Chris CoutinhoandClaude Opus 4.8
68c9e20636
- Failed deletes no longer bump astrolabe_documents_indexed_total: the outer except in process_document now gates doc_type on operation != "delete", so a delete error is counted as processed-error but not as an indexing event. Added test_failed_delete_is_processed_but_not_indexed. - registry parse span: pass record_exception=True explicitly (matches instrument_tool) and add a structured logger.warning on the parse-error path (processor/tier/byte_size/duration_ms) for a Loki-aggregatable failed-parse signal. - test_error_does_not_increment_throughput: snapshot-before/delta pattern instead of absolute 0.0 (counters are global singletons). - config: document the deliberate gateway asymmetry between get_embedding_model_name() (no gateway branch) and get_embedding_provider_family() (short-circuits on gateway). - Cleanup in touched scope: narrow `except (HTTPStatusError, Exception)` to `except Exception` (drop now-unused import); convert registry signatures from Optional[...] to `... | None`. Refs Deck #175, PR #831. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
1405 lines
58 KiB
Python
1405 lines
58 KiB
Python
import atexit
|
|
import logging
|
|
import logging.config
|
|
import os
|
|
import socket
|
|
import ssl
|
|
import tempfile
|
|
from dataclasses import dataclass
|
|
from pathlib import Path
|
|
from typing import Any
|
|
|
|
from dynaconf import Dynaconf, Validator
|
|
|
|
# Sentinel for "key not in dynaconf at all" vs "explicitly set to None".
|
|
_UNSET = object()
