Vasil Smutok, a Ukrainian Python developer, has created and released the ytscrape library—a lightweight, open-source tool for parsing YouTube. This solution enables the retrieval of subtitles, comments, and detailed metadata for videos and channels without the need for official API keys or resource-intensive browsers.
Vasil Smutok, a Ukrainian Python developer, has created and released the ytscrape library—a lightweight, open-source tool for parsing YouTube. This solution enables the retrieval of subtitles, comments, and detailed metadata for videos and channels without the need for official API keys or resource-intensive browsers.
Origin of the Idea
As Vasil shared on LinkedIn, the project’s journey began with a practical need: he required the implementation of in-depth analysis and the creation of textual summaries from YouTube videos. This went beyond a brief overview, demanding the ability to search for specific information within videos, work with transcripts, analyze comments, and gather extensive data about the channel itself.
While searching for existing tools, the developer encountered three primary limitations:
The YouTube Data API has strict quota limitations;
most available libraries addressed only isolated tasks (e.g., subtitles separately or metadata separately);
some tools required the execution of headless browsers (Playwright or Selenium), which significantly complicated and slowed down development.
Aiming to create a lightweight, free, and maximally simple-to-integrate solution, Vasil Smutok developed his own library, ytscrape. It functions by performing a single download of the InnerTube context and subsequently interacting through internal requests, eliminating the need for complex environment configuration.
ytscrape Capabilities
The tool offers a full range of capabilities for data collection from the platform:
Rapid navigation and collection of results for videos, channels, and playlists;
Extraction of detailed information about videos and author pages;
Parsing of comments and replies under videos;
Retrieval of subtitles in a convenient format;
Support for both synchronous and asynchronous APIs;
Automatic pagination and parsing of responses into frozen dataclasses.
Currently, ytscrape is in the status of a personal pet project and an experiment created for the author’s specific personal needs. Despite this, the library is fully ready for use by other developers who require quick and free access to YouTube data.
You can explore the project on GitHub.
Источник: www.vesti-ua.net