Commit graph

11 commits

Author SHA1 Message Date
Luc Mcgrady
f5285f359a
Feat/Add legacy evaluate config bool (#4149)
* Feat/Add legacy evaluate config bool

* Minor tweaks based on PR suggestions (dae)

New enabling command:

from anki.config import Config
mw.col.set_config_bool(Config.Bool.FSRS_LEGACY_EVALUATE, True)
2025-07-04 15:32:09 +07:00
Luc Mcgrady
55ecbc1125
Feat/Health check (#4047)
* Message on low log loss

* make console.log permanent

* Added: Health check option

* disable button

* change health check conditions

* i18n

* ./check

* Apply suggestions from code review

Co-authored-by: user1823 <92206575+user1823@users.noreply.github.com>

* delete shadowed fsrs

* Update ts/routes/deck-options/FsrsOptions.svelte

Co-authored-by: llama <gh@siid.sh>

* Update ftl/core/deck-config.ftl

Co-authored-by: user1823 <92206575+user1823@users.noreply.github.com>

* anon's suggestions

* snake_case

* capital slow

* make global

* on by default

* Adjusted loss values

* Show message on pass

* ./check

* ComputeParamsRequest

* update coefficients

* update thresholds

* fix thresholds

* Apply suggestions from code review

---------

Co-authored-by: user1823 <92206575+user1823@users.noreply.github.com>
Co-authored-by: llama <gh@siid.sh>
Co-authored-by: Damien Elmes <dae@users.noreply.github.com>
2025-06-06 12:43:33 +07:00
Jarrett Ye
e096c462fa
Feat/FSRS-6 (#3929)
* Feat/FSRS-6

* update comment

* add decay to Card

* ./ninja fix:minilints

* pass check

* fix NaN in evaluation

* remove console

* decay should fallback to 0.5 when it's None.

* Update SimulatorModal.svelte

* Update a few comments

* Update FSRS decay defaults to use constants for better maintainability and clarity

* Update rslib/src/storage/card/data.rs
2025-04-25 16:44:34 +10:00
llama
769f302ea8
Fix "discard changes" appearing on deck options (#3689)
* patch original config after a preset is selected for the first time

* update comment
2025-01-07 17:41:52 +03:00
llama
d01d83e2c6
convert floats to single-precision before comparing in isModified (#3686) 2025-01-07 17:13:43 +03:00
Abdo
4dcbcdb0ce
Escape special search characters in preset name (#3554) 2024-11-07 08:32:00 +10:00
Damien Elmes
c45fa518d2 Use separate field to store FSRS params
Will allow the user to keep using old params with older clients
2024-10-21 18:13:23 +10:00
Hikaru Y.
694a5089a9
Fix 'Discard changes' dialog appearing even when no changes are made (#3495)
* Fix 'Discard changes' dialog appearing even when no changes are made

https://forums.ankiweb.net/t/anki-24-10-beta/49989/166

* Fix geometry of deck options window not being saved

evt.accept() does not seem to trigger reject().
2024-10-16 00:47:33 +10:00
Hikaru Y.
407e2dc43b
Fix issues with 'Discard changes' confirmation dialog (#3478)
* Prevent memory leak

* Fix deck option changes not detected until focus is lost

* Accurately determine if there are any pending changes

This makes it so that the confirmation dialog appears when it should,
and not when it shouldn't.
2024-10-06 14:21:00 +10:00
Arthur Milchior
b08e454f57
If deck options are modified, ask before closing (#3410)
* If deck options are modified, ask before closing

This imitates the way the note editor behaves. If a user assumes by
error that chanhges are automatically saved, it ensures they won't
lose them.

Also, this will eventually allows to have the same feature on
AnkiDroid. While, currently, we always ask the user whether they want
to close the deck options, even when there are no modification, which
seems to regularly frustate users (including myself).

I'm new to Svelte, please let me know whether there is a better way to
obtain the information from Svelte state that I missed.

Note that I ensured that only a boolean can be obtained. I didn't
cause the whole state to be accessible. May be useful for some
add-ons, I guess, but risks breaking too much things.

Regarding the deckoptions.py, I tried to imitate addcards.py way to
check whether the add card view can be closed. Reusing the same
function and variable name when possible.

* Update qt/aqt/deckoptions.py (dae)
2024-09-22 18:07:24 +10:00
Damien Elmes
9f55cf26fc
Switch to SvelteKit (#3077)
* Update to latest Node LTS

* Add sveltekit

* Split tslib into separate @generated and @tslib components

SvelteKit's path aliases don't support multiple locations, so our old
approach of using @tslib to refer to both ts/lib and out/ts/lib will no
longer work. Instead, all generated sources and their includes are
placed in a separate out/ts/generated folder, and imported via @generated
instead. This also allows us to generate .ts files, instead of needing
to output separate .d.ts and .js files.

* Switch package.json to module type

* Avoid usage of baseUrl

Incompatible with SvelteKit

* Move sass into ts; use relative links

SvelteKit's default sass support doesn't allow overriding loadPaths

* jest->vitest, graphs example working with yarn dev

* most pages working in dev mode

* Some fixes after rebasing

* Fix/silence some svelte-check errors

* Get image-occlusion working with Fabric types

* Post-rebase lock changes

* Editor is now checked

* SvelteKit build integrated into ninja

* Use the new SvelteKit entrypoint for pages like congrats/deck options/etc

* Run eslint once for ts/**; fix some tests

* Fix a bunch of issues introduced when rebasing over latest main

* Run eslint fix

* Fix remaining eslint+pylint issues; tests now all pass

* Fix some issues with a clean build

* Latest bufbuild no longer requires @__PURE__ hack

* Add a few missed dependencies

* Add yarn.bat to fix Windows build

* Fix pages failing to show when ANKI_API_PORT not defined

* Fix svelte-check and vitest on Windows

* Set node path in ./yarn

* Move svelte-kit output to ts/.svelte-kit

Sadly, I couldn't figure out a way to store it in out/ if out/ is
a symlink, as it breaks module resolution when SvelteKit is run.

* Allow HMR inside Anki

* Skip SvelteKit build when HMR is defined

* Fix some post-rebase issues

I should have done a normal merge instead.
2024-03-31 09:16:31 +01:00
Renamed from ts/deck-options/lib.ts (Browse further)