Fimeg/letta-server
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity
Files
848aa962b6132fb3099958ca07acb9041f09bbf7
letta-server/letta/server/rest_api/routers/v1/folders.py
Kian Jones 848aa962b6 feat: add memory tracking to core (#6179) 
* add memory tracking to core

* move to asyncio from threading.Thread

* remove threading.thread all the way

* delay decorator monitoring initialization until after event loop is registered

* context manager to decorator

* add psutil
2025-11-24 19:09:32 -08:00

603 lines
26 KiB
Python
Raw Blame History

import mimetypes
import os
import tempfile
from pathlib import Path as PathLibPath
from typing import List, Literal, Optional
from fastapi import APIRouter, Depends, HTTPException, Query, UploadFile
from starlette import status
from starlette.responses import Response
import letta.constants as constants
from letta.errors import LettaInvalidArgumentError, LettaUnsupportedFileUploadError
from letta.helpers.pinecone_utils import (
delete_file_records_from_pinecone_index,
delete_source_records_from_pinecone_index,
should_use_pinecone,
)
from letta.helpers.tpuf_client import should_use_tpuf
from letta.log import get_logger
from letta.monitoring import track_operation
from letta.otel.tracing import trace_method
from letta.schemas.agent import AgentState
from letta.schemas.embedding_config import EmbeddingConfig
from letta.schemas.enums import DuplicateFileHandling, FileProcessingStatus
from letta.schemas.file import FileMetadata, FileMetadataBase
from letta.schemas.folder import BaseFolder, Folder
from letta.schemas.passage import Passage
from letta.schemas.source import BaseSource, Source, SourceCreate, SourceUpdate
from letta.schemas.source_metadata import OrganizationSourcesStats
from letta.schemas.user import User
from letta.server.rest_api.dependencies import HeaderParams, get_headers, get_letta_server
from letta.server.server import SyncServer
from letta.services.file_processor.embedder.openai_embedder import OpenAIEmbedder
from letta.services.file_processor.embedder.pinecone_embedder import PineconeEmbedder
from letta.services.file_processor.file_processor import FileProcessor
from letta.services.file_processor.file_types import get_allowed_media_types, get_extension_to_mime_type_map, register_mime_types
from letta.services.file_processor.parser.markitdown_parser import MarkitdownFileParser
from letta.services.file_processor.parser.mistral_parser import MistralFileParser
from letta.settings import settings
from letta.utils import safe_create_file_processing_task, safe_create_task, sanitize_filename
from letta.validators import FileId, FolderId
logger = get_logger(__name__)
# Register all supported file types with Python's mimetypes module
register_mime_types()
router = APIRouter(prefix="/folders", tags=["folders"])
@router.get("/count", response_model=int, operation_id="count_folders")
async def count_folders(
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Count all data folders created by a user.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.source_manager.size_async(actor=actor)
@router.get("/{folder_id}", response_model=Folder, operation_id="retrieve_folder")
async def retrieve_folder(
folder_id: FolderId,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Get a folder by ID
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
folder = await server.source_manager.get_source_by_id(source_id=folder_id, actor=actor)
return folder
@router.get("/name/{folder_name}", response_model=str, operation_id="get_folder_by_name", deprecated=True)
async def get_folder_by_name(
folder_name: str,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
**Deprecated**: Please use the list endpoint `GET /v1/folders?name=` instead.
Get a folder by name.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
folder = await server.source_manager.get_source_by_name(source_name=folder_name, actor=actor)
return folder.id
@router.get("/metadata", response_model=OrganizationSourcesStats, operation_id="retrieve_metadata")
async def retrieve_metadata(
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
include_detailed_per_source_metadata: bool = False,
):
"""
Get aggregated metadata for all folders in an organization.
