
+
+        except SQLAlchemyError as e:
+            current_app.logger.error(f'Error retrieving relevant documents: {e}')
+            db.session.rollback()
+            return []
+
+        return result
+
+    def _log_rag_tuning(self, query_obj, query_embedding):
+        """Log debug information for RAG tuning"""
+        current_app.rag_tuning_logger.debug("Debug: Query execution plan:")
+        current_app.rag_tuning_logger.debug(f"{query_obj.statement}")
+        if self._active_filters:
+            current_app.rag_tuning_logger.debug("Debug: Active metadata filters:")
+            current_app.rag_tuning_logger.debug(f"{self._active_filters}")
+
+    def _get_query_embedding(self, query: str):
+        """Get embedding for the query text"""
+        embedding_model = self.model_variables['embedding_model']
+        query_embedding = embedding_model.embed_query(query)
+        return query_embedding
+
+    @property
+    def model_variables(self) -> ModelVariables:
+        return self._model_variables
+
+    @property
+    def tenant_info(self) -> Dict[str, Any]:
+        return self._tenant_info

common/langchain/retrievers/eveai_retriever.py

@@ -0,0 +1,40 @@
+from pydantic import BaseModel, PrivateAttr
+from typing import Dict, Any
+
+from common.utils.model_utils import ModelVariables
+
+
+class EveAIRetriever(BaseModel):
+    _catalog_id: int = PrivateAttr()
+    _user_metadata: Dict[str, Any] = PrivateAttr()
+    _system_metadata: Dict[str, Any] = PrivateAttr()
+    _configuration: Dict[str, Any] = PrivateAttr()
+    _tenant_info: Dict[str, Any] = PrivateAttr()
+    _model_variables: ModelVariables = PrivateAttr()
+    _tuning: bool = PrivateAttr()
+
+    def __init__(self, catalog_id: int, user_metadata: Dict[str, Any], system_metadata: Dict[str, Any],
+                 configuration: Dict[str, Any]):
+        super().__init__()
+        self._catalog_id = catalog_id
+        self._user_metadata = user_metadata
+        self._system_metadata = system_metadata
+        self._configuration = configuration
+
+    @property
+    def catalog_id(self):
+        return self._catalog_id
+
+    @property
+    def user_metadata(self):
+        return self._user_metadata
+
+    @property
+    def system_metadata(self):
+        return self._system_metadata
+
+    @property
+    def configuration(self):
+        return self._configuration
+
+    # Any common methods that should be shared among retrievers can go here.

common/models/document.py

@@ -2,12 +2,70 @@ from common.extensions import db
 from .user import User, Tenant
 from pgvector.sqlalchemy import Vector
 from sqlalchemy.dialects.postgresql import JSONB
+from sqlalchemy.dialects.postgresql import ARRAY
+import sqlalchemy as sa
+
+
+class Catalog(db.Model):
+    id = db.Column(db.Integer, primary_key=True)
+    name = db.Column(db.String(50), nullable=False)
+    description = db.Column(db.Text, nullable=True)
+    type = db.Column(db.String(50), nullable=False, default="DEFAULT_CATALOG")
+
+    # Embedding variables
+    html_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'li'])
+    html_end_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'li'])
+    html_included_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
+    html_excluded_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
+    html_excluded_classes = db.Column(ARRAY(sa.String(200)), nullable=True)
+
+    min_chunk_size = db.Column(db.Integer, nullable=True, default=2000)
+    max_chunk_size = db.Column(db.Integer, nullable=True, default=3000)
+
+    # Chat variables ==> Move to Specialist?
+    chat_RAG_temperature = db.Column(db.Float, nullable=True, default=0.3)
+    chat_no_RAG_temperature = db.Column(db.Float, nullable=True, default=0.5)
+
+    # Tuning enablers
+    embed_tuning = db.Column(db.Boolean, nullable=True, default=False)
+
+    # Meta Data
+    user_metadata = db.Column(JSONB, nullable=True)
+    system_metadata = db.Column(JSONB, nullable=True)
+    configuration = db.Column(JSONB, nullable=True)
+
+    # Versioning Information
+    created_at = db.Column(db.DateTime, nullable=False, server_default=db.func.now())
+    created_by = db.Column(db.Integer, db.ForeignKey(User.id), nullable=True)
+    updated_at = db.Column(db.DateTime, nullable=False, server_default=db.func.now(), onupdate=db.func.now())
+    updated_by = db.Column(db.Integer, db.ForeignKey(User.id))
+
+
+class Retriever(db.Model):
+    id = db.Column(db.Integer, primary_key=True)
+    name = db.Column(db.String(50), nullable=False)
+    description = db.Column(db.Text, nullable=True)
+    catalog_id = db.Column(db.Integer, db.ForeignKey('catalog.id'), nullable=True)
+    type = db.Column(db.String(50), nullable=False, default="DEFAULT_RAG")
+    tuning = db.Column(db.Boolean, nullable=True, default=False)
+
+    # Meta Data
+    user_metadata = db.Column(JSONB, nullable=True)
+    system_metadata = db.Column(JSONB, nullable=True)
+    configuration = db.Column(JSONB, nullable=True)
+
+    # Versioning Information
+    created_at = db.Column(db.DateTime, nullable=False, server_default=db.func.now())
+    created_by = db.Column(db.Integer, db.ForeignKey(User.id), nullable=True)
+    updated_at = db.Column(db.DateTime, nullable=False, server_default=db.func.now(), onupdate=db.func.now())
+    updated_by = db.Column(db.Integer, db.ForeignKey(User.id))
 
 
 class Document(db.Model):
     id = db.Column(db.Integer, primary_key=True)
+    # tenant_id = db.Column(db.Integer, db.ForeignKey(Tenant.id), nullable=False)
+    catalog_id = db.Column(db.Integer, db.ForeignKey(Catalog.id), nullable=True)
     name = db.Column(db.String(100), nullable=False)
-    tenant_id = db.Column(db.Integer, db.ForeignKey(Tenant.id), nullable=False)
     valid_from = db.Column(db.DateTime, nullable=True)
     valid_to = db.Column(db.DateTime, nullable=True)
 
@@ -37,6 +95,7 @@ class DocumentVersion(db.Model):
     system_context = db.Column(db.Text, nullable=True)
     user_metadata = db.Column(JSONB, nullable=True)
     system_metadata = db.Column(JSONB, nullable=True)
+    catalog_properties = db.Column(JSONB, nullable=True)
 
     # Versioning Information
     created_at = db.Column(db.DateTime, nullable=False, server_default=db.func.now())

common/models/user.py

@@ -34,32 +34,32 @@ class Tenant(db.Model):
     embedding_model = db.Column(db.String(50), nullable=True)
     llm_model = db.Column(db.String(50), nullable=True)
 
-    # Embedding variables
-    html_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'li'])
-    html_end_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'li'])
-    html_included_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
-    html_excluded_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
-    html_excluded_classes = db.Column(ARRAY(sa.String(200)), nullable=True)
-
-    min_chunk_size = db.Column(db.Integer, nullable=True, default=2000)
-    max_chunk_size = db.Column(db.Integer, nullable=True, default=3000)
-
-    # Embedding search variables
-    es_k = db.Column(db.Integer, nullable=True, default=5)
-    es_similarity_threshold = db.Column(db.Float, nullable=True, default=0.7)
-
-    # Chat variables
-    chat_RAG_temperature = db.Column(db.Float, nullable=True, default=0.3)
-    chat_no_RAG_temperature = db.Column(db.Float, nullable=True, default=0.5)
+    # # Embedding variables ==> To be removed once all migrations (dev + prod) have been done
+    # html_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'li'])
+    # html_end_tags = db.Column(ARRAY(sa.String(10)), nullable=True, default=['p', 'li'])
+    # html_included_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
+    # html_excluded_elements = db.Column(ARRAY(sa.String(50)), nullable=True)
+    # html_excluded_classes = db.Column(ARRAY(sa.String(200)), nullable=True)
+    #
+    # min_chunk_size = db.Column(db.Integer, nullable=True, default=2000)
+    # max_chunk_size = db.Column(db.Integer, nullable=True, default=3000)
+    #
+    # # Embedding search variables
+    # es_k = db.Column(db.Integer, nullable=True, default=5)
+    # es_similarity_threshold = db.Column(db.Float, nullable=True, default=0.7)
+    #
+    # # Chat variables
+    # chat_RAG_temperature = db.Column(db.Float, nullable=True, default=0.3)
+    # chat_no_RAG_temperature = db.Column(db.Float, nullable=True, default=0.5)
     fallback_algorithms = db.Column(ARRAY(sa.String(50)), nullable=True)
 
     # Licensing Information
     encrypted_chat_api_key = db.Column(db.String(500), nullable=True)
     encrypted_api_key = db.Column(db.String(500), nullable=True)
 
-    # Tuning enablers
-    embed_tuning = db.Column(db.Boolean, nullable=True, default=False)
-    rag_tuning = db.Column(db.Boolean, nullable=True, default=False)
+    # # Tuning enablers
+    # embed_tuning = db.Column(db.Boolean, nullable=True, default=False)
+    # rag_tuning = db.Column(db.Boolean, nullable=True, default=False)
 
     # Entitlements
     currency = db.Column(db.String(20), nullable=True)
@@ -96,20 +96,7 @@ class Tenant(db.Model):
             'allowed_languages': self.allowed_languages,
             'embedding_model': self.embedding_model,
             'llm_model': self.llm_model,
-            'html_tags': self.html_tags,
-            'html_end_tags': self.html_end_tags,
-            'html_included_elements': self.html_included_elements,
-            'html_excluded_elements': self.html_excluded_elements,
-            'html_excluded_classes': self.html_excluded_classes,
-            'min_chunk_size': self.min_chunk_size,
-            'max_chunk_size': self.max_chunk_size,
-            'es_k': self.es_k,
-            'es_similarity_threshold': self.es_similarity_threshold,
-            'chat_RAG_temperature': self.chat_RAG_temperature,
-            'chat_no_RAG_temperature': self.chat_no_RAG_temperature,
             'fallback_algorithms': self.fallback_algorithms,
-            'embed_tuning': self.embed_tuning,
-            'rag_tuning': self.rag_tuning,
             'currency': self.currency,
             'usage_email': self.usage_email,
         }

common/utils/document_utils.py

@@ -17,15 +17,17 @@ from ..models.user import Tenant
 
 def create_document_stack(api_input, file, filename, extension, tenant_id):
     # Create the Document
-    new_doc = create_document(api_input, filename, tenant_id)
+    catalog_id = int(api_input.get('catalog_id'))
+    new_doc = create_document(api_input, filename, catalog_id)
     db.session.add(new_doc)
 
     # Create the DocumentVersion
-    new_doc_vers = create_version_for_document(new_doc,
+    new_doc_vers = create_version_for_document(new_doc, tenant_id,
                                                api_input.get('url', ''),
                                                api_input.get('language', 'en'),
                                                api_input.get('user_context', ''),
                                                api_input.get('user_metadata'),
+                                               api_input.get('catalog_properties')
                                                )
     db.session.add(new_doc_vers)
 
@@ -45,7 +47,7 @@ def create_document_stack(api_input, file, filename, extension, tenant_id):
     return new_doc, new_doc_vers
 
 
-def create_document(form, filename, tenant_id):
+def create_document(form, filename, catalog_id):
     new_doc = Document()
     if form['name'] == '':
         new_doc.name = filename.rsplit('.', 1)[0]
@@ -56,13 +58,13 @@ def create_document(form, filename, tenant_id):
         new_doc.valid_from = form['valid_from']
     else:
         new_doc.valid_from = dt.now(tz.utc)
-    new_doc.tenant_id = tenant_id
+    new_doc.catalog_id = catalog_id
     set_logging_information(new_doc, dt.now(tz.utc))
 
     return new_doc
 
 
-def create_version_for_document(document, url, language, user_context, user_metadata):
+def create_version_for_document(document, tenant_id, url, language, user_context, user_metadata, catalog_properties):
     new_doc_vers = DocumentVersion()
     if url != '':
         new_doc_vers.url = url
@@ -78,11 +80,14 @@ def create_version_for_document(document, url, language, user_context, user_meta
     if user_metadata != '' and user_metadata is not None:
         new_doc_vers.user_metadata = user_metadata
 
+    if catalog_properties != '' and catalog_properties is not None:
+        new_doc_vers.catalog_properties = catalog_properties
+
     new_doc_vers.document = document
 
     set_logging_information(new_doc_vers, dt.now(tz.utc))
 
-    mark_tenant_storage_dirty(document.tenant_id)
+    mark_tenant_storage_dirty(tenant_id)
 
     return new_doc_vers
 
@@ -272,9 +277,10 @@ def edit_document(document_id, name, valid_from, valid_to):
         return None, str(e)
 
 
-def edit_document_version(version_id, user_context):
+def edit_document_version(version_id, user_context, catalog_properties):
     doc_vers = DocumentVersion.query.get_or_404(version_id)
     doc_vers.user_context = user_context
+    doc_vers.catalog_properties = catalog_properties
     update_logging_information(doc_vers, dt.now(tz.utc))
 
     try:
@@ -286,7 +292,7 @@ def edit_document_version(version_id, user_context):
         return None, str(e)
 
 
-def refresh_document_with_info(doc_id, api_input):
+def refresh_document_with_info(doc_id, tenant_id, api_input):
     doc = Document.query.get_or_404(doc_id)
     old_doc_vers = DocumentVersion.query.filter_by(doc_id=doc_id).order_by(desc(DocumentVersion.id)).first()
 
@@ -294,11 +300,12 @@ def refresh_document_with_info(doc_id, api_input):
         return None, "This document has no URL. Only documents with a URL can be refreshed."
 
     new_doc_vers = create_version_for_document(
-        doc,
+        doc, tenant_id,
         old_doc_vers.url,
         api_input.get('language', old_doc_vers.language),
         api_input.get('user_context', old_doc_vers.user_context),
-        api_input.get('user_metadata', old_doc_vers.user_metadata)
+        api_input.get('user_metadata', old_doc_vers.user_metadata),
+        api_input.get('catalog_properties', old_doc_vers.catalog_properties),
     )
 
     set_logging_information(new_doc_vers, dt.now(tz.utc))
@@ -318,9 +325,9 @@ def refresh_document_with_info(doc_id, api_input):
     response.raise_for_status()
     file_content = response.content
 
-    upload_file_for_version(new_doc_vers, file_content, extension, doc.tenant_id)
+    upload_file_for_version(new_doc_vers, file_content, extension, tenant_id)
 
-    task = current_celery.send_task('create_embeddings', args=[doc.tenant_id, new_doc_vers.id,], queue='embeddings')
+    task = current_celery.send_task('create_embeddings', args=[tenant_id, new_doc_vers.id,], queue='embeddings')
     current_app.logger.info(f'Embedding creation started for document {doc_id} on version {new_doc_vers.id} '
                             f'with task id: {task.id}.')
 
@@ -328,7 +335,7 @@ def refresh_document_with_info(doc_id, api_input):
 
 
 # Update the existing refresh_document function to use the new refresh_document_with_info
-def refresh_document(doc_id):
+def refresh_document(doc_id, tenant_id):
     current_app.logger.info(f'Refreshing document {doc_id}')
     doc = Document.query.get_or_404(doc_id)
     old_doc_vers = DocumentVersion.query.filter_by(doc_id=doc_id).order_by(desc(DocumentVersion.id)).first()
@@ -336,14 +343,17 @@ def refresh_document(doc_id):
     api_input = {
         'language': old_doc_vers.language,
         'user_context': old_doc_vers.user_context,
-        'user_metadata': old_doc_vers.user_metadata
+        'user_metadata': old_doc_vers.user_metadata,
+        'catalog_properties': old_doc_vers.catalog_properties,
     }
 
-    return refresh_document_with_info(doc_id, api_input)
+    return refresh_document_with_info(doc_id, tenant_id, api_input)
 
 
 # Function triggered when a document_version is created or updated
 def mark_tenant_storage_dirty(tenant_id):
-    tenant = db.session.query(Tenant).filter_by(id=tenant_id).first()
+    tenant = db.session.query(Tenant).filter_by(id=int(tenant_id)).first()
     tenant.storage_dirty = True
     db.session.commit()
+
+

common/utils/model_utils.py

@@ -14,7 +14,7 @@ from portkey_ai.langchain.portkey_langchain_callback_handler import LangchainCal
 from common.langchain.llm_metrics_handler import LLMMetricsHandler
 from common.langchain.tracked_openai_embeddings import TrackedOpenAIEmbeddings
 from common.langchain.tracked_transcribe import tracked_transcribe
-from common.models.document import EmbeddingSmallOpenAI, EmbeddingLargeOpenAI
+from common.models.document import EmbeddingSmallOpenAI, EmbeddingLargeOpenAI, Catalog
 from common.models.user import Tenant
 from config.model_config import MODEL_CONFIG
 from common.utils.business_event_context import current_event
@@ -42,8 +42,9 @@ def set_language_prompt_template(cls, language_prompt):
 
 
 class ModelVariables(MutableMapping):
-    def __init__(self, tenant: Tenant):
+    def __init__(self, tenant: Tenant, catalog_id=None):
         self.tenant = tenant
+        self.catalog_id = catalog_id
         self._variables = self._initialize_variables()
         self._embedding_model = None
         self._llm = None
@@ -57,25 +58,32 @@ class ModelVariables(MutableMapping):
     def _initialize_variables(self):
         variables = {}
 
-        # We initialize the variables that are available knowing the tenant. For the other, we will apply 'lazy loading'
-        variables['k'] = self.tenant.es_k or 5
-        variables['similarity_threshold'] = self.tenant.es_similarity_threshold or 0.7
-        variables['RAG_temperature'] = self.tenant.chat_RAG_temperature or 0.3
-        variables['no_RAG_temperature'] = self.tenant.chat_no_RAG_temperature or 0.5
-        variables['embed_tuning'] = self.tenant.embed_tuning or False
-        variables['rag_tuning'] = self.tenant.rag_tuning or False
+        # Get the Catalog if catalog_id is passed
+        if self.catalog_id:
+            catalog = Catalog.query.get_or_404(self.catalog_id)
+
+            # We initialize the variables that are available knowing the tenant.
+            variables['embed_tuning'] = catalog.embed_tuning or False
+
+            # Set HTML Chunking Variables
+            variables['html_tags'] = catalog.html_tags
+            variables['html_end_tags'] = catalog.html_end_tags
+            variables['html_included_elements'] = catalog.html_included_elements
+            variables['html_excluded_elements'] = catalog.html_excluded_elements
+            variables['html_excluded_classes'] = catalog.html_excluded_classes
+
+            # Set Chunk Size variables
+            variables['min_chunk_size'] = catalog.min_chunk_size
+            variables['max_chunk_size'] = catalog.max_chunk_size
+
+        # Set the RAG Context (will have to change once specialists are defined
         variables['rag_context'] = self.tenant.rag_context or " "
-
-        # Set HTML Chunking Variables
-        variables['html_tags'] = self.tenant.html_tags
-        variables['html_end_tags'] = self.tenant.html_end_tags
-        variables['html_included_elements'] = self.tenant.html_included_elements
-        variables['html_excluded_elements'] = self.tenant.html_excluded_elements
-        variables['html_excluded_classes'] = self.tenant.html_excluded_classes
-
-        # Set Chunk Size variables
-        variables['min_chunk_size'] = self.tenant.min_chunk_size
-        variables['max_chunk_size'] = self.tenant.max_chunk_size
+        # Temporary setting until we have Specialists
+        variables['rag_tuning'] = False
+        variables['RAG_temperature'] = 0.3
+        variables['no_RAG_temperature'] = 0.5
+        variables['k'] = 8
+        variables['similarity_threshold'] = 0.4
 
