refac: kb sync

This commit is contained in:
Timothy Jaeryang Baek
2026-05-20 01:13:52 +04:00
parent d0b17f0569
commit 60c9db1cb8
6 changed files with 548 additions and 18 deletions
+6
View File
@@ -1,4 +1,5 @@
import asyncio
import hashlib
import json
import logging
import os
@@ -290,6 +291,10 @@ async def upload_file_handler(
},
)
# SHA-256 of raw uploaded bytes for incremental sync diffing.
# If the client pre-computed and sent file_hash, use that.
file_hash = file_metadata.get('file_hash') or hashlib.sha256(contents).hexdigest()
file_item = await Files.insert_new_file(
user.id,
FileForm(
@@ -304,6 +309,7 @@ async def upload_file_handler(
'name': name,
'content_type': (file.content_type if isinstance(file.content_type, str) else None),
'size': len(contents),
'file_hash': file_hash,
'data': file_metadata,
},
}
+178
View File
@@ -984,6 +984,184 @@ async def reset_knowledge_by_id(
return knowledge
############################
# SyncKnowledgeDiff
############################
class FileManifestEntry(BaseModel):
filename: str # basename: "readme.md"
path: str # relative dir: "docs/api" or "" for root
checksum: str # SHA-256 of raw bytes
size: int
class SyncDiffForm(BaseModel):
manifest: list[FileManifestEntry]
class SyncDiffResponse(BaseModel):
added: list[dict] # [{filename, path}] — new files
modified: list[dict] # [{filename, path, stale_file_id}] — changed files
deleted: list[dict] # [{file_id, filename}] — files to remove
mkdir: list[str] # directory paths to create
rmdir: list[str] # directory IDs to remove
unmodified_count: int
@router.post('/{id}/sync/diff', response_model=SyncDiffResponse)
async def sync_knowledge_diff(
id: str,
form_data: SyncDiffForm,
user=Depends(get_verified_user),
db: AsyncSession = Depends(get_async_session),
):
"""
Compare a local file manifest against the knowledge base to determine
which files need uploading, removing, and which directories to create/remove.
"""
await _verify_knowledge_write_access(id, user, db)
# ── Index existing state ──
knowledge_files = await Knowledges.get_files_with_directory_ids(id, db=db)
existing_directories = await Knowledges.get_all_directories(id, db=db)
# Build directory path lookups
directory_path_by_id: dict[str, str] = {}
directory_id_by_path: dict[str, str] = {}
for directory in existing_directories:
segments = [directory.name]
parent_id = directory.parent_id
while parent_id:
parent = next((d for d in existing_directories if d.id == parent_id), None)
if not parent:
break
segments.insert(0, parent.name)
parent_id = parent.parent_id
full_path = '/'.join(segments)
directory_path_by_id[directory.id] = full_path
directory_id_by_path[full_path] = directory.id
# Index existing files by (path, filename) → {file_id, checksum}
indexed_files: dict[tuple[str, str], dict] = {}
for file_model, directory_id in knowledge_files:
file_path = directory_path_by_id.get(directory_id, '') if directory_id else ''
stored_checksum = (file_model.meta or {}).get('file_hash')
indexed_files[(file_path, file_model.filename)] = {
'file_id': file_model.id,
'checksum': stored_checksum,
}
# ── Diff files ──
added: list[dict] = []
modified: list[dict] = []
deleted: list[dict] = []
unmodified_count = 0
manifest_keys: set[tuple[str, str]] = set()
for entry in form_data.manifest:
key = (entry.path, entry.filename)
manifest_keys.add(key)
if key not in indexed_files:
added.append({'filename': entry.filename, 'path': entry.path})
elif indexed_files[key]['checksum'] != entry.checksum:
modified.append({
'filename': entry.filename,
'path': entry.path,
'stale_file_id': indexed_files[key]['file_id'],
})
else:
unmodified_count += 1
for key, file_info in indexed_files.items():
if key not in manifest_keys:
deleted.append({'file_id': file_info['file_id'], 'filename': key[1]})
# ── Diff directories ──
required_directory_paths: set[str] = set()
for entry in form_data.manifest:
if entry.path:
segments = entry.path.split('/')
for depth in range(len(segments)):
required_directory_paths.add('/'.join(segments[:depth + 1]))
mkdir = sorted(
[p for p in required_directory_paths if p not in directory_id_by_path],
key=lambda p: p.count('/')
)
orphaned_directory_paths = set(directory_id_by_path) - required_directory_paths
rmdir = [directory_id_by_path[p] for p in orphaned_directory_paths]
return SyncDiffResponse(
added=added,
modified=modified,
deleted=deleted,
mkdir=mkdir,
rmdir=rmdir,
unmodified_count=unmodified_count,
)
############################
# SyncKnowledgeCleanup
############################
class SyncCleanupForm(BaseModel):
file_ids: list[str] # file IDs to delete
dir_ids: list[str] = [] # directory IDs to rmdir
@router.post('/{id}/sync/cleanup')
async def sync_knowledge_cleanup(
id: str,
form_data: SyncCleanupForm,
user=Depends(get_verified_user),
db: AsyncSession = Depends(get_async_session),
):
"""
Remove stale files and orphaned directories from a knowledge base
after an incremental sync.
"""
await _verify_knowledge_write_access(id, user, db)
# ── Remove deleted files ──
for file_id in form_data.file_ids:
file = await Files.get_file_by_id(file_id, db=db)
if not file:
continue
await Knowledges.remove_file_from_knowledge_by_id(id, file_id, db=db)
try:
await ASYNC_VECTOR_DB_CLIENT.delete(
collection_name=id, filter={'file_id': file_id}
)
await ASYNC_VECTOR_DB_CLIENT.delete(
collection_name=id, filter={'hash': file.hash}
)
except Exception:
pass
try:
collection_name = f'file-{file_id}'
if await ASYNC_VECTOR_DB_CLIENT.has_collection(collection_name):
await ASYNC_VECTOR_DB_CLIENT.delete_collection(collection_name)
except Exception:
pass
if file.user_id == user.id or user.role == 'admin':
await Files.delete_file_by_id(file_id, db=db)
# ── Remove orphaned directories (children before parents) ──
for dir_id in reversed(form_data.dir_ids):
await Knowledges.delete_directory(dir_id, move_files_to_parent=False, db=db)
return {'status': True}
############################
# AddFilesToKnowledge
############################