Documentation
¶
Overview ¶
Package chunker splits file content into overlapping token-limited chunks for embedding-based search.
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type Chunk ¶
type Chunk struct {
Index int // 0-based chunk index within the file.
Content string // Chunk text.
StartLine int // 1-based start line (inclusive).
EndLine int // 1-based end line (inclusive).
}
Chunk represents a piece of a file with positional metadata.
type Chunker ¶
type Chunker struct {
// contains filtered or unexported fields
}
Chunker splits file content into overlapping chunks.
func (*Chunker) Chunk ¶
Chunk splits content into token-limited chunks with overlap. Small files that fit within MaxTokens are returned as a single chunk.
type Config ¶
type Config struct {
Strategy Strategy // Chunking strategy (default: auto).
MaxTokens int // Target max tokens per chunk (default: 500).
OverlapTokens int // Overlap tokens between adjacent chunks (default: 75).
Estimate func(string) int // Token estimation function.
}
Config holds chunking parameters.
func (*Config) ApplyDefaults ¶
func (c *Config) ApplyDefaults()
ApplyDefaults sets defaults for zero-valued fields.
Click to show internal directories.
Click to hide internal directories.