godber/EmbeddingBuddy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 7 Wiki Activity

add-os-load #3

Merged
godber merged 3 commits from add-os-load into main 2025-08-14 19:07:24 -07:00
Conversation 0 Commits 3 Files Changed 16 +537 -246
7 changed files with 537 additions and 246 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit 1b6845774b - Show all commits

2
pyproject.toml
View File

@@ -1,6 +1,6 @@
[project]
name = "embeddingbuddy"
version = "0.2.0"
version = "0.3.0"
description = "A Python Dash application for interactive exploration and visualization of embedding vectors through dimensionality reduction techniques."
readme = "README.md"
requires-python = ">=3.11"

2
src/embeddingbuddy/app.py
View File

@@ -9,7 +9,7 @@ from .ui.callbacks.interactions import InteractionCallbacks
def create_app():
app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
# Allow callbacks to components that are dynamically created in tabs
app.config.suppress_callback_exceptions = True

66
src/embeddingbuddy/models/field_mapper.py
View File

@@ -25,7 +25,7 @@ class FieldMapper:
def suggest_mappings(field_analysis: Dict) -> Dict[str, List[str]]:
"""
Suggest field mappings based on field analysis.
Each dropdown will show ALL available fields, but ordered by relevance
with the most likely candidates first.
@@ -59,42 +59,70 @@ class FieldMapper:
# Embedding field suggestions (vector fields first, then name-based candidates, then all fields)
embedding_candidates = vector_fields.copy()
# Add fields that likely contain embeddings based on name
embedding_name_candidates = [f for f in all_fields if any(
keyword in f.lower() for keyword in ["embedding", "embeddings", "vector", "vectors", "embed"]
)]
embedding_name_candidates = [
f
for f in all_fields
if any(
keyword in f.lower()
for keyword in ["embedding", "embeddings", "vector", "vectors", "embed"]
)
]
# Add name-based candidates that aren't already in vector_fields
for candidate in embedding_name_candidates:
if candidate not in embedding_candidates:
embedding_candidates.append(candidate)
suggestions["embedding"] = create_ordered_suggestions(embedding_candidates, all_fields)
suggestions["embedding"] = create_ordered_suggestions(
embedding_candidates, all_fields
)
# Text field suggestions (text fields first, then all fields)
text_candidates = text_fields.copy()
suggestions["text"] = create_ordered_suggestions(text_candidates, all_fields)
# ID field suggestions (ID-like fields first, then all fields)
id_candidates = [f for f in keyword_fields if any(
keyword in f.lower() for keyword in ["id", "_id", "doc", "document"]
)]
id_candidates = [
f
for f in keyword_fields
if any(keyword in f.lower() for keyword in ["id", "_id", "doc", "document"])
]
id_candidates.append("_id") # _id is always available
suggestions["id"] = create_ordered_suggestions(id_candidates, all_fields)
# Category field suggestions (category-like fields first, then all fields)
category_candidates = [f for f in keyword_fields if any(
keyword in f.lower() for keyword in ["category", "class", "type", "label"]
)]
suggestions["category"] = create_ordered_suggestions(category_candidates, all_fields)
category_candidates = [
f
for f in keyword_fields
if any(
keyword in f.lower()
for keyword in ["category", "class", "type", "label"]
)
]
suggestions["category"] = create_ordered_suggestions(
category_candidates, all_fields
)
# Subcategory field suggestions (subcategory-like fields first, then all fields)
subcategory_candidates = [f for f in keyword_fields if any(
keyword in f.lower() for keyword in ["subcategory", "subclass", "subtype", "subtopic"]
)]
suggestions["subcategory"] = create_ordered_suggestions(subcategory_candidates, all_fields)
subcategory_candidates = [
f
for f in keyword_fields
if any(
keyword in f.lower()
for keyword in ["subcategory", "subclass", "subtype", "subtopic"]
)
]
suggestions["subcategory"] = create_ordered_suggestions(
subcategory_candidates, all_fields
)
# Tags field suggestions (tag-like fields first, then all fields)
tags_candidates = [f for f in keyword_fields if any(
keyword in f.lower() for keyword in ["tag", "tags", "keyword", "keywords"]
)]
tags_candidates = [
f
for f in keyword_fields
if any(
keyword in f.lower()
for keyword in ["tag", "tags", "keyword", "keywords"]
)
]
suggestions["tags"] = create_ordered_suggestions(tags_candidates, all_fields)
return suggestions

