Skip to content

STORE_LISTING.md

Latest commit

 

History

History
68 lines (39 loc) · 9.51 KB

STORE_LISTING.md

File metadata and controls

68 lines (39 loc) · 9.51 KB
Raw

Store listing source

Status: v1.3.1 release copy — sideload tested and approved for publication. This document describes the release; submission, review and public availability must be verified separately in each Store.

Name: GetTranscript — Video Transcript Exporter

Short description:

Export video transcripts as VTT, SRT, text, Markdown or JSON. Preserve available speaker names. Processed locally on click.

Full description

Keep the conversation in a format you can use.

GetTranscript exports captions already available on a video page. Open the extension, choose a format, and download. A Caption track selector appears only when the player provides multiple separate tracks. Transcripts can contain several spoken languages. Choose WebVTT for caption tools, SRT for video editors, plain text or Markdown for notes, and JSON for structured processing.

On supported Microsoft Stream pages, GetTranscript first reads the recording's structured transcript and native speaker names without scrolling the transcript panel. If needed, it uses the player's captions and matches panel labels by text and timestamps. It preserves existing VTT voice tags on other supported players. Unmatched captions remain unnamed.

Meeting split into several recordings? Open More options → Combine recordings. Add up to 20 links from the same website, arrange the parts and choose one combined transcript or a ZIP of separate files. Find links on source offers Stream links already present on the page; it does not guess filenames. Speaker names are matched separately for each part.

Join recordings end to end using their full video durations, set start times to preserve known gaps, or keep each part’s original times in Markdown, text and JSON. Separate files always retain original timestamps. Small caption overruns of up to 5 seconds are preserved with a timing warning while the next part keeps its video-duration offset. Larger timing conflicts block the shared timeline. Recordings with different spoken languages can be combined without changing or translating their text. Markdown, text and JSON can include links back to the recordings.

Reading visits each recording in your open source tab, then restores its starting page and player state when you have not navigated elsewhere. Keep that tab open. Close the popup or workspace whenever you like: reading and exports continue, and reopening restores progress and download status. Pause and continue collection, or retry a failed part. Slow Stream players get up to 45 seconds to become ready, and supported Transcript controls open automatically. If help is needed, Open recording brings the source tab forward so you can open Transcript before retrying without a reload. The player fallback reports incomplete speaker coverage when it cannot expose further rows.

Everything is processed locally. There is no developer server, analytics, account registration or AI processing. Results, recording links and progress checkpoints stay in temporary browser session memory. Refresh replaces cached results; source-tab close, browser restart or extension reload clears them. Only export preferences are stored persistently. The extension uses temporary access to the source tab without permanent access to every website.

Requirements: a readable caption track in a supported HTML5 player, or a standard Microsoft Stream recording page hosted on SharePoint with an accessible transcript. Caption languages are discovered without a language whitelist. A transcript containing multiple spoken languages remains together; no single language is assigned to the whole conversation in the interface. The interface is in English; captions retain their original language and text direction. For embedded videos, open the original player page. For pages without captions, GetTranscript cannot generate a transcript from audio. Live caption sources may expose only a rolling window.

GetTranscript is independent of Microsoft and Google. Export only content you are authorized to keep.

Permission justifications

  • activeTab: access the source tab after the user opens the extension, including user-selected recordings on the same origin during a collection.
  • scripting: retrieve the current Stream recording's structured transcript, native caption cues, page-created caption blobs, media duration, exposed recording links and transcript speaker labels; remember and restore player state in the selected tab.
  • downloads: save the user’s chosen transcript file and verify completion.
  • storage: remember output format and speaker display preferences locally; retain per-tab results, recording collections, source URLs, progress checkpoints and download status in temporary in-memory session storage. Refresh replaces cached content; closing the source tab clears its results and collection.

Single purpose: export available video captions and transcript labels into user-selected local text formats.

Remote code: none. All executable code is included in the package.

Data use: page titles, recording URLs, durations, captions and speaker names are processed locally for export. Do not declare that no data is accessed: it is accessed, but not collected by the developer or transmitted for processing. Complete Store privacy disclosures consistently with PRIVACY.md and the current form wording.

Stream requests use existing page access. When required, a temporary observer can reuse the player's authorization header only for the same origin and recording; it restores the observer and changed panel state after reading. Authentication values are never saved in extension storage or included in exports. No additional permission, account registration or cloud processing is introduced.

Disclose website content, personal communications and personally identifiable information (speaker names). These categories are processed locally, with no developer collection or cloud processing.

Reviewer test instructions

No extension account, payment or developer credentials are required. Open https://iandevlin.github.io/mdn/video-player-with-captions/video-with-captions.html in Chrome or Microsoft Edge, then open GetTranscript. Select English, German or Spanish and download VTT, SRT, text, Markdown or JSON. This public sample exposes 14 cues per language and has no speaker labels, so the speaker option is unavailable as expected. Close and reopen the popup during reading or after downloading to verify that it reconnects to the same result. Refresh starts a new scan; closing the video tab clears the session. For the Stream adapter, open a SharePoint-hosted recording with an accessible transcript using your own authorized Microsoft account. Begin with the transcript panel closed. Structured transcript data preserves native names without row scrolling; the fallback uses structural controls and locale-aware numeric timestamps. Allow up to 45 seconds for a slow player to become ready. For a failed part, Open recording brings the source tab forward; open Transcript if needed, then Retry this part. An already-open recording is retried without a reload. Missing or ambiguous names remain unnamed, and incomplete panel coverage is reported. The extension cannot generate captions from audio or provide access to recordings the user cannot open.

For multipart review, use two recordings that your own account can open on the same SharePoint origin. Open the first, select More options → Combine recordings, add the second link, confirm order and select Read recordings. The source tab visits both and returns to its starting page; keep it open. Close and reopen the workspace to check retained results, or Pause and Continue a longer collection. Export combined Markdown and VTT, then separate SRT files as a ZIP. Verify full-duration offsets, per-part names, and original timestamps in the ZIP. In Advanced options, try custom starts, original-time Markdown/TXT/JSON, recording links. Verify that recordings with different spoken languages combine without an additional language setting. Combined JSON uses schema version 2. Refresh collection retains its queue and options while clearing captured content. Closing the source tab clears all its temporary collection data.

Listing maintenance

Local sideload validation is complete and the user has approved v1.3.1 publication. Update the existing Chrome and Edge products with the same verified release ZIP. Verify the current publisher dashboard and public listing before recording a new publication result. The v1.3.1 copy here is listing content, not evidence of submission. Use the verified developer/support contact and public privacy-policy URL, review declarations, and use screenshots containing only fictional meeting content. Do not reuse another extension’s product ID.

Regenerate the two 1280 × 800 screenshots from current fictional demos with node scripts/store-artwork.mjs (set BROWSER_BIN to an installed Chromium browser when needed). The script opens a local Vite server, checks the format and advanced controls, and captures both views at 1× scale. It does not build or package the extension.

Public links