Types

Result

Extraction result containing text, metadata, and media information.

type Result struct {
    Text           string        `json:"text"`
    Title          string        `json:"title"`
    Images         []ImageInfo   `json:"images,omitempty"`
    Links          []LinkInfo    `json:"links,omitempty"`
    Videos         []VideoInfo   `json:"videos,omitempty"`
    Audios         []AudioInfo   `json:"audios,omitempty"`
    ProcessingTime time.Duration `json:"-"`       // Processing duration (excluded from standard serialization)
    WordCount      int           `json:"word_count"`
    ReadingTime    time.Duration `json:"-"`       // Estimated reading time (excluded from standard serialization)
}

Fields

Field	Type	Description
`Result.Text`	`string`	Extracted plain-text body
`Result.Title`	`string`	Document title (`<title>`; falls back to the first `<h1>`/`<h2>` when missing)
`Result.Images`	`[]ImageInfo`	Images in document order; empty when `PreserveImages=false`
`Result.Links`	`[]LinkInfo`	Links in document order; empty when `PreserveLinks=false`
`Result.Videos`	`[]VideoInfo`	Extracted video sources; empty when `PreserveVideos=false`
`Result.Audios`	`[]AudioInfo`	Extracted audio sources; empty when `PreserveAudios=false`
`Result.ProcessingTime`	`time.Duration`	Wall-clock time of this extraction (`processing_time_ms` in JSON output)
`Result.WordCount`	`int`	Word count of `Text` (whitespace tokenization; CJK characters counted one by one)
`Result.ReadingTime`	`time.Duration`	Reading time estimated from `WordCount` at 200 words per minute (`reading_time_ms` in JSON output)

MarshalJSON

Custom JSON serialization is implemented by Result.MarshalJSON: ProcessingTime and ReadingTime carry json:"-" tags (standard serialization skips them), but Result.MarshalJSON emits them as millisecond values.

func (r *Result) MarshalJSON() ([]byte, error)

WARNING

Result does not implement UnmarshalJSON. If you deserialize the output of MarshalJSON() back into a Result, duration fields such as ProcessingTime and ReadingTime will be lost — the JSON output key names (processing_time_ms, reading_time_ms) do not match the struct field names, so they cannot be restored.

This is intentional: the JSON format is designed for external consumption (e.g., API responses, logs, frontend display), not for bidirectional serialization.

ImageInfo

Element type of Result.Images ([]ImageInfo). ImageInfo holds the extraction information of a single <img>.

type ImageInfo struct {
    URL          string `json:"url"`           // Image URL
    Alt          string `json:"alt"`           // Alternative text
    Title        string `json:"title"`         // Title
    Width        string `json:"width"`         // Width
    Height       string `json:"height"`        // Height
    IsDecorative bool   `json:"is_decorative"` // Whether the image is decorative
    Position     int    `json:"position"`      // Position in the document
}

Fields

Field	Type	Description
`ImageInfo.URL`	`string`	Image URL (`src` attribute)
`ImageInfo.Alt`	`string`	Alternative text (`alt` attribute)
`ImageInfo.Title`	`string`	Advisory title (`title` attribute)
`ImageInfo.Width`	`string`	Raw width attribute (not parsed into a number)
`ImageInfo.Height`	`string`	Raw height attribute (not parsed into a number)
`ImageInfo.IsDecorative`	`bool`	`true` when `Alt` is empty (decorative image)
`ImageInfo.Position`	`int`	1-based index in the extracted content (`0` when not located)

Field Semantics

Field	Description
`URL`	The `src` attribute value of the image; only valid URLs (internal validation, not exported) are included — `<img>` with an invalid URL does not appear in the result
`Alt`	The original `alt` attribute; when empty, `IsDecorative` is `true`
`Title`	The original `title` attribute (not the page title)
`Width`/`Height`	The raw strings from the HTML attributes (e.g. `"640"`, `"50%"`), not parsed into numbers — different pages may use inconsistent formats
`IsDecorative`	`true` when `Alt` is empty; can be used to identify and skip decorative images
`Position`	1-based index in the document; when `PreserveImages = false`, the entire `Images` slice is empty

Width/Height are string types

Width and Height are string types, not int. They preserve the original representation from the HTML source (which may include units, percentages, etc.). Parse them into numbers yourself when needed.

