Merge branch 'main' into docs-20-nov-25

This commit is contained in:
April I. Murphy 2025-11-24 10:15:20 -08:00 committed by GitHub
commit 6dc8ba233a
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
11 changed files with 385 additions and 85 deletions

View file

@ -391,7 +391,7 @@ function SearchPage() {
<DeleteConfirmationDialog
open={showBulkDeleteDialog}
onOpenChange={setShowBulkDeleteDialog}
title="Delete Documents"
title={selectedRows.length > 1 ? "Delete Documents" : "Delete Document"}
description={`Are you sure you want to delete ${
selectedRows.length
} document${
@ -400,7 +400,7 @@ function SearchPage() {
Documents to be deleted:
${selectedRows.map((row) => `${row.filename}`).join("\n")}`}
confirmText="Delete All"
confirmText={selectedRows.length > 1 ? "Delete All" : "Delete"}
onConfirm={handleBulkDelete}
isLoading={deleteDocumentMutation.isPending}
/>

View file

@ -171,7 +171,7 @@ export function OnboardingContent({
/>
</OnboardingStep>
{/* Step 2 */}
{/* Step 3 */}
<OnboardingStep
isVisible={currentStep >= 2}
isCompleted={currentStep > 2 || !!selectedNudge}
@ -209,7 +209,7 @@ export function OnboardingContent({
/>
)}
{/* Step 3 */}
{/* Step 4 */}
<OnboardingStep
isVisible={currentStep >= 3 && !isLoading && !!displayMessage}
isCompleted={currentStep > 3}

View file

@ -43,6 +43,33 @@ import {
} from "@/lib/upload-utils";
import { cn } from "@/lib/utils";
// Supported file extensions - single source of truth
const SUPPORTED_EXTENSIONS = [
".pdf",
".doc",
".docx",
".pptx",
".ppt",
".xlsx",
".xls",
".csv",
".txt",
".md",
".html",
".htm",
".rtf",
".odt",
".asciidoc",
".adoc",
".png",
".jpg",
".jpeg",
".gif",
".bmp",
".tiff",
".webp",
];
export function KnowledgeDropdown() {
const { addTask } = useTask();
const { refetch: refetchTasks } = useGetTasksQuery();
@ -247,27 +274,18 @@ export function KnowledgeDropdown() {
try {
const fileList = Array.from(files);
const supportedExtensions = [
".pdf",
".doc",
".docx",
".txt",
".md",
".rtf",
".odt",
];
const filteredFiles = fileList.filter((file) => {
const ext = file.name
.substring(file.name.lastIndexOf("."))
.toLowerCase();
return supportedExtensions.includes(ext);
return SUPPORTED_EXTENSIONS.includes(ext);
});
if (filteredFiles.length === 0) {
toast.error("No supported files found", {
description:
"Please select a folder containing PDF, DOC, DOCX, TXT, MD, RTF, or ODT files.",
"Please select a folder containing supported document files (PDF, DOCX, PPTX, XLSX, CSV, HTML, images, etc.).",
});
return;
}
@ -527,7 +545,7 @@ export function KnowledgeDropdown() {
type="file"
onChange={handleFileChange}
className="hidden"
accept=".pdf,.doc,.docx,.txt,.md,.rtf,.odt"
accept={SUPPORTED_EXTENSIONS.join(",")}
/>
<input

View file

@ -28,7 +28,7 @@ export const UI_CONSTANTS = {
export const ANIMATION_DURATION = 0.4;
export const SIDEBAR_WIDTH = 280;
export const HEADER_HEIGHT = 54;
export const TOTAL_ONBOARDING_STEPS = 5;
export const TOTAL_ONBOARDING_STEPS = 4;
/**
* Local Storage Keys

View file

@ -334,7 +334,7 @@ class AppClients:
# Initialize Langflow HTTP client
self.langflow_http_client = httpx.AsyncClient(
base_url=LANGFLOW_URL, timeout=300.0
base_url=LANGFLOW_URL, timeout=1200.0
)
return self

View file

@ -365,6 +365,21 @@ class OpenRAGTUI(App):
self.container_manager = ContainerManager()
self.env_manager = EnvManager()
self.docling_manager = DoclingManager() # Initialize singleton instance
def notify(
self,
message: str,
*,
title: str = "",
severity: str = "information",
timeout: float | None = None,
markup: bool = True,
) -> None:
"""Override notify to make notifications last 20 seconds by default."""
# If timeout is None (default), make it 20 seconds
if timeout is None:
timeout = 20.0
super().notify(message, title=title, severity=severity, timeout=timeout, markup=markup)
def on_mount(self) -> None:
"""Initialize the application."""

