Currently, the text has to be extracted every time a search starts. It would be much faster for successive searches if the content is cached like what ripgrep-all does.
The extracted text could be stored in a SQLite database indexed by a hash of file path and modification date.
Currently, the text has to be extracted every time a search starts. It would be much faster for successive searches if the content is cached like what ripgrep-all does.
The extracted text could be stored in a SQLite database indexed by a hash of file path and modification date.