LinkInfo

Element type of Result.Links ([]LinkInfo). LinkInfo holds the extraction information of a single <a>.

type LinkInfo struct {
    URL        string `json:"url"`         // Link URL
    Text       string `json:"text"`        // Link text
    Title      string `json:"title"`       // Link title
    IsExternal bool   `json:"is_external"` // Whether it is an external link (determined by whether the URL itself is an absolute external URL, not by comparison with BaseURL)
    IsNoFollow bool   `json:"is_nofollow"` // Whether the link is nofollow
    Position   int    `json:"position"`    // Position in the document
}

Fields

Field	Type	Description
`LinkInfo.URL`	`string`	Link URL (`href` attribute)
`LinkInfo.Text`	`string`	Visible link text (including text inside nested tags)
`LinkInfo.Title`	`string`	Advisory title (`title` attribute, not the link text)
`LinkInfo.IsExternal`	`bool`	Whether the link is external
`LinkInfo.IsNoFollow`	`bool`	`true` when `rel` contains `nofollow`
`LinkInfo.Position`	`int`	1-based index in the extracted content (invalid `<a>` still consume index numbers, so values may be non-contiguous)

Field Semantics

Field	Description
`URL`	The `href` attribute value; only valid URLs (internal validation, not exported) are included — an `<a>` with an invalid URL still consumes a Position but is not added to the slice
`Text`	Recursive concatenation of all text nodes inside the `<a>` tag (including text inside nested tags)
`Title`	The original `title` attribute (not the link text)
`IsExternal`	Determined by whether the URL itself is an absolute external address, not by domain comparison with `BaseURL` — this differs from the internal/external link determination in `ExtractAllLinks`
`IsNoFollow`	`true` when the `rel` attribute contains `nofollow` (case-insensitive, ASCII-fold matching)
`Position`	1-based index in the document; invalid `<a>` (illegal or missing href) still consume index numbers but are not added to the slice, so Position values may be non-contiguous

VideoInfo

Element type of Result.Videos ([]VideoInfo). VideoInfo holds the extraction information of a single video source.

type VideoInfo struct {
    URL      string `json:"url"`      // Video URL
    Type     string `json:"type"`     // Video type
    Poster   string `json:"poster"`   // Poster image URL
    Width    string `json:"width"`    // Width
    Height   string `json:"height"`   // Height
    Duration string `json:"duration"` // Duration
}

Fields

Field	Type	Description
`VideoInfo.URL`	`string`	Video source URL
`VideoInfo.Type`	`string`	Video type (file-container MIME or `"embed"`; see the value rules below)
`VideoInfo.Poster`	`string`	Poster frame URL of the `<video>` (`poster` attribute)
`VideoInfo.Width`	`string`	Raw width attribute (not parsed into a number)
`VideoInfo.Height`	`string`	Raw height attribute (not parsed into a number)
`VideoInfo.Duration`	`string`	Raw duration attribute (not parsed into a number)

Type Field Value Rules

Type value	Meaning	When it occurs
`"embed"`	A video page referenced by an iframe	Embedded players such as YouTube, Vimeo, Youku, Bilibili
MIME type (e.g. `"video/mp4"`)	Video file container	Attribute value from `<source type="video/mp4">`
Empty string	No type detected	`<video src="...">` specifying the source directly with no `<source>` child element

Three sources of video extraction

Video extraction runs in three steps — raw HTML scan, DOM traversal, regex fallback — with deduplication at each step. See Media Extraction Guide.

