Name Description Size Coverage
DOMExtractor.sys.mjs @see {extractTextFromDOM} for a high level overview of this file. 38374 -
metrics.yaml 4131 -
moz.build 559 -
PageExtractor.d.ts Reader mode doesn't provide the types for the result. This a stub for making the results easier to interpret. 4015 -
PageExtractorChild.sys.mjs @import { GetTextOptions, CanvasSnapshot, ExtractionResult, PageMetadata, ReaderModeDocument } from './PageExtractor.d.ts' @import { PageExtractorParent } from './PageExtractorParent.sys.mjs' @import { PageExtractorPhase, TraceId } from './PageExtractorEvents.sys.mjs' 20903 93 %
PageExtractorEvents.sys.mjs Formats only the options the caller passed to getText(), so the marker shows what was selected, not defaults it never applied. `sourceUrl` is excluded: it goes on its own "url"-format field, which the profiler sanitizes. Anything added here lands in a "string" field that it does not, so keep URLs and paths out. @param {Record<string, any>} options 14171 -
PageExtractorParent.sys.mjs @import { HiddenFrame } from "resource://gre/modules/HiddenFrame.sys.mjs" @import { GetTextOptions, ExtractionResult, PageMetadata } from './PageExtractor.d.ts' @import { PageExtractorChild } from './PageExtractorChild.sys.mjs' @import { PageExtractorPhase, TraceId } from './PageExtractorEvents.sys.mjs' 16756 90 %
pings.yaml 1174 -
tests -
YouTubeExtraction.sys.mjs Extracts a YouTube watch page into a compact, labeled block. YouTube renders the transcript into an engagement panel only after the user opens it, so this module opens that panel (by activating the "Show transcript" control), reads the rendered segments, then closes the panel again. Reading the rendered DOM avoids having to intercept and decode YouTube's `get_transcript`/`get_panel` network payloads. Two transcript segment element generations currently coexist and are both supported: - New: `transcript-segment-view-model` - Legacy: `ytd-transcript-segment-renderer` 24598 -