Automatic file splitting, whether by time, size or hotkey will now correctly update the formatting too. This works by hijacking the default file splitting, as a result it might be slightly less precise in particular for the size one. Improved the handling of pyparsing import failures, with a descriptive message and a procedure to report such issues with relevant info. Possible fix of the pyparsing import on Macos.
Changed filename cropping to be performed only post-generation, and to behave correctly with multi-bytes characters Added zero-width spaces sanitisation (this can be changed with the remove_zws customisation parameter) Fixed default data arrays (counter) causing crashes : Electric bogaloo, this time I think I found out what the actual fuck is happening with them. Fixed ctypes wrapping on macos Added version number printing at script load From now on, to limit build actions, executables of the custom build needed for replay buffer capacities will be distributed in releases.
Added sanitization to Windows filenames : the characters * " < > : | ? will be remplaced by an underscore. Fixed a potential crash on reloading script after adding it.
* " < > : | ?
Added counters, which increase every time a file is saved. Added tokens for the spelled out month and day name, $Month$ and $Day$, as well as for abbreviated month and day, $month$ and $day$. Fixed a minor error when from_file wasn't defined for newly created text sources
$Month$
$Day$
$month$
$day$
from_file
Added v$[file_text]$ as token for text sources, for when the source is set to read from file. Added v$username$ as token, standing for the user's username. Added the ability to add tokens for specific sources without procs. Limited the generated filename to 250 characters, so as to not overwrite the extension.
v$[file_text]$
v$username$
Added v$user$ as token, standing for the user's Home directory.
v$user$