AudioInfo

Element type of Result.Audios ([]AudioInfo). AudioInfo holds the extraction information of a single audio source.

type AudioInfo struct {
    URL      string `json:"url"`      // Audio URL
    Type     string `json:"type"`     // Audio type
    Duration string `json:"duration"` // Duration
}

Fields

Field	Type	Description
`AudioInfo.URL`	`string`	Audio source URL
`AudioInfo.Type`	`string`	Audio type (file-container MIME; see the value rules below)
`AudioInfo.Duration`	`string`	Raw duration attribute (not parsed into a number)

Type Field Value Rules

Type value	When it occurs
MIME type (e.g. `"audio/mpeg"`)	Attribute value from `<source type="audio/mpeg">`
Empty string	`<audio src="...">` specifying the source directly with no `<source>` child element

Duality of the .ogg extension

An OGG container can carry video or audio; a .ogg URL will appear in both Videos and Audios. The audio-only variant .oga appears only in Audios.

LinkResource

Link resource, the element type of the ExtractAllLinks function family and GroupLinksByType.

type LinkResource struct {
    URL   string // Link URL (already resolved against BaseURL into an absolute address when ResolveRelativeURLs is enabled)
    Title string // Link title
    Type  string // Link type: "link", "image", "video", "audio", "css", "js", "icon", "media"
}

Fields

Field	Type	Description
`LinkResource.URL`	`string`	Resource URL; resolved into an absolute address against `BaseURL` when `ResolveRelativeURLs=true`
`LinkResource.Title`	`string`	Human-readable resource label
`LinkResource.Type`	`string`	Resource type category (`link`/`image`/`video`/`audio`/`css`/`js`/`icon`/`media`)

For the HTML tag sources and control switches corresponding to each Type value, see Link Extraction.

Statistics

Processing statistics, returned by Processor.GetStatistics.

type Statistics struct {
    TotalProcessed     int64         // Total processed count
    CacheHits          int64         // Cache hit count
    CacheMisses        int64         // Cache miss count
    ErrorCount         int64         // Error count
    AverageProcessTime time.Duration // Average processing time
}

Fields

Field	Type	Description
`Statistics.TotalProcessed`	`int64`	Number of extractions that completed without error (including cache hits)
`Statistics.CacheHits`	`int64`	Number of times served directly from the cache
`Statistics.CacheMisses`	`int64`	Number of misses requiring full processing
`Statistics.ErrorCount`	`int64`	Number of extractions that returned an error
`Statistics.AverageProcessTime`	`time.Duration`	Average wall-clock time per extraction

BatchResult

Batch processing result, the return type of the ExtractBatch function family and methods.

type BatchResult struct {
    Results   []*Result // Extraction results, nil on failure or cancellation
    Errors    []error   // Failed errors
    Success   int       // Success count
    Failed    int       // Failure count
    Cancelled int       // Cancelled count
}

Fields

Field	Type	Description
`BatchResult.Results`	`[]*Result`	Extraction results in one-to-one correspondence with the input; failed/cancelled items are `nil`
`BatchResult.Errors`	`[]error`	Errors in one-to-one correspondence with the input; successful items are `nil`
`BatchResult.Success`	`int`	Count of successful items
`BatchResult.Failed`	`int`	Count of failed items
`BatchResult.Cancelled`	`int`	Count of items left unprocessed due to context cancellation

For field semantics, cancellation behavior, and input validation see Batch Processing.

NodeAttr

HTML node attribute, the element type of ContentNode.Attrs ([]NodeAttr). NodeAttr has only two fields, Key and Value.

type NodeAttr struct {
    Key   string // Attribute name
    Value string // Attribute value
}

Fields

Field	Type	Description
`NodeAttr.Key`	`string`	Attribute name
`NodeAttr.Value`	`string`	Attribute value

Types ​

Result ​

Fields ​

MarshalJSON ​

ImageInfo ​

Fields ​

Field Semantics ​

LinkInfo ​

Fields ​

Field Semantics ​

VideoInfo ​

Fields ​

Type Field Value Rules ​

AudioInfo ​

Fields ​

Type Field Value Rules ​

LinkResource ​

Fields ​

Statistics ​

Fields ​

BatchResult ​

Fields ​

NodeAttr ​

Fields ​

Types

Result

Fields

MarshalJSON

ImageInfo

Fields

Field Semantics

LinkInfo

Fields

Field Semantics

VideoInfo

Fields

Type Field Value Rules

AudioInfo

Fields

Type Field Value Rules

LinkResource

Fields

Statistics

Fields

BatchResult

Fields

NodeAttr

Fields