marvnc / jp-resources Goto Github PK

first off, thank you for making that back filling application. I have been struggling for a while with a morphman that just would not apply any type of frequency ordering, so switching to a new system will be helpful.

Something I had difficulties with when trying to backfill my collection was the fact that my deck had furigana in its only expression field.
I was eventually able to hack together a solution that works for me, since I only had to remove anything that lies between [ these types of brackets].
If you are interested in doing a minor update to this program, could I ask that you make furigana ignoring backfill an official feature?
My implementation was:

rx_HTML = re.compile("<.*?>")
rx_Furigana = re.compile(r"\[.*?\]")

def normalize_expr(expression: str):
    # removes HTML and surrounding whitespace
    result = re.sub(rx_Furigana, '', expression).strip()
    result = re.sub(rx_HTML, '', result).strip()
    
    return result

but of course something like this should probably go behind a proper command-line flag instead of being applied automatically. I would make a pull request with this myself, except I have never made a program with a command-line interface in python before. there is also the issue of making this furigana support work across multiple types of formatting, as I remember seeing some decks that use other pairs of brackets for this instead.
Either way, many thanks!

I just noticed that all the words I mined today had 999999 freq even though I didnt though the handlebars and it was working normally before. i dont know when the handlebars stopped grabbing the frequencies if that makes sense. very common words arent getting sorted and just get 999999 freq. I tried updating ankiconnect and anki itself but same thing. help!