|
|
|
|
# Built-in defaults — declared in Python so env vars work without any settings
|
|
# file being present (e.g., `uvx` / `pip install` deployments). Mirrors the
|
|
# [default] section that used to live in settings.toml. Keys set here are
|
|
# "known" to dynaconf, which is required because we run with
|
|
# ignore_unknown_envvars=True. See ADR-024/025.
|
|
_DEFAULTS: dict[str, Any] = {
|
|
# Deployment mode (ADR-021)
|
|
"mcp_deployment_mode": None,
|
|
# Nextcloud core
|
|
"nextcloud_host": None,
|
|
"nextcloud_username": None,
|
|
"nextcloud_password": None,
|
|
"nextcloud_app_password": None,
|
|
"nextcloud_verify_ssl": True,
|
|
"nextcloud_ca_bundle": None,
|
|
"nextcloud_mcp_server_url": None,
|
|
"nextcloud_resource_uri": None,
|
|
"nextcloud_public_issuer_url": None,
|
|
"cookie_secure": None,
|
|
# OAuth/OIDC
|
|
"oidc_discovery_url": None,
|
|
"nextcloud_oidc_client_id": None,
|
|
"nextcloud_oidc_client_secret": None,
|
|
"oidc_issuer": None,
|
|
"jwks_uri": None,
|
|
"introspection_uri": None,
|
|
"userinfo_uri": None,
|
|
"oidc_resource_server_id": None,
|
|
# Mode flags
|
|
# NOTE: `enable_multi_user_basic_auth` and `enable_login_flow` are
|
|
# intentionally absent — they are derived from MCP_DEPLOYMENT_MODE in
|
|
# Settings.__post_init__ (ADR-022) and not read from the dynaconf store.
|
|
"enable_semantic_search": False,
|
|
"enable_background_operations": False,
|
|
"vector_sync_enabled": False,
|
|
"enable_offline_access": False,
|
|
"enable_token_exchange": False,
|
|
# Token storage
|
|
"token_encryption_key": None,
|
|
# None = ephemeral per-process tempfile (see get_token_db_path()).
|
|
# Set TOKEN_STORAGE_DB to persist tokens across restarts.
|
|
"token_storage_db": None,
|
|
# Centralized backend (any SQLAlchemy URL). Wins over TOKEN_STORAGE_DB
|
|
# when set. Use postgresql+asyncpg://user:pw@host/db for HA k8s
|
|
# deployments so pods can be stateless. See ADR-026.
|
|
"database_url": None,
|
|
# TLS for the Postgres backend (mirror NEXTCLOUD_VERIFY_SSL pattern).
|
|
# Default is None — preserve asyncpg's `prefer` mode so cluster-local
|
|
# Postgres without TLS works out of the box. Set to True for full
|
|
# verification or False to silence cert errors against self-signed
|
|
# homelab servers. DATABASE_CA_BUNDLE points at a private-CA PEM.
|
|
"database_verify_ssl": None,
|
|
"database_ca_bundle": None,
|
|
# Postgres connection pool sizing (ADR-026 → "Concurrency model and
|
|
# pool sizing"). Per-pod defaults to 2 + 5 overflow = 7 max
|
|
# connections. asyncpg connections are single-flight, so the pool
|
|
# only needs to cover typical multi-user MCP burst — not every
|
|
# potential in-flight tool call. Tune up with DATABASE_POOL_SIZE /
|
|
# DATABASE_MAX_OVERFLOW for high-traffic prod fleets.
|
|
"database_pool_size": 2,
|
|
"database_max_overflow": 5,
|
|
# Webhook delivery authentication (ADR-010): when set, registrations
|
|
# tell NC to add `Authorization: Bearer <secret>` to webhook deliveries
|
|
# and the receiver rejects unauthenticated requests.
|
|
"webhook_secret": None,
|
|
# Internal URL override for webhook registration; wins over
|
|
# NEXTCLOUD_MCP_SERVER_URL when set (e.g. split internal/external URLs).
|
|
"webhook_internal_url": None,
|
|
# Vector sync
|
|
"vector_sync_scan_interval": 300,
|
|
"vector_sync_processor_workers": 3,
|
|
"vector_sync_queue_max_size": 10000,
|
|
"vector_sync_user_poll_interval": 60,
|
|
# Orphan-sweep at Pod startup (card #101). When True, delete any
|
|
# placeholders carrying a different / absent ``instance_id`` before
|
|
# the scanner's first cycle, so a Pod restart mid-batch doesn't
|
|
# leave work stuck behind the 5x-scan-interval staleness gate.
|
|
# Escape hatch only — leave on by default.
|
|
"vector_sync_orphan_sweep_enabled": True,
|
|
# Verify-on-read concurrency cap (ADR-019)
|
|
"verification_concurrency": 20,
|
|
# Qdrant
|
|
"qdrant_url": None,
|
|
"qdrant_location": None,
|
|
"qdrant_api_key": None,
|
|
"qdrant_collection": "nextcloud_content",
|
|
# Ollama
|
|
"ollama_base_url": None,
|
|
"ollama_embedding_model": "nomic-embed-text",
|
|
"ollama_generation_model": None,
|
|
"ollama_verify_ssl": True,
|
|
# OpenAI
|
|
"openai_api_key": None,
|
|
"openai_base_url": None,
|
|
"openai_embedding_model": "text-embedding-3-small",
|
|
"openai_generation_model": None,
|
|
# Bedrock (AWS)
|
|
"aws_region": None,
|
|
"aws_access_key_id": None,
|
|
"aws_secret_access_key": None,
|
|
"bedrock_embedding_model": None,
|
|
"bedrock_generation_model": None,
|
|
# Mistral
|
|
"mistral_api_key": None,
|
|
"mistral_embedding_model": "mistral-embed",
|
|
"mistral_base_url": None,
|
|
# Simple (fallback) embedding dimension
|
|
"simple_embedding_dimension": 384,
|
|
# Document chunking
|
|
"document_chunk_size": 2048,
|
|
"document_chunk_overlap": 200,
|
|
# Observability
|
|
"metrics_enabled": True,
|
|
"metrics_port": 9090,
|
|
"otel_exporter_otlp_endpoint": None,
|
|
"otel_exporter_verify_ssl": False,
|
|
"otel_service_name": "nextcloud-mcp-server",
|
|
"otel_traces_sampler": "always_on",
|
|
"otel_traces_sampler_arg": 1.0,
|
|
"log_format": "text",
|
|
"log_level": "INFO",
|
|
"log_include_trace_context": True,
|
|
# Document processing
|
|
"enable_document_processing": False,
|
|
"document_processor": "unstructured",
|
|
"enable_unstructured": False,
|
|
"unstructured_api_url": "http://unstructured:8000",
|
|
"unstructured_timeout": 120,
|
|
"unstructured_strategy": "auto",
|
|
"unstructured_languages": "eng,deu",
|
|
"progress_interval": 10,
|
|
"enable_tesseract": False,
|
|
"tesseract_cmd": None,
|
|
"tesseract_lang": "eng",
|
|
"enable_pymupdf": True,
|
|
"pymupdf_extract_images": True,
|
|
"pymupdf_image_dir": None,
|
|
"enable_custom_processor": False,
|
|
"custom_processor_url": None,
|
|
"custom_processor_types": "application/pdf",
|
|
"custom_processor_name": "custom",
|
|
"custom_processor_api_key": None,
|
|
"custom_processor_timeout": 60,
|
|
# Tag-based file exclusion (issue #710): comma-separated list of
|
|
# Nextcloud system tag names. Files/folders carrying any of these tags
|
|
# are hidden from WebDAV MCP tools. Empty = feature off.
|
|
"excluded_tags": "",
|
|
# MCP decomposition hook points (design §10). Every default reproduces
|
|
# the current monolithic behavior; self-hosters who set none are
|
|
# unaffected. See docs/architecture/mcp-decomposition.md (sibling repo).
|
|
"embedding_provider": "autodetect", # autodetect | gateway
|
|
"ingest_mode": "local", # local | external
|
|
"status_backend": "local", # local | bus
|
|
"collection_metadata_source": "qdrant", # qdrant | api
|
|
# CP base URL for COLLECTION_METADATA_SOURCE=api (e.g. http://control-plane).
|
|
# Required only when the source is api.
|
|
"collection_metadata_api_url": None,
|
|
"fact_event_emitter": "none", # none | nats | stdout
|
|
"ingest_bus_url": None, # required when ingest_mode=external
|
|
"embedding_gateway_url": None, # required when embedding_provider=gateway
|
|
# Provider-namespaced model the gateway serves, "<provider>/<model>"
|
|
# (the gateway routes on the "/"-prefix; mistral/mistral-embed → Mistral
|
|
# for the MVP). Only consulted when embedding_provider=gateway.
|
|
"embedding_gateway_model": "mistral/mistral-embed",
|
|
# Gateway auth: the MCP server is an OIDC *client* in the gateway's own
|
|
# M2M realm (parallel to, and distinct from, the tenant realm it already
|
|
# serves). It obtains a client-credentials token and the gateway maps the
|
|
# client-id → the tenant's underlying provider API key. All four unset =
|
|
# call the gateway unauthenticated (matches today's not-yet-authed gateway).
|
|
"embedding_gateway_token_url": None, # M2M token endpoint
|
|
"embedding_gateway_client_id": None,
|
|
"embedding_gateway_client_secret": None,
|
|
"embedding_gateway_scope": None, # e.g. astrolabe-embedding-gateway/embed
|
|
"tenant_id": None, # NATS per-tenant subject token (UUID form)
|
|
"ingest_bus_num_replicas": 1, # JetStream stream replicas (prod: 3)
|
|
# Query-side ACL pre-filter (design §11). OFF by default: a Qdrant
|
|
# `match any` on `acl_hash` excludes points missing the key, so enabling
|
|
# this before a real ACL backfill would silently drop legacy results.
|
|
# verify-on-read remains the correctness backstop regardless.
|
|
"acl_prefilter_enabled": False,
|
|
}
|
|
|
|
|
|
def _resolve_settings_files() -> list[str]:
|
|
"""Find optional external settings files.
|
|
|
|
Priority:
|
|
1. NEXTCLOUD_MCP_SETTINGS_FILE env var (absolute or relative path).
|
|
If set but the file does not exist, raise FileNotFoundError —
|
|
silently falling back to defaults on a typo would be a footgun.
|
|
.secrets.toml is looked for alongside the explicit file.
|
|
2. Otherwise ./settings.toml in cwd (for docker / dev workflows),
|
|
with .secrets.toml also looked for in cwd.
|
|
|
|
Returns an empty list if nothing is configured — that's fine, defaults
|
|
and env vars still apply.
|
|
"""
|
|
files: list[str] = []
|
|
explicit = os.environ.get("NEXTCLOUD_MCP_SETTINGS_FILE")
|
|
if explicit:
|
|
p = Path(explicit)
|
|
if not p.exists():
|
|
raise FileNotFoundError(
|
|
f"NEXTCLOUD_MCP_SETTINGS_FILE points to a file that does "
|
|
f"not exist: {explicit}"
|
|
)
|
|
files.append(str(p))
|
|
secrets = p.parent / ".secrets.toml"
|
|
else:
|
|
cwd_settings = Path.cwd() / "settings.toml"
|
|
if cwd_settings.exists():
|
|
files.append(str(cwd_settings))
|
|
secrets = Path.cwd() / ".secrets.toml"
|
|
if secrets.exists():
|
|
files.append(str(secrets))
|
|
return files
|
|
|
|
|
|
# Dynaconf instance — env vars always win (12-factor). Settings files are
|
|
# optional; when absent the defaults above provide the full key schema so
|
|
# env vars still override correctly. See ADR-024/025 for architecture.
|
|
_dynaconf = Dynaconf(
|
|
settings_files=_resolve_settings_files(),
|
|
environments=True,
|
|
envvar_prefix=False,
|
|
env_switcher="MCP_DEPLOYMENT_MODE",
|
|
ignore_unknown_envvars=True,
|
|
load_dotenv=False,
|
|
**_DEFAULTS,
|
|
validators=[
|
|
# Port ranges
|
|
Validator("METRICS_PORT", gte=1, lte=65535),
|
|
# Positive integers
|
|
Validator("VECTOR_SYNC_SCAN_INTERVAL", gte=1),
|
|
Validator("VECTOR_SYNC_PROCESSOR_WORKERS", gte=1),
|
|
Validator("VECTOR_SYNC_QUEUE_MAX_SIZE", gte=1),
|
|
Validator("VECTOR_SYNC_USER_POLL_INTERVAL", gte=1),
|
|
Validator("VERIFICATION_CONCURRENCY", gte=1),
|
|
Validator("DOCUMENT_CHUNK_SIZE", gte=1),
|
|
# Non-negative
|
|
Validator("DOCUMENT_CHUNK_OVERLAP", gte=0),
|
|
# Enum constraints
|
|
Validator("LOG_FORMAT", is_in=["text", "json"]),
|
|
Validator(
|
|
"LOG_LEVEL",
|
|
is_in=["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"],
|
|
),
|
|
Validator(
|
|
"OTEL_TRACES_SAMPLER",
|
|
is_in=[
|
|
"always_on",
|
|
"always_off",
|
|
"traceidratio",
|
|
"parentbased_always_on",
|
|
"parentbased_always_off",
|
|
"parentbased_traceidratio",
|
|
],
|
|
),
|
|
# Float ranges
|
|
Validator("OTEL_TRACES_SAMPLER_ARG", gte=0.0, lte=1.0),
|
|
],
|
|
)
|
|
|
|
|
|
def _reload_config():
|
|
"""Reload dynaconf settings from files and environment.
|
|
|
|
Call this in tests after modifying os.environ to refresh the cache.
|
|
Re-validates all validators since reload() only checks unchecked ones.
|
|
"""
|
|
_dynaconf.reload()
|
|
_dynaconf.validators.validate_all()
|
|
|
|
|
|
_ephemeral_db_path: str | None = None
|
|
|
|
|
|
def get_token_db_path() -> str:
|
|
"""Resolve the token SQLite database path.
|
|
|
|
Priority:
|
|
1. TOKEN_STORAGE_DB if explicitly set — docker-compose pins
|
|
/app/data/tokens.db this way. Read via dynaconf, which picks up
|
|
the env var because TOKEN_STORAGE_DB is declared in _DEFAULTS.
|
|
2. Otherwise a per-process tempfile under tempfile.gettempdir(),
|
|
allocated lazily and deleted at interpreter exit via atexit.
|
|
Ephemeral: tokens are wiped on restart, matching the Qdrant
|
|
":memory:" default pattern used elsewhere in this project.
|
|
"""
|
|
explicit = _dynaconf.get("TOKEN_STORAGE_DB")
|
|
if explicit:
|
|
return str(explicit)
|
|
global _ephemeral_db_path
|
|
if _ephemeral_db_path is None:
|
|
fd, path = tempfile.mkstemp(
|
|
prefix=f"nextcloud-mcp-tokens-{os.getpid()}-", suffix=".db"
|
|
)
|
|
os.close(fd)
|
|
_ephemeral_db_path = path
|
|
|
|
def _cleanup(p: str = path) -> None:
|
|
try:
|
|
if os.path.exists(p):
|
|
os.unlink(p)
|
|
except OSError:
|
|
pass
|
|
|
|
atexit.register(_cleanup)
|
|
return _ephemeral_db_path
|
|
|
|
|
|
def is_ephemeral_token_db(path: str) -> bool:
|
|
"""Return True if the given path is the process-local ephemeral tempfile.
|
|
|
|
Precondition: `get_token_db_path()` must have been called at least once
|
|
in this process to allocate the tempfile. If called before allocation,
|
|
this returns False for any input (including the eventual tempfile path),
|
|
because there is nothing to compare against yet. In practice every call
|
|
site in this repo resolves the path via `get_token_db_path()` first.
|
|
"""
|
|
return path == _ephemeral_db_path
|
|
|
|
|
|
def get_database_url() -> str:
|
|
"""Resolve the SQLAlchemy database URL for token storage.
|
|
|
|
Priority:
|
|
1. ``DATABASE_URL`` if set — any SQLAlchemy URL is accepted; the primary
|
|
supported backends are ``postgresql+asyncpg://...`` for HA k8s
|
|
deployments and ``sqlite+aiosqlite:///...`` for development.