         # Set model providers
         variables['embedding_provider'], variables['embedding_model'] = self.tenant.embedding_model.rsplit('.', 1)
@@ -195,7 +203,12 @@ class ModelVariables(MutableMapping):
             return self.transcription_client
         elif key in self._variables.get('prompt_templates', []):
             return self.get_prompt_template(key)
-        return self._variables.get(key)
+        else:
+            value = self._variables.get(key)
+            if value is not None:
+                return value
+            else:
+                raise KeyError(f'Variable {key} does not exist in ModelVariables')
 
     def __setitem__(self, key: str, value: Any) -> None:
         self._variables[key] = value
@@ -225,8 +238,8 @@ class ModelVariables(MutableMapping):
         return self._variables.values()
 
 
-def select_model_variables(tenant):
-    model_variables = ModelVariables(tenant=tenant)
+def select_model_variables(tenant, catalog_id=None):
+    model_variables = ModelVariables(tenant=tenant, catalog_id=catalog_id)
     return model_variables
 
 

config/catalog_types.py

@@ -0,0 +1,53 @@
+# Catalog Types
+CATALOG_TYPES = {
+    "DEFAULT": {
+        "name": "Default Catalog",
+        "Description": "A Catalog with information in Evie's Library, to be considered as a whole",
+        "configuration": {}
+    },
+    "DOSSIER": {
+        "name": "Dossier Catalog",
+        "Description": "A Catalog with information in Evie's Library in which several Dossiers can be stored",
+        "configuration": {
+            "tagging_fields": {
+                "name": "Tagging Fields",
+                "type": "tagging_fields",
+                "description": """Define the metadata fields that will be used for tagging documents.
+                                Each field must have:
+                                - type: one of 'string', 'integer', 'float', 'date', 'enum'
+                                - required: boolean indicating if the field is mandatory
+                                - description: field description
+                                - allowed_values: list of values (for enum type only)
+                                - min_value/max_value: range limits (for numeric types only)""",
+                "required": True,
+                "default": {},
+                "field_properties": {
+                    "type": {
+                        "allowed_values": ["string", "integer", "float", "date", "enum"],
+                        "required": True
+                    },
+                    "required": {
+                        "type": "boolean",
+                        "default": False
+                    },
+                    "description": {
+                        "type": "string"
+                    },
+                    "allowed_values": {
+                        "type": "list",
+                        "description": "For enum type fields only"
+                    },
+                    "min_value": {
+                        "type": "number",
+                        "description": "For numeric fields only"
+                    },
+                    "max_value": {
+                        "type": "number",
+                        "description": "For numeric fields only"
+                    }
+                }
+            }
+        },
+        "document_version_configurations": ["tagging_fields"]
+    },
+}

config/logging_config.py

@@ -37,7 +37,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_app.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -45,7 +45,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_workers.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -53,7 +53,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_chat.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -61,7 +61,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_chat_workers.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -69,7 +69,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_api.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -77,7 +77,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_beat.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -85,7 +85,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/eveai_entitlements.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -93,7 +93,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/sqlalchemy.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -101,7 +101,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/mailman.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -109,7 +109,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/security.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -117,7 +117,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/rag_tuning.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -125,7 +125,7 @@ LOGGING = {
             'level': 'DEBUG',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/embed_tuning.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },
@@ -133,7 +133,7 @@ LOGGING = {
             'level': 'INFO',
             'class': 'logging.handlers.RotatingFileHandler',
             'filename': 'logs/business_events.log',
-            'maxBytes': 1024 * 1024 * 5,  # 5MB
+            'maxBytes': 1024 * 1024 * 1,  # 1MB
             'backupCount': 10,
             'formatter': 'standard',
         },

config/retriever_types.py

@@ -0,0 +1,23 @@
+# Retriever Types
+RETRIEVER_TYPES = {
+    "DEFAULT_RAG": {
+        "name": "Default RAG",
+        "description": "Retrieving all embeddings conform the query",
+        "configuration": {
+            "es_k": {
+                "name": "es_k",
+                "type": "int",
+                "description": "K-value to retrieve embeddings (max embeddings retrieved)",
+                "required": True,
+                "default": 8,
+            },
+            "es_similarity_threshold": {
+                "name": "es_similarity_threshold",
+                "type": "float",
+                "description": "Similarity threshold for retrieving embeddings",
+                "required": True,
+                "default": 0.3,
+            },
+        }
+    }
+}

docker/build_and_push_eveai.sh

@@ -169,4 +169,5 @@ for SERVICE in "${SERVICES[@]}"; do
     fi
 done
 
-echo "All specified services processed."
+echo -e "\033[35mAll specified services processed.\033[0m"
+echo -e "\033[35mFinished at $(date +"%d/%m/%Y %H:%M:%S")\033[0m"

docker/release_and_tag_eveai.sh

@@ -0,0 +1,62 @@
+#!/bin/bash
+
+# Initialize variables
+RELEASE_VERSION=""
+RELEASE_MESSAGE=""
+DOCKER_ACCOUNT="josakola"  # Your Docker account name
+
+# Parse input arguments
+while getopts r:m: flag
+do
+    case "${flag}" in
+        r) RELEASE_VERSION=${OPTARG};;
+        m) RELEASE_MESSAGE=${OPTARG};;
+        *)
+          echo "Usage: $0 -r <release_version> -m <release_message>"
+          exit 1 ;;
+    esac
+done
+
+# Ensure both version and message are provided
+if [ -z "$RELEASE_VERSION" ]; then
+  echo "Error: Release version not provided. Use -r <release_version>"
+  exit 1
+fi
+
+if [ -z "$RELEASE_MESSAGE" ]; then
+  echo "Error: Release message not provided. Use -m <release_message>"
+  exit 1
+fi
+
+# Path to your docker-compose file
+DOCKER_COMPOSE_FILE="compose_dev.yaml"
+
+# Get all the images defined in docker-compose
+IMAGES=$(docker compose -f $DOCKER_COMPOSE_FILE config | grep 'image:' | awk '{ print $2 }')
+
+# Start tagging only relevant images
+for DOCKER_IMAGE in $IMAGES; do
+    # Check if the image belongs to your Docker account and ends with :latest
+    if [[ $DOCKER_IMAGE == $DOCKER_ACCOUNT* && $DOCKER_IMAGE == *:latest ]]; then
+        # Remove the ":latest" tag to use the base image name
+        BASE_IMAGE=${DOCKER_IMAGE%:latest}
+
+        echo "Tagging Docker image: $BASE_IMAGE with version: $RELEASE_VERSION"
+
+        # Tag the 'latest' image with the new release version
+        docker tag $DOCKER_IMAGE $BASE_IMAGE:$RELEASE_VERSION
+
+        # Push the newly tagged image to Docker Hub
+        docker push $BASE_IMAGE:$RELEASE_VERSION
+    else
+        echo "Skipping image: $DOCKER_IMAGE (not part of $DOCKER_ACCOUNT or not tagged as latest)"
+    fi
+done
+
+# Step 3: Tag the Git repository with the release version
+echo "Tagging Git repository with version: $RELEASE_VERSION"
+git tag -a v$RELEASE_VERSION -m "Release $RELEASE_VERSION: $RELEASE_MESSAGE"
+git push origin v$RELEASE_VERSION
+
+echo -e "\033[35mRelease process completed for version: $RELEASE_VERSION \033[0m"
+echo -e "\033[35mFinished at $(date +"%d/%m/%Y %H:%M:%S")\033[0m"

eveai_api/__init__.py

@@ -56,13 +56,6 @@ def create_app(config_file=None):
         app.logger.debug(f'Request URL: {request.url}')
         app.logger.debug(f'Request headers: {dict(request.headers)}')
 
-        # Log request arguments
-        app.logger.debug(f'Request args: {request.args}')
-
-        # Log form data if it's a POST request
-        if request.method == 'POST':
-            app.logger.debug(f'Form data: {request.form}')
-
         # Log JSON data if the content type is application/json
         if request.is_json:
             app.logger.debug(f'JSON data: {request.json}')
@@ -95,6 +88,10 @@ def create_app(config_file=None):
                 # Don't raise the exception here, let the request continue
                 # The appropriate error handling will be done in the specific endpoints
 
+    @app.route('/api/v1')
+    def swagger():
+        return api_rest.render_doc()
+
     return app
 
 

eveai_api/api/document_api.py

@@ -33,6 +33,7 @@ document_ns = Namespace('documents', description='Document related operations')
 
 # Define models for request parsing and response serialization
 upload_parser = reqparse.RequestParser()
+upload_parser.add_argument('catalog_id', location='form', type=int, required=True, help='The catalog to add the file to')
 upload_parser.add_argument('file', location='files', type=FileStorage, required=True, help='The file to upload')
 upload_parser.add_argument('name', location='form', type=str, required=False, help='Name of the document')
 upload_parser.add_argument('language', location='form', type=str, required=True, help='Language of the document')
@@ -42,6 +43,9 @@ upload_parser.add_argument('valid_from', location='form', type=validate_date, re
                            help='Valid from date for the document (ISO format)')
 upload_parser.add_argument('user_metadata', location='form', type=validate_json, required=False,
                            help='User metadata for the document (JSON format)')
+upload_parser.add_argument('catalog_properties', location='form', type=validate_json, required=False,
+                           help='The catalog configuration to be passed along (JSON format). Validity is against catalog requirements '
+                                'is not checked, and is the responsibility of the calling client.')
 
 add_document_response = document_ns.model('AddDocumentResponse', {
     'message': fields.String(description='Status message'),
@@ -75,11 +79,13 @@ class AddDocument(Resource):
             validate_file_type(extension)
 
             api_input = {
+                'catalog_id': args.get('catalog_id'),
                 'name': args.get('name') or filename,
                 'language': args.get('language'),
                 'user_context': args.get('user_context'),
                 'valid_from': args.get('valid_from'),
                 'user_metadata': args.get('user_metadata'),
+                'catalog_properties': args.get('catalog_properties'),
             }
 
             new_doc, new_doc_vers = create_document_stack(api_input, file, filename, extension, tenant_id)
@@ -102,13 +108,18 @@ class AddDocument(Resource):
 
 # Models for AddURL
 add_url_model = document_ns.model('AddURL', {
+    'catalog_id': fields.Integer(required='True', description='ID of the catalog the URL needs to be added to'),
     'url': fields.String(required=True, description='URL of the document to add'),
     'name': fields.String(required=False, description='Name of the document'),
     'language': fields.String(required=True, description='Language of the document'),
     'user_context': fields.String(required=False, description='User context for the document'),
     'valid_from': fields.String(required=False, description='Valid from date for the document'),
     'user_metadata': fields.String(required=False, description='User metadata for the document'),
-    'system_metadata': fields.String(required=False, description='System metadata for the document')
+    'system_metadata': fields.String(required=False, description='System metadata for the document'),
+    'catalog_properties': fields.String(required=False, description='The catalog configuration to be passed along (JSON '
+                                                                    'format). Validity is against catalog requirements '
+                                                                    'is not checked, and is the responsibility of the '
+                                                                    'calling client.'),
 })
 
 add_url_response = document_ns.model('AddURLResponse', {
@@ -138,12 +149,14 @@ class AddURL(Resource):
             file_content, filename, extension = process_url(args['url'], tenant_id)
 
             api_input = {
+                'catalog_id': args['catalog_id'],
                 'url': args['url'],
                 'name': args.get('name') or filename,
                 'language': args['language'],
                 'user_context': args.get('user_context'),
                 'valid_from': args.get('valid_from'),
                 'user_metadata': args.get('user_metadata'),
+                'catalog_properties': args.get('catalog_properties'),
             }
 
             new_doc, new_doc_vers = create_document_stack(api_input, file_content, filename, extension, tenant_id)
@@ -213,7 +226,8 @@ class DocumentResource(Resource):
     @document_ns.response(200, 'Document refreshed successfully')
     def post(self, document_id):
         """Refresh a document"""
-        new_version, result = refresh_document(document_id)
+        tenant_id = get_jwt_identity()
+        new_version, result = refresh_document(document_id, tenant_id)
         if new_version:
             return {'message': f'Document refreshed. New version: {new_version.id}. Task ID: {result}'}, 200
         else:
@@ -222,6 +236,7 @@ class DocumentResource(Resource):
 
 edit_document_version_model = document_ns.model('EditDocumentVersion', {
     'user_context': fields.String(required=True, description='New user context for the document version'),
+    'catalog_properties': fields.String(required=True, description='New catalog properties for the document version'),
 })
 
 
@@ -234,7 +249,7 @@ class DocumentVersionResource(Resource):
     def put(self, version_id):
         """Edit a document version"""
         data = request.json
-        updated_version, error = edit_document_version(version_id, data['user_context'])
+        updated_version, error = edit_document_version(version_id, data['user_context'], data.get('catalog_properties'))
         if updated_version:
             return {'message': f'Document Version {updated_version.id} updated successfully'}, 200
         else:
@@ -246,7 +261,8 @@ refresh_document_model = document_ns.model('RefreshDocument', {
     'name': fields.String(required=False, description='New name for the document'),
     'language': fields.String(required=False, description='Language of the document'),
     'user_context': fields.String(required=False, description='User context for the document'),
-    'user_metadata': fields.Raw(required=False, description='User metadata for the document')
+    'user_metadata': fields.Raw(required=False, description='User metadata for the document'),
+    'catalog_properties': fields.Raw(required=False, description='Catalog properties for the document'),
 })
 
 
@@ -263,7 +279,7 @@ class RefreshDocument(Resource):
         current_app.logger.info(f'Refreshing document {document_id} for tenant {tenant_id}')
 
         try:
-            new_version, result = refresh_document(document_id)
+            new_version, result = refresh_document(document_id, tenant_id)
 
             if new_version:
                 return {
@@ -296,7 +312,7 @@ class RefreshDocumentWithInfo(Resource):
 
         try:
             api_input = request.json
-            new_version, result = refresh_document_with_info(document_id, api_input)
+            new_version, result = refresh_document_with_info(document_id, tenant_id, api_input)
 
             if new_version:
                 return {

eveai_app/templates/document/add_document.html

@@ -11,9 +11,17 @@
         {{ form.hidden_tag() }}
         {%  set disabled_fields = [] %}
         {%  set exclude_fields = [] %}
-        {% for field in form %}
+        {% for field in form.get_static_fields() %}
             {{ render_field(field, disabled_fields, exclude_fields) }}
         {% endfor %}
+        {% for collection_name, fields in form.get_dynamic_fields().items() %}
+            {% if fields|length > 0 %}
+                <h4 class="mt-4">{{ collection_name }}</h4>
+            {% endif %}
+            {% for field in fields %}
+                {{ render_field(field, disabled_fields, exclude_fields) }}
+            {% endfor %}
+        {% endfor %}
         <button type="submit" class="btn btn-primary">Add Document</button>
     </form>
 {% endblock %}

eveai_app/templates/document/catalog.html

@@ -0,0 +1,23 @@
+{% extends 'base.html' %}
+{% from "macros.html" import render_field %}
+
+{% block title %}Catalog Registration{% endblock %}
+
+{% block content_title %}Register Catalog{% endblock %}
+{% block content_description %}Define a new catalog of documents in Evie's Library{% endblock %}
+
+{% block content %}
+    <form method="post">
+        {{ form.hidden_tag() }}
+        {%  set disabled_fields = [] %}
+        {%  set exclude_fields = [] %}
+        {% for field in form %}
+            {{ render_field(field, disabled_fields, exclude_fields) }}
+        {% endfor %}
+        <button type="submit" class="btn btn-primary">Register Catalog</button>
+    </form>
+{% endblock %}
+
+{% block content_footer %}
+
+{% endblock %}

eveai_app/templates/document/catalogs.html

@@ -0,0 +1,24 @@
+{% extends 'base.html' %}
+{%  from 'macros.html' import render_selectable_table, render_pagination %}
+
+{% block title %}Documents{% endblock %}
+
+{% block content_title %}Catalogs{% endblock %}
+{% block content_description %}View Catalogs for Tenant{% endblock %}
+{% block content_class %}<div class="col-xl-12 col-lg-5 col-md-7 mx-auto"></div>{% endblock %}
+
+{% block content %}
+<div class="container">
+    <form method="POST" action="{{ url_for('document_bp.handle_catalog_selection') }}">
+        {{ render_selectable_table(headers=["Catalog ID", "Name"], rows=rows, selectable=True, id="catalogsTable") }}
+        <div class="form-group mt-3">
+            <button type="submit" name="action" value="set_session_catalog" class="btn btn-primary">Set Session Catalog</button>
+            <button type="submit" name="action" value="edit_catalog" class="btn btn-primary">Edit Catalog</button>
+        </div>
+    </form>
+</div>
+{% endblock %}
+
+{% block content_footer %}
+    {{ render_pagination(pagination, 'document_bp.catalogs') }}
+{% endblock %}

eveai_app/templates/document/document_versions_list_view.html

@@ -23,15 +23,23 @@
 
     {{ render_collapsible_section('Filter', 'Filter Options', filter_form) }}
 
-    <!-- Document Versions Table -->
-    {{ render_selectable_sortable_table(
-        headers=["ID", "File Type", "Processing", "Processing Start", "Processing Finish", "Processing Error"],
-        rows=rows,
-        selectable=True,
-        id="documentVersionsTable",
-        sort_by=sort_by,
-        sort_order=sort_order
-    ) }}
+    <div class="form-group mt-3">
+        <form method="POST" action="{{ url_for('document_bp.handle_document_version_selection') }}">
+            <!-- Document Versions Table -->
+            {{ render_selectable_sortable_table(
+                headers=["ID", "File Type", "Processing", "Processing Start", "Processing Finish", "Processing Error"],
+                rows=rows,
+                selectable=True,
+                id="documentVersionsTable",
+                sort_by=sort_by,
+                sort_order=sort_order
+            ) }}
+            <div class="form-group mt-4">
+                <button type="submit" name="action" value="edit_document_version" class="btn btn-primary">Edit Document Version</button>
+                <button type="submit" name="action" value="process_document_version" class="btn btn-danger">Process Document Version</button>
+            </div>
+        </form>
+    </div>
 {% endblock %}
 