View file

@ -35,6 +35,13 @@ class EnvConfig:
langflow_ingest_flow_id: str = "5488df7c-b93f-4f87-a446-b67028bc0813"
langflow_url_ingest_flow_id: str = "72c3d17c-2dac-4a73-b48a-6518473d7830"
# Provider API keys and endpoints
anthropic_api_key: str = ""
ollama_endpoint: str = ""
watsonx_api_key: str = ""
watsonx_endpoint: str = ""
watsonx_project_id: str = ""
# OAuth settings
google_oauth_client_id: str = ""
google_oauth_client_secret: str = ""
@ -109,57 +116,75 @@ class EnvManager:
return f"'{escaped_value}'"
def load_existing_env(self) -> bool:
"""Load existing .env file if it exists."""
if not self.env_file.exists():
return False
"""Load existing .env file if it exists, or fall back to environment variables."""
import os
# Map env vars to config attributes
attr_map = {
"OPENAI_API_KEY": "openai_api_key",
"ANTHROPIC_API_KEY": "anthropic_api_key",
"OLLAMA_ENDPOINT": "ollama_endpoint",
"WATSONX_API_KEY": "watsonx_api_key",
"WATSONX_ENDPOINT": "watsonx_endpoint",
"WATSONX_PROJECT_ID": "watsonx_project_id",
"OPENSEARCH_PASSWORD": "opensearch_password",
"LANGFLOW_SECRET_KEY": "langflow_secret_key",
"LANGFLOW_SUPERUSER": "langflow_superuser",
"LANGFLOW_SUPERUSER_PASSWORD": "langflow_superuser_password",
"LANGFLOW_CHAT_FLOW_ID": "langflow_chat_flow_id",
"LANGFLOW_INGEST_FLOW_ID": "langflow_ingest_flow_id",
"LANGFLOW_URL_INGEST_FLOW_ID": "langflow_url_ingest_flow_id",
"NUDGES_FLOW_ID": "nudges_flow_id",
"GOOGLE_OAUTH_CLIENT_ID": "google_oauth_client_id",
"GOOGLE_OAUTH_CLIENT_SECRET": "google_oauth_client_secret",
"MICROSOFT_GRAPH_OAUTH_CLIENT_ID": "microsoft_graph_oauth_client_id",
"MICROSOFT_GRAPH_OAUTH_CLIENT_SECRET": "microsoft_graph_oauth_client_secret",
"WEBHOOK_BASE_URL": "webhook_base_url",
"AWS_ACCESS_KEY_ID": "aws_access_key_id",
"AWS_SECRET_ACCESS_KEY": "aws_secret_access_key",
"LANGFLOW_PUBLIC_URL": "langflow_public_url",
"OPENRAG_DOCUMENTS_PATHS": "openrag_documents_paths",
"OPENSEARCH_DATA_PATH": "opensearch_data_path",
"LANGFLOW_AUTO_LOGIN": "langflow_auto_login",
"LANGFLOW_NEW_USER_IS_ACTIVE": "langflow_new_user_is_active",
"LANGFLOW_ENABLE_SUPERUSER_CLI": "langflow_enable_superuser_cli",
"DISABLE_INGEST_WITH_LANGFLOW": "disable_ingest_with_langflow",
}
loaded_from_file = False
# Try to load from .env file first
if self.env_file.exists():
try:
with open(self.env_file, "r") as f:
for line in f:
line = line.strip()
if not line or line.startswith("#"):
continue
try:
with open(self.env_file, "r") as f:
for line in f:
line = line.strip()
if not line or line.startswith("#"):
continue
if "=" in line:
key, value = line.split("=", 1)
key = key.strip()
value = sanitize_env_value(value)
if "=" in line:
key, value = line.split("=", 1)
key = key.strip()
value = sanitize_env_value(value)
if key in attr_map:
setattr(self.config, attr_map[key], value)
# Map env vars to config attributes
attr_map = {
"OPENAI_API_KEY": "openai_api_key",
"OPENSEARCH_PASSWORD": "opensearch_password",
"LANGFLOW_SECRET_KEY": "langflow_secret_key",
"LANGFLOW_SUPERUSER": "langflow_superuser",
"LANGFLOW_SUPERUSER_PASSWORD": "langflow_superuser_password",
"LANGFLOW_CHAT_FLOW_ID": "langflow_chat_flow_id",
"LANGFLOW_INGEST_FLOW_ID": "langflow_ingest_flow_id",
"LANGFLOW_URL_INGEST_FLOW_ID": "langflow_url_ingest_flow_id",
"NUDGES_FLOW_ID": "nudges_flow_id",
"GOOGLE_OAUTH_CLIENT_ID": "google_oauth_client_id",
"GOOGLE_OAUTH_CLIENT_SECRET": "google_oauth_client_secret",
"MICROSOFT_GRAPH_OAUTH_CLIENT_ID": "microsoft_graph_oauth_client_id",
"MICROSOFT_GRAPH_OAUTH_CLIENT_SECRET": "microsoft_graph_oauth_client_secret",
"WEBHOOK_BASE_URL": "webhook_base_url",
"AWS_ACCESS_KEY_ID": "aws_access_key_id",
"AWS_SECRET_ACCESS_KEY": "aws_secret_access_key",
"LANGFLOW_PUBLIC_URL": "langflow_public_url",
"OPENRAG_DOCUMENTS_PATHS": "openrag_documents_paths",
