extract

type Attr struct {
	// The HTML attribute to extract from each part.
	Attr string
	//BaseURL specifies the base URL to use for all relative URLs contained within a document.
	BaseURL string
	// By default, if there is only a single attribute extracted, AttrExtractor
	// will return the match itself (as opposed to an array containing the single
	// match). Set AlwaysReturnList to true to disable this behaviour, ensuring
	// that the Extract function always returns an array.
	AlwaysReturnList bool

	// If no parts with this attribute are found, then return the empty list from
	// Extract, instead of  'nil'.  This signals that the result of this
	// Part should be included to the results, as opposed to omitting
	// the empty list.
	IncludeIfEmpty bool
	//Filters are used to manipulate HTML attribute when extracting.
	//Currently the following filters are available:
	//upperCase makes all of the letters in the Attr uppercase.
	//lowerCase makes all of the letters in the Attr lowercase.
	//capitalizes the first letter of each word in the Attr
	//trim returns a copy of the Attr, with all leading and trailing white space removed
	Filters []string
}

type Const struct {
	// The value to return when the Extract() function is called.
	Val interface{}
}

type Count struct {
	// If no parts with this attribute are found, then return a number from
	// Extract, instead of 'nil'.  This signals that the result of this
	// Part should be included to the results, as opposed to omitting
	// the empty list.
	IncludeIfEmpty bool
}

type Extractor interface {
	// Extract some data from the given Selection and return it.  The returned
	// data should be encodable - i.e. passing it to json.Marshal should succeed.
	// If the returned data is nil, then the output from this part will not be
	// included.
	//
	// If this function returns an error, then the scrape is aborted.
	Extract(*goquery.Selection) (interface{}, error)
}

type OuterHtml struct{}

type Regex struct {
	// The regular expression to match.  This regular expression must define
	// exactly one parenthesized subexpression (sometimes known as a "capturing
	// group"), which will be extracted.
	Regex *regexp.Regexp
	// The subexpression of the regex to match.  If this value is not set, and if
	// the given regex has more than one subexpression, an error will be thrown.
	Subexpression int

	// When OnlyText is true, only run the given regex over the text contents of
	// each part in the selection, as opposed to the HTML contents.
	OnlyText bool

	// By default, if there is only a single match, Regex will return
	// the match itself (as opposed to an array containing the single match).
	// Set AlwaysReturnList to true to disable this behaviour, ensuring that the
	// Extract function always returns an array.
	AlwaysReturnList bool

	// If no matches of the provided regex could be extracted, then return the empty list
	// from Extract, instead of 'nil'.  This signals that the result of
	// this Part should be included to the results, as opposed to
	// omitting the empty list.
	IncludeIfEmpty bool
}

type Text struct {
	// If text is empty in the selection, then return the empty string from Extract,
	// instead of 'nil'.  This signals that the result of this Part
	// should be included to the results, as opposed to omitting the
	// empty string.
	IncludeIfEmpty bool
	//Filters are used to manipulate Text data when extracting.
	//Currently the following filters are available:
	//upperCase makes all of the letters in the selected text  uppercase.
	//lowerCase makes all of the letters in the selected text lowercase.
	//capitalize capitalizes the first letter of each word in the selected text
	//trim returns a copy of the text, with all leading and trailing white space removed
	Filters []string
}