127
src/embeddingbuddy/ui/callbacks/data_processing.py
View File

@@ -97,10 +97,9 @@ class DataProcessingCallbacks:
# Register collapsible section callbacks
self._register_collapse_callbacks()
def _register_opensearch_callbacks(self, section_type, opensearch_client):
"""Register callbacks for a specific section (data or prompts)."""
@callback(
Output(f"{section_type}-auth-collapse", "is_open"),
[Input(f"{section_type}-auth-toggle", "n_clicks")],
@@ -144,9 +143,23 @@ class DataProcessingCallbacks:
],
prevent_initial_call=True,
)
def test_opensearch_connection(n_clicks, url, index_name, username, password, api_key):
def test_opensearch_connection(
n_clicks, url, index_name, username, password, api_key
):
if not n_clicks or not url or not index_name:
return no_update, no_update, no_update, no_update, no_update, no_update, no_update, no_update, no_update, no_update, no_update
return (
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
no_update,
)
# Test connection
success, message = opensearch_client.connect(
@@ -173,7 +186,9 @@ class DataProcessingCallbacks:
)
# Analyze fields
success, field_analysis, analysis_message = opensearch_client.analyze_fields(index_name)
success, field_analysis, analysis_message = (
opensearch_client.analyze_fields(index_name)
)
if not success:
return (
@@ -194,8 +209,11 @@ class DataProcessingCallbacks:
field_suggestions = FieldMapper.suggest_mappings(field_analysis)
from ...ui.components.datasource import DataSourceComponent
datasource = DataSourceComponent()
field_mapping_ui = datasource.create_field_mapping_interface(field_suggestions, section_type)
field_mapping_ui = datasource.create_field_mapping_interface(
field_suggestions, section_type
)
return (
self._create_status_alert(f"✅ {message}", "success"),
@@ -203,16 +221,36 @@ class DataProcessingCallbacks:
{"display": "block"},
{"display": "block"},
False,
[{"label": field, "value": field} for field in field_suggestions.get("embedding", [])],
[{"label": field, "value": field} for field in field_suggestions.get("text", [])],
[{"label": field, "value": field} for field in field_suggestions.get("id", [])],
[{"label": field, "value": field} for field in field_suggestions.get("category", [])],
[{"label": field, "value": field} for field in field_suggestions.get("subcategory", [])],
[{"label": field, "value": field} for field in field_suggestions.get("tags", [])],
[
{"label": field, "value": field}
for field in field_suggestions.get("embedding", [])
],
[
{"label": field, "value": field}
for field in field_suggestions.get("text", [])
],
[
{"label": field, "value": field}
for field in field_suggestions.get("id", [])
],
[
{"label": field, "value": field}
for field in field_suggestions.get("category", [])
],
[
{"label": field, "value": field}
for field in field_suggestions.get("subcategory", [])
],
[
{"label": field, "value": field}
for field in field_suggestions.get("tags", [])
],
)
# Determine output target based on section type
output_target = "processed-data" if section_type == "data" else "processed-prompts"
output_target = (
"processed-data" if section_type == "data" else "processed-prompts"
)
@callback(
[
@@ -235,8 +273,17 @@ class DataProcessingCallbacks:
],
prevent_initial_call=True,