 {% block content_footer %}

eveai_app/templates/document/documents.html

@@ -1,5 +1,5 @@
 {% extends 'base.html' %}
-{%  from 'macros.html' import render_selectable_table, render_pagination %}
+{% from 'macros.html' import render_selectable_table, render_pagination, render_filter_field, render_date_filter_field, render_collapsible_section, render_selectable_sortable_table_with_dict_headers %}
 
 {% block title %}Documents{% endblock %}
 
@@ -8,18 +8,88 @@
 {% block content_class %}<div class="col-xl-12 col-lg-5 col-md-7 mx-auto"></div>{% endblock %}
 
 {% block content %}
-<div class="container">
-    <form method="POST" action="{{ url_for('document_bp.handle_document_selection') }}">
-        {{ render_selectable_table(headers=["Document ID", "Name", "Valid From", "Valid To"], rows=rows, selectable=True, id="documentsTable") }}
-        <div class="form-group mt-3">
-            <button type="submit" name="action" value="edit_document" class="btn btn-primary">Edit Document</button>
-            <button type="submit" name="action" value="document_versions" class="btn btn-secondary">Show Document Versions</button>
-            <button type="submit" name="action" value="refresh_document" class="btn btn-secondary">Refresh Document (new version)</button>
-        </div>
-    </form>
-</div>
+    <!-- Filter Form -->
+    {% set filter_form %}
+        <form method="GET" action="{{ url_for('document_bp.documents') }}">
+            {{ render_filter_field('catalog_id', 'Catalog', filter_options['catalog_id'], filters.get('catalog_id', [])) }}
+            {{ render_filter_field('validity', 'Validity', filter_options['validity'], filters.get('validity', [])) }}
+
+            <button type="submit" class="btn btn-primary">Apply Filters</button>
+        </form>
+    {% endset %}
+
+    {{ render_collapsible_section('Filter', 'Filter Options', filter_form) }}
+
+    <div class="form-group mt-3">
+        <form method="POST" action="{{ url_for('document_bp.handle_document_selection') }}">
+            <!-- Documents Table -->
+            {{ render_selectable_sortable_table_with_dict_headers(
+                headers=[
+                    {"text": "ID", "sort": "id"},
+                    {"text": "Name", "sort": "name"},
+                    {"text": "Catalog", "sort": "catalog_name"},
+                    {"text": "Valid From", "sort": "valid_from"},
+                    {"text": "Valid To", "sort": "valid_to"}
+                ],
+                rows=rows,
+                selectable=True,
+                id="documentsTable",
+                sort_by=sort_by,
+                sort_order=sort_order
+            ) }}
+            <div class="form-group mt-4">
+                <button type="submit" name="action" value="edit_document" class="btn btn-primary">Edit Document</button>
+                <button type="submit" name="action" value="document_versions" class="btn btn-secondary">Show Document Versions</button>
+                <button type="submit" name="action" value="refresh_document" class="btn btn-secondary">Refresh Document (new version)</button>
+            </div>
+        </form>
+    </div>
 {% endblock %}
 
 {% block content_footer %}
     {{ render_pagination(pagination, 'document_bp.documents') }}
+{% endblock %}
+
+{% block scripts %}
+<script>
+document.addEventListener('DOMContentLoaded', function() {
+    const table = document.getElementById('documentsTable');
+    const headers = table.querySelectorAll('th.sortable');
+
+    headers.forEach(header => {
+        header.addEventListener('click', function() {
+            const sortBy = this.dataset.sort;
+            let sortOrder = 'asc';
+
+            if (this.querySelector('.fa-sort-up')) {
+                sortOrder = 'desc';
+            } else if (this.querySelector('.fa-sort-down')) {
+                sortOrder = 'none';
+            }
+
+            window.location.href = updateQueryStringParameter(window.location.href, 'sort_by', sortBy);
+            window.location.href = updateQueryStringParameter(window.location.href, 'sort_order', sortOrder);
+        });
+    });
+
+    function updateQueryStringParameter(uri, key, value) {
+        var re = new RegExp("([?&])" + key + "=.*?(&|$)", "i");
+        var separator = uri.indexOf('?') !== -1 ? "&" : "?";
+        if (uri.match(re)) {
+            return uri.replace(re, '$1' + key + "=" + value + '$2');
+        }
+        else {
+            return uri + separator + key + "=" + value;
+        }
+    }
+
+    table.addEventListener('change', function(event) {
+        if (event.target.type === 'radio') {
+            var selectedRow = event.target.closest('tr');
+            var documentId = selectedRow.cells[1].textContent;
+            console.log('Selected Document ID:', documentId);
+        }
+    });
+});
+</script>
 {% endblock %}

eveai_app/templates/document/edit_catalog.html

@@ -0,0 +1,35 @@
+{% extends 'base.html' %}
+{% from "macros.html" import render_field %}
+
+{% block title %}Edit Catalog{% endblock %}
+
+{% block content_title %}Edit Catalog{% endblock %}
+{% block content_description %}Edit a catalog of documents in Evie's Library.
+When you change chunking of embedding information, you'll need to manually refresh the library if you want immediate impact.
+{% endblock %}
+
+{% block content %}
+    <form method="post">
+        {{ form.hidden_tag() }}
+        {% set disabled_fields = ['type'] %}
+        {% set exclude_fields = [] %}
+        <!-- Render Static Fields -->
+        {% for field in form.get_static_fields() %}
+            {{ render_field(field, disabled_fields, exclude_fields) }}
+        {% endfor %}
+        <!-- Render Dynamic Fields -->
+        {% for collection_name, fields in form.get_dynamic_fields().items() %}
+            {% if fields|length > 0 %}
+                <h4 class="mt-4">{{ collection_name }}</h4>
+            {% endif %}
+            {% for field in fields %}
+                {{ render_field(field, disabled_fields, exclude_fields) }}
+            {% endfor %}
+        {% endfor %}
+        <button type="submit" class="btn btn-primary">Save Retriever</button>
+    </form>
+{% endblock %}
+
+{% block content_footer %}
+
+{% endblock %}

eveai_app/templates/document/edit_document.html

@@ -8,11 +8,17 @@
 {% block content %}
     <form method="post">
         {{ form.hidden_tag() }}
-        {%  set disabled_fields = [] %}
-        {%  set exclude_fields = [] %}
-        {% for field in form %}
-            {{ render_field(field, disabled_fields, exclude_fields) }}
-        {% endfor %}
+        {% set disabled_fields = [] %}
+        {% set exclude_fields = [] %}
+
+        {{ render_field(form.name, disabled_fields, exclude_fields) }}
+        {{ render_field(form.valid_from, disabled_fields, exclude_fields) }}
+        {{ render_field(form.valid_to, disabled_fields, exclude_fields) }}
+
+        <div class="form-group">
+            <label for="catalog_name">Catalog</label>
+            <input type="text" class="form-control" id="catalog_name" value="{{ catalog_name }}" readonly>
+        </div>
         <button type="submit" class="btn btn-primary">Update Document</button>
     </form>
-{% endblock %}
+{% endblock %}

eveai_app/templates/document/edit_retriever.html

@@ -0,0 +1,33 @@
+{% extends 'base.html' %}
+{% from "macros.html" import render_field %}
+
+{% block title %}Edit Retriever{% endblock %}
+
+{% block content_title %}Edit Retriever{% endblock %}
+{% block content_description %}Edit a Retriever (for a Catalog){% endblock %}
+
+{% block content %}
+    <form method="post">
+        {{ form.hidden_tag() }}
+        {% set disabled_fields = ['type'] %}
+        {% set exclude_fields = [] %}
+        <!-- Render Static Fields -->
+        {% for field in form.get_static_fields() %}
+            {{ render_field(field, disabled_fields, exclude_fields) }}
+        {% endfor %}
+        <!-- Render Dynamic Fields -->
+        {% for collection_name, fields in form.get_dynamic_fields().items() %}
+            {% if fields|length > 0 %}
+                <h4 class="mt-4">{{ collection_name }}</h4>
+            {% endif %}
+            {% for field in fields %}
+                {{ render_field(field, disabled_fields, exclude_fields) }}
+            {% endfor %}
+        {% endfor %}
+        <button type="submit" class="btn btn-primary">Save Retriever</button>
+    </form>
+{% endblock %}
+
+{% block content_footer %}
+
+{% endblock %}

eveai_app/templates/document/retriever.html

@@ -0,0 +1,23 @@
+{% extends 'base.html' %}
+{% from "macros.html" import render_field %}
+
+{% block title %}Retriever Registration{% endblock %}
+
+{% block content_title %}Register Retriever{% endblock %}
+{% block content_description %}Define a new retriever (for a catalog){% endblock %}
+
+{% block content %}
+    <form method="post">
+        {{ form.hidden_tag() }}
+        {%  set disabled_fields = [] %}
+        {%  set exclude_fields = [] %}
+        {% for field in form %}
+            {{ render_field(field, disabled_fields, exclude_fields) }}
+        {% endfor %}
+        <button type="submit" class="btn btn-primary">Register Retriever</button>
+    </form>
+{% endblock %}
+
+{% block content_footer %}
+
+{% endblock %}

eveai_app/templates/document/retrievers.html

@@ -0,0 +1,23 @@
+{% extends 'base.html' %}
+{%  from 'macros.html' import render_selectable_table, render_pagination %}
+
+{% block title %}Retrievers{% endblock %}
+
+{% block content_title %}Retrievers{% endblock %}
+{% block content_description %}View Retrieers for Tenant{% endblock %}
+{% block content_class %}<div class="col-xl-12 col-lg-5 col-md-7 mx-auto"></div>{% endblock %}
+
+{% block content %}
+<div class="container">
+    <form method="POST" action="{{ url_for('document_bp.handle_retriever_selection') }}">
+        {{ render_selectable_table(headers=["Retriever ID", "Name", "Type", "Catalog ID"], rows=rows, selectable=True, id="retrieverssTable") }}
+        <div class="form-group mt-3">
+            <button type="submit" name="action" value="edit_retriever" class="btn btn-primary">Edit Retriever</button>
+        </div>
+    </form>
+</div>
+{% endblock %}
+
+{% block content_footer %}
+    {{ render_pagination(pagination, 'document_bp.retrievers') }}
+{% endblock %}

eveai_app/templates/header.html

@@ -1,5 +1,5 @@
 <header class="header-2">
-    <div class="page-header min-vh-25" style="background-image: url({{url_for('static', filename='/assets/img/EveAI_bg.jpg')}})" loading="lazy">
+    <div class="page-header min-vh-25" style="background-image: url({{url_for('static', filename='/assets/img/EveAI_bg.jpg')}}); background-position: top left; background-repeat: no-repeat; background-size: cover;" loading="lazy">
         <span class="mask bg-gradient-primary opacity-4"></span>
         <div class="container">
             <div class="row">
@@ -10,4 +10,4 @@
             </div>
         </div>
     </div>
-</header>
+</header>

eveai_app/templates/interaction/view_chat_session.html

@@ -54,7 +54,7 @@
                             {% if embedding.url %}
                             <a href="{{ embedding.url }}" target="_blank">{{ embedding.url }}</a>
                             {% else %}
-                            {{ embedding.file_name }}
+                            {{ embedding.object_name }}
                             {% endif %}
                         </li>
                         {% endfor %}

eveai_app/templates/macros.html

@@ -1,18 +1,136 @@
-{% macro render_field(field, disabled_fields=[], exclude_fields=[], class='') %}
+<!--{% macro render_field(field, disabled_fields=[], exclude_fields=[], class='') %}-->
+<!--    {% set disabled = field.name in disabled_fields %}-->
+<!--    {% set exclude_fields = exclude_fields + ['csrf_token', 'submit'] %}-->
+<!--    {% if field.name not in exclude_fields %}-->
+<!--        {% if field.type == 'BooleanField' %}-->
+<!--            <div class="form-check">-->
+<!--                {{ field(class="form-check-input " + class, type="checkbox", id="flexSwitchCheckDefault") }}-->
+<!--                {{ field.label(class="form-check-label", for="flexSwitchCheckDefault", disabled=disabled) }}-->
+<!--            </div>-->
+<!--        {% else %}-->
+<!--            <div class="form-group">-->
+<!--                {{ field.label(class="form-label") }}-->
+<!--                {{ field(class="form-control " + class, disabled=disabled) }}-->
+<!--                {% if field.errors %}-->
+<!--                    <div class="invalid-feedback">-->
+<!--                        {% for error in field.errors %}-->
+<!--                            {{ error }}-->
+<!--                        {% endfor %}-->
+<!--                    </div>-->
+<!--                {% endif %}-->
+<!--            </div>-->
+<!--        {% endif %}-->
+<!--    {% endif %}-->
+<!--{% endmacro %}-->
+
+{% macro render_field_old(field, disabled_fields=[], exclude_fields=[], class='') %}
+    <!-- Debug info -->
+    <!-- Field name: {{ field.name }}, Field type: {{ field.__class__.__name__ }} -->
+
     {% set disabled = field.name in disabled_fields %}
     {% set exclude_fields = exclude_fields + ['csrf_token', 'submit'] %}
     {% if field.name not in exclude_fields %}
         {% if field.type == 'BooleanField' %}
-            <div class="form-check">
-                {{ field(class="form-check-input " + class, type="checkbox", id="flexSwitchCheckDefault") }}
-                {{ field.label(class="form-check-label", for="flexSwitchCheckDefault", disabled=disabled) }}
+            <div class="form-group">
+                <div class="form-check form-switch">
+                    {{ field(class="form-check-input " + class, disabled=disabled) }}
+                    {% if field.description %}
+                        {{ field.label(class="form-check-label",
+                                     **{'data-bs-toggle': 'tooltip',
+                                        'data-bs-placement': 'right',
+                                        'title': field.description}) }}
+                    {% else %}
+                        {{ field.label(class="form-check-label") }}
+                    {% endif %}
+                </div>
+                {% if field.errors %}
+                    <div class="invalid-feedback d-block">
+                        {% for error in field.errors %}
+                            {{ error }}
+                        {% endfor %}
+                    </div>
+                {% endif %}
             </div>
         {% else %}
             <div class="form-group">
-                {{ field.label(class="form-label") }}
-                {{ field(class="form-control " + class, disabled=disabled) }}
+                {% if field.description %}
+                    {{ field.label(class="form-label",
+                                 **{'data-bs-toggle': 'tooltip',
+                                    'data-bs-placement': 'right',
+                                    'title': field.description}) }}
+                {% else %}
+                    {{ field.label(class="form-label") }}
+                {% endif %}
+
+                {% if field.type == 'TextAreaField' and 'json-editor' in class %}
+                    <div id="{{ field.id }}-editor" class="json-editor-container"></div>
+                    {{ field(class="form-control d-none " + class, disabled=disabled) }}
+                {% else %}
+                    {{ field(class="form-control " + class, disabled=disabled) }}
+                {% endif %}
+
                 {% if field.errors %}
-                    <div class="invalid-feedback">
+                    <div class="invalid-feedback d-block">
+                        {% for error in field.errors %}
+                            {{ error }}
+                        {% endfor %}
+                    </div>
+                {% endif %}
+            </div>
+        {% endif %}
+    {% endif %}
+{% endmacro %}
+
+{% macro render_field(field, disabled_fields=[], exclude_fields=[], class='') %}
+    <!-- Debug info -->
+    <!-- Field name: {{ field.name }}, Field type: {{ field.__class__.__name__ }} -->
+
+    {% set disabled = field.name in disabled_fields %}
+    {% set exclude_fields = exclude_fields + ['csrf_token', 'submit'] %}
+    {% if field.name not in exclude_fields %}
+        {% if field.type == 'BooleanField' %}
+            <div class="form-group">
+                <div class="form-check form-switch">
+                    {{ field(class="form-check-input " + class, disabled=disabled) }}
+                    {% if field.description %}
+                        {{ field.label(class="form-check-label",
+                                     **{'data-bs-toggle': 'tooltip',
+                                        'data-bs-placement': 'right',
+                                        'title': field.description}) }}
+                    {% else %}
+                        {{ field.label(class="form-check-label") }}
+                    {% endif %}
+                </div>
+                {% if field.errors %}
+                    <div class="invalid-feedback d-block">
+                        {% for error in field.errors %}
+                            {{ error }}
+                        {% endfor %}
+                    </div>
+                {% endif %}
+            </div>
+        {% else %}
+            <div class="form-group">
+                {% if field.description %}
+                    {{ field.label(class="form-label",
+                                 **{'data-bs-toggle': 'tooltip',
+                                    'data-bs-placement': 'right',
+                                    'title': field.description}) }}
+                {% else %}
+                    {{ field.label(class="form-label") }}
+                {% endif %}
+
+                {% if field.type == 'TextAreaField' and 'json-editor' in class %}
+                    <div id="{{ field.id }}-editor" class="json-editor-container"></div>
+                    {{ field(class="form-control d-none " + class, disabled=disabled) }}
+                {% elif field.type == 'SelectField' %}
+                    {{ field(class="form-control form-select " + class, disabled=disabled) }}
+                {% else %}
+                    {{ field(class="form-control " + class, disabled=disabled) }}
+                {% endif %}
+
+                {% if field.errors %}
+                    <div class="invalid-feedback d-block">
                         {% for error in field.errors %}
                             {{ error }}
                         {% endfor %}
@@ -177,6 +295,48 @@
 </div>
 {% endmacro %}
 
