indexer

type AtomicWriter struct {
	// contains filtered or unexported fields
}

type Chunk struct {
	ID        string                 `json:"id"`
	ChunkType ChunkType              `json:"chunk_type"`
	Title     string                 `json:"title"`
	Text      string                 `json:"text"`
	Embedding []float32              `json:"embedding"`
	Tags      []string               `json:"tags"`
	Metadata  map[string]interface{} `json:"metadata"`
	CreatedAt time.Time              `json:"created_at"`
	UpdatedAt time.Time              `json:"updated_at"`
}

type ChunkFile struct {
	Metadata ChunkFileMetadata `json:"_metadata"`
	Chunks   []Chunk           `json:"chunks"`
}

type ChunkFileMetadata struct {
	Model      string    `json:"model"`
	Dimensions int       `json:"dimensions"`
	ChunkType  ChunkType `json:"chunk_type"`
	Generated  time.Time `json:"generated"`
	Version    string    `json:"version"`
}

type ChunkType string

type Chunker interface {
	// ChunkDocument splits a markdown file into semantic chunks.
	// Returns a slice of DocumentationChunk.
	ChunkDocument(ctx context.Context, filePath string, content string) ([]DocumentationChunk, error)
}

type CodeExtraction struct {
	// Symbols contains high-level overview (package, imports count, type/function names)
	Symbols *extraction.SymbolsData

	// Definitions contains full type definitions and function signatures
	Definitions *extraction.DefinitionsData

	// Data contains constants, global variables, and configuration
	Data *extraction.DataData

	// Metadata about the extraction
	Language  string
	FilePath  string
	StartLine int
	EndLine   int
}

type Config struct {
	// Root directory of the codebase to index
	RootDir string

	// Paths configuration
	CodePatterns   []string
	DocsPatterns   []string
	IgnorePatterns []string

	// Chunking configuration
	ChunkStrategies []string // ["symbols", "definitions", "data"]
	DocChunkSize    int      // tokens
	CodeChunkSize   int      // characters
	Overlap         int      // tokens

	// Output configuration
	OutputDir string // .cortex/chunks/

	// Embedding configuration
	EmbeddingProvider string
	EmbeddingModel    string
	EmbeddingDims     int
	EmbeddingEndpoint string
	EmbeddingBinary   string
}

type DocumentationChunk struct {
	FilePath         string
	SectionIndex     int
	ChunkIndex       int
	Text             string
	StartLine        int
	EndLine          int
	IsLargeParagraph bool
	IsSplitParagraph bool
}

type FileDiscovery struct {
	// contains filtered or unexported fields
}

type Formatter interface {
	// FormatSymbols converts SymbolsData into natural language text.
	FormatSymbols(data *extraction.SymbolsData, language string) string

	// FormatDefinitions converts DefinitionsData into formatted code with line comments.
	FormatDefinitions(data *extraction.DefinitionsData, language string) string

	// FormatData converts DataData into formatted code with line comments.
	FormatData(data *extraction.DataData, language string) string

	// FormatDocumentation formats a documentation chunk (may add context).
	FormatDocumentation(chunk *DocumentationChunk) string
}

type GeneratorMetadata struct {
	Version       string               `json:"version"`
	GeneratedAt   time.Time            `json:"generated_at"`
	FileChecksums map[string]string    `json:"file_checksums"`
	FileMtimes    map[string]time.Time `json:"file_mtimes"`
	Stats         ProcessingStats      `json:"stats"`
}

type Indexer interface {
	// Index processes all files in the codebase and generates chunk files.
	// Returns statistics about the indexing process.
	Index(ctx context.Context) (*ProcessingStats, error)

	// IndexIncremental processes only changed files based on checksums.
	// Returns statistics about the indexing process.
	IndexIncremental(ctx context.Context) (*ProcessingStats, error)

	// Watch starts watching for file changes and reindexes incrementally.
	// Blocks until context is cancelled.
	Watch(ctx context.Context) error

	// Close releases all resources held by the indexer.
	Close() error
}

type IndexerWatcher struct {
	// contains filtered or unexported fields
}

type NoOpProgressReporter struct{}

type Parser interface {
	// ParseFile extracts code structure from a source file.
	// Returns CodeExtraction containing symbols, definitions, and data.
	ParseFile(ctx context.Context, filePath string) (*CodeExtraction, error)

	// SupportsLanguage checks if this parser supports the given language.
	SupportsLanguage(language string) bool
}

type ProcessingStats struct {
	DocsProcessed         int     `json:"docs_processed"`
	CodeFilesProcessed    int     `json:"code_files_processed"`
	TotalDocChunks        int     `json:"total_doc_chunks"`
	TotalCodeChunks       int     `json:"total_code_chunks"`
	ProcessingTimeSeconds float64 `json:"processing_time_seconds"`
}

type ProgressReporter interface {
	// OnDiscoveryStart is called when file discovery begins.
	OnDiscoveryStart()

	// OnDiscoveryComplete is called when file discovery finishes.
	OnDiscoveryComplete(codeFiles, docFiles int)

	// OnFileProcessingStart is called before processing files.
	OnFileProcessingStart(totalFiles int)

	// OnFileProcessed is called after each file is processed.
	OnFileProcessed(fileName string)

	// OnEmbeddingStart is called before generating embeddings.
	OnEmbeddingStart(totalChunks int)

	// OnEmbeddingProgress is called after each batch of embeddings.
	OnEmbeddingProgress(processedChunks int)

	// OnWritingChunks is called when writing chunk files begins.
	OnWritingChunks()

	// OnComplete is called when indexing completes successfully.
	OnComplete(stats *ProcessingStats)
}