)
def load_opensearch_data(n_clicks, index_name, query_size, embedding_field, text_field,
id_field, category_field, subcategory_field, tags_field):
def load_opensearch_data(
n_clicks,
index_name,
query_size,
embedding_field,
text_field,
id_field,
category_field,
subcategory_field,
tags_field,
):
if not n_clicks or not index_name or not embedding_field or not text_field:
return no_update, no_update, no_update, no_update, no_update
@@ -248,14 +295,16 @@ class DataProcessingCallbacks:
query_size = 1000 # Cap at reasonable maximum
# Create field mapping
field_mapping = FieldMapper.create_mapping_from_dict({
"embedding": embedding_field,
"text": text_field,
"id": id_field,
"category": category_field,
"subcategory": subcategory_field,
"tags": tags_field
})
field_mapping = FieldMapper.create_mapping_from_dict(
{
"embedding": embedding_field,
"text": text_field,
"id": id_field,
"category": category_field,
"subcategory": subcategory_field,
"tags": tags_field,
}
)
# Fetch data from OpenSearch
success, raw_documents, message = opensearch_client.fetch_data(
@@ -268,11 +317,13 @@ class DataProcessingCallbacks:
"",
False,
f"❌ Failed to fetch {section_type}: {message}",
True
True,
)
# Process the data
processed_data = self.processor.process_opensearch_data(raw_documents, field_mapping)
processed_data = self.processor.process_opensearch_data(
raw_documents, field_mapping
)
if processed_data.error:
return (
@@ -280,7 +331,7 @@ class DataProcessingCallbacks:
"",
False,
f"❌ {section_type.title()} processing error: {processed_data.error}",
True
True,
)
success_message = f"✅ Successfully loaded {len(processed_data.documents)} {section_type} from OpenSearch"
@@ -290,27 +341,29 @@ class DataProcessingCallbacks:
return (
{
"documents": [
self._document_to_dict(doc) for doc in processed_data.documents
self._document_to_dict(doc)
for doc in processed_data.documents
],
"embeddings": processed_data.embeddings.tolist(),
},
success_message,
True,
"",
False
False,
)
else: # prompts
return (
{
"prompts": [
self._document_to_dict(doc) for doc in processed_data.documents
self._document_to_dict(doc)
for doc in processed_data.documents
],
"embeddings": processed_data.embeddings.tolist(),
},
success_message,
True,
"",
False
False,
)
except Exception as e:
@@ -367,7 +420,7 @@ class DataProcessingCallbacks:
def _register_collapse_callbacks(self):
"""Register callbacks for collapsible sections."""
# Data section collapse callback
@callback(
[
@@ -381,7 +434,11 @@ class DataProcessingCallbacks:
def toggle_data_collapse(n_clicks, is_open):
if n_clicks:
new_state = not is_open
icon_class = "fas fa-chevron-down me-2" if new_state else "fas fa-chevron-right me-2"
icon_class = (
"fas fa-chevron-down me-2"
if new_state
else "fas fa-chevron-right me-2"
)
return new_state, icon_class
return is_open, "fas fa-chevron-down me-2"
@@ -398,7 +455,11 @@ class DataProcessingCallbacks:
def toggle_prompts_collapse(n_clicks, is_open):
if n_clicks:
new_state = not is_open
icon_class = "fas fa-chevron-down me-2" if new_state else "fas fa-chevron-right me-2"
icon_class = (
"fas fa-chevron-down me-2"
if new_state
else "fas fa-chevron-right me-2"
)
return new_state, icon_class
return is_open, "fas fa-chevron-down me-2"