|
|
2. Otherwise build ``sqlite+aiosqlite:///{get_token_db_path()}`` so the
|
|
legacy ``TOKEN_STORAGE_DB`` env var and the ephemeral-tempfile
|
|
fallback both keep working unchanged.
|
|
"""
|
|
explicit = _dynaconf.get("DATABASE_URL")
|
|
if explicit:
|
|
return str(explicit)
|
|
return f"sqlite+aiosqlite:///{get_token_db_path()}"
|
|
|
|
|
|
def is_sqlite_url(url: str) -> bool:
|
|
"""Return True for SQLite SQLAlchemy URLs (used to gate sqlite-only logic
|
|
like file-permission hardening and ``sqlite_master`` legacy lookups).
|
|
|
|
Recognizes both file-backed (``sqlite+aiosqlite:///path/to/db``) and
|
|
in-memory (``sqlite+aiosqlite:///:memory:``) URLs. The caller is
|
|
responsible for handling ``:memory:`` as a magic value where a real
|
|
filesystem path is expected.
|
|
"""
|
|
return url.lower().startswith("sqlite")
|
|
|
|
|
|
def mask_db_password(url: str) -> str:
|
|
"""Return a logger-safe rendering of a SQLAlchemy URL.
|
|
|
|
DATABASE_URL routinely carries a password (e.g.
|
|
``postgresql+asyncpg://mcp:secret@db/mcp``); logging it raw leaks the
|
|
secret to stdout/stderr and any aggregator. SQLAlchemy's
|
|
:func:`make_url` + ``render_as_string(hide_password=True)`` substitutes
|
|
a fixed ``***`` placeholder while keeping the rest of the URL intact
|
|
so operators can still see which host / driver they're hitting.
|
|
"""
|
|
try:
|
|
from sqlalchemy.engine.url import make_url # noqa: PLC0415
|
|
|
|
return make_url(url).render_as_string(hide_password=True)
|
|
except Exception:
|
|
# If parsing fails (e.g. an explicit ssl-disable test URL with an
|
|
# exotic shape), fall back to a regex that scrubs any
|
|
# ``://user:password@`` pattern. Never raise from a logging path.
|
|
import re # noqa: PLC0415
|
|
|
|
return re.sub(r"(://[^:/]+):[^@]*@", r"\1:***@", url)
|
|
|
|
|
|
LOGGING_CONFIG = {
|
|
"version": 1,
|
|
"disable_existing_loggers": False,
|
|
"handlers": {
|
|
"default": {
|
|
"class": "logging.StreamHandler",
|
|
"formatter": "http",
|
|
},
|
|
},
|
|
"formatters": {
|
|
"http": {
|
|
"format": "%(levelname)s [%(asctime)s] %(name)s - %(message)s",
|
|
"datefmt": "%Y-%m-%d %H:%M:%S",
|
|
},
|
|
},
|
|
"loggers": {
|
|
"": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
},
|
|
"httpx": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
"propagate": False, # Prevent propagation to root logger
|
|
},
|
|
"httpcore": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
"propagate": False, # Prevent propagation to root logger
|
|
},
|
|
"uvicorn": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
"propagate": False,
|
|
},
|
|
"uvicorn.access": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
"propagate": False,
|
|
},
|
|
"uvicorn.error": {
|
|
"handlers": ["default"],
|
|
"level": "INFO",
|
|
"propagate": False,
|
|
},
|
|
},
|
|
}
|
|
|
|
|
|
def setup_logging():
|
|
logging.config.dictConfig(LOGGING_CONFIG)
|
|
|
|
|
|
# Document Processing Configuration
|
|
|
|
|
|
def get_document_processor_config() -> dict[str, Any]:
|
|
"""Get document processor configuration from dynaconf.
|
|
|
|
Returns:
|
|
Dict with processor configs:
|
|
{
|
|
"enabled": bool,
|
|
"default_processor": str,
|
|
"processors": {
|
|
"unstructured": {...},
|
|
"tesseract": {...},
|
|
"custom": {...},
|
|
}
|
|
}
|
|
"""
|
|
config: dict[str, Any] = {
|
|
"enabled": _dynaconf.get("ENABLE_DOCUMENT_PROCESSING"),
|
|
"default_processor": _dynaconf.get("DOCUMENT_PROCESSOR"),
|
|
"processors": {},
|
|
}
|
|
|
|
# Unstructured configuration
|
|
if _dynaconf.get("ENABLE_UNSTRUCTURED"):
|
|
languages_str = _dynaconf.get("UNSTRUCTURED_LANGUAGES")
|
|
config["processors"]["unstructured"] = {
|
|
"api_url": _dynaconf.get("UNSTRUCTURED_API_URL"),
|
|
"timeout": _dynaconf.get("UNSTRUCTURED_TIMEOUT"),
|
|
"strategy": _dynaconf.get("UNSTRUCTURED_STRATEGY"),
|
|
"languages": [
|
|
lang.strip() for lang in languages_str.split(",") if lang.strip()
|
|
],
|
|
"progress_interval": _dynaconf.get("PROGRESS_INTERVAL"),
|
|
}
|
|
|
|
# Tesseract configuration
|
|
if _dynaconf.get("ENABLE_TESSERACT"):
|
|
config["processors"]["tesseract"] = {
|
|
"tesseract_cmd": _dynaconf.get("TESSERACT_CMD"), # None = auto-detect
|
|
"lang": _dynaconf.get("TESSERACT_LANG"),
|
|
}
|
|
|
|
# PyMuPDF configuration (local PDF processing)
|
|
if _dynaconf.get("ENABLE_PYMUPDF"): # Enabled by default
|
|
config["processors"]["pymupdf"] = {
|
|
"extract_images": _dynaconf.get("PYMUPDF_EXTRACT_IMAGES"),
|
|
"image_dir": _dynaconf.get(
|
|
"PYMUPDF_IMAGE_DIR"
|
|
), # None = use temp directory
|
|
}
|
|
|
|
# Custom processor (via HTTP API)
|
|
if _dynaconf.get("ENABLE_CUSTOM_PROCESSOR"):
|
|
custom_url = _dynaconf.get("CUSTOM_PROCESSOR_URL")
|
|
if custom_url:
|
|
supported_types_str = _dynaconf.get("CUSTOM_PROCESSOR_TYPES")
|
|
supported_types = {
|
|
t.strip() for t in supported_types_str.split(",") if t.strip()
|
|
}
|
|
|
|
config["processors"]["custom"] = {
|
|
"name": _dynaconf.get("CUSTOM_PROCESSOR_NAME"),
|
|
"api_url": custom_url,
|
|
"api_key": _dynaconf.get("CUSTOM_PROCESSOR_API_KEY"),
|
|
"timeout": _dynaconf.get("CUSTOM_PROCESSOR_TIMEOUT"),
|
|
"supported_types": supported_types,
|
|
}
|
|
|
|
return config
|
|
|
|
|
|
@dataclass
|
|
class Settings:
|
|
"""Application settings from environment variables."""