+{% macro render_selectable_sortable_table_with_dict_headers(headers, rows, selectable, id, sort_by, sort_order) %}
+<div class="card">
+  <div class="table-responsive">
+    <table class="table align-items-center mb-0" id="{{ id }}">
+      <thead>
+        <tr>
+          {% if selectable %}
+            <th class="text-uppercase text-secondary text-xxs font-weight-bolder opacity-7">Select</th>
+          {% endif %}
+          {% for header in headers %}
+            <th class="text-uppercase text-secondary text-xxs font-weight-bolder opacity-7 sortable" data-sort="{{ header['sort'] }}">
+              {{ header['text'] }}
+              {% if sort_by == header['sort'] %}
+                {% if sort_order == 'asc' %}
+                  <i class="fas fa-sort-up"></i>
+                {% elif sort_order == 'desc' %}
+                  <i class="fas fa-sort-down"></i>
+                {% endif %}
+              {% else %}
+                <i class="fas fa-sort"></i>
+              {% endif %}
+            </th>
+          {% endfor %}
+        </tr>
+      </thead>
+      <tbody>
+        {% for row in rows %}
+          <tr>
+            {% if selectable %}
+              <td><input type="radio" name="selected_row" value="{{ row[0].value }}"></td>
+            {% endif %}
+            {% for cell in row %}
+              <td>{{ cell.value }}</td>
+            {% endfor %}
+          </tr>
+        {% endfor %}
+      </tbody>
+    </table>
+  </div>
+</div>
+{% endmacro %}
+
 {% macro render_accordion(accordion_id, accordion_items, header_title, header_description) %}
 <div class="accordion-1">
   <div class="container">

eveai_app/templates/navbar.html

@@ -81,6 +81,10 @@
                             {% endif %}
                             {% if current_user.is_authenticated %}
                                 {{ dropdown('Document Mgmt', 'note_stack', [
+                                    {'name': 'Add Catalog', 'url': '/document/catalog', 'roles': ['Super User', 'Tenant Admin']},
+                                    {'name': 'All Catalogs', 'url': '/document/catalogs', 'roles': ['Super User', 'Tenant Admin']},
+                                    {'name': 'Add Retriever', 'url': '/document/retriever', 'roles': ['Super User', 'Tenant Admin']},
+                                    {'name': 'All Retrievers', 'url': '/document/retrievers', 'roles': ['Super User', 'Tenant Admin']},
                                     {'name': 'Add Document', 'url': '/document/add_document', 'roles': ['Super User', 'Tenant Admin']},
                                     {'name': 'Add URL', 'url': '/document/add_url', 'roles': ['Super User', 'Tenant Admin']},
                                     {'name': 'Add a list of URLs', 'url': '/document/add_urls', 'roles': ['Super User', 'Tenant Admin']},
@@ -114,6 +118,17 @@
                             {% endif %}
                         </ul>
                         {% if current_user.is_authenticated %}
+                            <ul class="navbar-nav d-lg-block d-none">
+                                <li class="nav-item">
+                                    <a href="/document/catalogs" class="btn btn-sm  bg-gradient-primary  mb-0 me-2">
+                                        {% if 'catalog_name' in session %}
+                                            CATALOG: {{ session['catalog_name'] }}
+                                        {% else %}
+                                            CHOOSE CATALOG
+                                        {% endif %}
+                                    </a>
+                                </li>
+                            </ul>
                             <ul class="navbar-nav d-lg-block d-none">
                                 <li class="nav-item">
                                     <a href="/session_defaults" class="btn btn-sm  bg-gradient-primary  mb-0">

eveai_app/templates/scripts.html

@@ -14,4 +14,58 @@
         <script src="{{url_for('static', filename='assets/js/material-kit-pro.min.js')}}?v=3.0.4 type="text/javascript"></script>
         <script src="https://cdnjs.cloudflare.com/ajax/libs/bootstrap/5.3.3/js/bootstrap.bundle.min.js"></script>
         <script src="https://cdnjs.cloudflare.com/ajax/libs/select2/4.0.13/js/select2.min.js"></script>
+        <link href="https://cdnjs.cloudflare.com/ajax/libs/jsoneditor/10.1.0/jsoneditor.min.css" rel="stylesheet" type="text/css">
+        <script src="https://cdnjs.cloudflare.com/ajax/libs/jsoneditor/10.1.0/jsoneditor.min.js"></script>
 
+<script>
+document.addEventListener('DOMContentLoaded', function() {
+    // Initialize tooltips
+    var tooltipTriggerList = [].slice.call(document.querySelectorAll('[data-bs-toggle="tooltip"]'))
+    var tooltipList = tooltipTriggerList.map(function (tooltipTriggerEl) {
+        return new bootstrap.Tooltip(tooltipTriggerEl)
+    });
+
+    // Initialize JSON editors
+    document.querySelectorAll('.json-editor').forEach(function(textarea) {
+        // Create container for editor
+        var container = document.getElementById(textarea.id + '-editor');
+
+        // Initialize the editor
+        var editor = new JSONEditor(container, {
+            mode: 'code',
+            modes: ['code', 'tree'],
+            onChangeText: function(jsonString) {
+                textarea.value = jsonString;
+            }
+        });
+
+        // Set initial value
+        try {
+            const initialValue = textarea.value ? JSON.parse(textarea.value) : {};
+            editor.set(initialValue);
+        } catch (e) {
+            console.error('Error parsing initial JSON:', e);
+            editor.set({});
+        }
+
+        // Add validation indicator
+        editor.validate().then(function(errors) {
+            if (errors.length) {
+                container.style.border = '2px solid red';
+            } else {
+                container.style.border = '1px solid #ccc';
+            }
+        });
+    });
+});
+</script>
+
+<style>
+.json-editor-container {
+    height: 400px;
+    margin-bottom: 1rem;
+}
+.tooltip {
+    position: fixed;
+}
+</style>

eveai_app/templates/user/tenant.html

@@ -10,7 +10,7 @@
     <form method="post">
         {{ form.hidden_tag() }}
         <!-- Main Tenant Information -->
-        {% set main_fields = ['name', 'website', 'default_language', 'allowed_languages', 'rag_context', 'type'] %}
+        {% set main_fields = ['name', 'website', 'default_language', 'allowed_languages', 'timezone','rag_context', 'type'] %}
         {% for field in form %}
             {{ render_included_field(field, disabled_fields=[], include_fields=main_fields) }}
         {% endfor %}
@@ -30,21 +30,6 @@
                                 License Information
                             </a>
                         </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#chunking-tab" role="tab" aria-controls="chunking" aria-selected="false">
-                                Chunking
-                            </a>
-                        </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#embedding-search-tab" role="tab" aria-controls="html-chunking" aria-selected="false">
-                                Embedding Search
-                            </a>
-                        </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#tuning-tab" role="tab" aria-controls="html-chunking" aria-selected="false">
-                                Tuning
-                            </a>
-                        </li>
                     </ul>
                 </div>
                 <div class="tab-content tab-space">
@@ -62,8 +47,10 @@
                             {{ render_included_field(field, disabled_fields=[], include_fields=license_fields) }}
                         {% endfor %}
                         <!-- Register API Key Button -->
-                        <button type="button" class="btn btn-primary" onclick="generateNewChatApiKey()">Register Chat API Key</button>
-                        <button type="button" class="btn btn-primary" onclick="generateNewApiKey()">Register API Key</button>
+                        <div class="form-group">
+                            <button type="button" class="btn btn-primary" onclick="generateNewChatApiKey()">Register Chat API Key</button>
+                            <button type="button" class="btn btn-primary" onclick="generateNewApiKey()">Register API Key</button>
+                        </div>
                         <!-- API Key Display Field -->
                         <div id="chat-api-key-field" style="display:none;">
                             <label for="chat-api-key">Chat API Key:</label>
@@ -78,27 +65,6 @@
                             <p id="copy-message" style="display:none;color:green;">API key copied to clipboard</p>
                         </div>
                     </div>
-                    <!-- Chunking Settings Tab -->
-                    <div class="tab-pane fade" id="chunking-tab" role="tabpanel">
-                        {% set html_fields = ['html_tags', 'html_end_tags', 'html_included_elements', 'html_excluded_elements', 'html_excluded_classes', 'min_chunk_size', 'max_chunk_size'] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=[], include_fields=html_fields) }}
-                        {% endfor %}
-                    </div>
-                    <!-- Embedding Search Settings Tab -->
-                    <div class="tab-pane fade" id="embedding-search-tab" role="tabpanel">
-                        {% set es_fields = ['es_k', 'es_similarity_threshold', ] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=[], include_fields=es_fields) }}
-                        {% endfor %}
-                    </div>
-                    <!-- Tuning Settings Tab -->
-                    <div class="tab-pane fade" id="tuning-tab" role="tabpanel">
-                        {% set tuning_fields = ['embed_tuning', 'rag_tuning', ] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=[], include_fields=tuning_fields) }}
-                        {% endfor %}
-                    </div>
                 </div>
             </div>
         </div>

eveai_app/templates/user/tenant_overview.html

@@ -30,21 +30,6 @@
                                 License Information
                             </a>
                         </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#chunking-tab" role="tab" aria-controls="chunking" aria-selected="false">
-                                Chunking
-                            </a>
-                        </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#embedding-search-tab" role="tab" aria-controls="html-chunking" aria-selected="false">
-                                Embedding Search
-                            </a>
-                        </li>
-                        <li class="nav-item">
-                            <a class="nav-link mb-0 px-0 py-1" data-toggle="tab" href="#tuning-tab" role="tab" aria-controls="html-chunking" aria-selected="false">
-                                Tuning
-                            </a>
-                        </li>
                     </ul>
                 </div>
                 <div class="tab-content tab-space">
@@ -78,27 +63,6 @@
                             <p id="copy-message" style="display:none;color:green;">API key copied to clipboard</p>
                         </div>
                     </div>
-                    <!-- Chunking Settings Tab -->
-                    <div class="tab-pane fade" id="chunking-tab" role="tabpanel">
-                        {% set html_fields = ['html_tags', 'html_end_tags', 'html_included_elements', 'html_excluded_elements', 'html_excluded_classes', 'min_chunk_size', 'max_chunk_size'] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=html_fields, include_fields=html_fields) }}
-                        {% endfor %}
-                    </div>
-                    <!-- Embedding Search Settings Tab -->
-                    <div class="tab-pane fade" id="embedding-search-tab" role="tabpanel">
-                        {% set es_fields = ['es_k', 'es_similarity_threshold', ] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=es_fields, include_fields=es_fields) }}
-                        {% endfor %}
-                    </div>
-                    <!-- Tuning Settings Tab -->
-                    <div class="tab-pane fade" id="tuning-tab" role="tabpanel">
-                        {% set tuning_fields = ['embed_tuning', 'rag_tuning', ] %}
-                        {% for field in form %}
-                            {{ render_included_field(field, disabled_fields=tuning_fields, include_fields=tuning_fields) }}
-                        {% endfor %}
-                    </div>
                 </div>
             </div>
         </div>

eveai_app/views/document_forms.py

@@ -1,11 +1,20 @@
-from flask import session, current_app
+from flask import session, current_app, request
 from flask_wtf import FlaskForm
-from wtforms import (StringField, BooleanField, SubmitField, DateField,
+from wtforms import (StringField, BooleanField, SubmitField, DateField, IntegerField, FloatField, SelectMultipleField,
                      SelectField, FieldList, FormField, TextAreaField, URLField)
-from wtforms.validators import DataRequired, Length, Optional, URL, ValidationError
+from wtforms.validators import DataRequired, Length, Optional, URL, ValidationError, NumberRange
 from flask_wtf.file import FileField, FileAllowed, FileRequired
 import json
 
+from wtforms_sqlalchemy.fields import QuerySelectField
+
+from common.extensions import db
+from common.models.document import Catalog
+
+from config.catalog_types import CATALOG_TYPES
+from config.retriever_types import RETRIEVER_TYPES
+from .dynamic_form_base import DynamicFormBase
+
 
 def allowed_file(form, field):
     if field.data:
@@ -23,7 +32,126 @@ def validate_json(form, field):
             raise ValidationError('Invalid JSON format')
 
 
-class AddDocumentForm(FlaskForm):
+class CatalogForm(FlaskForm):
+    name = StringField('Name', validators=[DataRequired(), Length(max=50)])
+    description = TextAreaField('Description', validators=[Optional()])
+
+    # Select Field for Catalog Type (Uses the CATALOG_TYPES defined in config)
+    type = SelectField('Catalog Type', validators=[DataRequired()])
+
+    # Metadata fields
+    user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
+    system_metadata = TextAreaField('System Metadata', validators=[Optional(), validate_json])
+    configuration = TextAreaField('Configuration', validators=[Optional(), validate_json])
+
+    # HTML Embedding Variables
+    html_tags = StringField('HTML Tags', validators=[DataRequired()],
+                            default='p, h1, h2, h3, h4, h5, h6, li, , tbody, tr, td')
+    html_end_tags = StringField('HTML End Tags', validators=[DataRequired()],
+                                default='p, li')
+    html_included_elements = StringField('HTML Included Elements', validators=[Optional()], default='article, main')
+    html_excluded_elements = StringField('HTML Excluded Elements', validators=[Optional()],
+                                         default='header, footer, nav, script')
+    html_excluded_classes = StringField('HTML Excluded Classes', validators=[Optional()])
+    min_chunk_size = IntegerField('Minimum Chunk Size (2000)', validators=[NumberRange(min=0), Optional()],
+                                  default=2000)
+    max_chunk_size = IntegerField('Maximum Chunk Size (3000)', validators=[NumberRange(min=0), Optional()],
+                                  default=3000)
+    # Chat Variables
+    chat_RAG_temperature = FloatField('RAG Temperature', default=0.3, validators=[NumberRange(min=0, max=1)])
+    chat_no_RAG_temperature = FloatField('No RAG Temperature', default=0.5, validators=[NumberRange(min=0, max=1)])
+    # Tuning variables
+    embed_tuning = BooleanField('Enable Embedding Tuning', default=False)
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        # Dynamically populate the 'type' field using the constructor
+        self.type.choices = [(key, value['name']) for key, value in CATALOG_TYPES.items()]
+
+
+class EditCatalogForm(DynamicFormBase):
+    name = StringField('Name', validators=[DataRequired(), Length(max=50)])
+    description = TextAreaField('Description', validators=[Optional()])
+
+    # Select Field for Catalog Type (Uses the CATALOG_TYPES defined in config)
+    type = StringField('Catalog Type', validators=[DataRequired()], render_kw={'readonly': True})
+
+    # Metadata fields
+    user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
+    system_metadata = TextAreaField('System Metadata', validators=[Optional(), validate_json],)
+
+    # HTML Embedding Variables
+    html_tags = StringField('HTML Tags', validators=[DataRequired()],
+                            default='p, h1, h2, h3, h4, h5, h6, li, , tbody, tr, td')
+    html_end_tags = StringField('HTML End Tags', validators=[DataRequired()],
+                                default='p, li')
+    html_included_elements = StringField('HTML Included Elements', validators=[Optional()], default='article, main')
+    html_excluded_elements = StringField('HTML Excluded Elements', validators=[Optional()],
+                                         default='header, footer, nav, script')
+    html_excluded_classes = StringField('HTML Excluded Classes', validators=[Optional()])
+    min_chunk_size = IntegerField('Minimum Chunk Size (2000)', validators=[NumberRange(min=0), Optional()],
+                                  default=2000)
+    max_chunk_size = IntegerField('Maximum Chunk Size (3000)', validators=[NumberRange(min=0), Optional()],
+                                  default=3000)
+    # Chat Variables
+    chat_RAG_temperature = FloatField('RAG Temperature', default=0.3, validators=[NumberRange(min=0, max=1)])
+    chat_no_RAG_temperature = FloatField('No RAG Temperature', default=0.5, validators=[NumberRange(min=0, max=1)])
+    # Tuning variables
+    embed_tuning = BooleanField('Enable Embedding Tuning', default=False)
+
+
+class RetrieverForm(FlaskForm):
+    name = StringField('Name', validators=[DataRequired(), Length(max=50)])
+    description = TextAreaField('Description', validators=[Optional()])
+    # Catalog for the Retriever
+    catalog = QuerySelectField(
+        'Catalog ID',
+        query_factory=lambda: Catalog.query.all(),
+        allow_blank=True,
+        get_label='name',
+        validators=[Optional()],
+    )
+    # Select Field for Retriever Type (Uses the RETRIEVER_TYPES defined in config)
+    type = SelectField('Retriever Type', validators=[DataRequired()])
+    tuning = BooleanField('Enable Tuning', default=False)
+
+    # Metadata fields
+    user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
+    system_metadata = TextAreaField('System Metadata', validators=[Optional(), validate_json])
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        # Dynamically populate the 'type' field using the constructor
+        self.type.choices = [(key, value['name']) for key, value in RETRIEVER_TYPES.items()]
+
+
+class EditRetrieverForm(DynamicFormBase):
+    name = StringField('Name', validators=[DataRequired(), Length(max=50)])
+    description = TextAreaField('Description', validators=[Optional()])
+    # Catalog for the Retriever
+    catalog = QuerySelectField(
+        'Catalog ID',
+        query_factory=lambda: Catalog.query.all(),
+        allow_blank=True,
+        get_label='name',
+        validators=[Optional()],
+    )
+    # Select Field for Retriever Type (Uses the RETRIEVER_TYPES defined in config)
+    type = SelectField('Retriever Type', validators=[DataRequired()], render_kw={'readonly': True})
+    tuning = BooleanField('Enable Tuning', default=False)
+
+    # Metadata fields
+    user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
+    system_metadata = TextAreaField('System Metadata', validators=[Optional(), validate_json])
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+
+        # Set the retriever type choices (loaded from config)
+        self.type.choices = [(key, value['name']) for key, value in RETRIEVER_TYPES.items()]
+
+
+class AddDocumentForm(DynamicFormBase):
     file = FileField('File', validators=[FileRequired(), allowed_file])
     name = StringField('Name', validators=[Length(max=100)])
     language = SelectField('Language', choices=[], validators=[Optional()])
@@ -31,17 +159,15 @@ class AddDocumentForm(FlaskForm):
     valid_from = DateField('Valid from', id='form-control datepicker', validators=[Optional()])
     user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
 
-    submit = SubmitField('Submit')
-
-    def __init__(self):
-        super().__init__()
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
         self.language.choices = [(language, language) for language in
                                  session.get('tenant').get('allowed_languages')]
         if not self.language.data:
             self.language.data = session.get('tenant').get('default_language')
 
 
-class AddURLForm(FlaskForm):
+class AddURLForm(DynamicFormBase):
     url = URLField('URL', validators=[DataRequired(), URL()])
     name = StringField('Name', validators=[Length(max=100)])
     language = SelectField('Language', choices=[], validators=[Optional()])
@@ -49,10 +175,8 @@ class AddURLForm(FlaskForm):
     valid_from = DateField('Valid from', id='form-control datepicker', validators=[Optional()])
     user_metadata = TextAreaField('User Metadata', validators=[Optional(), validate_json])
 
-    submit = SubmitField('Submit')
-
-    def __init__(self):
-        super().__init__()
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
         self.language.choices = [(language, language) for language in
                                  session.get('tenant').get('allowed_languages')]
         if not self.language.data:
@@ -84,7 +208,7 @@ class EditDocumentForm(FlaskForm):
     submit = SubmitField('Submit')
 
