Ksable's Nest

Code is endless to write, and bugs are also endless to fix.

This article is a practical tutorial on creating a crawler for a comic website, detailing how to use Python to scrape comic resources from the site. The tutorial starts by analyzing the website's dynamic loading mechanism, then step by step explains the process of obtaining the comic ID (galleryid), parsing comic information (galleryinfo), and generating image download links. It focuses on how to construct image URLs by processing JavaScript functions (such as `url_from_url_from_hash`) and variables, and provides a complete Python script with features including exception handling, proxy settings, and safe file naming. Finally, it achieves batch downloading and local saving of comic images (supporting WebP format and metadata recording), making it suitable for crawler enthusiasts with some Python foundation to refer to and learn from.

I've spent half a day finally getting the website's Service Worker cache configuration set up. Previously, I always thought the website's speed wasn't fast enough, often having to wait several seconds for it to load. Background images would always take a delay before loading properly. I've already done my best to maximize the network loading speed, and the first load has been optimized to the fastest possible speed. I've also added cache-control response headers to all resources I can control. However, when opening the site again, some images and JS files still aren't cached, causing them to briefly appear white before loading. Today I went through the hassle of configuring the Service Worker...

Open-source project Alist seems to have been sold, with the author selling it without a word to a "black-market" company. That "black-market" company is said to have implanted a lot of spam code, collecting user information, and even hiring penetration testers. Now the original project's issue section is filled with complaints.转载（点我展开） Answer a few questions: 0. The project was acquired by a trash company, and only one person knew about it, while other project members were unaware. Spam ads were inserted, and private information was collected! Please stop...

Today is the last day of the college entrance exam. May the candidates have smooth sailing and answer every question they know and guess every one they don't. Next year, when this day arrives, it will be me taking the exam. (捂脸) Now my homework isn't done yet, but I'll go back to school tomorrow. I scraped the m3u8 file for "Linglong Season 2 Episode 4" from a certain website, used Python to download all the TS files, and then spent a lot of effort to find a website that could merge the TS videos. The entire process was extremely troublesome, with none of the three steps—scraping, downloading, and merging—going smoothly. A certain website has an anti-crawling restriction mechanism, and Python's... 2025-06-09 Rambling | Read more

A photo taken in the evening during the Strongest Voice Concert on the 11th.

When removing HTML tags in JavaScript, you must absolutely make sure to not only remove <> but also eliminate escaped HTML entities like & and <. Otherwise, a very strange BUG may occur, where HTML tags unexpectedly appear, making it difficult to troubleshoot for half an hour (what the browser displays is <>). Without this experience, it's really hard to identify such a BUG.

From Wang Miao's perspective, Yang Dong, although a woman he admired, was no different from other scientists in "Science Boundary"—she committed suicide due to the deception of the Zhi Zi's schemes. But is the truth really that simple?

Let everyone see the gifts I received for my birthday.

Today is April 20th, yesterday was April 19th. The lunar calendar date of yesterday was March 10th, and my mom said it was my birthday. On Saturday,...