|
|
|
|
# Deployment mode (ADR-021: explicit mode selection; updated by ADR-022)
|
|
# Optional: If not set, mode is auto-detected from other settings
|
|
# Valid values: single_user_basic, multi_user_basic, login_flow
|
|
# (ADR-022: `oauth_single_audience` was renamed to `login_flow`.)
|
|
deployment_mode: str | None = None
|
|
|
|
# OAuth/OIDC settings
|
|
oidc_discovery_url: str | None = None
|
|
oidc_client_id: str | None = None
|
|
oidc_client_secret: str | None = None
|
|
oidc_issuer: str | None = None
|
|
oidc_resource_server_id: str | None = None
|
|
|
|
# Nextcloud settings
|
|
nextcloud_host: str | None = None
|
|
nextcloud_username: str | None = None
|
|
nextcloud_password: str | None = None
|
|
nextcloud_app_password: str | None = None # Preferred over nextcloud_password
|
|
|
|
# Browser-reachable public URL for OAuth/Login-Flow-v2 redirects when
|
|
# NEXTCLOUD_HOST is an internal Docker hostname. Falls back to
|
|
# nextcloud_host when unset.
|
|
nextcloud_public_issuer_url: str | None = None
|
|
|
|
# Browser cookie Secure flag. None = auto-detect from nextcloud_host
|
|
# scheme (https → True, else False). Set COOKIE_SECURE=true/false to
|
|
# override.
|
|
cookie_secure: bool | None = None
|
|
|
|
# Nextcloud SSL/TLS settings
|
|
nextcloud_verify_ssl: bool = True
|
|
nextcloud_ca_bundle: str | None = None
|
|
|
|
# Postgres backend TLS settings (ADR-026). Default verify_ssl is None,
|
|
# not True: when DATABASE_URL is unset there's nothing to verify, and
|
|
# when it is set we don't want to break cluster-internal Postgres that
|
|
# commonly runs without TLS. Operators opt in to verify-full with True
|
|
# or supply a private-CA bundle.
|
|
database_verify_ssl: bool | None = None
|
|
database_ca_bundle: str | None = None
|
|
# Postgres connection pool sizing — DEPRECATED, retained for
|
|
# backward compatibility. The asyncpg engine switched to NullPool
|
|
# in #799 (cross-event-loop crashes under anyio TaskGroups made
|
|
# the original QueuePool + pool_pre_ping setup unsafe). These
|
|
# fields no longer affect the Postgres engine; the validators
|
|
# below still reject invalid values so misconfigured deploys
|
|
# fail loudly rather than silently. See ADR-026 § Connection
|
|
# pool and docs/configuration.md.
|
|
database_pool_size: int = 2
|
|
database_max_overflow: int = 5
|
|
|
|
# ADR-005: Token Audience Validation (required for OAuth mode)
|
|
nextcloud_mcp_server_url: str | None = None # MCP server URL (used as audience)
|
|
nextcloud_resource_uri: str | None = None # Nextcloud resource identifier
|
|
|
|
# Token verification endpoints
|
|
jwks_uri: str | None = None
|
|
introspection_uri: str | None = None
|
|
userinfo_uri: str | None = None
|
|
|
|
# Progressive Consent settings (always enabled - no flag needed)
|
|
enable_offline_access: bool = False
|
|
|
|
# Multi-user BasicAuth pass-through mode (ADR-019 interim solution).
|
|
# Internal — not user-settable; the ENABLE_MULTI_USER_BASIC_AUTH env-var
|
|
# alias was removed in the ADR-022 follow-up. Auto-set by
|
|
# Settings.__post_init__ when MCP_DEPLOYMENT_MODE=multi_user_basic. When True,
|
|
# the MCP server extracts BasicAuth credentials from request headers and
|
|
# passes them through to Nextcloud APIs (no storage, stateless). Kept
|
|
# as a field for backward compat with the runtime call sites that read it.
|
|
enable_multi_user_basic_auth: bool = False
|
|
|
|
# Login Flow v2 derived flag (ADR-022). Internal — not user-settable.
|
|
# Auto-set by Settings.__post_init__ when the resolved deployment mode is
|
|
# LOGIN_FLOW. Kept as a field for backward compat with the runtime call
|
|
# sites that read it (app.py, context.py, scope_authorization.py).
|
|
enable_login_flow: bool = False
|
|
|
|
# Token and webhook storage settings
|
|
# TOKEN_ENCRYPTION_KEY: Optional - Only required for OAuth token storage operations.
|
|
# Webhook tracking works without encryption key.
|
|
# If set, must be a valid base64-encoded Fernet key (32 bytes).
|
|
# TOKEN_STORAGE_DB: Path to SQLite database for persistent storage.
|
|
# Used for webhook tracking (all modes) and OAuth token storage.
|
|
# Defaults to /tmp/tokens.db
|
|
token_encryption_key: str | None = None
|
|
token_storage_db: str | None = None
|
|
|
|
# Webhook delivery authentication (ADR-010).
|
|
# When set, the registrar passes Authorization: Bearer <secret> as the
|
|
# webhook authData and the receiver validates the same header on each
|
|
# delivery. When unset, registration uses authMethod="none" and the
|
|
# receiver accepts unauthenticated POSTs (backward-compatible).
|
|
webhook_secret: str | None = None
|
|
# Internal URL override for webhook registration. Highest-priority
|
|
# source for the URL we register with NC (above
|
|
# nextcloud_mcp_server_url and the docker-detection fallback).
|
|
webhook_internal_url: str | None = None
|
|
|
|
# Vector sync settings (ADR-007)
|
|
vector_sync_enabled: bool = False
|
|
vector_sync_scan_interval: int = 300 # seconds (5 minutes)
|
|
vector_sync_processor_workers: int = 3
|
|
vector_sync_queue_max_size: int = 10000
|
|
vector_sync_user_poll_interval: int = 60 # seconds - OAuth mode user discovery
|
|
vector_sync_orphan_sweep_enabled: bool = True # card #101
|
|
|
|
# Verify-on-read concurrency (ADR-019). Cap on parallel Nextcloud
|
|
# round-trips during search-result verification fan-out. Lower this if the
|
|
# Nextcloud backend struggles with the parallel load; raise it on a
|
|
# healthy connection to speed up large result pages.
|
|
verification_concurrency: int = 20
|
|
|
|
# Qdrant settings (mutually exclusive modes)
|
|
qdrant_url: str | None = None # Network mode: http://qdrant:6333
|
|
qdrant_location: str | None = None # Local mode: :memory: or /path/to/data
|
|
qdrant_api_key: str | None = None
|
|
qdrant_collection: str = "nextcloud_content"
|
|
|
|
# Ollama settings (embeddings + optional generation)
|
|
ollama_base_url: str | None = None
|
|
ollama_embedding_model: str = "nomic-embed-text"
|
|
ollama_generation_model: str | None = None
|
|
ollama_verify_ssl: bool = True
|
|
|
|
# OpenAI settings (embeddings + optional generation)
|
|
openai_api_key: str | None = None
|
|
openai_base_url: str | None = None
|
|
openai_embedding_model: str = "text-embedding-3-small"
|
|
openai_generation_model: str | None = None
|
|
|
|
# Bedrock (AWS) settings — boto3 also reads these from its credential chain
|
|
aws_region: str | None = None
|
|
aws_access_key_id: str | None = None
|
|
aws_secret_access_key: str | None = None
|
|
bedrock_embedding_model: str | None = None
|
|
bedrock_generation_model: str | None = None
|
|
|
|
# Mistral settings (embeddings only)
|
|
mistral_api_key: str | None = None
|
|
mistral_embedding_model: str = "mistral-embed"
|
|
mistral_base_url: str | None = None
|
|
|
|
# Simple (fallback) provider — dimension when no real provider configured
|
|
simple_embedding_dimension: int = 384
|
|
|
|
# Document chunking settings (for vector embeddings)
|
|
document_chunk_size: int = 2048 # Characters per chunk
|
|
document_chunk_overlap: int = 200 # Overlapping characters between chunks
|
|
|
|
# Observability settings
|
|
metrics_enabled: bool = True
|
|
metrics_port: int = 9090
|
|
otel_exporter_otlp_endpoint: str | None = None
|
|
otel_exporter_verify_ssl: bool = False
|
|
otel_service_name: str = "nextcloud-mcp-server"
|
|
otel_traces_sampler: str = "always_on"
|
|
otel_traces_sampler_arg: float = 1.0
|
|
log_format: str = "text" # "json" or "text"
|
|
log_level: str = "INFO"
|
|
log_include_trace_context: bool = True
|
|
|
|
# Tag-based file exclusion (issue #710): comma-separated list of
|
|
# Nextcloud system tag names. Files/folders carrying any of these tags
|
|
# are hidden from WebDAV MCP tools.
|
|
excluded_tags: str = ""
|
|
|
|
# MCP decomposition hook points (design §10, opt-in). All defaults
|
|
# reproduce the current monolith; validated in __post_init__.
|
|
embedding_provider: str = "autodetect" # autodetect | gateway
|
|
ingest_mode: str = "local" # local | external
|
|
status_backend: str = "local" # local | bus
|
|
collection_metadata_source: str = "qdrant" # qdrant | api
|
|
collection_metadata_api_url: str | None = None # CP URL when source=api
|
|
fact_event_emitter: str = "none" # none | nats | stdout
|
|
ingest_bus_url: str | None = None # required when ingest_mode=external
|
|
embedding_gateway_url: str | None = None # required when provider=gateway
|
|
embedding_gateway_model: str = (
|
|
"mistral/mistral-embed" # provider-namespaced id the gateway routes on
|
|
)
|
|
# Gateway M2M OIDC client creds (separate realm; see _DEFAULTS comment).
|
|
embedding_gateway_token_url: str | None = None
|
|
embedding_gateway_client_id: str | None = None
|
|
embedding_gateway_client_secret: str | None = None
|
|
embedding_gateway_scope: str | None = None
|
|
tenant_id: str | None = None # NATS per-tenant subject token (UUID form)
|
|
ingest_bus_num_replicas: int = 1 # JetStream stream replicas (prod: 3)
|
|
acl_prefilter_enabled: bool = False # query-side ACL pre-filter (§11); OFF
|
|
|
|
def __post_init__(self):
|
|
"""Validate configuration and set defaults."""
|
|
logger = logging.getLogger(__name__)
|
|
|
|
# Validate SSL/TLS configuration
|
|
if not self.nextcloud_verify_ssl:
|
|
logger.warning(
|
|
"NEXTCLOUD_VERIFY_SSL is disabled. "
|
|
"TLS certificate verification is turned off for all Nextcloud connections. "
|
|
"This is insecure and should only be used for development/testing."
