PageIndexProduct update
PageIndex · v0.5.1
PageIndex retrieves over a document tree without a vector database. v0.5.1 makes the default Flash parser keep one page in memory at a time on low-CPU machines or Type-3-font documents, cutting a 1665-page textbook peak from about 3.1 GB to 434 MB with an unchanged tree, and PDFs with stray brackets in font ToUnicode maps no longer hang until memory runs out.
Flash's parser now holds one page at a time instead of the whole document: a 1,665-page textbook peaks at 434 MB instead of 3.1 GB. The tree is unchanged.
Saved . A shortened excerpt from GitHub Releases; the original publisher retains authorship.