 
-class EditDocumentVersionForm(FlaskForm):
+class EditDocumentVersionForm(DynamicFormBase):
     language = StringField('Language')
     user_context = TextAreaField('User Context', validators=[Optional()])
     system_context = TextAreaField('System Context', validators=[Optional()])

eveai_app/views/document_list_view.py

@@ -0,0 +1,102 @@
+from datetime import datetime
+from flask import request, render_template, session
+from sqlalchemy import desc, asc, or_, and_, cast, Integer
+from common.models.document import Document, Catalog
+from common.utils.filtered_list_view import FilteredListView
+from common.utils.view_assistants import prepare_table_for_macro
+
+
+class DocumentListView(FilteredListView):
+    allowed_filters = ['catalog_id', 'validity']
+    allowed_sorts = ['id', 'name', 'catalog_name', 'valid_from', 'valid_to']
+
+    def get_query(self):
+        return Document.query.join(Catalog).add_columns(
+            Document.id,
+            Document.name,
+            Catalog.name.label('catalog_name'),
+            Document.valid_from,
+            Document.valid_to
+        )
+
+    def apply_filters(self, query):
+        filters = request.args.to_dict(flat=False)
+
+        if 'catalog_id' in filters:
+            catalog_ids = filters['catalog_id']
+            if catalog_ids:
+                # Convert catalog_ids to a list of integers
+                catalog_ids = [int(cid) for cid in catalog_ids if cid.isdigit()]
+                if catalog_ids:
+                    query = query.filter(Document.catalog_id.in_(catalog_ids))
+
+        if 'validity' in filters:
+            now = datetime.utcnow().date()
+            if 'valid' in filters['validity']:
+                query = query.filter(
+                    and_(
+                        or_(Document.valid_from.is_(None), Document.valid_from <= now),
+                        or_(Document.valid_to.is_(None), Document.valid_to >= now)
+                    )
+                )
+
+        return query
+
+    def apply_sorting(self, query):
+        sort_by = request.args.get('sort_by', 'id')
+        sort_order = request.args.get('sort_order', 'asc')
+
+        if sort_by in self.allowed_sorts:
+            if sort_by == 'catalog_name':
+                column = Catalog.name
+            else:
+                column = getattr(Document, sort_by)
+
+            if sort_order == 'asc':
+                query = query.order_by(asc(column))
+            elif sort_order == 'desc':
+                query = query.order_by(desc(column))
+
+        return query
+
+    def get(self):
+        query = self.get_query()
+        query = self.apply_filters(query)
+        query = self.apply_sorting(query)
+        pagination = self.paginate(query)
+
+        def format_date(date):
+            if isinstance(date, datetime):
+                return date.strftime('%Y-%m-%d')
+            elif isinstance(date, str):
+                return date
+            else:
+                return ''
+
+        rows = [
+            [
+                {'value': item.id, 'class': '', 'type': 'text'},
+                {'value': item.name, 'class': '', 'type': 'text'},
+                {'value': item.catalog_name, 'class': '', 'type': 'text'},
+                {'value': format_date(item.valid_from), 'class': '', 'type': 'text'},
+                {'value': format_date(item.valid_to), 'class': '', 'type': 'text'}
+            ] for item in pagination.items
+        ]
+
+        catalogs = Catalog.query.all()
+
+        context = {
+            'rows': rows,
+            'pagination': pagination,
+            'filters': request.args.to_dict(flat=False),
+            'sort_by': request.args.get('sort_by', 'id'),
+            'sort_order': request.args.get('sort_order', 'asc'),
+            'filter_options': self.get_filter_options(catalogs)
+        }
+        return render_template(self.template, **context)
+
+    def get_filter_options(self, catalogs):
+        return {
+            'catalog_id': [(str(cat.id), cat.name) for cat in catalogs],
+            'validity': [('valid', 'Valid'), ('all', 'All')]
+        }

eveai_app/views/document_version_list_view.py

@@ -12,7 +12,7 @@ class DocumentVersionListView(FilteredListView):
     allowed_sorts = ['id', 'processing_started_at', 'processing_finished_at', 'processing_error']
 
     def get_query(self):
-        return DocumentVersion.query.join(Document).filter(Document.tenant_id == session.get('tenant', {}).get('id'))
+        return DocumentVersion.query.join(Document)
 
     def apply_filters(self, query):
         filters = request.args.to_dict()

eveai_app/views/document_views.py

@@ -1,9 +1,11 @@
 import ast
 from datetime import datetime as dt, timezone as tz
 
+from babel.messages.setuptools_frontend import update_catalog
 from flask import request, redirect, flash, render_template, Blueprint, session, current_app
 from flask_security import roles_accepted, current_user
 from sqlalchemy import desc
+from sqlalchemy.orm import aliased
 from werkzeug.utils import secure_filename
 from sqlalchemy.exc import SQLAlchemyError
 import requests
@@ -12,19 +14,23 @@ from urllib.parse import urlparse, unquote
 import io
 import json
 
-from common.models.document import Document, DocumentVersion
+from common.models.document import Document, DocumentVersion, Catalog, Retriever
 from common.extensions import db, minio_client
 from common.utils.document_utils import validate_file_type, create_document_stack, start_embedding_task, process_url, \
     process_multiple_urls, get_documents_list, edit_document, \
     edit_document_version, refresh_document
 from common.utils.eveai_exceptions import EveAIInvalidLanguageException, EveAIUnsupportedFileType, \
     EveAIDoubleURLException
-from .document_forms import AddDocumentForm, AddURLForm, EditDocumentForm, EditDocumentVersionForm, AddURLsForm
+from .document_forms import AddDocumentForm, AddURLForm, EditDocumentForm, EditDocumentVersionForm, AddURLsForm, \
+    CatalogForm, EditCatalogForm, RetrieverForm, EditRetrieverForm
 from common.utils.middleware import mw_before_request
 from common.utils.celery_utils import current_celery
 from common.utils.nginx_utils import prefixed_url_for
 from common.utils.view_assistants import form_validation_failed, prepare_table_for_macro, form_to_dict
+from .document_list_view import DocumentListView
 from .document_version_list_view import DocumentVersionListView
+from config.catalog_types import CATALOG_TYPES
+from config.retriever_types import RETRIEVER_TYPES
 
 document_bp = Blueprint('document_bp', __name__, url_prefix='/document')
 
@@ -52,13 +58,260 @@ def before_request():
         raise
 
 
+@document_bp.route('/catalog', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def catalog():
+    form = CatalogForm()
+
+    if form.validate_on_submit():
+        tenant_id = session.get('tenant').get('id')
+        new_catalog = Catalog()
+        form.populate_obj(new_catalog)
+        # Handle Embedding Variables
+        new_catalog.html_tags = [tag.strip() for tag in form.html_tags.data.split(',')] if form.html_tags.data else []
+        new_catalog.html_end_tags = [tag.strip() for tag in form.html_end_tags.data.split(',')] \
+            if form.html_end_tags.data else []
+        new_catalog.html_included_elements = [tag.strip() for tag in form.html_included_elements.data.split(',')] \
+            if form.html_included_elements.data else []
+        new_catalog.html_excluded_elements = [tag.strip() for tag in form.html_excluded_elements.data.split(',')] \
+            if form.html_excluded_elements.data else []
+        new_catalog.html_excluded_classes = [cls.strip() for cls in form.html_excluded_classes.data.split(',')] \
+            if form.html_excluded_classes.data else []
+        set_logging_information(new_catalog, dt.now(tz.utc))
+
+        try:
+            db.session.add(new_catalog)
+            db.session.commit()
+            flash('Catalog successfully added!', 'success')
+            current_app.logger.info(f'Catalog {new_catalog.name} successfully added for tenant {tenant_id}!')
+        except SQLAlchemyError as e:
+            db.session.rollback()
+            flash(f'Failed to add catalog. Error: {e}', 'danger')
+            current_app.logger.error(f'Failed to add catalog {new_catalog.name}'
+                                     f'for tenant {tenant_id}. Error: {str(e)}')
+
+    return render_template('document/catalog.html', form=form)
+
+
+@document_bp.route('/catalogs', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def catalogs():
+    page = request.args.get('page', 1, type=int)
+    per_page = request.args.get('per_page', 10, type=int)
+
+    query = Catalog.query.order_by(Catalog.id)
+
+    pagination = query.paginate(page=page, per_page=per_page)
+    the_catalogs = pagination.items
+
+    # prepare table data
+    rows = prepare_table_for_macro(the_catalogs, [('id', ''), ('name', ''), ('type', '')])
+
+    # Render the catalogs in a template
+    return render_template('document/catalogs.html', rows=rows, pagination=pagination)
+
+
+@document_bp.route('/handle_catalog_selection', methods=['POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def handle_catalog_selection():
+    catalog_identification = request.form.get('selected_row')
+    catalog_id = ast.literal_eval(catalog_identification).get('value')
+    action = request.form['action']
+    catalog = Catalog.query.get_or_404(catalog_id)
+
+    if action == 'set_session_catalog':
+        current_app.logger.info(f'Setting session catalog to {catalog.name}')
+        session['catalog_id'] = catalog_id
+        session['catalog_name'] = catalog.name
+        current_app.logger.info(f'Finished setting session catalog to {catalog.name}')
+    elif action == 'edit_catalog':
+        return redirect(prefixed_url_for('document_bp.edit_catalog', catalog_id=catalog_id))
+
+    return redirect(prefixed_url_for('document_bp.catalogs'))
+
+
+@document_bp.route('/catalog/<int:catalog_id>', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def edit_catalog(catalog_id):
+    catalog = Catalog.query.get_or_404(catalog_id)
+    tenant_id = session.get('tenant').get('id')
+
+    form = EditCatalogForm(request.form, obj=catalog)
+    configuration_config = CATALOG_TYPES[catalog.type]["configuration"]
+    form.add_dynamic_fields("configuration", configuration_config, catalog.configuration)
+
+    # Convert arrays to comma-separated strings for display
+    if request.method == 'GET':
+        form.html_tags.data = ', '.join(catalog.html_tags or '')
+        form.html_end_tags.data = ', '.join(catalog.html_end_tags or '')
+        form.html_included_elements.data = ', '.join(catalog.html_included_elements or '')
+        form.html_excluded_elements.data = ', '.join(catalog.html_excluded_elements or '')
+        form.html_excluded_classes.data = ', '.join(catalog.html_excluded_classes or '')
+
+    if request.method == 'POST' and form.validate_on_submit():
+        form.populate_obj(catalog)
+        # Handle Embedding Variables
+        catalog.html_tags = [tag.strip() for tag in form.html_tags.data.split(',')] if form.html_tags.data else []
+        catalog.html_end_tags = [tag.strip() for tag in form.html_end_tags.data.split(',')] \
+            if form.html_end_tags.data else []
+        catalog.html_included_elements = [tag.strip() for tag in form.html_included_elements.data.split(',')] \
+            if form.html_included_elements.data else []
+        catalog.html_excluded_elements = [tag.strip() for tag in form.html_excluded_elements.data.split(',')] \
+            if form.html_excluded_elements.data else []
+        catalog.html_excluded_classes = [cls.strip() for cls in form.html_excluded_classes.data.split(',')] \
+            if form.html_excluded_classes.data else []
+
+        catalog.configuration = form.get_dynamic_data('configuration')
+        update_logging_information(catalog, dt.now(tz.utc))
+        try:
+            db.session.add(catalog)
+            db.session.commit()
+            flash('Catalog successfully updated successfully!', 'success')
+            current_app.logger.info(f'Catalog {catalog.name} successfully updated for tenant {tenant_id}')
+        except SQLAlchemyError as e:
+            db.session.rollback()
+            flash(f'Failed to update catalog. Error: {e}', 'danger')
+            current_app.logger.error(f'Failed to update catalog {catalog_id} for tenant {tenant_id}. Error: {str(e)}')
+
+        return redirect(prefixed_url_for('document_bp.catalogs'))
+    else:
+        form_validation_failed(request, form)
+
+    return render_template('document/edit_catalog.html', form=form, catalog_id=catalog_id)
+
+
+@document_bp.route('/retriever', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def retriever():
+    form = RetrieverForm()
+
+    if form.validate_on_submit():
+        tenant_id = session.get('tenant').get('id')
+        new_retriever = Retriever()
+        form.populate_obj(new_retriever)
+        new_retriever.catalog_id = form.catalog.data.id
+
+        set_logging_information(new_retriever, dt.now(tz.utc))
+
+        try:
+            db.session.add(new_retriever)
+            db.session.commit()
+            flash('Retriever successfully added!', 'success')
+            current_app.logger.info(f'Catalog {new_retriever.name} successfully added for tenant {tenant_id}!')
+        except SQLAlchemyError as e:
+            db.session.rollback()
+            flash(f'Failed to add retriever. Error: {e}', 'danger')
+            current_app.logger.error(f'Failed to add retriever {new_retriever.name}'
+                                     f'for tenant {tenant_id}. Error: {str(e)}')
+
+        # Enable step 2 of creation of retriever - add configuration of the retriever (dependent on type)
+        return redirect(prefixed_url_for('document_bp.retriever', retriever_id=new_retriever.id))
+
+    return render_template('document/retriever.html', form=form)
+
+
+@document_bp.route('/retriever/<int:retriever_id>', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def edit_retriever(retriever_id):
+    """Edit an existing retriever configuration."""
+    # Get the retriever or return 404
+    retriever = Retriever.query.get_or_404(retriever_id)
+
+    if retriever.catalog_id:
+        # If catalog_id is just an ID, fetch the Catalog object
+        retriever.catalog = Catalog.query.get(retriever.catalog_id)
+    else:
+        retriever.catalog = None
+
+    # Create form instance with the retriever
+    form = EditRetrieverForm(request.form, obj=retriever)
+
+    configuration_config = RETRIEVER_TYPES[retriever.type]["configuration"]
+    form.add_dynamic_fields("configuration", configuration_config, retriever.configuration)
+    if request.method == 'POST':
+        current_app.logger.debug(f'Received POST request with {request.form}')
+
+    if form.validate_on_submit():
+        # Update basic fields
+        form.populate_obj(retriever)
+        retriever.configuration = form.get_dynamic_data('configuration')
+
+        # Update catalog relationship
+        retriever.catalog_id = form.catalog.data.id if form.catalog.data else None
+
+        # Update logging information
+        update_logging_information(retriever, dt.now(tz.utc))
+
+        # Save changes to database
+        try:
+            db.session.add(retriever)
+            db.session.commit()
+            flash('Retriever updated successfully!', 'success')
+            current_app.logger.info(f'Retriever {retriever.id} updated successfully')
+        except SQLAlchemyError as e:
+            db.session.rollback()
+            flash(f'Failed to update retriever. Error: {str(e)}', 'danger')
+            current_app.logger.error(f'Failed to update retriever {retriever_id}. Error: {str(e)}')
+            return render_template('document/edit_retriever.html', form=form, retriever_id=retriever_id)
+
+        return redirect(prefixed_url_for('document_bp.retrievers'))
+    else:
+        form_validation_failed(request, form)
+
+    current_app.logger.debug(f"Rendering Template for {retriever_id}")
+    return render_template('document/edit_retriever.html', form=form, retriever_id=retriever_id)
+
+
+@document_bp.route('/retrievers', methods=['GET', 'POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def retrievers():
+    page = request.args.get('page', 1, type=int)
+    per_page = request.args.get('per_page', 10, type=int)
+
+    query = Retriever.query.order_by(Retriever.id)
+
+    pagination = query.paginate(page=page, per_page=per_page)
+    the_retrievers = pagination.items
+
+    # prepare table data
+    rows = prepare_table_for_macro(the_retrievers,
+                                   [('id', ''), ('name', ''), ('type', ''), ('catalog_id', '')])
+
+    # Render the catalogs in a template
+    return render_template('document/retrievers.html', rows=rows, pagination=pagination)
+
+
+@document_bp.route('/handle_retriever_selection', methods=['POST'])
+@roles_accepted('Super User', 'Tenant Admin')
+def handle_retriever_selection():
+    retriever_identification = request.form.get('selected_row')
+    retriever_id = ast.literal_eval(retriever_identification).get('value')
+    action = request.form['action']
+
+    if action == 'edit_retriever':
+        return redirect(prefixed_url_for('document_bp.edit_retriever', retriever_id=retriever_id))
+
+    return redirect(prefixed_url_for('document_bp.retrievers'))
+
+
 @document_bp.route('/add_document', methods=['GET', 'POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def add_document():
-    form = AddDocumentForm()
+    form = AddDocumentForm(request.form)
+    catalog_id = session.get('catalog_id', None)
+    if catalog_id is None:
+        flash('You need to set a Session Catalog before adding Documents or URLs')
+        return redirect(prefixed_url_for('document_bp.catalogs'))
+
+    catalog = Catalog.query.get_or_404(catalog_id)
+    if catalog.configuration and len(catalog.configuration) > 0:
+        document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+        for config in document_version_configurations:
+            form.add_dynamic_fields(config, catalog.configuration[config])
 
     if form.validate_on_submit():
         try:
+            current_app.logger.info(f'Adding Document for {catalog_id}')
             tenant_id = session['tenant']['id']
             file = form.file.data
             filename = secure_filename(file.filename)
@@ -66,13 +319,19 @@ def add_document():
 
             validate_file_type(extension)
 
-            current_app.logger.debug(f'Language on form: {form.language.data}')
+            catalog_properties = {}
+            document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+            for config in document_version_configurations:
+                catalog_properties[config] = form.get_dynamic_data(config)
+
             api_input = {
+                'catalog_id': catalog_id,
                 'name': form.name.data,
                 'language': form.language.data,
                 'user_context': form.user_context.data,
                 'valid_from': form.valid_from.data,
                 'user_metadata': json.loads(form.user_metadata.data) if form.user_metadata.data else None,
+                'catalog_properties': catalog_properties,
             }
             current_app.logger.debug(f'Creating document stack with input {api_input}')
 
@@ -95,7 +354,17 @@ def add_document():
 @document_bp.route('/add_url', methods=['GET', 'POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def add_url():
-    form = AddURLForm()
+    form = AddURLForm(request.form)
+    catalog_id = session.get('catalog_id', None)
+    if catalog_id is None:
+        flash('You need to set a Session Catalog before adding Documents or URLs')
+        return redirect(prefixed_url_for('document_bp.catalogs'))
+
+    catalog = Catalog.query.get_or_404(catalog_id)
+    if catalog.configuration and len(catalog.configuration) > 0:
+        document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+        for config in document_version_configurations:
+            form.add_dynamic_fields(config, catalog.configuration[config])
 
     if form.validate_on_submit():
         try:
@@ -104,13 +373,20 @@ def add_url():
 
             file_content, filename, extension = process_url(url, tenant_id)
 
+            catalog_properties = {}
+            document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+            for config in document_version_configurations:
+                catalog_properties[config] = form.get_dynamic_data(config)
+
             api_input = {
+                'catalog_id': catalog_id,
                 'name': form.name.data or filename,
                 'url': url,
                 'language': form.language.data,
                 'user_context': form.user_context.data,
                 'valid_from': form.valid_from.data,
                 'user_metadata': json.loads(form.user_metadata.data) if form.user_metadata.data else None,
+                'catalog_properties': catalog_properties,
             }
 
             new_doc, new_doc_vers = create_document_stack(api_input, file_content, filename, extension, tenant_id)
@@ -131,62 +407,26 @@ def add_url():
     return render_template('document/add_url.html', form=form)
 