|
|
)
|
|
if self.nextcloud_ca_bundle:
|
|
if not os.path.isfile(self.nextcloud_ca_bundle):
|
|
raise ValueError(
|
|
f"NEXTCLOUD_CA_BUNDLE path does not exist: {self.nextcloud_ca_bundle}"
|
|
)
|
|
logger.info("Using custom CA bundle: %s", self.nextcloud_ca_bundle)
|
|
|
|
# Validate Postgres backend TLS configuration (ADR-026)
|
|
if self.database_verify_ssl is False:
|
|
logger.warning(
|
|
"DATABASE_VERIFY_SSL is disabled. "
|
|
"TLS certificate verification is turned off for the Postgres "
|
|
"backend. Only acceptable for homelab / self-signed setups; "
|
|
"prefer DATABASE_CA_BUNDLE for production."
|
|
)
|
|
if self.database_ca_bundle:
|
|
if not os.path.isfile(self.database_ca_bundle):
|
|
raise ValueError(
|
|
f"DATABASE_CA_BUNDLE path does not exist: {self.database_ca_bundle}"
|
|
)
|
|
logger.info(
|
|
"Using custom CA bundle for Postgres backend: %s",
|
|
self.database_ca_bundle,
|
|
)
|
|
|
|
# Pool sizing must be sensible — guard against operators accidentally
|
|
# setting 0 / negative via env (would deadlock at first request).
|
|
if self.database_pool_size < 1:
|
|
raise ValueError(
|
|
f"DATABASE_POOL_SIZE must be >= 1; got {self.database_pool_size}"
|
|
)
|
|
if self.database_max_overflow < 0:
|
|
raise ValueError(
|
|
f"DATABASE_MAX_OVERFLOW must be >= 0; got {self.database_max_overflow}"
|
|
)
|
|
|
|
# Ensure mutual exclusivity
|
|
if self.qdrant_url and self.qdrant_location:
|
|
raise ValueError(
|
|
"Cannot set both QDRANT_URL and QDRANT_LOCATION. "
|
|
"Use QDRANT_URL for network mode or QDRANT_LOCATION for local mode."
|
|
)
|
|
|
|
# Default to :memory: if neither set
|
|
if not self.qdrant_url and not self.qdrant_location:
|
|
self.qdrant_location = ":memory:"
|
|
logger.debug("Using default Qdrant mode: in-memory (:memory:)")
|
|
|
|
# Warn if API key set in local mode
|
|
if self.qdrant_location and self.qdrant_api_key:
|
|
logger.warning(
|
|
"QDRANT_API_KEY is set but QDRANT_LOCATION is used (local mode). "
|
|
"API key is only relevant for network mode and will be ignored."
|
|
)
|
|
|
|
# Validate chunking configuration
|
|
if self.document_chunk_overlap >= self.document_chunk_size:
|
|
raise ValueError(
|
|
f"DOCUMENT_CHUNK_OVERLAP ({self.document_chunk_overlap}) must be less than "
|
|
f"DOCUMENT_CHUNK_SIZE ({self.document_chunk_size}). "
|
|
f"Overlap should be 10-20% of chunk size for optimal results."
|
|
)
|
|
|
|
if self.document_chunk_size < 512:
|
|
logger.warning(
|
|
"DOCUMENT_CHUNK_SIZE is set to %s characters, which is quite small. Smaller chunks may lose context. Consider using at least 1024 characters.",
|
|
self.document_chunk_size,
|
|
)
|
|
|
|
# --- MCP decomposition hook points (design §10) ---
|
|
# Normalize + validate the opt-in enum settings. Defaults reproduce
|
|
# the monolith, so deployments that set none of these pass through.
|
|
_enum_fields = {
|
|
"embedding_provider": {"autodetect", "gateway"},
|
|
"ingest_mode": {"local", "external"},
|
|
"status_backend": {"local", "bus"},
|
|
"collection_metadata_source": {"qdrant", "api"},
|
|
"fact_event_emitter": {"none", "nats", "stdout"},
|
|
}
|
|
for _field, _allowed in _enum_fields.items():
|
|
_val = (getattr(self, _field) or "").strip().lower()
|
|
setattr(self, _field, _val)
|
|
if _val not in _allowed:
|
|
raise ValueError(
|
|
f"{_field.upper()} must be one of {sorted(_allowed)}; got {_val!r}"
|
|
)
|
|
|
|
# Fail-fast: external ingest sources its status from the bus. With the
|
|
# in-process state machine empty, STATUS_BACKEND=local would leave
|
|
# status streams silently empty — crash loudly instead (design §10.1).
|
|
if self.status_backend == "local" and self.ingest_mode == "external":
|
|
raise RuntimeError(
|
|
"STATUS_BACKEND=local is incompatible with INGEST_MODE=external; "
|
|
"set STATUS_BACKEND=bus"
|
|
)
|
|
|
|
# Conditional-required settings for the active hook points.
|
|
if self.ingest_mode == "external":
|
|
if not self.ingest_bus_url:
|
|
raise ValueError("INGEST_BUS_URL is required when INGEST_MODE=external")
|
|
if not self.tenant_id:
|
|
raise ValueError("TENANT_ID is required when INGEST_MODE=external")
|
|
if self.embedding_provider == "gateway" and not self.embedding_gateway_url:
|
|
raise ValueError(
|
|
"EMBEDDING_GATEWAY_URL is required when EMBEDDING_PROVIDER=gateway"
|
|
)
|
|
if (
|
|
self.collection_metadata_source == "api"
|
|
and not self.collection_metadata_api_url
|
|
):
|
|
raise ValueError(
|
|
"COLLECTION_METADATA_API_URL is required when "
|
|
"COLLECTION_METADATA_SOURCE=api"
|
|
)
|
|
|
|
# Gateway M2M OIDC creds are all-or-nothing: a partial set (e.g. a
|
|
# client_id with no token endpoint) is a misconfiguration that would
|
|
# silently fall back to unauthenticated calls. scope is optional.
|
|
_gw_creds = (
|
|
self.embedding_gateway_token_url,
|
|
self.embedding_gateway_client_id,
|
|
self.embedding_gateway_client_secret,
|
|
)
|
|
if any(_gw_creds) and not all(_gw_creds):
|
|
raise ValueError(
|
|
"EMBEDDING_GATEWAY_TOKEN_URL, EMBEDDING_GATEWAY_CLIENT_ID, and "
|
|
"EMBEDDING_GATEWAY_CLIENT_SECRET must be set together (M2M OIDC "
|
|
"client-credentials) or all left unset (unauthenticated gateway)"
|
|
)
|
|
|
|
# TENANT_ID is a NATS subject token; '.', '*', '>', and whitespace are
|
|
# reserved/illegal there and would silently break subscriptions (§3.4).
|
|
if self.tenant_id and (
|
|
any(c in self.tenant_id for c in ".*>")
|
|
or any(c.isspace() for c in self.tenant_id)
|
|
):
|
|
raise ValueError(
|
|
"TENANT_ID must not contain '.', '*', '>', or whitespace "
|
|
"(it is used as a NATS subject token)"
|
|
)
|
|
|
|
if self.ingest_bus_num_replicas < 1:
|
|
raise ValueError(
|
|
f"INGEST_BUS_NUM_REPLICAS must be >= 1; "
|
|
f"got {self.ingest_bus_num_replicas}"
|
|
)
|
|
|
|
# --- ADR-022 follow-up: deployment mode is the single source of truth ---
|
|
# The ENABLE_MULTI_USER_BASIC_AUTH and ENABLE_LOGIN_FLOW env vars were
|
|
# removed in favour of MCP_DEPLOYMENT_MODE. We do TWO things here:
|
|
#
|
|
# 1. Loud-fail if a user still has either legacy env var set to a
|
|
# truthy value (silent removal would have flipped them into the
|
|
# wrong runtime mode). Only fires for truthy strings, so an
|
|
# explicit `ENABLE_LOGIN_FLOW=false` in a leftover .env passes
|
|
# through harmlessly.
|
|
# 2. Derive `enable_login_flow` and `enable_multi_user_basic_auth`
|
|
# from the resolved deployment mode here, in __post_init__, so
|
|
# every Settings instance carries correct flags. (`get_settings()`
|
|
# builds a fresh Settings on each call — without this, the
|
|
# mutation that used to live in detect_auth_mode would only stick
|
|
# on the startup Settings instance, leaving per-request handlers
|
|
# with default False values.)
|
|
_truthy = {"1", "true", "yes", "on"}
|
|
for _legacy, _replacement in (
|
|
("ENABLE_MULTI_USER_BASIC_AUTH", "multi_user_basic"),
|
|
("ENABLE_LOGIN_FLOW", "login_flow"),
|
|
):
|
|
if os.environ.get(_legacy, "").strip().lower() in _truthy:
|
|
raise ValueError(
|
|
f"{_legacy} is no longer read from the environment. "
|
|
f"Set MCP_DEPLOYMENT_MODE={_replacement} instead "
|
|
"(ADR-022). The deployment mode is the single source "
|
|
"of truth for selecting an auth flow."
