ankiclient: add four new aggregate frequency markers #238
Conversation
|
This needs to be fixed. I recommend you download Qt Creator rather than editing .ui files directly.
|
Fixed in c5671f0. For some reason, Qt Creator mangled the whole file, rendering the diff unreadable so I manually post-processed the XML file to format it in the same order as originally. Additionally, since the description labels have no word wrapping, I omitted the very last part of the descriptions listed on Yomitan's page. |
precondition
force-pushed
the
aggregate-frequency-markers
branch
from
4e1a940 to
c5671f0
Compare
|
Looks good. My only request is that you squash all your changes into one commit. You can just I do have a quick question though. It looks like the only difference between rank and occurrence tags is the default value. Is that all it's supposed to be? |
This commit adds support for 4 new markers:
1. {frequency-harmonic-rank} (Default: 9999999)
2. {frequency-harmonic-occurrence} (Default: 0)
3. {frequency-average-rank} (Default: 9999999)
4. {frequency-average-occurrence} (Default: 0)
These markers aggregate the term frequency information of multiple
dictionaries into a single numeric value, often used for sorting Anki
cards by frequency.
precondition
force-pushed
the
aggregate-frequency-markers
branch
from
c5671f0 to
6fb7b0e
Compare
It is funny you say that because I wondered the same thing. When I asked the question in the At the moment, it is expected that the user enables exclusively rank-based or occurrence-based dictionaries or else the measure will be totally wrong due to averaging numbers from different scales. A possible extension one can consider is to infer the frequency sorting mode (a simple heuristic checking whether the frequency of の is a single digit number or a multi-digit number would be sufficient I think) and automatically exclude dictionaries that are in a different scale from the computation. |
I think trusting the user to figure it out on their own is better than employing imperfect heuristics. Everything looks good. Merging this. |
This commit adds support for 4 new markers (introduced in Yomitan in yomidevs/yomitan#600):
These markers aggregate the term frequency information of multiple dictionaries into a single numeric value, often used for sorting Anki cards by frequency.