 
-@document_bp.route('/add_urls', methods=['GET', 'POST'])
-@roles_accepted('Super User', 'Tenant Admin')
-def add_urls():
-    form = AddURLsForm()
-
-    if form.validate_on_submit():
-        try:
-            tenant_id = session['tenant']['id']
-            urls = form.urls.data.split('\n')
-            urls = [url.strip() for url in urls if url.strip()]
-
-            api_input = {
-                'name': form.name.data,
-                'language': form.language.data,
-                'user_context': form.user_context.data,
-                'valid_from': form.valid_from.data
-            }
-
-            results = process_multiple_urls(urls, tenant_id, api_input)
-
-            for result in results:
-                if result['status'] == 'success':
-                    flash(
-                        f"Processed URL: {result['url']} - Document ID: {result['document_id']}, Version ID: {result['document_version_id']}",
-                        'success')
-                else:
-                    flash(f"Error processing URL: {result['url']} - {result['message']}", 'error')
-
-            return redirect(prefixed_url_for('document_bp.documents'))
-
-        except Exception as e:
-            current_app.logger.error(f'Error adding multiple URLs: {str(e)}')
-            flash('An error occurred while adding the URLs.', 'error')
-
-    return render_template('document/add_urls.html', form=form)
-
-
 @document_bp.route('/documents', methods=['GET', 'POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def documents():
-    page = request.args.get('page', 1, type=int)
-    per_page = request.args.get('per_page', 10, type=int)
-
-    pagination = get_documents_list(page, per_page)
-    docs = pagination.items
-
-    rows = prepare_table_for_macro(docs, [('id', ''), ('name', ''), ('valid_from', ''), ('valid_to', '')])
-
-    return render_template('document/documents.html', rows=rows, pagination=pagination)
+    view = DocumentListView(Document, 'document/documents.html', per_page=10)
+    return view.get()
 
 
 @document_bp.route('/handle_document_selection', methods=['POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def handle_document_selection():
     document_identification = request.form['selected_row']
-    doc_id = ast.literal_eval(document_identification).get('value')
+    if isinstance(document_identification, int) or document_identification.isdigit():
+        doc_id = int(document_identification)
+    else:
+        # If it's not an integer, assume it's a string representation of a dictionary
+        try:
+            doc_id = ast.literal_eval(document_identification).get('value')
+        except (ValueError, AttributeError):
+            flash('Invalid document selection.', 'error')
+            return redirect(prefixed_url_for('document_bp.documents'))
 
     action = request.form['action']
 
@@ -208,9 +448,25 @@ def handle_document_selection():
 @document_bp.route('/edit_document/<int:document_id>', methods=['GET', 'POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def edit_document_view(document_id):
-    doc = Document.query.get_or_404(document_id)
+    # Use an alias for the Catalog to avoid column name conflicts
+    CatalogAlias = aliased(Catalog)
+
+    # Query for the document and its catalog
+    result = db.session.query(Document, CatalogAlias.name.label('catalog_name')) \
+        .join(CatalogAlias, Document.catalog_id == CatalogAlias.id) \
+        .filter(Document.id == document_id) \
+        .first_or_404()
+
+    doc, catalog_name = result
+
     form = EditDocumentForm(obj=doc)
 
+    if request.method == 'GET':
+        # Populate form with current values
+        form.name.data = doc.name
+        form.valid_from.data = doc.valid_from
+        form.valid_to.data = doc.valid_to
+
     if form.validate_on_submit():
         updated_doc, error = edit_document(
             document_id,
@@ -226,19 +482,36 @@ def edit_document_view(document_id):
     else:
         form_validation_failed(request, form)
 
-    return render_template('document/edit_document.html', form=form, document_id=document_id)
+    return render_template('document/edit_document.html', form=form, document_id=document_id, catalog_name=catalog_name)
 
 
 @document_bp.route('/edit_document_version/<int:document_version_id>', methods=['GET', 'POST'])
 @roles_accepted('Super User', 'Tenant Admin')
 def edit_document_version_view(document_version_id):
     doc_vers = DocumentVersion.query.get_or_404(document_version_id)
-    form = EditDocumentVersionForm(obj=doc_vers)
+    form = EditDocumentVersionForm(request.form, obj=doc_vers)
+
+    catalog_id = session.get('catalog_id', None)
+    if catalog_id is None:
+        flash('You need to set a Session Catalog before adding Documents or URLs')
+        return redirect(prefixed_url_for('document_bp.catalogs'))
+
+    catalog = Catalog.query.get_or_404(catalog_id)
+    if catalog.configuration and len(catalog.configuration) > 0:
+        document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+        for config in document_version_configurations:
+            form.add_dynamic_fields(config, catalog.configuration[config], doc_vers.catalog_properties[config])
 
     if form.validate_on_submit():
+        catalog_properties = {}
+        document_version_configurations = CATALOG_TYPES[catalog.type]['document_version_configurations']
+        for config in document_version_configurations:
+            catalog_properties[config] = form.get_dynamic_data(config)
+
         updated_version, error = edit_document_version(
             document_version_id,
-            form.user_context.data
+            form.user_context.data,
+            catalog_properties,
         )
         if updated_version:
             flash(f'Document Version {updated_version.id} updated successfully', 'success')
@@ -280,7 +553,15 @@ def document_versions(document_id):
 @roles_accepted('Super User', 'Tenant Admin')
 def handle_document_version_selection():
     document_version_identification = request.form['selected_row']
-    doc_vers_id = ast.literal_eval(document_version_identification).get('value')
+    if isinstance(document_version_identification, int) or document_version_identification.isdigit():
+        doc_vers_id = int(document_version_identification)
+    else:
+        # If it's not an integer, assume it's a string representation of a dictionary
+        try:
+            doc_vers_id = ast.literal_eval(document_version_identification).get('value')
+        except (ValueError, AttributeError):
+            flash('Invalid document version selection.', 'error')
+            return redirect(prefixed_url_for('document_bp.document_versions_list'))
 
     action = request.form['action']
 
@@ -332,7 +613,7 @@ def refresh_all_documents():
 
 
 def refresh_document_view(document_id):
-    new_version, result = refresh_document(document_id)
+    new_version, result = refresh_document(document_id, session['tenant']['id'])
     if new_version:
         flash(f'Document refreshed. New version: {new_version.id}. Task ID: {result}', 'success')
     else:
@@ -395,49 +676,3 @@ def fetch_html(url):
 
     response.raise_for_status()  # Will raise an exception for bad requests
     return response.content
-
-
-def prepare_document_data(docs):
-    rows = []
-    for doc in docs:
-        doc_row = [{'value': doc.name, 'class': '', 'type': 'text'},
-                   {'value': doc.created_at.strftime("%Y-%m-%d %H:%M:%S"), 'class': '', 'type': 'text'}]
-        # Document basic details
-        if doc.valid_from:
-            doc_row.append({'value': doc.valid_from.strftime("%Y-%m-%d"), 'class': '', 'type': 'text'})
-        else:
-            doc_row.append({'value': '', 'class': '', 'type': 'text'})
-
-        # Nested languages and versions
-        languages_rows = []
-        for lang in doc.languages:
-            lang_row = [{'value': lang.language, 'class': '', 'type': 'text'}]
-
-            # Latest version details if available (should be available ;-) )
-            if lang.latest_version:
-                lang_row.append({'value': lang.latest_version.created_at.strftime("%Y-%m-%d %H:%M:%S"),
-                                 'class': '', 'type': 'text'})
-                if lang.latest_version.url:
-                    lang_row.append({'value': lang.latest_version.url,
-                                     'class': '', 'type': 'link', 'href': lang.latest_version.url})
-                else:
-                    lang_row.append({'value': '', 'class': '', 'type': 'text'})
-
-                if lang.latest_version.file_name:
-                    lang_row.append({'value': lang.latest_version.file_name, 'class': '', 'type': 'text'})
-                else:
-                    lang_row.append({'value': '', 'class': '', 'type': 'text'})
-
-                if lang.latest_version.file_type:
-                    lang_row.append({'value': lang.latest_version.file_type, 'class': '', 'type': 'text'})
-                else:
-                    lang_row.append({'value': '', 'class': '', 'type': 'text'})
-                    # Include other details as necessary
-
-            languages_rows.append(lang_row)
-
-        doc_row.append({'is_group': True, 'colspan': '5',
-                        'headers': ['Language', 'Latest Version', 'URL', 'File Name', 'Type'],
-                        'sub_rows': languages_rows})
-        rows.append(doc_row)
-    return rows

eveai_app/views/dynamic_form_base.py

@@ -0,0 +1,216 @@
+from flask_wtf import FlaskForm
+from wtforms import IntegerField, FloatField, BooleanField, StringField, TextAreaField, validators, ValidationError
+from flask import current_app
+import json
+
+from wtforms.fields.choices import SelectField
+from wtforms.fields.datetime import DateField
+
+
+class DynamicFormBase(FlaskForm):
+    def __init__(self, formdata=None, *args, **kwargs):
+        super(DynamicFormBase, self).__init__(*args, **kwargs)
+        # Maps collection names to lists of field names
+        self.dynamic_fields = {}
+        # Store formdata for later use
+        self.formdata = formdata
+
+    def _create_field_validators(self, field_def):
+        """Create validators based on field definition"""
+        validators_list = []
+
+        # Required validator
+        if field_def.get('required', False):
+            validators_list.append(validators.InputRequired())
+        else:
+            validators_list.append(validators.Optional())
+
+        # Type-specific validators
+        field_type = field_def.get('type')
+        if field_type in ['integer', 'float']:
+            min_value = field_def.get('min_value')
+            max_value = field_def.get('max_value')
+            if min_value is not None or max_value is not None:
+                validators_list.append(
+                    validators.NumberRange(
+                        min=min_value if min_value is not None else -float('inf'),
+                        max=max_value if max_value is not None else float('inf'),
+                        message=f"Value must be between {min_value or '-∞'} and {max_value or '∞'}"
+                    )
+                )
+
+        return validators_list
+
+    def add_dynamic_fields(self, collection_name, config, initial_data=None):
+        """Add dynamic fields to the form based on the configuration."""
+        self.dynamic_fields[collection_name] = []
+        for field_name, field_def in config.items():
+            current_app.logger.debug(f"{field_name}: {field_def}")
+            # Prefix the field name with the collection name
+            full_field_name = f"{collection_name}_{field_name}"
+            label = field_def.get('name', field_name)
+            field_type = field_def.get('type')
+            description = field_def.get('description', '')
+            default = field_def.get('default')
+
+            # Determine standard validators
+            field_validators = self._create_field_validators(field_def)
+
+            # Handle special case for tagging_fields
+            if field_type == 'tagging_fields':
+                field_class = TextAreaField
+                field_validators.append(validate_tagging_fields)
+                extra_classes = 'json-editor'
+                field_kwargs = {}
+            elif field_type == 'enum':
+                field_class = SelectField
+                allowed_values = field_def.get('allowed_values', [])
+                choices = [(str(val), str(val)) for val in allowed_values]
+                extra_classes = ''
+                field_kwargs = {'choices': choices}
+            else:
+                extra_classes = ''
+                field_class = {
+                    'integer': IntegerField,
+                    'float': FloatField,
+                    'boolean': BooleanField,
+                    'string': StringField,
+                    'date': DateField,
+                }.get(field_type, StringField)
+                field_kwargs = {}
+
+            # Prepare field data
+            field_data = None
+            if initial_data and field_name in initial_data:
+                field_data = initial_data[field_name]
+                if field_type == 'tagging_fields' and isinstance(field_data, dict):
+                    try:
+                        field_data = json.dumps(field_data, indent=2)
+                    except (TypeError, ValueError) as e:
+                        current_app.logger.error(f"Error converting initial data to JSON: {e}")
+                        field_data = "{}"
+            elif default is not None:
+                field_data = default
+
+            # Create render_kw with classes and any other HTML attributes
+            render_kw = {'class': extra_classes} if extra_classes else {}
+            if description:
+                render_kw['title'] = description  # For tooltip
+                render_kw['data-bs-toggle'] = 'tooltip'
+                render_kw['data-bs-placement'] = 'right'
+
+            # Create the field
+            field_kwargs.update({
+                'label': label,
+                'description': description,
+                'validators': field_validators,
+                'default': field_data,
+                'render_kw': render_kw
+            })
+
+            unbound_field = field_class(**field_kwargs)
+
+            # Bind the field to the form
+            bound_field = unbound_field.bind(form=self, name=full_field_name)
+
+            # Process the field with formdata
+            if self.formdata and full_field_name in self.formdata:
+                bound_field.process(self.formdata)
+            else:
+                bound_field.process(formdata=None, data=field_data)  # Use prepared field_data
+
+            # Add the field to the form
+            setattr(self, full_field_name, bound_field)
+            self._fields[full_field_name] = bound_field
+            self.dynamic_fields[collection_name].append(full_field_name)
+
+    def get_static_fields(self):
+        """Return a list of static field instances."""
+        # Get names of dynamic fields
+        dynamic_field_names = set()
+        for field_list in self.dynamic_fields.values():
+            dynamic_field_names.update(field_list)
+
+        # Return all fields that are not dynamic
+        return [field for name, field in self._fields.items() if name not in dynamic_field_names]
+
+    def get_dynamic_fields(self):
+        """Return a dictionary of dynamic fields per collection."""
+        result = {}
+        for collection_name, field_names in self.dynamic_fields.items():
+            result[collection_name] = [getattr(self, name) for name in field_names]
+        return result
+
+    def get_dynamic_data(self, collection_name):
+        """Retrieve the data from dynamic fields of a specific collection."""
+        data = {}
+        current_app.logger.debug(f"{collection_name} in {self.dynamic_fields}?")
+        if collection_name not in self.dynamic_fields:
+            return data
+        prefix_length = len(collection_name) + 1  # +1 for the underscore
+        for full_field_name in self.dynamic_fields[collection_name]:
+            current_app.logger.debug(f"{full_field_name}: {full_field_name}")
+            original_field_name = full_field_name[prefix_length:]
+            current_app.logger.debug(f"{original_field_name}: {original_field_name}")
+            field = getattr(self, full_field_name)
+            current_app.logger.debug(f"{field}: {field}")
+            # Parse JSON for tagging_fields type
+            if isinstance(field, TextAreaField) and field.data:
+                try:
+                    data[original_field_name] = json.loads(field.data)
+                except json.JSONDecodeError:
+                    # Validation should catch this, but just in case
+                    data[original_field_name] = field.data
+            else:
+                data[original_field_name] = field.data
+        return data
+
+
+def validate_tagging_fields(form, field):
+    """Validate the tagging fields structure"""
+    if not field.data:
+        return
+
+    try:
+        # Parse JSON data
+        fields_data = json.loads(field.data)
+
+        # Validate it's a dictionary
+        if not isinstance(fields_data, dict):
+            raise ValidationError("Tagging fields must be a dictionary")
+
+        # Validate each field definition
+        for field_name, field_def in fields_data.items():
+            if not isinstance(field_def, dict):
+                raise ValidationError(f"Field definition for {field_name} must be a dictionary")
+
+            # Check required properties
+            if 'type' not in field_def:
+                raise ValidationError(f"Field {field_name} missing required 'type' property")
+
+            # Validate type
+            if field_def['type'] not in ['string', 'integer', 'float', 'date', 'enum']:
+                raise ValidationError(f"Field {field_name} has invalid type: {field_def['type']}")
+
+            # Validate enum fields have allowed_values
+            if field_def['type'] == 'enum':
+                if 'allowed_values' not in field_def:
+                    raise ValidationError(f"Enum field {field_name} missing required 'allowed_values' list")
+                if not isinstance(field_def['allowed_values'], list):
+                    raise ValidationError(f"Field {field_name} allowed_values must be a list")
+
+            # Validate numeric fields
+            if field_def['type'] in ['integer', 'float']:
+                if 'min_value' in field_def and 'max_value' in field_def:
+                    min_val = float(field_def['min_value'])
+                    max_val = float(field_def['max_value'])
+                    if min_val >= max_val:
+                        raise ValidationError(f"Field {field_name} min_value must be less than max_value")
+
+    except json.JSONDecodeError:
+        raise ValidationError("Invalid JSON format")
+    except (TypeError, ValueError) as e:
+        raise ValidationError(f"Invalid field definition: {str(e)}")
+
+
+

eveai_app/views/interaction_views.py

@@ -93,17 +93,17 @@ def view_chat_session(chat_session_id):
     # Fetch all related embeddings for the interactions in this session
     embedding_query = (db.session.query(InteractionEmbedding.interaction_id,
                                         DocumentVersion.url,
-                                        DocumentVersion.file_name)
+                                        DocumentVersion.object_name)
                        .join(Embedding, InteractionEmbedding.embedding_id == Embedding.id)
                        .join(DocumentVersion, Embedding.doc_vers_id == DocumentVersion.id)
                        .filter(InteractionEmbedding.interaction_id.in_([i.id for i in interactions])))
 
     # Create a dictionary to store embeddings for each interaction
     embeddings_dict = {}
-    for interaction_id, url, file_name in embedding_query:
+    for interaction_id, url, object_name in embedding_query:
         if interaction_id not in embeddings_dict:
             embeddings_dict[interaction_id] = []
-        embeddings_dict[interaction_id].append({'url': url, 'file_name': file_name})
+        embeddings_dict[interaction_id].append({'url': url, 'object_name': object_name})
 
     return render_template('interaction/view_chat_session.html',
                            chat_session=chat_session,

eveai_app/views/user_views.py

@@ -48,22 +48,6 @@ def tenant():
         new_tenant = Tenant()
         form.populate_obj(new_tenant)
 