|
|
)
|
|
|
|
# NOTE: this block mirrors the resolution logic in
|
|
# `config_validators.detect_auth_mode` (which works on strings via a
|
|
# `mode_map`). Both call sites resolve the deployment mode
|
|
# independently — the canonical AuthMode enum in detect_auth_mode,
|
|
# and the boolean derived flags here. **Keep them in sync when
|
|
# adding a new mode**: a new entry must be added in both places, in
|
|
# addition to `mode_map` (`config_validators.py`) and any
|
|
# MODE_REQUIREMENTS entry.
|
|
resolved_mode = (self.deployment_mode or "").strip().lower()
|
|
if not resolved_mode:
|
|
if self.nextcloud_username and self.nextcloud_password:
|
|
resolved_mode = "single_user_basic"
|
|
else:
|
|
# Default multi-user mode is Login Flow v2 (browser-based
|
|
# app-password acquisition); the un-augmented OAuth bearer
|
|
# pass-through it replaced needed unmerged Nextcloud
|
|
# user_oidc patches and is no longer supported.
|
|
resolved_mode = "login_flow"
|
|
|
|
self.enable_multi_user_basic_auth = resolved_mode == "multi_user_basic"
|
|
self.enable_login_flow = resolved_mode == "login_flow"
|
|
|
|
def get_embedding_model_name(self) -> str:
|
|
"""
|
|
Get the active embedding model name based on provider priority.
|
|
|
|
Priority order (same as ProviderRegistry):
|
|
1. Bedrock - if AWS_REGION or BEDROCK_EMBEDDING_MODEL is set
|
|
2. OpenAI - if OPENAI_API_KEY is set
|
|
3. Mistral - if MISTRAL_API_KEY is set
|
|
4. Ollama - if OLLAMA_BASE_URL is set
|
|
5. Simple - fallback (returns "simple-{dimension}")
|
|
|
|
Returns:
|
|
Active embedding model name
|
|
"""
|
|
# NOTE: there is intentionally no "gateway" branch here. When
|
|
# EMBEDDING_PROVIDER=gateway this falls through to the underlying
|
|
# provider's model (used for the Qdrant collection name), whereas
|
|
# get_embedding_provider_family() short-circuits to the gateway-routed
|
|
# family. Keep that asymmetry in mind before joining metrics/labels
|
|
# derived from these two methods.
|
|
if (
|
|
self.aws_region
|
|
or self.bedrock_embedding_model
|
|
or self.bedrock_generation_model
|
|
):
|
|
return self.bedrock_embedding_model or "bedrock-default"
|
|
|
|
if self.openai_api_key:
|
|
return self.openai_embedding_model
|
|
|
|
if self.mistral_api_key:
|
|
return self.mistral_embedding_model
|
|
|
|
if self.ollama_base_url:
|
|
return self.ollama_embedding_model
|
|
|
|
return f"simple-{self.simple_embedding_dimension}"
|
|
|
|
def get_embedding_provider_family(self) -> str:
|
|
"""
|
|
Get the active dense-embedding provider family (a low-cardinality label).
|
|
|
|
This is the single source of truth for the ``provider`` metric label and
|
|
the ``embedding.provider`` span attribute. It returns the provider
|
|
*family* (e.g. "bedrock"), never the model name, to keep metric
|
|
cardinality bounded.
|
|
|
|
Priority mirrors ``get_embedding_model_name`` / ProviderRegistry:
|
|
1. Gateway - if EMBEDDING_PROVIDER=gateway (family from the model prefix,
|
|
e.g. "mistral/mistral-embed" -> "mistral")
|
|
2. Bedrock - if AWS_REGION or BEDROCK_EMBEDDING_MODEL is set
|
|
3. OpenAI - if OPENAI_API_KEY is set
|
|
4. Mistral - if MISTRAL_API_KEY is set
|
|
5. Ollama - if OLLAMA_BASE_URL is set
|
|
6. Simple - fallback
|
|
|
|
Returns:
|
|
Provider family: gateway-routed family | bedrock | openai | mistral
|
|
| ollama | simple
|
|
"""
|
|
if self.embedding_provider == "gateway":
|
|
model = self.embedding_gateway_model or ""
|
|
return model.split("/", 1)[0] if "/" in model else "gateway"
|
|
|
|
if (
|
|
self.aws_region
|
|
or self.bedrock_embedding_model
|
|
or self.bedrock_generation_model
|
|
):
|
|
return "bedrock"
|
|
|
|
if self.openai_api_key:
|
|
return "openai"
|
|
|
|
if self.mistral_api_key:
|
|
return "mistral"
|
|
|
|
if self.ollama_base_url:
|
|
return "ollama"
|
|
|
|
return "simple"
|
|
|
|
def get_collection_name(self) -> str:
|
|
"""
|
|
Get Qdrant collection name.
|
|
|
|
Auto-generates from deployment ID + model name unless explicitly set.
|
|
Deployment ID uses OTEL_SERVICE_NAME if configured, otherwise hostname.
|
|
|
|
This enables:
|
|
- Safe embedding model switching (new model → new collection)
|
|
- Multi-server deployments (unique deployment IDs)
|
|
- Clear collection naming (shows deployment and model)
|
|
|
|
Format: {deployment-id}-{model-name}
|
|
|
|
Examples:
|
|
- "my-deployment-nomic-embed-text" (Ollama)
|
|
- "my-deployment-text-embedding-3-small" (OpenAI)
|
|
- "mcp-container-openai-text-embedding-3-small" (hostname fallback)
|
|
|
|
Returns:
|
|
Collection name string
|
|
"""
|
|
|
|
# Use explicit override if user configured non-default value
|
|
if self.qdrant_collection != "nextcloud_content":
|
|
return self.qdrant_collection
|
|
|
|
# Determine deployment ID (OTEL service name or hostname fallback)
|
|
if self.otel_service_name != "nextcloud-mcp-server": # Non-default
|
|
deployment_id = self.otel_service_name
|
|
else:
|
|
# Fallback to hostname for simple Docker deployments without OTEL config
|
|
deployment_id = socket.gethostname()
|
|
|
|
# Sanitize deployment ID and model name
|
|
deployment_id = deployment_id.lower().replace(" ", "-").replace("_", "-")
|
|
model_name = self.get_embedding_model_name().replace("/", "-").replace(":", "-")
|
|
|
|
return f"{deployment_id}-{model_name}"
|
|
|
|
# ADR-021: Property aliases for new naming convention
|
|
# These provide the new names while maintaining backward compatibility with old field names
|
|
|
|
@property
|
|
def enable_semantic_search(self) -> bool:
|
|
"""Semantic search enabled (ADR-021 alias for vector_sync_enabled)."""
|
|
return self.vector_sync_enabled
|
|
|
|
@property
|
|
def enable_background_operations(self) -> bool:
|
|
"""Background operations enabled (ADR-021 alias for enable_offline_access)."""
|
|
return self.enable_offline_access
|
|
|
|
|
|
def _get_semantic_search_enabled() -> bool:
|
|
"""Get semantic search enabled status, supporting both old and new variable names.
|
|
|
|
Supports:
|
|
- ENABLE_SEMANTIC_SEARCH (new, preferred)
|
|
- VECTOR_SYNC_ENABLED (old, deprecated)
|
|
|
|
Returns:
|
|
True if semantic search should be enabled
|
|
"""
|
|
logger = logging.getLogger(__name__)
|
|
|
|
new_value = _dynaconf.get("ENABLE_SEMANTIC_SEARCH", False)
|
|
old_value = _dynaconf.get("VECTOR_SYNC_ENABLED", False)
|
|
|
|
if new_value and old_value:
|
|
logger.warning(
|
|
"Both ENABLE_SEMANTIC_SEARCH and VECTOR_SYNC_ENABLED are set. "
|
|
"Using ENABLE_SEMANTIC_SEARCH. "
|
|
"VECTOR_SYNC_ENABLED is deprecated and will be removed in v1.0.0."
|
|
)
|
|
elif old_value and not new_value:
|
|
logger.warning(
|
|
"VECTOR_SYNC_ENABLED is deprecated. "
|
|
"Please use ENABLE_SEMANTIC_SEARCH instead. "
|
|
"Support for VECTOR_SYNC_ENABLED will be removed in v1.0.0."
|
|
)
|
|
|
|
return new_value or old_value
|
|
|
|
|
|
def _is_multi_user_mode() -> bool:
|
|
"""Detect if this is a multi-user deployment mode.
|
|
|
|
Runs early in config setup (before Settings is fully built) for
|
|
mode-conditional defaults. Must match the canonical detection in
|
|
`config_validators.detect_auth_mode`, but works directly against the
|
|
raw dynaconf store since Settings doesn't exist yet.
