Documentation
¶
Overview ¶
Package related holds code to help finding related content.
Index ¶
Constants ¶
const ( TypeBasic = "basic" TypeFragments = "fragments" )
Variables ¶
var ( // DefaultConfig is the default related config. DefaultConfig = Config{ Threshold: 80, Indices: IndicesConfig{ IndexConfig{Name: "keywords", Weight: 100, Type: TypeBasic}, IndexConfig{Name: "date", Weight: 10, Type: TypeBasic}, }, } )
Functions ¶
This section is empty.
Types ¶
type Config ¶
type Config struct {
// Only include matches >= threshold, a normalized rank between 0 and 100.
Threshold int
// To get stable "See also" sections we, by default, exclude newer related pages.
IncludeNewer bool
// Will lower case all string values and queries to the indices.
// May get better results, but at a slight performance cost.
ToLower bool
Indices IndicesConfig
}
Config is the top level configuration element used to configure how to retrieve related content in Hugo.
func DecodeConfig ¶
DecodeConfig decodes a slice of map into Config.
type Document ¶
type Document interface {
// RelatedKeywords returns a list of keywords for the given index config.
RelatedKeywords(cfg IndexConfig) ([]Keyword, error)
// When this document was or will be published.
PublishDate() time.Time
// Name is used as an tiebreaker if both Weight and PublishDate are
// the same.
Name() string
}
Document is the interface an indexable document in Hugo must fulfill.
type FragmentKeyword ¶
type FragmentKeyword string
FragmentKeyword represents a document fragment.
func (FragmentKeyword) String ¶
func (f FragmentKeyword) String() string
type FragmentProvider ¶
type FragmentProvider interface {
Fragments(context.Context) *tableofcontents.Fragments
// For internal use.
ApplyFilterToHeadings(context.Context, func(*tableofcontents.Heading) bool) Document
}
FragmentProvider is an optional interface that can be implemented by a Document.
type IndexConfig ¶
type IndexConfig struct {
// The index name. This directly maps to a field or Param name.
Name string
// The index type.
Type string
// Enable to apply a type specific filter to the results.
// This is currently only used for the "fragments" type.
ApplyFilter bool
// Contextual pattern used to convert the Param value into a string.
// Currently only used for dates. Can be used to, say, bump posts in the same
// time frame when searching for related documents.
// For dates it follows Go's time.Format patterns, i.e.
// "2006" for YYYY and "200601" for YYYYMM.
Pattern string
// This field's weight when doing multi-index searches. Higher is "better".
Weight int
// A percentage (0-100) used to remove common keywords from the index.
// As an example, setting this to 50 will remove all keywords that are
// used in more than 50% of the documents in the index.
CardinalityThreshold int
// Will lower case all string values in and queries tothis index.
// May get better accurate results, but at a slight performance cost.
ToLower bool
}
IndexConfig configures an index.
func (IndexConfig) StringsToKeywords ¶
func (cfg IndexConfig) StringsToKeywords(s ...string) []Keyword
StringsToKeywords converts the given slice of strings to a slice of Keyword.
func (IndexConfig) ToKeywords ¶
func (cfg IndexConfig) ToKeywords(v any) ([]Keyword, error)
ToKeywords returns a Keyword slice of the given input.
type IndicesConfig ¶
type IndicesConfig []IndexConfig
IndicesConfig holds a set of index configurations.
type InvertedIndex ¶
type InvertedIndex struct {
// contains filtered or unexported fields
}
InvertedIndex holds an inverted index, also sometimes named posting list, which lists, for every possible search term, the documents that contain that term.
func NewInvertedIndex ¶
func NewInvertedIndex(cfg Config) *InvertedIndex
NewInvertedIndex creates a new InvertedIndex. Documents to index must be added in Add.
func (*InvertedIndex) Add ¶
func (idx *InvertedIndex) Add(ctx context.Context, docs ...Document) error
Add documents to the inverted index. The value must support == and !=.
func (*InvertedIndex) Search ¶
func (idx *InvertedIndex) Search(ctx context.Context, opts SearchOpts) ([]Document, error)
Search finds the documents matching any of the keywords in the given indices against query options in opts. The resulting document set will be sorted according to number of matches and the index weights, and any matches with a rank below the configured threshold (normalize to 0..100) will be removed. If an index name is provided, only that index will be queried.
type Keyword ¶
type Keyword interface {
String() string
}
Keyword is the interface a keyword in the search index must implement.
type SearchOpts ¶
type SearchOpts struct {
// The Document to search for related content for.
Document Document
// The keywords to search for.
NamedSlices []types.KeyValues
// The indices to search in.
Indices []string
// Fragments holds a a list of special keywords that is used
// for indices configured as type "fragments".
// This will match the fragment identifiers of the documents.
Fragments []string
}
SearchOpts holds the options for a related search.
type StringKeyword ¶
type StringKeyword string
StringKeyword is a string search keyword.
func (StringKeyword) String ¶
func (s StringKeyword) String() string
Source Files