Comic Book ZIP (CBZ) converter

A CBZ file is an ordinary ZIP archive full of page images, renamed so that comic reader apps recognise it. There is no special structure: readers sort the images by file name and show them one by one. Most CBZ files hold JPEG scans, newer ones PNG or WebP, often inside a folder named after the issue. Many also carry a ComicInfo.xml with the series, issue number, credits and whether the book reads right to left.

Page order is the classic trap. Names like page1.jpg, page2.jpg, page10.jpg sort as 1, 10, 2 in tools that compare letters, and archives made on a Mac often include __MACOSX folders of resource-fork files that some readers show as blank pages; Windows adds Thumbs.db. We sort by the numbers inside names, keep only real images and use ComicInfo.xml for the title.

Converting is mostly about where you want to read. PDF opens on any device and suits printing; we copy JPEG scans into it without recompressing them, so the PDF looks exactly like the archive. A fixed-layout EPUB 3 suits reading apps that shelve EPUB but ignore CBZ; it keeps one scan per screen, marks the cover and respects right-to-left manga flags. Neither format can make the speech balloons editable, because a scan has no text layer.

Full issues are big: the public-domain issues we tested were 5 to 37 MB each. Archives are read in memory with limits on the number of files, the size each file expands to, and the pixel count of each page, so a ZIP bomb or a 20000 × 20000 pixel page is refused instead of exhausting the server.