|
|
|
|
Multi-user modes are:
|
|
- Multi-user BasicAuth (MCP_DEPLOYMENT_MODE=multi_user_basic)
|
|
- Login Flow v2 / default OAuth (MCP_DEPLOYMENT_MODE=login_flow, or no
|
|
username/password and no explicit mode)
|
|
- OAuth Token Exchange (ENABLE_TOKEN_EXCHANGE=true)
|
|
|
|
Single-user mode is:
|
|
- Single-user BasicAuth (username and password both set)
|
|
|
|
Returns:
|
|
True if multi-user mode detected
|
|
"""
|
|
# Explicit deployment mode wins. The ENABLE_MULTI_USER_BASIC_AUTH env-var
|
|
# alias was removed in the ADR-022 follow-up; selection is now via
|
|
# MCP_DEPLOYMENT_MODE.
|
|
explicit_mode = str(_dynaconf.get("MCP_DEPLOYMENT_MODE", "") or "").lower().strip()
|
|
if explicit_mode in {"multi_user_basic", "login_flow"}:
|
|
return True
|
|
if explicit_mode == "single_user_basic":
|
|
return False
|
|
|
|
# Token exchange implies OAuth multi-user
|
|
if _dynaconf.get("ENABLE_TOKEN_EXCHANGE", False):
|
|
return True
|
|
|
|
# If both username and password are set, it's single-user BasicAuth
|
|
has_username = bool(_dynaconf.get("NEXTCLOUD_USERNAME"))
|
|
has_password = bool(_dynaconf.get("NEXTCLOUD_PASSWORD"))
|
|
if has_username and has_password:
|
|
return False
|
|
|
|
# Otherwise, assume multi-user (default when no credentials provided)
|
|
return True
|
|
|
|
|
|
# Per-process guard for the three advisory log messages emitted by
|
|
# `_get_background_operations_enabled()`. The function runs on every
|
|
# `get_settings()` call (per ADR-024 / dynaconf design `get_settings()` is
|
|
# intentionally non-cached), so unguarded `logger.info`/`logger.warning`
|
|
# calls spam every MCP tool invocation. Mirrors the precedent at
|
|
# `nextcloud_mcp_server/vector/webhook_receiver.py:_warn_missing_secret_once`.
|
|
_bg_ops_advisories_logged: bool = False
|
|
|
|
|
|
def _log_bg_ops_advisories_once(
|
|
explicit: bool, legacy: bool, auto_enabled: bool
|
|
) -> None:
|
|
"""Emit ENABLE_BACKGROUND_OPERATIONS advisory logs at most once per process."""
|
|
global _bg_ops_advisories_logged
|
|
if _bg_ops_advisories_logged:
|
|
return
|
|
_bg_ops_advisories_logged = True
|
|
|
|
logger = logging.getLogger(__name__)
|
|
if explicit and legacy:
|
|
logger.warning(
|
|
"Both ENABLE_BACKGROUND_OPERATIONS and ENABLE_OFFLINE_ACCESS are set. "
|
|
"Using ENABLE_BACKGROUND_OPERATIONS. "
|
|
"ENABLE_OFFLINE_ACCESS is deprecated and will be removed in v1.0.0."
|
|
)
|
|
elif legacy and not explicit:
|
|
logger.warning(
|
|
"ENABLE_OFFLINE_ACCESS is deprecated. "
|
|
"Please use ENABLE_BACKGROUND_OPERATIONS instead. "
|
|
"Support for ENABLE_OFFLINE_ACCESS will be removed in v1.0.0."
|
|
)
|
|
if auto_enabled and not (explicit or legacy):
|
|
logger.info(
|
|
"Automatically enabled background operations for semantic search in multi-user mode. "
|
|
"Set ENABLE_BACKGROUND_OPERATIONS=false to disable (this will also disable semantic search)."
|
|
)
|
|
|
|
|
|
def _get_background_operations_enabled() -> bool:
|
|
"""Get background operations enabled status with auto-enablement for semantic search.
|
|
|
|
Supports:
|
|
- ENABLE_BACKGROUND_OPERATIONS (new, preferred)
|
|
- ENABLE_OFFLINE_ACCESS (old, deprecated)
|
|
- Auto-enabled if ENABLE_SEMANTIC_SEARCH=true in multi-user modes
|
|
|
|
Returns:
|
|
True if background operations should be enabled
|
|
"""
|
|
explicit = _dynaconf.get("ENABLE_BACKGROUND_OPERATIONS", False)
|
|
legacy = _dynaconf.get("ENABLE_OFFLINE_ACCESS", False)
|
|
semantic_search_enabled = _get_semantic_search_enabled()
|
|
is_multi_user = _is_multi_user_mode()
|
|
auto_enabled = semantic_search_enabled and is_multi_user
|
|
|
|
_log_bg_ops_advisories_once(explicit, legacy, auto_enabled)
|
|
|
|
return explicit or legacy or auto_enabled
|
|
|
|
|
|
def _dget(key):
|
|
"""Get a value from dynaconf if configured, otherwise return _UNSET.
|
|
|
|
Distinguishes "explicitly set to None" (via @none in TOML or env var)
|
|
from "not configured at all". When _UNSET is returned, callers should
|
|
let the Settings dataclass default apply.
|
|
"""
|
|
return _dynaconf[key] if key in _dynaconf else _UNSET
|
|
|
|
|
|
def get_settings() -> Settings:
|
|
"""Get application settings from dynaconf configuration.
|
|
|
|
Settings are loaded from (last wins):
|
|
1. settings.toml [default] section
|
|
2. settings.toml [<mode>] section (via MCP_DEPLOYMENT_MODE)
|
|
3. .secrets.toml (if present)
|
|
4. settings.local.toml (if present)
|
|
5. Environment variables (highest priority)
|
|
|
|
Values not found in any source are omitted, letting Settings dataclass
|
|
defaults apply. This ensures the server starts correctly even without
|
|
settings.toml (e.g., env-var-only deployments).
|
|
|
|
Returns:
|
|
Settings object with configuration values
|
|
"""
|
|
# Get consolidated values with smart dependency resolution
|
|
enable_semantic_search = _get_semantic_search_enabled()
|
|
enable_background_operations = _get_background_operations_enabled()