"OPENSEARCH_DATA_PATH": "opensearch_data_path",
"LANGFLOW_AUTO_LOGIN": "langflow_auto_login",
"LANGFLOW_NEW_USER_IS_ACTIVE": "langflow_new_user_is_active",
"LANGFLOW_ENABLE_SUPERUSER_CLI": "langflow_enable_superuser_cli",
"DISABLE_INGEST_WITH_LANGFLOW": "disable_ingest_with_langflow",
}
if key in attr_map:
setattr(self.config, attr_map[key], value)
loaded_from_file = True
except Exception as e:
logger.error("Error loading .env file", error=str(e))
# Fall back to environment variables if .env file doesn't exist or failed to load
if not loaded_from_file:
logger.info("No .env file found, loading from environment variables")
for env_key, attr_name in attr_map.items():
value = os.environ.get(env_key, "")
if value:
setattr(self.config, attr_name, value)
return True
except Exception as e:
logger.error("Error loading .env file", error=str(e))
return False
return loaded_from_file
def setup_secure_defaults(self) -> None:
"""Set up secure default values for passwords and keys."""
@ -197,6 +222,30 @@ class EnvManager:
"Invalid OpenAI API key format (should start with sk-)"
)
# Import validation functions for new provider fields
from ..utils.validation import validate_anthropic_api_key
# Validate Anthropic API key format if provided
if self.config.anthropic_api_key:
if not validate_anthropic_api_key(self.config.anthropic_api_key):
self.config.validation_errors["anthropic_api_key"] = (
"Invalid Anthropic API key format (should start with sk-ant-)"
)
# Validate Ollama endpoint if provided
if self.config.ollama_endpoint:
if not validate_url(self.config.ollama_endpoint):
self.config.validation_errors["ollama_endpoint"] = (
"Invalid Ollama endpoint URL format"
)
# Validate IBM watsonx.ai endpoint if provided
if self.config.watsonx_endpoint:
if not validate_url(self.config.watsonx_endpoint):
self.config.validation_errors["watsonx_endpoint"] = (
"Invalid IBM watsonx.ai endpoint URL format"
)
# Validate documents paths only if provided (optional)
if self.config.openrag_documents_paths:
is_valid, error_msg, _ = validate_documents_paths(
@ -289,9 +338,6 @@ class EnvManager:
f.write(f"LANGFLOW_URL_INGEST_FLOW_ID={self._quote_env_value(self.config.langflow_url_ingest_flow_id)}\n")
f.write(f"NUDGES_FLOW_ID={self._quote_env_value(self.config.nudges_flow_id)}\n")
f.write(f"OPENSEARCH_PASSWORD={self._quote_env_value(self.config.opensearch_password)}\n")
# Only write OpenAI API key if provided (can be set during onboarding instead)
if self.config.openai_api_key:
f.write(f"OPENAI_API_KEY={self._quote_env_value(self.config.openai_api_key)}\n")
f.write(
f"OPENRAG_DOCUMENTS_PATHS={self._quote_env_value(self.config.openrag_documents_paths)}\n"
)
@ -300,6 +346,27 @@ class EnvManager:
)
f.write("\n")
# Provider API keys and endpoints (optional - can be set during onboarding)
provider_vars = []
if self.config.openai_api_key:
provider_vars.append(("OPENAI_API_KEY", self.config.openai_api_key))
if self.config.anthropic_api_key:
provider_vars.append(("ANTHROPIC_API_KEY", self.config.anthropic_api_key))
if self.config.ollama_endpoint:
provider_vars.append(("OLLAMA_ENDPOINT", self.config.ollama_endpoint))
if self.config.watsonx_api_key:
provider_vars.append(("WATSONX_API_KEY", self.config.watsonx_api_key))
if self.config.watsonx_endpoint:
provider_vars.append(("WATSONX_ENDPOINT", self.config.watsonx_endpoint))
if self.config.watsonx_project_id:
provider_vars.append(("WATSONX_PROJECT_ID", self.config.watsonx_project_id))
if provider_vars:
f.write("# AI Provider API Keys and Endpoints\n")
for var_name, var_value in provider_vars:
f.write(f"{var_name}={self._quote_env_value(var_value)}\n")
f.write("\n")
# Ingestion settings
f.write("# Ingestion settings\n")
f.write(f"DISABLE_INGEST_WITH_LANGFLOW={self._quote_env_value(self.config.disable_ingest_with_langflow)}\n")