-        # Handle Embedding Variables
-        new_tenant.html_tags = [tag.strip() for tag in form.html_tags.data.split(',')] if form.html_tags.data else []
-        new_tenant.html_end_tags = [tag.strip() for tag in form.html_end_tags.data.split(',')] \
-            if form.html_end_tags.data else []
-        new_tenant.html_included_elements = [tag.strip() for tag in form.html_included_elements.data.split(',')] \
-            if form.html_included_elements.data else []
-        new_tenant.html_excluded_elements = [tag.strip() for tag in form.html_excluded_elements.data.split(',')] \
-            if form.html_excluded_elements.data else []
-        new_tenant.html_excluded_classes = [cls.strip() for cls in form.html_excluded_classes.data.split(',')] \
-            if form.html_excluded_classes.data else []
-
-        current_app.logger.debug(f'html_tags: {new_tenant.html_tags},'
-                                 f'html_end_tags: {new_tenant.html_end_tags},'
-                                 f'html_included_elements: {new_tenant.html_included_elements},'
-                                 f'html_excluded_elements: {new_tenant.html_excluded_elements}')
-
         # Handle Timestamps
         timestamp = dt.now(tz.utc)
         new_tenant.created_at = timestamp
@@ -105,30 +89,11 @@ def edit_tenant(tenant_id):
     if request.method == 'GET':
         # Populate the form with tenant data
         form.populate_obj(tenant)
-        if tenant.html_tags:
-            form.html_tags.data = ', '.join(tenant.html_tags)
-        if tenant.html_end_tags:
-            form.html_end_tags.data = ', '.join(tenant.html_end_tags)
-        if tenant.html_included_elements:
-            form.html_included_elements.data = ', '.join(tenant.html_included_elements)
-        if tenant.html_excluded_elements:
-            form.html_excluded_elements.data = ', '.join(tenant.html_excluded_elements)
-        if tenant.html_excluded_classes:
-            form.html_excluded_classes.data = ', '.join(tenant.html_excluded_classes)
 
     if form.validate_on_submit():
         current_app.logger.debug(f'Updating tenant {tenant_id}')
         # Populate the tenant with form data
         form.populate_obj(tenant)
-        # Then handle the special fields manually
-        tenant.html_tags = [tag.strip() for tag in form.html_tags.data.split(',') if tag.strip()]
-        tenant.html_end_tags = [tag.strip() for tag in form.html_end_tags.data.split(',') if tag.strip()]
-        tenant.html_included_elements = [elem.strip() for elem in form.html_included_elements.data.split(',') if
-                                         elem.strip()]
-        tenant.html_excluded_elements = [elem.strip() for elem in form.html_excluded_elements.data.split(',') if
-                                         elem.strip()]
-        tenant.html_excluded_classes = [elem.strip() for elem in form.html_excluded_classes.data.split(',') if
-                                        elem.strip()]
 
         db.session.commit()
         flash('Tenant updated successfully.', 'success')
@@ -266,10 +231,16 @@ def handle_tenant_selection():
     tenant_identification = request.form['selected_row']
     tenant_id = ast.literal_eval(tenant_identification).get('value')
     the_tenant = Tenant.query.get(tenant_id)
+
+    # set tenant information in the session
     session['tenant'] = the_tenant.to_dict()
     session['default_language'] = the_tenant.default_language
     session['embedding_model'] = the_tenant.embedding_model
     session['llm_model'] = the_tenant.llm_model
+    # remove catalog-related items from the session
+    session.pop('catalog_id', None)
+    session.pop('catalog_name', None)
+
     action = request.form['action']
 
     match action:

eveai_chat_workers/tasks.py

@@ -1,29 +1,22 @@
 from datetime import datetime as dt, timezone as tz
-from flask import current_app, session
+from flask import current_app
 from langchain_core.output_parsers import StrOutputParser
 from langchain_core.runnables import RunnableParallel, RunnablePassthrough
-from langchain.globals import set_debug
 from sqlalchemy.exc import SQLAlchemyError
-from celery import states
-from celery.exceptions import Ignore
-import os
 
 # OpenAI imports
-from langchain_openai import OpenAIEmbeddings, ChatOpenAI
 from langchain_core.prompts import ChatPromptTemplate
-from langchain.chains.summarize import load_summarize_chain
-from langchain.text_splitter import CharacterTextSplitter
 from langchain_core.exceptions import LangChainException
 
 from common.utils.database import Database
-from common.models.document import DocumentVersion, EmbeddingMistral, EmbeddingSmallOpenAI, Embedding
+from common.models.document import Embedding
 from common.models.user import Tenant
 from common.models.interaction import ChatSession, Interaction, InteractionEmbedding
 from common.extensions import db
 from common.utils.celery_utils import current_celery
 from common.utils.model_utils import select_model_variables, create_language_template, replace_variable_in_template
-from common.langchain.eveai_retriever import EveAIRetriever
-from common.langchain.eveai_history_retriever import EveAIHistoryRetriever
+from common.langchain.retrievers.eveai_default_rag_retriever import EveAIDefaultRagRetriever
+from common.langchain.retrievers.eveai_history_retriever import EveAIHistoryRetriever
 from common.utils.business_event import BusinessEvent
 from common.utils.business_event_context import current_event
 
@@ -36,7 +29,7 @@ def ping():
 
 def detail_question(question, language, model_variables, session_id):
     current_app.logger.debug(f'Detail question: {question}')
-    current_app.logger.debug(f'model_varialbes: {model_variables}')
+    current_app.logger.debug(f'model_variables: {model_variables}')
     current_app.logger.debug(f'session_id: {session_id}')
     retriever = EveAIHistoryRetriever(model_variables=model_variables, session_id=session_id)
     llm = model_variables['llm']
@@ -93,12 +86,6 @@ def ask_question(tenant_id, question, language, session_id, user_timezone, room)
                     current_app.logger.error(f'ask_question: Error initializing chat session in database: {e}')
                     raise
 
-            if tenant.rag_tuning:
-                current_app.rag_tuning_logger.debug(f'Received question for tenant {tenant_id}:\n{question}. Processing...')
-                current_app.rag_tuning_logger.debug(f'Tenant Information: \n{tenant.to_dict()}')
-                current_app.rag_tuning_logger.debug(f'===================================================================')
-                current_app.rag_tuning_logger.debug(f'===================================================================')
-
             with current_event.create_span("RAG Answer"):
                 result, interaction = answer_using_tenant_rag(question, language, tenant, chat_session)
                 result['algorithm'] = current_app.config['INTERACTION_ALGORITHMS']['RAG_TENANT']['name']
@@ -138,22 +125,21 @@ def answer_using_tenant_rag(question, language, tenant, chat_session):
 
     with current_event.create_span("Detail Question"):
         detailed_question = detail_question(question, language, model_variables, chat_session.session_id)
-        current_app.logger.debug(f'Original question:\n {question}\n\nDetailed question: {detailed_question}')
-        if tenant.rag_tuning:
+        if model_variables['rag_tuning']:
             current_app.rag_tuning_logger.debug(f'Detailed Question for tenant {tenant.id}:\n{question}.')
             current_app.rag_tuning_logger.debug(f'-------------------------------------------------------------------')
         new_interaction.detailed_question = detailed_question
         new_interaction.detailed_question_at = dt.now(tz.utc)
 
     with current_event.create_span("Generate Answer using RAG"):
-        retriever = EveAIRetriever(model_variables, tenant_info)
+        retriever = EveAIDefaultRagRetriever(model_variables, tenant_info)
         llm = model_variables['llm']
         template = model_variables['rag_template']
         language_template = create_language_template(template, language)
         full_template = replace_variable_in_template(language_template, "{tenant_context}", model_variables['rag_context'])
         rag_prompt = ChatPromptTemplate.from_template(full_template)
         setup_and_retrieval = RunnableParallel({"context": retriever, "question": RunnablePassthrough()})
-        if tenant.rag_tuning:
+        if model_variables['rag_tuning']:
             current_app.rag_tuning_logger.debug(f'Full prompt for tenant {tenant.id}:\n{full_template}.')
             current_app.rag_tuning_logger.debug(f'-------------------------------------------------------------------')
 
@@ -181,7 +167,7 @@ def answer_using_tenant_rag(question, language, tenant, chat_session):
             current_app.logger.debug(f'ask_question: result answer: {result['answer']}')
             current_app.logger.debug(f'ask_question: result citations: {result["citations"]}')
             current_app.logger.debug(f'ask_question: insufficient information: {result["insufficient_info"]}')
-            if tenant.rag_tuning:
+            if model_variables['rag_tuning']:
                 current_app.rag_tuning_logger.debug(f'ask_question: result answer: {result['answer']}')
                 current_app.rag_tuning_logger.debug(f'ask_question: result citations: {result["citations"]}')
                 current_app.rag_tuning_logger.debug(f'ask_question: insufficient information: {result["insufficient_info"]}')
@@ -197,7 +183,7 @@ def answer_using_tenant_rag(question, language, tenant, chat_session):
             )
             existing_embedding_ids = [emb.id for emb in embeddings]
             urls = list(set(emb.document_version.url for emb in embeddings))
-            if tenant.rag_tuning:
+            if model_variables['rag_tuning']:
                 current_app.rag_tuning_logger.debug(f'Referenced documents for answer for tenant {tenant.id}:\n')
                 current_app.rag_tuning_logger.debug(f'{urls}')
                 current_app.rag_tuning_logger.debug(f'-------------------------------------------------------------------')
@@ -250,7 +236,7 @@ def answer_using_llm(question, language, tenant, chat_session):
         new_interaction.detailed_question_at = dt.now(tz.utc)
 
     with current_event.create_span("Detail Answer using LLM"):
-        retriever = EveAIRetriever(model_variables, tenant_info)
+        retriever = EveAIDefaultRagRetriever(model_variables, tenant_info)
         llm = model_variables['llm_no_rag']
         template = model_variables['encyclopedia_template']
         language_template = create_language_template(template, language)

eveai_workers/Processors/html_processor.py

@@ -15,6 +15,7 @@ class HTMLProcessor(Processor):
         self.html_end_tags = model_variables['html_end_tags']
         self.html_included_elements = model_variables['html_included_elements']
         self.html_excluded_elements = model_variables['html_excluded_elements']
+        self.html_excluded_classes = model_variables['html_excluded_classes']
         self.chunk_size = model_variables['processing_chunk_size']  # Adjust this based on your LLM's optimal input size
         self.chunk_overlap = model_variables[
             'processing_chunk_overlap']  # Adjust for context preservation between chunks
@@ -45,7 +46,7 @@ class HTMLProcessor(Processor):
         self._log(f'Parsing HTML for tenant {self.tenant.id}')
         soup = BeautifulSoup(html_content, 'html.parser')
         extracted_html = ''
-        excluded_classes = self._parse_excluded_classes(self.tenant.html_excluded_classes)
+        excluded_classes = self._parse_excluded_classes(self.html_excluded_classes)
 
         if self.html_included_elements:
             elements_to_parse = soup.find_all(self.html_included_elements)

eveai_workers/tasks.py

@@ -13,7 +13,7 @@ from langchain_core.runnables import RunnablePassthrough
 from sqlalchemy.exc import SQLAlchemyError
 
 from common.extensions import db, minio_client
-from common.models.document import DocumentVersion, Embedding
+from common.models.document import DocumentVersion, Embedding, Document
 from common.models.user import Tenant
 from common.utils.celery_utils import current_celery
 from common.utils.database import Database
@@ -50,8 +50,12 @@ def create_embeddings(tenant_id, document_version_id):
         if document_version is None:
             raise Exception(f'Document version {document_version_id} not found')
 
+        # Retrieve the Catalog ID
+        doc = Document.query.get_or_404(document_version.doc_id)
+        catalog_id = doc.catalog_id
+
         # Select variables to work with depending on tenant and model
-        model_variables = select_model_variables(tenant)
+        model_variables = select_model_variables(tenant, catalog_id=catalog_id)
         current_app.logger.debug(f'Model variables: {model_variables}')
 
     except Exception as e:

integrations/Wordpress/eveai_sync/README.md

@@ -51,6 +51,10 @@ No additional configuration is needed; the plugin will automatically detect the
 
 ## Versions
 
+### 1.1.1 - Add Reinitialisation functionality
+
+### 1.1.0 - Add Catalog Functionality
+
 ### 1.0.x - Bugfixing Releases
 
 ### 1.0.0 - Initial Release

integrations/Wordpress/eveai_sync/eveai_sync.php

@@ -3,7 +3,7 @@
  * Plugin Name: EveAI Sync
  * Plugin URI: https://askeveai.com/
  * Description: Synchronizes WordPress content with EveAI API.
- * Version: 1.0.16
+ * Version: 1.1.1
  * Author: Josako, Pieter Laroy
  * Author URI: https://askeveai.com/about/
  * License: GPL v2 or later
@@ -17,7 +17,7 @@ if (!defined('ABSPATH')) {
 }
 
 // Define plugin constants
-define('EVEAI_SYNC_VERSION', '1.0.0');
+define('EVEAI_SYNC_VERSION', '1.1.1');
 define('EVEAI_SYNC_PLUGIN_DIR', plugin_dir_path(__FILE__));
 define('EVEAI_SYNC_PLUGIN_URL', plugin_dir_url(__FILE__));
 
@@ -50,6 +50,30 @@ function eveai_delete_post_meta($post_id) {
 }
 add_action('before_delete_post', 'eveai_delete_post_meta');
 
+// Clean metadata from Wordpress site
+function eveai_reinitialize_site() {
+    check_ajax_referer('eveai_reinitialize_site', 'nonce');
+
+    if (!current_user_can('manage_options')) {
+        wp_send_json_error('You do not have permission to perform this action.');
+        return;
+    }
+
+    global $wpdb;
+
+    // Remove all EveAI-related post meta
+    $wpdb->query("DELETE FROM $wpdb->postmeta WHERE meta_key LIKE '_eveai_%'");
+
+    // Remove all EveAI-related options
+    delete_option('eveai_last_sync_time');
+    delete_option('eveai_sync_status');
+
+    // Optionally, you might want to clear any custom tables if you have any
+
+    wp_send_json_success('Site reinitialized. All EveAI metadata has been removed.');
+}
+add_action('wp_ajax_eveai_reinitialize_site', 'eveai_reinitialize_site');
+
 // Display sync info in post
 function eveai_display_sync_info($post) {
     $document_id = get_post_meta($post->ID, '_eveai_document_id', true);

integrations/Wordpress/eveai_sync/includes/class-eveai-admin.php

@@ -16,6 +16,7 @@ class EveAI_Admin {
         register_setting('eveai_settings', 'eveai_excluded_categories');
         register_setting('eveai_settings', 'eveai_access_token');
         register_setting('eveai_settings', 'eveai_token_expiry');
+        register_setting('eveai_settings', 'eveai_catalog_id');
     }
 
     public function add_admin_menu() {
@@ -50,6 +51,10 @@ class EveAI_Admin {
                         <th scope="row">API Key</th>
                         <td><input type="text" name="eveai_api_key" value="<?php echo esc_attr(get_option('eveai_api_key')); ?>" style="width: 100%;" /></td>
                     </tr>
+                    <tr valign="top">
+                        <th scope="row">Catalog ID</th>
+                        <td><input type="text" name="eveai_catalog_id" value="<?php echo esc_attr(get_option('eveai_catalog_id')); ?>" style="width: 100%;" /></td>
+                    </tr>
                     <tr valign="top">
                         <th scope="row">Default Language</th>
                         <td><input type="text" name="eveai_default_language" value="<?php echo esc_attr(get_option('eveai_default_language', 'en')); ?>" style="width: 100%;" /></td>
@@ -71,6 +76,11 @@ class EveAI_Admin {
                 <?php wp_nonce_field('eveai_bulk_sync', 'eveai_bulk_sync_nonce'); ?>
                 <input type="submit" name="eveai_bulk_sync" class="button button-primary" value="Start Bulk Sync">
             </form>
+
+            <h2>Reinitialize Site</h2>
+            <p>Click the button below to remove all EveAI metadata from your site. This will reset the sync status for all posts and pages.</p>
+            <button id="eveai-reinitialize" class="button button-secondary">Reinitialize Site</button>
+
             <div id="eveai-sync-results" style="margin-top: 20px;"></div>
         </div>
         <script>
@@ -105,6 +115,29 @@ class EveAI_Admin {
                     });
                 }
             });
+            $('#eveai-reinitialize').on('click', function(e) {
+                e.preventDefault();
+                if (confirm('Are you sure you want to reinitialize? This will remove all EveAI metadata from your site.')) {
+                    $.ajax({
+                        url: ajaxurl,
+                        type: 'POST',
+                        data: {
+                            action: 'eveai_reinitialize_site',
+                            nonce: '<?php echo wp_create_nonce('eveai_reinitialize_site'); ?>'
+                        },
+                        success: function(response) {
+                            if (response.success) {
+                                alert(response.data);
+                            } else {
+                                alert('Error: ' + response.data);
+                            }
+                        },
+                        error: function() {
+                            alert('An error occurred. Please try again.');
+                        }
+                    });
+                }
+            });
         });
         </script>
         <?php

integrations/Wordpress/eveai_sync/includes/class-eveai-api.php

@@ -13,6 +13,7 @@ class EveAI_API {
         $this->api_key = get_option('eveai_api_key');
         $this->access_token = get_option('eveai_access_token');
         $this->token_expiry = get_option('eveai_token_expiry', 0);
+        $this->catalog_id = get_option('eveai_catalog_id');
     }
 
     private function ensure_valid_token() {
@@ -111,6 +112,7 @@ class EveAI_API {
     }
 
     public function add_url($data) {
+        $data['catalog_id'] = get_option('eveai_catalog_id');  // Include catalog_id
         return $this->make_request('POST', '/api/v1/documents/add_url', $data);
     }
 

migrations/tenant/env.py

@@ -1,5 +1,7 @@
 import inspect
 import logging
+import sys
+import os
 from logging.config import fileConfig
 
 from flask import current_app
@@ -19,8 +21,26 @@ config = context.config
 
 # Interpret the config file for Python logging.
 # This line sets up loggers basically.
-fileConfig(config.config_file_name)
-logger = logging.getLogger('alembic.env')
+# fileConfig(config.config_file_name)
+# logger = logging.getLogger('alembic.env')
+
+logging.basicConfig(
+    stream=sys.stdout,
+    level=logging.INFO,
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
+)
+
+logger = logging.getLogger()
+
+# Reset handlers to avoid issues with Alembic overriding them
+for handler in logger.handlers:
+    logger.removeHandler(handler)
+
+# Add a stream handler to output logs to the console
+console_handler = logging.StreamHandler(sys.stdout)
+console_handler.setFormatter(logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s'))
+logger.addHandler(console_handler)
+logger.setLevel(logging.INFO)
 
 
 def get_engine():
@@ -125,6 +145,7 @@ def run_migrations_online():
         tenants = get_tenant_ids()
         for tenant in tenants:
             try:
+                os.environ['TENANT_ID'] = str(tenant)
                 logger.info(f"Migrating tenant: {tenant}")
                 # set search path on the connection, which ensures that
                 # PostgreSQL will emit all CREATE / ALTER / DROP statements

migrations/tenant/versions/0f5932ff3051_adding_catalogs_to_document_domain.py

@@ -0,0 +1,56 @@
+"""Adding Catalogs to Document Domain
+
+Revision ID: 0f5932ff3051
+Revises: 5a75fb6da7b8
+Create Date: 2024-10-14 15:20:45.058329
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = '0f5932ff3051'
+down_revision = '5a75fb6da7b8'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table('catalog',
+                    sa.Column('id', sa.Integer(), nullable=False),
+                    sa.Column('name', sa.String(length=50), nullable=False),
+                    sa.Column('description', sa.Text(), nullable=True),
+                    sa.Column('html_tags', postgresql.ARRAY(sa.String(length=10)), nullable=True),
+                    sa.Column('html_end_tags', postgresql.ARRAY(sa.String(length=10)), nullable=True),
+                    sa.Column('html_included_elements', postgresql.ARRAY(sa.String(length=50)), nullable=True),
+                    sa.Column('html_excluded_elements', postgresql.ARRAY(sa.String(length=50)), nullable=True),
+                    sa.Column('html_excluded_classes', postgresql.ARRAY(sa.String(length=200)), nullable=True),
+                    sa.Column('min_chunk_size', sa.Integer(), nullable=True),
+                    sa.Column('max_chunk_size', sa.Integer(), nullable=True),
+                    sa.Column('es_k', sa.Integer(), nullable=True),
+                    sa.Column('es_similarity_threshold', sa.Float(), nullable=True),
+                    sa.Column('chat_RAG_temperature', sa.Float(), nullable=True),
+                    sa.Column('chat_no_RAG_temperature', sa.Float(), nullable=True),
+                    sa.Column('embed_tuning', sa.Boolean(), nullable=True),
+                    sa.Column('rag_tuning', sa.Boolean(), nullable=True),
+                    sa.Column('created_at', sa.DateTime(), server_default=sa.text('now()'), nullable=False),
+                    sa.Column('created_by', sa.Integer(), nullable=True),
+                    sa.Column('updated_at', sa.DateTime(), server_default=sa.text('now()'), nullable=False),
+                    sa.Column('updated_by', sa.Integer(), nullable=True),
+                    sa.ForeignKeyConstraint(['created_by'], ['public.user.id'], ),
+                    sa.ForeignKeyConstraint(['updated_by'], ['public.user.id'], ),
+                    sa.PrimaryKeyConstraint('id')
+                    )
+    op.add_column('document', sa.Column('catalog_id', sa.Integer(), nullable=True))
+    op.create_foreign_key(None, 'document', 'catalog', ['catalog_id'], ['id'])
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column('document', 'catalog_id')
+    op.drop_table('catalog')
+    # ### end Alembic commands ###

migrations/tenant/versions/28984b05d396_upgrading_existing_documents_to_default_.py