494
src/embeddingbuddy/ui/components/datasource.py
View File

@@ -43,194 +43,314 @@ class DataSourceComponent:
return html.Div(
[
# Data Section
dbc.Card([
dbc.CardHeader([
dbc.Button(
dbc.Card(
[
dbc.CardHeader(
[
html.I(className="fas fa-chevron-down me-2", id="data-collapse-icon"),
"📄 Documents/Data"
],
id="data-collapse-toggle",
color="link",
className="text-start p-0 w-100 text-decoration-none",
style={"border": "none", "font-size": "1.25rem", "font-weight": "500"}
dbc.Button(
[
html.I(
className="fas fa-chevron-down me-2",
id="data-collapse-icon",
),
"📄 Documents/Data",
],
id="data-collapse-toggle",
color="link",
className="text-start p-0 w-100 text-decoration-none",
style={
"border": "none",
"font-size": "1.25rem",
"font-weight": "500",
},
),
]
),
]),
dbc.Collapse([
dbc.CardBody([
self._create_opensearch_section("data")
])
], id="data-collapse", is_open=True)
], className="mb-4"),
dbc.Collapse(
[dbc.CardBody([self._create_opensearch_section("data")])],
id="data-collapse",
is_open=True,
),
],
className="mb-4",
),
# Prompts Section
dbc.Card([
dbc.CardHeader([
dbc.Button(
dbc.Card(
[
dbc.CardHeader(
[
html.I(className="fas fa-chevron-down me-2", id="prompts-collapse-icon"),
"💬 Prompts"
],
id="prompts-collapse-toggle",
color="link",
className="text-start p-0 w-100 text-decoration-none",
style={"border": "none", "font-size": "1.25rem", "font-weight": "500"}
dbc.Button(
[
html.I(
className="fas fa-chevron-down me-2",
id="prompts-collapse-icon",
),
"💬 Prompts",
],
id="prompts-collapse-toggle",
color="link",
className="text-start p-0 w-100 text-decoration-none",
style={
"border": "none",
"font-size": "1.25rem",
"font-weight": "500",
},
),
]
),
]),
dbc.Collapse([
dbc.CardBody([
self._create_opensearch_section("prompts")
])
], id="prompts-collapse", is_open=True)
], className="mb-4"),
dbc.Collapse(
[
dbc.CardBody(
[self._create_opensearch_section("prompts")]
)
],
id="prompts-collapse",
is_open=True,
),
],
className="mb-4",
),
# Hidden dropdowns to prevent callback errors (for both sections)
html.Div([
# Data dropdowns (hidden sync targets)
dcc.Dropdown(id="data-embedding-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="data-text-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="data-id-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="data-category-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="data-subcategory-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="data-tags-field-dropdown", style={"display": "none"}),
# Data UI dropdowns (hidden placeholders)
dcc.Dropdown(id="data-embedding-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="data-text-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="data-id-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="data-category-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="data-subcategory-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="data-tags-field-dropdown-ui", style={"display": "none"}),
# Prompts dropdowns (hidden sync targets)
dcc.Dropdown(id="prompts-embedding-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="prompts-text-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="prompts-id-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="prompts-category-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="prompts-subcategory-field-dropdown", style={"display": "none"}),
dcc.Dropdown(id="prompts-tags-field-dropdown", style={"display": "none"}),
# Prompts UI dropdowns (hidden placeholders)
dcc.Dropdown(id="prompts-embedding-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="prompts-text-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="prompts-id-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="prompts-category-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="prompts-subcategory-field-dropdown-ui", style={"display": "none"}),
dcc.Dropdown(id="prompts-tags-field-dropdown-ui", style={"display": "none"}),
], style={"display": "none"}),
html.Div(
[
# Data dropdowns (hidden sync targets)
dcc.Dropdown(
id="data-embedding-field-dropdown",
style={"display": "none"},
),
dcc.Dropdown(
id="data-text-field-dropdown", style={"display": "none"}
),
dcc.Dropdown(
id="data-id-field-dropdown", style={"display": "none"}
),
dcc.Dropdown(
id="data-category-field-dropdown", style={"display": "none"}