View file

@ -20,7 +20,13 @@ from rich.text import Text
from pathlib import Path
from ..managers.env_manager import EnvManager
from ..utils.validation import validate_openai_api_key, validate_documents_paths
from ..utils.validation import (
validate_openai_api_key,
validate_anthropic_api_key,
validate_ollama_endpoint,
validate_watsonx_endpoint,
validate_documents_paths,
)
from pathlib import Path
@ -37,6 +43,45 @@ class OpenAIKeyValidator(Validator):
return self.failure("Invalid OpenAI API key format (should start with sk-)")
class AnthropicKeyValidator(Validator):
"""Validator for Anthropic API keys."""
def validate(self, value: str) -> ValidationResult:
if not value:
return self.success()
if validate_anthropic_api_key(value):
return self.success()
else:
return self.failure("Invalid Anthropic API key format (should start with sk-ant-)")
class OllamaEndpointValidator(Validator):
"""Validator for Ollama endpoint URLs."""
def validate(self, value: str) -> ValidationResult:
if not value:
return self.success()
if validate_ollama_endpoint(value):
return self.success()
else:
return self.failure("Invalid Ollama endpoint URL format")
class WatsonxEndpointValidator(Validator):
"""Validator for IBM watsonx.ai endpoint URLs."""
def validate(self, value: str) -> ValidationResult:
if not value:
return self.success()
if validate_watsonx_endpoint(value):
return self.success()
else:
return self.failure("Invalid watsonx.ai endpoint URL format")
class DocumentsPathValidator(Validator):
"""Validator for documents paths."""
@ -100,6 +145,9 @@ class ConfigScreen(Screen):
self.mode = mode # "no_auth" or "full"
self.env_manager = EnvManager()
self.inputs = {}
# Check if .env file exists
self.has_env_file = self.env_manager.env_file.exists()
# Load existing config if available
self.env_manager.load_existing_env()
@ -111,12 +159,15 @@ class ConfigScreen(Screen):
with ScrollableContainer(id="config-scroll"):
with Vertical(id="config-form"):
yield from self._create_all_fields()
yield Horizontal(
# Create button row - conditionally include Back button
buttons = [
Button("Generate Passwords", variant="default", id="generate-btn"),
Button("Save Configuration", variant="success", id="save-btn"),
Button("Back", variant="default", id="back-btn"),
classes="button-row",
)
]
# Only show Back button if .env file exists
if self.has_env_file:
buttons.append(Button("Back", variant="default", id="back-btn"))
yield Horizontal(*buttons, classes="button-row")
yield Footer()
def _create_header_text(self) -> Text:
@ -203,14 +254,13 @@ class ConfigScreen(Screen):
yield Static(" ")
# OpenAI API Key
yield Label("OpenAI API Key")
# Where to create OpenAI keys (helper above the box)
yield Label("OpenAI API Key (optional)")
yield Static(
Text("Get a key: https://platform.openai.com/api-keys", style="dim"),
classes="helper-text",
)
yield Static(
Text("Can also be provided during onboarding", style="dim italic"),
Text("Can be configured later in the UI", style="dim italic"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "openai_api_key", "")
@ -227,6 +277,107 @@ class ConfigScreen(Screen):
yield Button("Show", id="toggle-openai-key", variant="default")
yield Static(" ")
# Anthropic API Key
yield Label("Anthropic API Key (optional)")
yield Static(
Text("Get a key: https://console.anthropic.com/settings/keys", style="dim"),
classes="helper-text",
)
yield Static(