Returns structured metadata including:
- Total number of folders
- Total number of files across all folders
- Total size of all files
- Per-source breakdown with file details (file_name, file_size per file) if include_detailed_per_source_metadata is True
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.file_manager.get_organization_sources_metadata(
actor=actor, include_detailed_per_source_metadata=include_detailed_per_source_metadata
)
@router.get("/", response_model=List[Folder], operation_id="list_folders")
async def list_folders(
before: Optional[str] = Query(
None, description="Folder ID cursor for pagination. Returns folders that come before this folder ID in the specified sort order"
),
after: Optional[str] = Query(
None, description="Folder ID cursor for pagination. Returns folders that come after this folder ID in the specified sort order"
),
limit: Optional[int] = Query(50, description="Maximum number of folders to return"),
order: Literal["asc", "desc"] = Query(
"asc", description="Sort order for folders by creation time. 'asc' for oldest first, 'desc' for newest first"
),
order_by: Literal["created_at"] = Query("created_at", description="Field to sort by"),
name: Optional[str] = Query(None, description="Folder name to filter by"),
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
List all data folders created by a user.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.source_manager.list_sources(
actor=actor, before=before, after=after, limit=limit, ascending=(order == "asc"), name=name
)
@router.post("/", response_model=Folder, operation_id="create_folder")
async def create_folder(
folder_create: SourceCreate,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Create a new data folder.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
# TODO: need to asyncify this
if not folder_create.embedding_config:
if not folder_create.embedding:
if settings.default_embedding_handle is None:
raise LettaInvalidArgumentError(
"Must specify either embedding or embedding_config in request", argument_name="default_embedding_handle"
)
else:
folder_create.embedding = settings.default_embedding_handle
folder_create.embedding_config = await server.get_embedding_config_from_handle_async(
handle=folder_create.embedding,
embedding_chunk_size=folder_create.embedding_chunk_size or constants.DEFAULT_EMBEDDING_CHUNK_SIZE,
actor=actor,
)
folder = Source(
name=folder_create.name,
embedding_config=folder_create.embedding_config,
description=folder_create.description,
instructions=folder_create.instructions,
metadata=folder_create.metadata,
)
return await server.source_manager.create_source(source=folder, actor=actor)
@router.patch("/{folder_id}", response_model=Folder, operation_id="modify_folder")
async def modify_folder(
folder: SourceUpdate,
folder_id: FolderId,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Update the name or documentation of an existing data folder.
"""
# TODO: allow updating the handle/embedding config
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
await server.source_manager.get_source_by_id(source_id=folder_id, actor=actor)
return await server.source_manager.update_source(source_id=folder_id, source_update=folder, actor=actor)
@router.delete("/{folder_id}", response_model=None, operation_id="delete_folder")
async def delete_folder(
folder_id: FolderId,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Delete a data folder.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
folder = await server.source_manager.get_source_by_id(source_id=folder_id, actor=actor)
agent_states = await server.source_manager.list_attached_agents(source_id=folder_id, actor=actor)
files = await server.file_manager.list_files(folder_id, actor)
file_ids = [f.id for f in files]
if should_use_tpuf():
logger.info(f"Deleting folder {folder_id} from Turbopuffer")
from letta.helpers.tpuf_client import TurbopufferClient
tpuf_client = TurbopufferClient()
await tpuf_client.delete_source_passages(source_id=folder_id, organization_id=actor.organization_id)
elif should_use_pinecone():
logger.info(f"Deleting folder {folder_id} from pinecone index")
await delete_source_records_from_pinecone_index(source_id=folder_id, actor=actor)
for agent_state in agent_states:
await server.remove_files_from_context_window(agent_state=agent_state, file_ids=file_ids, actor=actor)
if agent_state.enable_sleeptime:
block = await server.agent_manager.get_block_with_label_async(agent_id=agent_state.id, block_label=folder.name, actor=actor)
if block:
await server.block_manager.delete_block_async(block.id, actor)
await server.delete_source(source_id=folder_id, actor=actor)
@router.post("/{folder_id}/upload", response_model=FileMetadata, operation_id="upload_file_to_folder")
@track_operation("file_upload_to_folder")
async def upload_file_to_folder(
file: UploadFile,
folder_id: FolderId,
duplicate_handling: DuplicateFileHandling = Query(DuplicateFileHandling.SUFFIX, description="How to handle duplicate filenames"),
name: Optional[str] = Query(None, description="Optional custom name to override the uploaded file's name"),
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Upload a file to a data folder.