@@ -0,0 +1,99 @@
+"""Upgrading Existing documents to default catalog
+
+Revision ID: 28984b05d396
+Revises: 0f5932ff3051
+Create Date: 2024-10-15 09:02:23.355660
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.orm import Session
+import pgvector
+from common.models.document import Catalog, Document
+from common.models.user import Tenant
+from flask import current_app
+import logging
+import os
+
+# Set up logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+# Create a console handler
+console_handler = logging.StreamHandler()
+console_handler.setLevel(logging.INFO)
+
+# Create a logging format
+formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
+console_handler.setFormatter(formatter)
+
+# Add the handler to the logger
+logger.addHandler(console_handler)
+
+logger.info("Starting revision upgrade 28984b05d396")
+
+# revision identifiers, used by Alembic.
+revision = '28984b05d396'
+down_revision = '0f5932ff3051'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    logger.info("Starting migration: Creating default catalog and updating documents.")
+
+    bind = op.get_bind()
+    session = Session(bind=bind)
+
+    try:
+        tenant_id = int(os.getenv('TENANT_ID'))
+        logger.info(f"In migration: tenant_id = {tenant_id}")
+
+        # Step 1: Create a new Default Catalog
+        logger.info("Creating default catalog")
+        default_catalog = Catalog(
+            name='Default Catalog',
+            description=None,
+        )
+        tenant = Tenant.query.get_or_404(tenant_id)
+        default_catalog.html_tags = tenant.html_tags
+        default_catalog.html_end_tags = tenant.html_end_tags
+        default_catalog.html_included_elements = tenant.html_included_elements
+        default_catalog.html_excluded_elements = tenant.html_excluded_elements
+        default_catalog.html_excluded_classes = tenant.html_excluded_classes
+        default_catalog.min_chunk_size = tenant.min_chunk_size
+        default_catalog.max_chunk_size = tenant.max_chunk_size
+        default_catalog.es_k = tenant.es_k
+        default_catalog.es_similarity_threshold = tenant.es_similarity_threshold
+        default_catalog.chat_RAG_temperature = tenant.chat_RAG_temperature
+        default_catalog.chat_no_RAG_temperature = tenant.chat_no_RAG_temperature
+        default_catalog.embed_tuning = tenant.embed_tuning
+        default_catalog.rag_tuning = tenant.rag_tuning
+
+        session.add(default_catalog)
+        session.commit()
+
+        new_catalog_id = default_catalog.id
+        logger.info(f"Default catalog created with ID: {new_catalog_id}")
+
+        # Step 2: Update all documents to use this new catalog
+        logger.info("Updating documents with the new catalog ID.")
+        documents = session.query(Document).all()
+        for document in documents:
+            document.catalog_id = new_catalog_id
+        session.commit()
+
+        logger.info(f"Updated {len(documents)} documents with new catalog ID.")
+
+    except Exception as e:
+        logger.error(f"An error occurred during migration: {e}")
+        session.rollback()
+        raise
+    finally:
+        session.close()
+
+    logger.info("Migration completed successfully.")
+
+
+def downgrade():
+    pass

migrations/tenant/versions/322d3cf1f17b_documentversion_update_to_bucket_name_.py

@@ -26,55 +26,55 @@ def upgrade():
     op.add_column('document_version', sa.Column('object_name', sa.String(length=200), nullable=True))
     op.add_column('document_version', sa.Column('file_size', sa.Float(), nullable=True))
 
-    # ### Upgrade values for bucket_name, object_name and file_size to reflect minio reality ###
-    from common.models.document import DocumentVersion
-    from common.extensions import minio_client
-    from minio.error import S3Error
-
-    # Create a connection
-    connection = op.get_bind()
-    session = Session(bind=connection)
-
-    # Get the current schema name (which should be the tenant ID)
-    current_schema = connection.execute(text("SELECT current_schema()")).scalar()
-    tenant_id = int(current_schema)
-
-    doc_versions = session.query(DocumentVersion).all()
-    for doc_version in doc_versions:
-        try:
-            object_name = minio_client.generate_object_name(doc_version.doc_id,
-                                                            doc_version.language,
-                                                            doc_version.id,
-                                                            doc_version.file_name)
-            bucket_name = minio_client.generate_bucket_name(tenant_id)
-            doc_version.object_name = object_name
-            doc_version.bucket_name = bucket_name
-
-            try:
-                stat = minio_client.client.stat_object(
-                    bucket_name=bucket_name,
-                    object_name=object_name
-                )
-                doc_version.file_size = stat.size / 1048576
-                current_app.logger.info(f"Processed Upgrade for  DocumentVersion {doc_version.id} for Tenant {tenant_id}")
-            except S3Error as e:
-                if e.code == "NoSuchKey":
-                    current_app.logger.warning(
-                        f"Object {doc_version.object_name} not found in bucket {doc_version.bucket_name}. Skipping.")
-                    continue  # Move to the next item
-                else:
-                    raise e  # Handle other types of S3 errors
-        except Exception as e:
-            session.rollback()
-            current_app.logger.error(f"Couldn't process upgrade for DocumentVersion {doc_version.id} for "
-                                     f"Tenant {tenant_id}. Error: {str(e)}")
-
-    try:
-        session.commit()
-        current_app.logger.info(f"Successfully updated file sizes for tenant schema {current_schema}")
-    except Exception as e:
-        session.rollback()
-        current_app.logger.error(f"Error committing changes for tenant schema {current_schema}: {str(e)}")
+    # # ### Upgrade values for bucket_name, object_name and file_size to reflect minio reality ###
+    # from common.models.document import DocumentVersion
+    # from common.extensions import minio_client
+    # from minio.error import S3Error
+    #
+    # # Create a connection
+    # connection = op.get_bind()
+    # session = Session(bind=connection)
+    #
+    # # Get the current schema name (which should be the tenant ID)
+    # current_schema = connection.execute(text("SELECT current_schema()")).scalar()
+    # tenant_id = int(current_schema)
+    #
+    # doc_versions = session.query(DocumentVersion).all()
+    # for doc_version in doc_versions:
+    #     try:
+    #         object_name = minio_client.generate_object_name(doc_version.doc_id,
+    #                                                         doc_version.language,
+    #                                                         doc_version.id,
+    #                                                         doc_version.file_name)
+    #         bucket_name = minio_client.generate_bucket_name(tenant_id)
+    #         doc_version.object_name = object_name
+    #         doc_version.bucket_name = bucket_name
+    #
+    #         try:
+    #             stat = minio_client.client.stat_object(
+    #                 bucket_name=bucket_name,
+    #                 object_name=object_name
+    #             )
+    #             doc_version.file_size = stat.size / 1048576
+    #             current_app.logger.info(f"Processed Upgrade for  DocumentVersion {doc_version.id} for Tenant {tenant_id}")
+    #         except S3Error as e:
+    #             if e.code == "NoSuchKey":
+    #                 current_app.logger.warning(
+    #                     f"Object {doc_version.object_name} not found in bucket {doc_version.bucket_name}. Skipping.")
+    #                 continue  # Move to the next item
+    #             else:
+    #                 raise e  # Handle other types of S3 errors
+    #     except Exception as e:
+    #         session.rollback()
+    #         current_app.logger.error(f"Couldn't process upgrade for DocumentVersion {doc_version.id} for "
+    #                                  f"Tenant {tenant_id}. Error: {str(e)}")
+    #
+    # try:
+    #     session.commit()
+    #     current_app.logger.info(f"Successfully updated file sizes for tenant schema {current_schema}")
+    # except Exception as e:
+    #     session.rollback()
+    #     current_app.logger.error(f"Error committing changes for tenant schema {current_schema}: {str(e)}")
 
     # ### commands auto generated by Alembic - Remove old fields ###
     # op.drop_column('document_version', 'file_location')

migrations/tenant/versions/331f8100eb87_add_tuning_to_retriever.py

@@ -0,0 +1,29 @@
+"""Add tuning to Retriever
+
+Revision ID: 331f8100eb87
+Revises: 6c5ca750e60c
+Create Date: 2024-10-31 07:17:22.579376
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+
+
+# revision identifiers, used by Alembic.
+revision = '331f8100eb87'
+down_revision = '6c5ca750e60c'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('retriever', sa.Column('tuning', sa.Boolean(), nullable=True))
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column('retriever', 'tuning')
+    # ### end Alembic commands ###

migrations/tenant/versions/3717364e6429_add_retriever_model.py

@@ -0,0 +1,56 @@
+"""Add Retriever Model
+
+Revision ID: 3717364e6429
+Revises: 7b7b566e667f
+Create Date: 2024-10-21 14:22:30.258679
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = '3717364e6429'
+down_revision = '7b7b566e667f'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.create_table('retriever',
+    sa.Column('id', sa.Integer(), nullable=False),
+    sa.Column('name', sa.String(length=50), nullable=False),
+    sa.Column('description', sa.Text(), nullable=True),
+    sa.Column('catalog_id', sa.Integer(), nullable=True),
+    sa.Column('type', sa.String(length=50), nullable=False),
+    sa.Column('user_metadata', postgresql.JSONB(astext_type=sa.Text()), nullable=True),
+    sa.Column('system_metadata', postgresql.JSONB(astext_type=sa.Text()), nullable=True),
+    sa.Column('configuration', postgresql.JSONB(astext_type=sa.Text()), nullable=True),
+    sa.Column('created_at', sa.DateTime(), server_default=sa.text('now()'), nullable=False),
+    sa.Column('created_by', sa.Integer(), nullable=True),
+    sa.Column('updated_at', sa.DateTime(), server_default=sa.text('now()'), nullable=False),
+    sa.Column('updated_by', sa.Integer(), nullable=True),
+    sa.ForeignKeyConstraint(['catalog_id'], ['catalog.id'], ),
+    sa.ForeignKeyConstraint(['created_by'], ['public.user.id'], ),
+    sa.ForeignKeyConstraint(['updated_by'], ['public.user.id'], ),
+    sa.PrimaryKeyConstraint('id')
+    )
+    op.drop_column('catalog', 'es_similarity_threshold')
+    op.drop_column('catalog', 'es_k')
+    op.drop_column('catalog', 'rag_tuning')
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('catalog', sa.Column('rag_tuning', sa.BOOLEAN(), autoincrement=False, nullable=True))
+    op.add_column('catalog', sa.Column('es_k', sa.INTEGER(), autoincrement=False, nullable=True))
+    op.add_column('catalog', sa.Column('es_similarity_threshold', sa.DOUBLE_PRECISION(precision=53), autoincrement=False, nullable=True))
+    op.drop_constraint(None, 'catalog', type_='foreignkey')
+    op.drop_constraint(None, 'catalog', type_='foreignkey')
+    op.create_foreign_key('catalog_updated_by_fkey', 'catalog', 'user', ['updated_by'], ['id'])
+    op.create_foreign_key('catalog_created_by_fkey', 'catalog', 'user', ['created_by'], ['id'])
+    op.drop_table('retriever')
+    # ### end Alembic commands ###

migrations/tenant/versions/6c5ca750e60c_add_catalog_properties_to_.py

@@ -0,0 +1,29 @@
+"""Add catalog_properties to DocumentVersion
+
+Revision ID: 6c5ca750e60c
+Revises: b64d5cf32c7a
+Create Date: 2024-10-29 08:33:54.663211
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = '6c5ca750e60c'
+down_revision = 'b64d5cf32c7a'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('document_version', sa.Column('catalog_properties', postgresql.JSONB(astext_type=sa.Text()), nullable=True))
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column('document_version', 'catalog_properties')
+    # ### end Alembic commands ###

migrations/tenant/versions/7b7b566e667f_extensions_for_more_catalog_types_in_.py

@@ -0,0 +1,46 @@
+"""Extensions for more Catalog Types in Catalog Model
+
+Revision ID: 7b7b566e667f
+Revises: 28984b05d396
+Create Date: 2024-10-21 07:39:52.260054
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = '7b7b566e667f'
+down_revision = '28984b05d396'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('catalog', sa.Column('parent_id', sa.Integer(), nullable=True))
+
+    op.add_column('catalog', sa.Column('type', sa.String(length=50), nullable=False, server_default='DEFAULT'))
+    # Update all existing rows to have the 'type' set to 'DEFAULT'
+    op.execute("UPDATE catalog SET type='DEFAULT' WHERE type IS NULL")
+    # Remove the server default (optional but recommended)
+    op.alter_column('catalog', 'type', server_default=None)
+
+    op.add_column('catalog', sa.Column('user_metadata', postgresql.JSONB(astext_type=sa.Text()), nullable=True))
+    op.add_column('catalog', sa.Column('system_metadata', postgresql.JSONB(astext_type=sa.Text()), nullable=True))
+    op.add_column('catalog', sa.Column('configuration', postgresql.JSONB(astext_type=sa.Text()), nullable=True))
+    op.create_foreign_key(None, 'catalog', 'catalog', ['parent_id'], ['id'])
+
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_constraint(None, 'catalog', type_='foreignkey')
+    op.drop_column('catalog', 'configuration')
+    op.drop_column('catalog', 'system_metadata')
+    op.drop_column('catalog', 'user_metadata')
+    op.drop_column('catalog', 'type')
+    op.drop_column('catalog', 'parent_id')
+    # ### end Alembic commands ###

migrations/tenant/versions/b64d5cf32c7a_removing_parent_catalog_id_from_catalog_.py

@@ -0,0 +1,33 @@
+"""Removing Parent Catalog ID from Catalog ==> use tags in user_metadata instead
+
+Revision ID: b64d5cf32c7a
+Revises: 3717364e6429
+Create Date: 2024-10-28 07:48:04.624298
+
+"""
+from alembic import op
+import sqlalchemy as sa
+import pgvector
+
+
+# revision identifiers, used by Alembic.
+revision = 'b64d5cf32c7a'
+down_revision = '3717364e6429'
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_constraint('catalog_parent_id_fkey', 'catalog', type_='foreignkey')
+    op.drop_column('catalog', 'parent_id')
+    op.drop_constraint('chat_session_user_id_fkey', 'chat_session', type_='foreignkey')
+
+    # ### end Alembic commands ###
+
+
+def downgrade():
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column('catalog', sa.Column('parent_id', sa.INTEGER(), autoincrement=False, nullable=True))
+    op.create_foreign_key('catalog_parent_id_fkey', 'catalog', 'catalog', ['parent_id'], ['id'])
+    # ### end Alembic commands ###

nginx/nginx.conf

@@ -48,6 +48,9 @@ http {
 
         #access_log  logs/host.access.log  main;
 
+        location = / {
+            return 301 /admin/;
+        }
         location / {
             root   /etc/nginx/public;
             index  index.html index.htm;

nginx/public/chat_evie.html

@@ -17,7 +17,7 @@
     document.addEventListener('DOMContentLoaded', function() {
       const eveAI = new EveAI(
         '2',
-        'EveAI-CHAT-9079-8604-7441-6496-7604',
+        'EveAI-CHAT-6525-5692-6412-5828-7546',
         'http://macstudio.ask-eve-ai-local.com',
         'en',
         'en,fr,nl',

requirements.txt

@@ -82,3 +82,4 @@ prometheus-client~=0.20.0
 flower~=2.0.1
 psutil~=6.0.0
 celery-redbeat~=2.2.0
+WTForms-SQLAlchemy~=0.4.1

scripts/db_squash.sh

@@ -0,0 +1,42 @@
+#!/usr/bin/env bash
+
+# Usage: ./db_migrate.sh -m "Your migration message" -d migrations/public
+
+# Load environment variables
+set -a
+source .env
+set +a
+
+project_dir=$(pwd)
+# Ensure you're in the correct directory
+cd "$project_dir" || exit 1
+
+# Load environment variables (including database connection details)
+source .env
+
+# Set Flask app and Python path
+export FLASK_APP=scripts/run_eveai_app.py
+export PYTHONPATH="$PYTHONPATH:$project_dir"
+
+while getopts m: flag
+do
+    case "${flag}" in
+        m) message=${OPTARG};;
+        *)
+            echo "Invalid option: ${flag}"
+            exit 1
+        ;;
+    esac
+done
+
+# Check if the message and directory are provided
+if [ -z "$message" ]; then
+    echo "Message is required."
+    echo "Usage: ./db_migrate.sh -m \"Your migration message\" -d migrations/public"
+    exit 1
+fi
+
+# Run the alembic commands
+flask db revision --autogenerate -m "Consolidated Public Migrations - $message" -d migrations/public
+flask db revision --autogenerate -m "Consolidated Tenant Migrations - $message" -d migrations/tenant
+

scripts/start_eveai_app.sh

@@ -32,10 +32,12 @@ export PYTHONPATH="$PROJECT_DIR/patched_packages:$PYTHONPATH:$PROJECT_DIR"  # In
 # Run Alembic upgrade for the public schema
 echo "Applying migrations to the public schema..."
 flask db upgrade -d "${PROJECT_DIR}/migrations/public"
+echo "Finished applying migrations to the public schema..."
 
 # Run Alembic upgrade for the tenant schema
 echo "Applying migrations to the tenant schema..."
 flask db upgrade -d "${PROJECT_DIR}/migrations/tenant"
+echo "Finished applying migrations to the tenant schema..."
 
 # Set flask environment variables
 #export FLASK_ENV=development  # Use 'production' as appropriate