Text("Can be configured later in the UI", style="dim italic"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "anthropic_api_key", "")
with Horizontal(id="anthropic-key-row"):
input_widget = Input(
placeholder="sk-ant-...",
value=current_value,
password=True,
validators=[AnthropicKeyValidator()],
id="input-anthropic_api_key",
)
yield input_widget
self.inputs["anthropic_api_key"] = input_widget
yield Button("Show", id="toggle-anthropic-key", variant="default")
yield Static(" ")
# Ollama Endpoint
yield Label("Ollama Base URL (optional)")
yield Static(
Text("Endpoint of your Ollama server", style="dim"),
classes="helper-text",
)
yield Static(
Text("Can be configured later in the UI", style="dim italic"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "ollama_endpoint", "")
input_widget = Input(
placeholder="http://localhost:11434",
value=current_value,
validators=[OllamaEndpointValidator()],
id="input-ollama_endpoint",
)
yield input_widget
self.inputs["ollama_endpoint"] = input_widget
yield Static(" ")
# IBM watsonx.ai API Key
yield Label("IBM watsonx.ai API Key (optional)")
yield Static(
Text("Get a key: https://cloud.ibm.com/iam/apikeys", style="dim"),
classes="helper-text",
)
yield Static(
Text("Can be configured later in the UI", style="dim italic"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "watsonx_api_key", "")
with Horizontal(id="watsonx-key-row"):
input_widget = Input(
placeholder="",
value=current_value,
password=True,
id="input-watsonx_api_key",
)
yield input_widget
self.inputs["watsonx_api_key"] = input_widget
yield Button("Show", id="toggle-watsonx-key", variant="default")
yield Static(" ")
# IBM watsonx.ai Endpoint
yield Label("IBM watsonx.ai Endpoint")
yield Static(
Text("Example: https://us-south.ml.cloud.ibm.com", style="dim"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "watsonx_endpoint", "")
input_widget = Input(
placeholder="https://us-south.ml.cloud.ibm.com",
value=current_value,
validators=[WatsonxEndpointValidator()],
id="input-watsonx_endpoint",
)
yield input_widget
self.inputs["watsonx_endpoint"] = input_widget
yield Static(" ")
# IBM watsonx.ai Project ID
yield Label("IBM watsonx.ai Project ID")
yield Static(
Text("Find in your IBM Cloud project settings", style="dim"),
classes="helper-text",
)
current_value = getattr(self.env_manager.config, "watsonx_project_id", "")
input_widget = Input(
placeholder="",
value=current_value,
id="input-watsonx_project_id",
)
yield input_widget
self.inputs["watsonx_project_id"] = input_widget
yield Static(" ")
# Add OAuth fields only in full mode
if self.mode == "full":
# Google OAuth Client ID
@ -550,6 +701,18 @@ class ConfigScreen(Screen):
if input_widget:
input_widget.password = not input_widget.password
event.button.label = "🙈" if not input_widget.password else "👁"
elif event.button.id == "toggle-anthropic-key":
# Toggle Anthropic API key visibility
input_widget = self.inputs.get("anthropic_api_key")
if input_widget:
input_widget.password = not input_widget.password
event.button.label = "Hide" if not input_widget.password else "Show"
elif event.button.id == "toggle-watsonx-key":
# Toggle watsonx API key visibility
input_widget = self.inputs.get("watsonx_api_key")
if input_widget:
input_widget.password = not input_widget.password
event.button.label = "Hide" if not input_widget.password else "Show"
def action_generate(self) -> None:
"""Generate secure passwords for admin accounts."""