"""
# NEW: Cloud based file processing
# Determine file's MIME type
mimetypes.guess_type(file.filename)[0] or "application/octet-stream"
allowed_media_types = get_allowed_media_types()
# Normalize incoming Content-Type header (strip charset or any parameters).
raw_ct = file.content_type or ""
media_type = raw_ct.split(";", 1)[0].strip().lower()
# If client didn't supply a Content-Type or it's not one of the allowed types,
# attempt to infer from filename extension.
if media_type not in allowed_media_types and file.filename:
guessed, _ = mimetypes.guess_type(file.filename)
media_type = (guessed or "").lower()
if media_type not in allowed_media_types:
ext = PathLibPath(file.filename).suffix.lower()
ext_map = get_extension_to_mime_type_map()
media_type = ext_map.get(ext, media_type)
# If still not allowed, reject with 415.
if media_type not in allowed_media_types:
raise LettaUnsupportedFileUploadError(
message=(
f"Unsupported file type: {media_type or 'unknown'} "
f"(filename: {file.filename}). "
f"Supported types: PDF, text files (.txt, .md), JSON, and code files (.py, .js, .java, etc.)."
),
)
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
# Read file bytes once
file_bytes = await file.read()
# If enabled, delegate to Temporal workflow (Lettuce) and return its result
if settings.use_lettuce_for_file_uploads:
from letta.services.lettuce import LettuceClient
lettuce_client = await LettuceClient.create()
result = await lettuce_client.upload_file_to_folder(
folder_id=folder_id,
actor_id=actor.id,
file_name=file.filename,
content=file_bytes,
content_type=raw_ct or None,
duplicate_handling=duplicate_handling,
override_name=name,
)
if result is not None:
return result.file_metadata
folder = await server.source_manager.get_source_by_id(source_id=folder_id, actor=actor)
content = file_bytes
file_size_mb = len(content) / (1024 * 1024)
from letta.log import get_logger
logger = get_logger(__name__)
logger.info(f"File upload to folder: loaded {file_size_mb:.2f} MB into memory, filename: {file.filename}")
# Store original filename and handle duplicate logic
# Use custom name if provided, otherwise use the uploaded file's name
# If custom name is provided, use it directly (it's just metadata, not a filesystem path)
# Otherwise, sanitize the uploaded filename for security
original_filename = name if name else sanitize_filename(file.filename) # Basic sanitization only
# Check if duplicate exists
existing_file = await server.file_manager.get_file_by_original_name_and_source(
original_filename=original_filename, source_id=folder_id, actor=actor
)
unique_filename = None
if existing_file:
# Duplicate found, handle based on strategy
if duplicate_handling == DuplicateFileHandling.ERROR:
raise LettaInvalidArgumentError(
message=f"File '{original_filename}' already exists in folder '{folder.name}'",
argument_name="duplicate_handling",
)
elif duplicate_handling == DuplicateFileHandling.SKIP:
# Return existing file metadata with custom header to indicate it was skipped
response = Response(
content=existing_file.model_dump_json(), media_type="application/json", headers={"X-Upload-Result": "skipped"}
)
return response
elif duplicate_handling == DuplicateFileHandling.REPLACE:
# delete the file
deleted_file = await server.file_manager.delete_file(file_id=existing_file.id, actor=actor)
unique_filename = original_filename
if not unique_filename:
# For SUFFIX, continue to generate unique filename
# Generate unique filename (adds suffix if needed)
unique_filename = await server.file_manager.generate_unique_filename(
original_filename=original_filename, source=folder, organization_id=actor.organization_id
)
# create file metadata
file_metadata = FileMetadata(
source_id=folder_id,
file_name=unique_filename,
original_file_name=original_filename,
file_path=None,
file_type=mimetypes.guess_type(original_filename)[0] or file.content_type or "unknown",
file_size=file.size if file.size is not None else None,
processing_status=FileProcessingStatus.PARSING,
)
file_metadata = await server.file_manager.create_file(file_metadata, actor=actor)
# TODO: Do we need to pull in the full agent_states? Can probably simplify here right?
agent_states = await server.source_manager.list_attached_agents(source_id=folder_id, actor=actor)
# Use cloud processing for all files (simple files always, complex files with Mistral key)
logger.info("Running experimental cloud based file processing...")