),
dcc.Dropdown(
id="data-subcategory-field-dropdown",
style={"display": "none"},
),
dcc.Dropdown(
id="data-tags-field-dropdown", style={"display": "none"}
),
# Data UI dropdowns (hidden placeholders)
dcc.Dropdown(
id="data-embedding-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="data-text-field-dropdown-ui", style={"display": "none"}
),
dcc.Dropdown(
id="data-id-field-dropdown-ui", style={"display": "none"}
),
dcc.Dropdown(
id="data-category-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="data-subcategory-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="data-tags-field-dropdown-ui", style={"display": "none"}
),
# Prompts dropdowns (hidden sync targets)
dcc.Dropdown(
id="prompts-embedding-field-dropdown",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-text-field-dropdown", style={"display": "none"}
),
dcc.Dropdown(
id="prompts-id-field-dropdown", style={"display": "none"}
),
dcc.Dropdown(
id="prompts-category-field-dropdown",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-subcategory-field-dropdown",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-tags-field-dropdown", style={"display": "none"}
),
# Prompts UI dropdowns (hidden placeholders)
dcc.Dropdown(
id="prompts-embedding-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-text-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-id-field-dropdown-ui", style={"display": "none"}
),
dcc.Dropdown(
id="prompts-category-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-subcategory-field-dropdown-ui",
style={"display": "none"},
),
dcc.Dropdown(
id="prompts-tags-field-dropdown-ui",
style={"display": "none"},
),
],
style={"display": "none"},
),
]
)
def _create_opensearch_section(self, section_type):
"""Create a complete OpenSearch section for either 'data' or 'prompts'."""
section_id = section_type # 'data' or 'prompts'
return html.Div([
# Connection section
html.H6("Connection", className="mb-2"),
dbc.Row([
dbc.Col([
dbc.Label("OpenSearch URL:"),
dbc.Input(
id=f"{section_id}-opensearch-url",
type="text",
placeholder="https://opensearch.example.com:9200",
className="mb-2",
),
], width=12),
]),
dbc.Row([
dbc.Col([
dbc.Label("Index Name:"),
dbc.Input(
id=f"{section_id}-opensearch-index",
type="text",
placeholder="my-embeddings-index",
className="mb-2",
),
], width=6),
dbc.Col([
dbc.Label("Query Size:"),
dbc.Input(
id=f"{section_id}-opensearch-query-size",
type="number",
value=100,
min=1,
max=1000,
placeholder="100",
className="mb-2",
),
], width=6),
]),
dbc.Row([
dbc.Col([
dbc.Button(
"Test Connection",
id=f"{section_id}-test-connection-btn",
color="primary",
className="mb-3",
),
], width=12),
]),
# Authentication section (collapsible)
dbc.Collapse([
html.Hr(),
html.H6("Authentication (Optional)", className="mb-2"),
dbc.Row([
dbc.Col([
dbc.Label("Username:"),
return html.Div(
[
# Connection section
html.H6("Connection", className="mb-2"),
dbc.Row(
[
dbc.Col(
[
dbc.Label("OpenSearch URL:"),
dbc.Input(
id=f"{section_id}-opensearch-url",
type="text",
placeholder="https://opensearch.example.com:9200",
className="mb-2",
),
],
width=12,
),
]
),
dbc.Row(
[
dbc.Col(
[
dbc.Label("Index Name:"),
dbc.Input(
id=f"{section_id}-opensearch-index",
type="text",
placeholder="my-embeddings-index",
className="mb-2",
),
],
width=6,
),
dbc.Col(
[
dbc.Label("Query Size:"),
dbc.Input(
id=f"{section_id}-opensearch-query-size",
type="number",
value=100,
min=1,
max=1000,
placeholder="100",
className="mb-2",
),
],
width=6,
),
]
),
dbc.Row(
[
dbc.Col(
[
dbc.Button(
"Test Connection",
id=f"{section_id}-test-connection-btn",
color="primary",
className="mb-3",
),
],
width=12,
),
]
),
# Authentication section (collapsible)
dbc.Collapse(
[
html.Hr(),
html.H6("Authentication (Optional)", className="mb-2"),
dbc.Row(
[
dbc.Col(
[
dbc.Label("Username:"),
dbc.Input(
id=f"{section_id}-opensearch-username",
type="text",
className="mb-2",
),
],
width=6,
),
dbc.Col(
[
dbc.Label("Password:"),
dbc.Input(
id=f"{section_id}-opensearch-password",
type="password",
className="mb-2",
),
],
width=6,
),
]
),
dbc.Label("OR"),
dbc.Input(
id=f"{section_id}-opensearch-username",
id=f"{section_id}-opensearch-api-key",
type="text",
placeholder="API Key",
className="mb-2",
),
], width=6),
dbc.Col([
dbc.Label("Password:"),
dbc.Input(
id=f"{section_id}-opensearch-password",
