tsar

README

TSAR - Text Search All Rows

A library for specific search use case

tsar is a go library that creates a sorted reverse word index that uses binary search to look up. It also provides a small DSL for querying files directly

tsar works directly on text files and creates an index, this can be saved directly to a file and used for querying the text.

see example/simple.go

Documentation

Index

• Constants
• func MarshalIndex(i *Index) []byte
• type Entry
• type EntryList
  • func NewEntryList() EntryList
  • func (l EntryList) Append(key string, ptr uint32) error
  • func (l EntryList) Remove(key string)
  • func (l EntryList) Set(key string, pointers []uint32) error
  • func (l EntryList) ToIndex() *Index
• type Index
  • func UnmarshalIndex(data []byte) (*Index, error)
  • func UnmarshalIndexLazy(data []byte) (*Index, error)
  • func UnmarshalIndexLazyReader(reader io.ReadSeeker) (*Index, error)
  • func (i *Index) Find(needle string, match Matcher) ([]*Entry, error)
• type Matcher

Constants

const CheckpointSize = 8 // bytes 0:4 unused (was row number), bytes 4:8 are entry offset

const MaxEntryPointers = math.MaxUint32

const MaxKeyLen = math.MaxUint8

const PartitionSize = 20

const PointerSize = 4

Functions

func MarshalIndex

func MarshalIndex(i *Index) []byte

Types

type Entry

type Entry struct {
	Key      string
	Pointers []uint32
}

type EntryList

type EntryList map[string][]uint32

func NewEntryList

func NewEntryList() EntryList

func (EntryList) Append

func (l EntryList) Append(key string, ptr uint32) error

func (EntryList) Remove

func (l EntryList) Remove(key string)

func (EntryList) Set

func (l EntryList) Set(key string, pointers []uint32) error

func (EntryList) ToIndex

func (l EntryList) ToIndex() *Index

type Index

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

func UnmarshalIndex

func UnmarshalIndex(data []byte) (*Index, error)

func UnmarshalIndexLazy

func UnmarshalIndexLazy(data []byte) (*Index, error)

func UnmarshalIndexLazyReader

func UnmarshalIndexLazyReader(reader io.ReadSeeker) (*Index, error)

func (*Index) Find

func (i *Index) Find(needle string, match Matcher) ([]*Entry, error)

type Matcher

type Matcher func(candidate string, needle string) bool

var MatchEqual Matcher = func(a, b string) bool { return a == b }

var MatchPrefix Matcher = strings.HasPrefix