parser

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

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

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

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

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

type ExtractedData struct {
	URL         string            `json:"url"`
	Title       string            `json:"title"`
	Description string            `json:"description,omitempty"`
	Type        string            `json:"type"` // product, article, listing, generic
	Data        []map[string]any  `json:"data"`
	Meta        map[string]string `json:"meta,omitempty"`
	Links       []ExtractedLink   `json:"links,omitempty"`
	Images      []ExtractedImage  `json:"images,omitempty"`
	Tables      []ExtractedTable  `json:"tables,omitempty"`
	JSONLD      []map[string]any  `json:"json_ld,omitempty"`
}

type ExtractedImage struct {
	URL string `json:"url"`
	Alt string `json:"alt,omitempty"`
}

type ExtractedLink struct {
	Text string `json:"text"`
	URL  string `json:"url"`
	Rel  string `json:"rel,omitempty"`
}

type ExtractedTable struct {
	Headers []string            `json:"headers"`
	Rows    []map[string]string `json:"rows"`
}

type Parser interface {
	// Parse extracts items and follow-up URLs from a response.
	// It returns scraped items, discovered links, and any error.
	Parse(resp *types.Response, rules []config.ParseRule) ([]*types.Item, []string, error)
}

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

type SelectorCandidate struct {
	Selector    string  `json:"selector"`
	Specificity int     `json:"specificity"` // Higher = more specific
	MatchCount  int     `json:"match_count"` // How many elements this matches
	Score       float64 `json:"score"`       // Confidence score (0-1)
}

type SimilarElement struct {
	Selector   string  `json:"selector"`
	Text       string  `json:"text"`
	Similarity float64 `json:"similarity"` // 0 to 1
	Tag        string  `json:"tag"`
}

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

type StructuredData struct {
	Type StructuredDataType `json:"type"`
	Data map[string]any     `json:"data"`
	Raw  string             `json:"raw,omitempty"`
}

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

type StructuredDataType string

type TraversalResult struct {
	Text      string
	HTML      string
	Attribute string
	Tag       string
	Children  []TraversalResult
}

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