type="password",
className="mb-2",
),
], width=6),
]),
dbc.Label("OR"),
dbc.Input(
id=f"{section_id}-opensearch-api-key",
type="text",
placeholder="API Key",
className="mb-2",
],
id=f"{section_id}-auth-collapse",
is_open=False,
),
], id=f"{section_id}-auth-collapse", is_open=False),
dbc.Button(
"Show Authentication",
id=f"{section_id}-auth-toggle",
color="link",
size="sm",
className="p-0 mb-3",
),
# Connection status
html.Div(id=f"{section_id}-connection-status", className="mb-3"),
# Field mapping section (hidden initially)
html.Div(id=f"{section_id}-field-mapping-section", style={"display": "none"}),
# Load data button (hidden initially)
html.Div([
dbc.Button(
f"Load {section_type.title()}",
id=f"{section_id}-load-opensearch-data-btn",
color="success",
className="mb-2",
disabled=True,
"Show Authentication",
id=f"{section_id}-auth-toggle",
color="link",
size="sm",
className="p-0 mb-3",
),
], id=f"{section_id}-load-data-section", style={"display": "none"}),
# OpenSearch status/results
html.Div(id=f"{section_id}-opensearch-status", className="mb-3"),
])
# Connection status
html.Div(id=f"{section_id}-connection-status", className="mb-3"),
# Field mapping section (hidden initially)
html.Div(
id=f"{section_id}-field-mapping-section", style={"display": "none"}
),
# Load data button (hidden initially)
html.Div(
[
dbc.Button(
f"Load {section_type.title()}",
id=f"{section_id}-load-opensearch-data-btn",
color="success",
className="mb-2",
disabled=True,
),
],
id=f"{section_id}-load-data-section",
style={"display": "none"},
),
# OpenSearch status/results
html.Div(id=f"{section_id}-opensearch-status", className="mb-3"),
]
)
def create_field_mapping_interface(self, field_suggestions, section_type="data"):
"""Create field mapping interface based on detected fields."""
@@ -254,9 +374,13 @@ class DataSourceComponent:
id=f"{section_type}-embedding-field-dropdown-ui",
options=[
{"label": field, "value": field}
for field in field_suggestions.get("embedding", [])
for field in field_suggestions.get(
"embedding", []
)
],
value=field_suggestions.get("embedding", [None])[0], # Default to first suggestion
value=field_suggestions.get("embedding", [None])[
0
], # Default to first suggestion
placeholder="Select embedding field...",
className="mb-2",
),
@@ -274,7 +398,9 @@ class DataSourceComponent:
{"label": field, "value": field}
for field in field_suggestions.get("text", [])
],
value=field_suggestions.get("text", [None])[0], # Default to first suggestion
value=field_suggestions.get("text", [None])[
0
], # Default to first suggestion
placeholder="Select text field...",
className="mb-2",
),
@@ -296,7 +422,9 @@ class DataSourceComponent:
{"label": field, "value": field}
for field in field_suggestions.get("id", [])
],
value=field_suggestions.get("id", [None])[0], # Default to first suggestion
value=field_suggestions.get("id", [None])[
0
], # Default to first suggestion
placeholder="Select ID field...",
className="mb-2",
),
@@ -310,9 +438,13 @@ class DataSourceComponent:
id=f"{section_type}-category-field-dropdown-ui",
options=[
{"label": field, "value": field}
for field in field_suggestions.get("category", [])
for field in field_suggestions.get(
"category", []
)
],
value=field_suggestions.get("category", [None])[0], # Default to first suggestion
value=field_suggestions.get("category", [None])[
0
], # Default to first suggestion
placeholder="Select category field...",
className="mb-2",
),
@@ -330,9 +462,13 @@ class DataSourceComponent:
id=f"{section_type}-subcategory-field-dropdown-ui",
options=[
{"label": field, "value": field}
for field in field_suggestions.get("subcategory", [])
for field in field_suggestions.get(
"subcategory", []
)
],
value=field_suggestions.get("subcategory", [None])[0], # Default to first suggestion
value=field_suggestions.get("subcategory", [None])[
0
], # Default to first suggestion
placeholder="Select subcategory field...",
className="mb-2",
),
@@ -348,7 +484,9 @@ class DataSourceComponent:
{"label": field, "value": field}
for field in field_suggestions.get("tags", [])
],
value=field_suggestions.get("tags", [None])[0], # Default to first suggestion
value=field_suggestions.get("tags", [None])[
0
], # Default to first suggestion
placeholder="Select tags field...",
className="mb-2",
),