|
|
|
|
# Mapping from Settings field name to dynaconf key
|
|
_field_map = {
|
|
# Deployment mode (ADR-021)
|
|
"deployment_mode": "MCP_DEPLOYMENT_MODE",
|
|
# OAuth/OIDC settings
|
|
"oidc_discovery_url": "OIDC_DISCOVERY_URL",
|
|
"oidc_client_id": "NEXTCLOUD_OIDC_CLIENT_ID",
|
|
"oidc_client_secret": "NEXTCLOUD_OIDC_CLIENT_SECRET",
|
|
"oidc_issuer": "OIDC_ISSUER",
|
|
"oidc_resource_server_id": "OIDC_RESOURCE_SERVER_ID",
|
|
# Nextcloud settings
|
|
"nextcloud_host": "NEXTCLOUD_HOST",
|
|
"nextcloud_username": "NEXTCLOUD_USERNAME",
|
|
"nextcloud_password": "NEXTCLOUD_PASSWORD",
|
|
"nextcloud_app_password": "NEXTCLOUD_APP_PASSWORD",
|
|
"nextcloud_public_issuer_url": "NEXTCLOUD_PUBLIC_ISSUER_URL",
|
|
"cookie_secure": "COOKIE_SECURE",
|
|
# Nextcloud SSL/TLS settings
|
|
"nextcloud_verify_ssl": "NEXTCLOUD_VERIFY_SSL",
|
|
"nextcloud_ca_bundle": "NEXTCLOUD_CA_BUNDLE",
|
|
# Postgres backend TLS (ADR-026)
|
|
"database_verify_ssl": "DATABASE_VERIFY_SSL",
|
|
"database_ca_bundle": "DATABASE_CA_BUNDLE",
|
|
# Postgres backend pool sizing (ADR-026)
|
|
"database_pool_size": "DATABASE_POOL_SIZE",
|
|
"database_max_overflow": "DATABASE_MAX_OVERFLOW",
|
|
# ADR-005: Token Audience Validation
|
|
"nextcloud_mcp_server_url": "NEXTCLOUD_MCP_SERVER_URL",
|
|
"nextcloud_resource_uri": "NEXTCLOUD_RESOURCE_URI",
|
|
# Token verification endpoints
|
|
"jwks_uri": "JWKS_URI",
|
|
"introspection_uri": "INTROSPECTION_URI",
|
|
"userinfo_uri": "USERINFO_URI",
|
|
# NOTE: `enable_multi_user_basic_auth` and `enable_login_flow` no
|
|
# longer have env-var aliases — both are derived from the resolved
|
|
# MCP_DEPLOYMENT_MODE in detect_auth_mode() so users only configure
|
|
# the mode (ADR-022 follow-up).
|
|
# Token and webhook storage settings
|
|
"token_encryption_key": "TOKEN_ENCRYPTION_KEY",
|
|
"token_storage_db": "TOKEN_STORAGE_DB",
|
|
# Webhook auth (ADR-010)
|
|
"webhook_secret": "WEBHOOK_SECRET",
|
|
"webhook_internal_url": "WEBHOOK_INTERNAL_URL",
|
|
# Vector sync settings (ADR-007)
|
|
"vector_sync_scan_interval": "VECTOR_SYNC_SCAN_INTERVAL",
|
|
"vector_sync_processor_workers": "VECTOR_SYNC_PROCESSOR_WORKERS",
|
|
"vector_sync_queue_max_size": "VECTOR_SYNC_QUEUE_MAX_SIZE",
|
|
"vector_sync_user_poll_interval": "VECTOR_SYNC_USER_POLL_INTERVAL",
|
|
"vector_sync_orphan_sweep_enabled": "VECTOR_SYNC_ORPHAN_SWEEP_ENABLED",
|
|
# Verify-on-read (ADR-019)
|
|
"verification_concurrency": "VERIFICATION_CONCURRENCY",
|
|
# Qdrant settings
|
|
"qdrant_url": "QDRANT_URL",
|
|
"qdrant_location": "QDRANT_LOCATION",
|
|
"qdrant_api_key": "QDRANT_API_KEY",
|
|
"qdrant_collection": "QDRANT_COLLECTION",
|
|
# Ollama settings
|
|
"ollama_base_url": "OLLAMA_BASE_URL",
|
|
"ollama_embedding_model": "OLLAMA_EMBEDDING_MODEL",
|
|
"ollama_generation_model": "OLLAMA_GENERATION_MODEL",
|
|
"ollama_verify_ssl": "OLLAMA_VERIFY_SSL",
|
|
# OpenAI settings
|
|
"openai_api_key": "OPENAI_API_KEY",
|
|
"openai_base_url": "OPENAI_BASE_URL",
|
|
"openai_embedding_model": "OPENAI_EMBEDDING_MODEL",
|
|
"openai_generation_model": "OPENAI_GENERATION_MODEL",
|
|
# Bedrock (AWS) settings
|
|
"aws_region": "AWS_REGION",
|
|
"aws_access_key_id": "AWS_ACCESS_KEY_ID",
|
|
"aws_secret_access_key": "AWS_SECRET_ACCESS_KEY",
|
|
"bedrock_embedding_model": "BEDROCK_EMBEDDING_MODEL",
|
|
"bedrock_generation_model": "BEDROCK_GENERATION_MODEL",
|
|
# Mistral settings
|
|
"mistral_api_key": "MISTRAL_API_KEY",
|
|
"mistral_embedding_model": "MISTRAL_EMBEDDING_MODEL",
|
|
"mistral_base_url": "MISTRAL_BASE_URL",
|
|
# Simple provider
|
|
"simple_embedding_dimension": "SIMPLE_EMBEDDING_DIMENSION",
|
|
# Document chunking settings
|
|
"document_chunk_size": "DOCUMENT_CHUNK_SIZE",
|
|
"document_chunk_overlap": "DOCUMENT_CHUNK_OVERLAP",
|
|
# Observability settings
|
|
"metrics_enabled": "METRICS_ENABLED",
|
|
"metrics_port": "METRICS_PORT",
|
|
"otel_exporter_otlp_endpoint": "OTEL_EXPORTER_OTLP_ENDPOINT",
|
|
"otel_exporter_verify_ssl": "OTEL_EXPORTER_VERIFY_SSL",
|
|
"otel_service_name": "OTEL_SERVICE_NAME",
|
|
"otel_traces_sampler": "OTEL_TRACES_SAMPLER",
|
|
"otel_traces_sampler_arg": "OTEL_TRACES_SAMPLER_ARG",
|
|
"log_format": "LOG_FORMAT",
|
|
"log_level": "LOG_LEVEL",
|
|
"log_include_trace_context": "LOG_INCLUDE_TRACE_CONTEXT",
|
|
"excluded_tags": "EXCLUDED_TAGS",
|
|
# MCP decomposition hook points (design §10)
|
|
"embedding_provider": "EMBEDDING_PROVIDER",
|
|
"ingest_mode": "INGEST_MODE",
|
|
"status_backend": "STATUS_BACKEND",
|
|
"collection_metadata_source": "COLLECTION_METADATA_SOURCE",
|
|
"collection_metadata_api_url": "COLLECTION_METADATA_API_URL",
|
|
"fact_event_emitter": "FACT_EVENT_EMITTER",
|
|
"ingest_bus_url": "INGEST_BUS_URL",
|
|
"embedding_gateway_url": "EMBEDDING_GATEWAY_URL",
|
|
"embedding_gateway_model": "EMBEDDING_GATEWAY_MODEL",
|
|
"embedding_gateway_token_url": "EMBEDDING_GATEWAY_TOKEN_URL",
|
|
"embedding_gateway_client_id": "EMBEDDING_GATEWAY_CLIENT_ID",
|
|
"embedding_gateway_client_secret": "EMBEDDING_GATEWAY_CLIENT_SECRET",
|
|
"embedding_gateway_scope": "EMBEDDING_GATEWAY_SCOPE",
|
|
"tenant_id": "TENANT_ID",
|
|
"ingest_bus_num_replicas": "INGEST_BUS_NUM_REPLICAS",
|
|
"acl_prefilter_enabled": "ACL_PREFILTER_ENABLED",
|
|
}
|
|
|
|
# Only pass values that dynaconf actually has; omit unset keys so
|
|
# the Settings dataclass defaults apply.
|
|
kwargs = {
|
|
field: val
|
|
for field, key in _field_map.items()
|
|
if (val := _dget(key)) is not _UNSET
|
|
}
|
|
|
|
# Smart dependency overrides (always set, regardless of dynaconf)
|
|
kwargs["vector_sync_enabled"] = enable_semantic_search
|
|
kwargs["enable_offline_access"] = enable_background_operations
|
|
|
|
return Settings(**kwargs)
|
|
|
|
|
|
def get_nextcloud_ssl_verify() -> bool | ssl.SSLContext:
|
|
"""Return the SSL verification setting for Nextcloud connections.
|
|
|
|
Returns:
|
|
- False if NEXTCLOUD_VERIFY_SSL=false (disable verification)
|
|
- ssl.SSLContext if NEXTCLOUD_CA_BUNDLE is set (custom CA)
|
|
- True otherwise (default system CA verification)
|
|
"""
|
|
settings = get_settings()
|
|
if not settings.nextcloud_verify_ssl:
|
|
return False
|
|
if settings.nextcloud_ca_bundle:
|
|
ctx = ssl.create_default_context(cafile=settings.nextcloud_ca_bundle)
|
|
return ctx
|
|
return True
|
|
|
|
|
|
def get_database_ssl() -> bool | ssl.SSLContext | None:
|
|
"""Return the asyncpg ``ssl`` arg for the Postgres backend (ADR-026).
|
|
|
|
Returns:
|
|
- ``None`` when both DATABASE_VERIFY_SSL and DATABASE_CA_BUNDLE are
|
|
unset — caller skips passing ``ssl`` so asyncpg keeps its default
|
|
(``prefer``). Preserves PR #798 behavior for cluster-local
|
|
Postgres without TLS.
|
|
- ``False`` if DATABASE_VERIFY_SSL=false (silence cert errors).
|
|
- ``ssl.SSLContext`` if DATABASE_CA_BUNDLE is set (custom private
|
|
CA, implies verify-full).
|
|
- ``True`` if DATABASE_VERIFY_SSL=true and no bundle (verify-full
|
|
against system trust store).
|
|
|
|
DATABASE_VERIFY_SSL=false wins over DATABASE_CA_BUNDLE so an operator
|
|
can quickly silence cert errors during incident response without
|
|
having to delete the bundle path from their secret store. Matches the
|
|
Nextcloud-pattern precedence for symmetry with
|
|
:func:`get_nextcloud_ssl_verify`.
|
|
"""
|
|
settings = get_settings()
|
|
if settings.database_verify_ssl is False:
|
|
# Operator-explicit opt-out (DATABASE_VERIFY_SSL=false) — semantics
|
|
# are documented in the docstring above and ADR-026 TLS section.
|
|
# Bare NOSONAR silences any cert-verification-required rule that
|
|
# may fire on this branch (defensive; no such rule fires today).
|
|
return False # NOSONAR
|
|
if settings.database_ca_bundle:
|
|
# ``ssl.create_default_context()`` on Python 3.10+ already negotiates
|
|
# the strongest available protocol (TLS 1.2+ with secure ciphers);
|
|
# we pin Python 3.11+ in pyproject.toml. ``purpose=SERVER_AUTH`` is
|
|
# the default but spelt out here so the intent is visible to
|
|
# static analysers and human readers alike.
|
|
return ssl.create_default_context( # NOSONAR
|
|
purpose=ssl.Purpose.SERVER_AUTH,
|
|
cafile=settings.database_ca_bundle,
|
|
)
|
|
if settings.database_verify_ssl is True:
|
|
return True
|
|
return None
|