Potential improvement to 7z extraction

I have a 1.4GB 7zip archive that consists of a single solid block. In the root of the archive are 11 directories.

When I extract the archive using the 7zip context menu or such, the operation takes around 15 seconds. When I open the archive in fman then select and copy all the files (this is the only way I could figure out how to extract within fman), the operation takes around three minutes. When I use a plugin I'm working on to extract the entire archive, the operation takes around 15 seconds.

Looking at the code I couldn't work out exactly what goes on when using the open-select-copy method. But I'm guessing that each selected object is extracted with a separate call to 7za, which is inefficient due to the nature of solid archives.

If this is the case, a possible alternative would be to provide 7za with a "list file" containing the selected files, so they can all be extracted with one call. I tested this with a modification to my plugin and extracting the same files as above took around 15 seconds.

Trace the open-select-copy extraction entry point and how it invokes 7za; the issue provides no file or test names. Compare that path with the plugin's single-call extraction approach, then verify that extracting all selected files from a solid archive uses one efficient operation and preserves existing extraction behavior.