90
tests/test_opensearch.py
View File

@@ -99,22 +99,57 @@ class TestFieldMapper:
"text_fields": ["content", "description"],
"keyword_fields": ["doc_id", "category", "type", "tags"],
"numeric_fields": ["count"],
"all_fields": ["embedding", "content", "description", "doc_id", "category", "type", "tags", "count"],
"all_fields": [
"embedding",
"content",
"description",
"doc_id",
"category",
"type",
"tags",
"count",
],
}
suggestions = FieldMapper.suggest_mappings(field_analysis)
# Check that all dropdowns contain all fields
all_fields = ["embedding", "content", "description", "doc_id", "category", "type", "tags", "count"]
for field_type in ["embedding", "text", "id", "category", "subcategory", "tags"]:
all_fields = [
"embedding",
"content",
"description",
"doc_id",
"category",
"type",
"tags",
"count",
]
for field_type in [
"embedding",
"text",
"id",
"category",
"subcategory",
"tags",
]:
for field in all_fields:
assert field in suggestions[field_type], f"Field '{field}' missing from {field_type} suggestions"
assert field in suggestions[field_type], (
f"Field '{field}' missing from {field_type} suggestions"
)
# Check that best candidates are first
assert suggestions["embedding"][0] == "embedding" # vector field should be first
assert suggestions["text"][0] in ["content", "description"] # text fields should be first
assert (
suggestions["embedding"][0] == "embedding"
) # vector field should be first
assert suggestions["text"][0] in [
"content",
"description",
] # text fields should be first
assert suggestions["id"][0] == "doc_id" # ID-like field should be first
assert suggestions["category"][0] in ["category", "type"] # category-like field should be first
assert suggestions["category"][0] in [
"category",
"type",
] # category-like field should be first
assert suggestions["tags"][0] == "tags" # tags field should be first
def test_suggest_mappings_name_based_embedding(self):
@@ -124,19 +159,48 @@ class TestFieldMapper:
"text_fields": ["content", "description"],
"keyword_fields": ["doc_id", "category", "type", "tags"],
"numeric_fields": ["count"],
"all_fields": ["content", "description", "doc_id", "category", "embedding", "type", "tags", "count"],
"all_fields": [
"content",
"description",
"doc_id",
"category",
"embedding",
"type",
"tags",
"count",
],
}
suggestions = FieldMapper.suggest_mappings(field_analysis)
# Check that 'embedding' field is prioritized despite not being detected as vector type
assert suggestions["embedding"][0] == "embedding", "Field named 'embedding' should be first priority"
assert suggestions["embedding"][0] == "embedding", (
"Field named 'embedding' should be first priority"
)
# Check that all fields are still available
all_fields = ["content", "description", "doc_id", "category", "embedding", "type", "tags", "count"]
for field_type in ["embedding", "text", "id", "category", "subcategory", "tags"]:
all_fields = [
"content",
"description",
"doc_id",
"category",
"embedding",
"type",
"tags",
"count",
]
for field_type in [
"embedding",
"text",
"id",
"category",
"subcategory",
"tags",
]:
for field in all_fields:
assert field in suggestions[field_type], f"Field '{field}' missing from {field_type} suggestions"
assert field in suggestions[field_type], (
f"Field '{field}' missing from {field_type} suggestions"
)
def test_validate_mapping_success(self):
mapping = FieldMapping(

2
uv.lock generated
View File

@@ -412,7 +412,7 @@ wheels = [
[[package]]
name = "embeddingbuddy"
version = "0.2.0"
version = "0.3.0"
source = { editable = "." }
dependencies = [
{ name = "dash" },