View file

@ -22,13 +22,6 @@ class WelcomeScreen(Screen):
BINDINGS = [
("q", "quit", "Quit"),
("enter", "default_action", "Continue"),
("1", "no_auth_setup", "Basic Setup"),
("2", "full_setup", "Advanced Setup"),
("3", "monitor", "Status"),
("4", "diagnostics", "Diagnostics"),
("5", "start_stop_services", "Start/Stop Services"),
("6", "open_app", "Open App"),
]
def __init__(self):
@ -41,6 +34,9 @@ class WelcomeScreen(Screen):
self.has_oauth_config = False
self.default_button_id = "basic-setup-btn"
self._state_checked = False
# Check if .env file exists
self.has_env_file = self.env_manager.env_file.exists()
# Load .env file if it exists
load_dotenv()
@ -161,6 +157,23 @@ class WelcomeScreen(Screen):
buttons = []
# If no .env file exists, only show setup buttons
if not self.has_env_file:
if has_oauth:
# If OAuth is configured, only show advanced setup
buttons.append(
Button("Advanced Setup", variant="success", id="advanced-setup-btn")
)
else:
# If no OAuth, show both options with basic as primary
buttons.append(
Button("Basic Setup", variant="success", id="basic-setup-btn")
)
buttons.append(
Button("Advanced Setup", variant="default", id="advanced-setup-btn")
)
return Horizontal(*buttons, classes="button-row")
# Check if all services (native + container) are running
all_services_running = self.services_running and self.docling_running
@ -189,7 +202,7 @@ class WelcomeScreen(Screen):
)
buttons.append(
Button("Start All Services", variant="primary", id="start-all-services-btn")
Button("Start OpenRAG", variant="primary", id="start-all-services-btn")
)
# Always show status option
@ -253,6 +266,9 @@ class WelcomeScreen(Screen):
async def on_screen_resume(self) -> None:
"""Called when returning from another screen (e.g., config screen)."""
# Check if .env file exists (may have been created)
self.has_env_file = self.env_manager.env_file.exists()
# Reload environment variables
load_dotenv(override=True)

View file

@ -63,6 +63,27 @@ def validate_openai_api_key(key: str) -> bool:
return key.startswith("sk-") and len(key) > 20
def validate_anthropic_api_key(key: str) -> bool:
"""Validate Anthropic API key format."""
if not key:
return False
return key.startswith("sk-ant-") and len(key) > 20
def validate_ollama_endpoint(endpoint: str) -> bool:
"""Validate Ollama endpoint URL format."""
if not endpoint:
return False
return validate_url(endpoint)
def validate_watsonx_endpoint(endpoint: str) -> bool:
"""Validate IBM watsonx.ai endpoint URL format."""
if not endpoint:
return False
return validate_url(endpoint)
def validate_google_oauth_client_id(client_id: str) -> bool:
"""Validate Google OAuth client ID format."""
if not client_id:

4
uv.lock generated
View file

@ -1,5 +1,5 @@
version = 1
revision = 2
revision = 3
requires-python = ">=3.13"
resolution-markers = [
"platform_machine == 'x86_64' and sys_platform == 'linux'",
@ -2352,7 +2352,7 @@ wheels = [
[[package]]
name = "openrag"
version = "0.1.35"
version = "0.1.37"
source = { editable = "." }
dependencies = [
{ name = "agentd" },