What happens to a comic with a corrupt page
Very little. The comic still imports and opens in Readwyn, and every other page shows as usual. Where the broken one should be, you get a box with a broken-image icon, the words “This page can’t be shown”, and the page's file name inside the archive, so you know which file to replace.
What counts as corrupt
Readwyn reads each page's size from the header of its image (JPEG, PNG, GIF, or WebP). A page is corrupt if those bytes give no size, or if its entry in the archive can't be read. It keeps its place in the reading order and still counts as a page. Readwyn never repairs or rewrites the archive, so the broken file is still there for you to replace.
A page can also fail later, when the reader asks for its image and extracting it goes wrong. That page gets the placeholder too, and Readwyn tries it again if you turn to another spread and come back.
What the placeholder looks like
It's a plain box with a thin outline, standing where the page would be, so the other page of a pair stays where it was. A corrupt page has no size of its own, so the box borrows the shape of the nearest page that has one, or 2:3 if none does. The footer counts it as usual (“p. 7 of 24”), and progress moves past it like any other page.
Oversized pages and archives
A small archive can declare entries that unpack to far more data, by accident or on purpose. Readwyn has limits for that:
- A page over 100 megapixels, going by its header, shows the placeholder.
- A page larger than 256 MiB isn't read.
- An archive whose pages unpack to more than 16 GiB in total isn't scanned or copied.
- An entry that unpacks to more than the archive says it holds is malformed, and reading stops one byte past the declared size.
Files that aren't pages
Files that aren't images never turn up as broken pages. ComicInfo.xml, text files, and Thumbs.db are skipped, and so is anything with a part of its path starting with a dot, such as ._page1.jpg, or inside the __MACOSX folder.
If that leaves an archive with no page images at all, it isn't added. The toast after “Add books” names the file and says “it has no page images.” The other books of the same import still arrive.
A broken ComicInfo.xml doesn't stop the import. The comic comes in under its file name, as one without a ComicInfo.xml does.