safe_create_file_processing_task(
load_file_to_source_cloud(server, agent_states, content, folder_id, actor, folder.embedding_config, file_metadata),
file_metadata=file_metadata,
server=server,
actor=actor,
logger=logger,
label="file_processor.process",
)
safe_create_task(sleeptime_document_ingest_async(server, folder_id, actor), label="sleeptime_document_ingest_async")
return file_metadata
@router.get("/{folder_id}/agents", response_model=List[str], operation_id="list_agents_for_folder")
async def list_agents_for_folder(
folder_id: FolderId,
before: Optional[str] = Query(
None,
description="Agent ID cursor for pagination. Returns agents that come before this agent ID in the specified sort order",
),
after: Optional[str] = Query(
None,
description="Agent ID cursor for pagination. Returns agents that come after this agent ID in the specified sort order",
),
limit: Optional[int] = Query(50, description="Maximum number of agents to return"),
order: Literal["asc", "desc"] = Query(
"desc", description="Sort order for agents by creation time. 'asc' for oldest first, 'desc' for newest first"
),
order_by: Literal["created_at"] = Query("created_at", description="Field to sort by"),
server: SyncServer = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Get all agent IDs that have the specified folder attached.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.source_manager.get_agents_for_source_id(
source_id=folder_id,
before=before,
after=after,
limit=limit,
ascending=(order == "asc"),
actor=actor,
)
@router.get("/{folder_id}/passages", response_model=List[Passage], operation_id="list_folder_passages")
async def list_folder_passages(
folder_id: FolderId,
before: Optional[str] = Query(
None,
description="Passage ID cursor for pagination. Returns passages that come before this passage ID in the specified sort order",
),
after: Optional[str] = Query(
None,
description="Passage ID cursor for pagination. Returns passages that come after this passage ID in the specified sort order",
),
limit: Optional[int] = Query(100, description="Maximum number of passages to return"),
order: Literal["asc", "desc"] = Query(
"desc", description="Sort order for passages by creation time. 'asc' for oldest first, 'desc' for newest first"
),
order_by: Literal["created_at"] = Query("created_at", description="Field to sort by"),
server: SyncServer = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
List all passages associated with a data folder.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.agent_manager.query_source_passages_async(
actor=actor,
source_id=folder_id,
after=after,
before=before,
limit=limit,
ascending=(order == "asc"),
)
@router.get("/{folder_id}/files", response_model=List[FileMetadata], operation_id="list_folder_files")
async def list_folder_files(
folder_id: FolderId,
before: Optional[str] = Query(
None,
description="File ID cursor for pagination. Returns files that come before this file ID in the specified sort order",
),
after: Optional[str] = Query(
None,
description="File ID cursor for pagination. Returns files that come after this file ID in the specified sort order",
),
limit: Optional[int] = Query(1000, description="Maximum number of files to return"),
order: Literal["asc", "desc"] = Query(
"desc", description="Sort order for files by creation time. 'asc' for oldest first, 'desc' for newest first"
),
order_by: Literal["created_at"] = Query("created_at", description="Field to sort by"),
include_content: bool = Query(False, description="Whether to include full file content"),
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
List paginated files associated with a data folder.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
return await server.file_manager.list_files(
source_id=folder_id,
before=before,
after=after,
limit=limit,
ascending=(order == "asc"),
actor=actor,
include_content=include_content,
strip_directory_prefix=True, # TODO: Reconsider this. This is purely for aesthetics.
)
