Documentation
¶
Overview ¶
@index Graph building service that orchestrates parsing, persistence, and search indexing.
Index ¶
- func CheckParseFileSize(relPath string, size int64, maxFileBytes int64) error
- func CheckTotalParsedBytes(relPath string, current int64, next int64, maxTotalBytes int64) error
- func ExistingGraphFiles(ctx context.Context, db *gorm.DB) ([]string, error)
- func RefreshSearchDocuments(ctx context.Context, db *gorm.DB) (int, error)
- func RefreshSearchDocumentsFor(ctx context.Context, db *gorm.DB, nodeIDs []uint) (int, error)
- type BuildOptions
- type BuildStats
- type GraphService
- type IncrementalSyncer
- type NamespacePurger
- type Parser
- type UnreadableFilesError
- type UpdateOptions
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func CheckParseFileSize ¶
@intent reject individual files that exceed the configured per-file parse budget before loading them into memory.
func CheckTotalParsedBytes ¶
@intent stop one build or update pass once cumulative parsed bytes would exceed the configured safety limit.
func ExistingGraphFiles ¶
ExistingGraphFiles returns namespace-scoped graph file paths currently stored. @intent share deletion-scope discovery across CLI and MCP incremental updates
func RefreshSearchDocuments ¶
RefreshSearchDocuments rebuilds namespace-scoped search_documents from current graph nodes. @intent keep derived search documents consistent with graph state before FTS rebuilds
Types ¶
type BuildOptions ¶
type BuildOptions struct {
Dir string
NoRecursive bool
ExcludePatterns []string
IncludePaths []string
MaxFileBytes int64
MaxTotalParsedBytes int64
SkipSearchRebuild bool
FallbackCalls bool
}
BuildOptions configures one graph build run. @intent 빌드 대상 경로와 탐색 범위를 호출자에서 제어하게 한다.
type BuildStats ¶
type BuildStats struct {
TotalFiles int
TotalNodes int
TotalEdges int
Unresolved edgeresolve.FilterResolvedDiagnostics
}
BuildStats reports how much content a build processed. @intent CLI와 호출자가 빌드 결과 규모를 사용자에게 보여줄 수 있게 한다.
type GraphService ¶
type GraphService struct {
Store store.GraphStore
DB *gorm.DB
SearchBackend search.Backend
Walkers map[string]*treesitter.Walker
Parsers map[string]Parser
Logger *slog.Logger
}
GraphService orchestrates graph building and search document refresh. @intent 파싱 결과 저장과 검색 인덱스 재구성을 하나의 서비스로 묶는다.
func (*GraphService) Build ¶
func (s *GraphService) Build(ctx context.Context, opts BuildOptions) (BuildStats, error)
Build walks source files, stores parsed graph data, and rebuilds search docs. @intent perform a full graph build from the specified directory. @sideEffect 파일 시스템을 읽고 그래프 저장소·DB·검색 인덱스를 갱신한다.
func (*GraphService) Update ¶
func (s *GraphService) Update(ctx context.Context, opts UpdateOptions) (*incremental.SyncStats, error)
Update incrementally syncs changed files into the graph and optionally rebuilds search. @intent centralize file collection, include path, parse limit, and search policy for update callers
type IncrementalSyncer ¶
type IncrementalSyncer interface {
SyncWithExisting(ctx context.Context, files map[string]incremental.FileInfo, existingFiles []string) (*incremental.SyncStats, error)
}
IncrementalSyncer defines the sync operation GraphService.Update delegates to. @intent keep filesystem/build policy in service while preserving the existing incremental engine
type NamespacePurger ¶
type NamespacePurger struct {
// contains filtered or unexported fields
}
NamespacePurger는 단일 namespace의 모든 영속 상태를 일관되게 제거하는 단일 진입점이다. @intent gormstore.DeleteGraph(노드/엣지/어노테이션/멤버십/검색문서)와 그동안 MCP handler 안에 흩어져 있던 Community/Flow/orphan-membership/검색 백엔드 정리를 한 곳으로 모아 namespace-scoped 테이블 추가 시의 drift 위험을 줄인다. @sideEffect graph store와 service 레벨 테이블, 검색 백엔드를 하나의 DB 트랜잭션으로 정리한다.
func NewNamespacePurger ¶
func NewNamespacePurger(graphStore store.NodeWriter, db *gorm.DB, backend storesearch.Backend) *NamespacePurger
NewNamespacePurger는 NamespacePurger를 생성한다. @param graphStore namespace 단위 graph 정리를 수행할 저장소. @param db Community/Flow 등 service-level 테이블을 정리할 DB. nil이면 해당 단계는 skip 한다. @param backend FTS 등 검색 백엔드. nil이면 해당 단계는 skip 한다. @intent assemble one namespace purge entry point that can remove graph, service, and search state together.
func (*NamespacePurger) Purge ¶
func (p *NamespacePurger) Purge(ctx context.Context) error
Purge는 ctx의 namespace에 속한 모든 그래프/서비스/검색 상태를 제거한다. @intent workspace 삭제와 같은 namespace 단위 purge 경로의 단일 진입점을 제공한다. @requires ctx에 ctxns.WithNamespace로 namespace가 설정되어 있어야 한다. @sideEffect 노드, 엣지, 어노테이션, 멤버십, 검색 문서, 커뮤니티, 플로우, 검색 인덱스를 삭제한다.
type Parser ¶
type Parser interface {
Parse(filePath string, content []byte) ([]model.Node, []model.Edge, error)
ParseWithContext(ctx context.Context, filePath string, content []byte) ([]model.Node, []model.Edge, error)
}
Parser defines the minimum parser behavior required by graph builds. @intent let CLI, MCP, and tests share the same build policy with their parser implementations
type UnreadableFilesError ¶
type UnreadableFilesError struct {
Files []string
}
UnreadableFilesError signals that the update path encountered files it could not stat or read while FailOnUnreadable was enabled. @intent give webhook/server callers a structured failure they can surface instead of silent partial sync
func (*UnreadableFilesError) Error ¶
func (e *UnreadableFilesError) Error() string
Error formats the unreadable file failure with a count and a representative sample path. @intent give operators a stable, single-line summary they can grep instead of dumping every path.
type UpdateOptions ¶
type UpdateOptions struct {
BuildOptions
Syncer IncrementalSyncer
Replace bool
FailOnUnreadable bool
}
UpdateOptions configures one incremental graph sync run. @intent reuse GraphService traversal and parse limit policy for CLI and MCP updates
Source Files