# @router.get("/{folder_id}/files/{file_id}", response_model=FileMetadata, operation_id="get_file_metadata")
# async def get_file_metadata(
# folder_id: str,
# file_id: str,
# include_content: bool = Query(False, description="Whether to include full file content"),
# server: "SyncServer" = Depends(get_letta_server),
# headers: HeaderParams = Depends(get_headers),
# ):
# """
# Retrieve metadata for a specific file by its ID.
# """
# actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
#
# # Get file metadata using the file manager
# file_metadata = await server.file_manager.get_file_by_id(
# file_id=file_id, actor=actor, include_content=include_content, strip_directory_prefix=True
# )
#
# if not file_metadata:
# raise HTTPException(status_code=404, detail=f"File with id={file_id} not found.")
#
# # Verify the file belongs to the specified folder
# if file_metadata.source_id != folder_id:
# raise HTTPException(status_code=404, detail=f"File with id={file_id} not found in folder {folder_id}.")
#
# if should_use_pinecone() and file_metadata.processing_status == FileProcessingStatus.EMBEDDING:
# ids = await list_pinecone_index_for_files(file_id=file_id, actor=actor)
# logger.info(
# f"Embedded chunks {len(ids)}/{file_metadata.total_chunks} for {file_id} ({file_metadata.file_name}) in organization {actor.organization_id}"
# )
#
# if len(ids) != file_metadata.chunks_embedded or len(ids) == file_metadata.total_chunks:
# if len(ids) != file_metadata.total_chunks:
# file_status = file_metadata.processing_status
# else:
# file_status = FileProcessingStatus.COMPLETED
# try:
# file_metadata = await server.file_manager.update_file_status(
# file_id=file_metadata.id, actor=actor, chunks_embedded=len(ids), processing_status=file_status
# )
# except ValueError as e:
# # state transition was blocked - this is a race condition
# # log it but don't fail the request since we're just reading metadata
# logger.warning(f"Race condition detected in get_file_metadata: {str(e)}")
# # return the current file state without updating
#
# return file_metadata
# it's redundant to include /delete in the URL path. The HTTP verb DELETE already implies that action.
# it's still good practice to return a status indicating the success or failure of the deletion
@router.delete("/{folder_id}/{file_id}", status_code=204, operation_id="delete_file_from_folder")
async def delete_file_from_folder(
folder_id: FolderId,
file_id: FileId,
server: "SyncServer" = Depends(get_letta_server),
headers: HeaderParams = Depends(get_headers),
):
"""
Delete a file from a folder.
"""
actor = await server.user_manager.get_actor_or_default_async(actor_id=headers.actor_id)
deleted_file = await server.file_manager.delete_file(file_id=file_id, actor=actor)
await server.remove_file_from_context_windows(source_id=folder_id, file_id=deleted_file.id, actor=actor)
if should_use_tpuf():
logger.info(f"Deleting file {file_id} from Turbopuffer")
from letta.helpers.tpuf_client import TurbopufferClient
tpuf_client = TurbopufferClient()
await tpuf_client.delete_file_passages(source_id=folder_id, file_id=file_id, organization_id=actor.organization_id)
elif should_use_pinecone():
logger.info(f"Deleting file {file_id} from pinecone index")
await delete_file_records_from_pinecone_index(file_id=file_id, actor=actor)
safe_create_task(sleeptime_document_ingest_async(server, folder_id, actor, clear_history=True), label="document_ingest_after_delete")
async def load_file_to_source_async(server: SyncServer, source_id: str, job_id: str, filename: str, bytes: bytes, actor: User):
# Create a temporary directory (deleted after the context manager exits)
with tempfile.TemporaryDirectory() as tmpdirname:
file_path = os.path.join(tmpdirname, filename)
# Write the file to the sanitized path
with open(file_path, "wb") as buffer:
buffer.write(bytes)
# Pass the file to load_file_to_source
await server.load_file_to_source(source_id, file_path, job_id, actor)
async def sleeptime_document_ingest_async(server: SyncServer, source_id: str, actor: User, clear_history: bool = False):
source = await server.source_manager.get_source_by_id(source_id=source_id)
agents = await server.source_manager.list_attached_agents(source_id=source_id, actor=actor)
for agent in agents:
if agent.enable_sleeptime:
await server.sleeptime_document_ingest_async(agent, source, actor, clear_history)
@trace_method
async def load_file_to_source_cloud(
server: SyncServer,
agent_states: List[AgentState],
content: bytes,
source_id: str,
actor: User,
embedding_config: EmbeddingConfig,
file_metadata: FileMetadata,
):
# Choose parser based on mistral API key availability
if settings.mistral_api_key:
file_parser = MistralFileParser()
else:
file_parser = MarkitdownFileParser()
# determine which embedder to use - turbopuffer takes precedence
if should_use_tpuf():
from letta.services.file_processor.embedder.turbopuffer_embedder import TurbopufferEmbedder
embedder = TurbopufferEmbedder(embedding_config=embedding_config)
elif should_use_pinecone():
embedder = PineconeEmbedder(embedding_config=embedding_config)
else:
embedder = OpenAIEmbedder(embedding_config=embedding_config)
file_processor = FileProcessor(file_parser=file_parser, embedder=embedder, actor=actor)
await file_processor.process(agent_states=agent_states, source_id=source_id, content=content, file_metadata=file_metadata)
Reference in New Issue View Git Blame Copy Permalink