GitHub/Anki
1
0
Fork 0
mirror of https://github.com/ankitects/anki.git synced 2026-01-19 09:22:51 -05:00
Code Issues Projects Releases Packages 1 Wiki Activity Actions
7662 commits 12 branches 166 tags 146 MiB
Commit graph

209 commits

Author SHA1 Message Date
Damien Elmes
a93145c4c7 handle preview repeat in deck list 2020-07-30 09:11:13 +10:00
Damien Elmes
71dbbd9fa8 set schema on collection creation 2020-07-03 11:55:27 +10:00
Damien Elmes
618fc65282 handle revlog entries with negative review times 2020-06-29 15:48:01 +10:00
Damien Elmes
4d88b6521d cards added graph 2020-06-29 15:48:01 +10:00
Damien Elmes
1e9901e9bd experiment with exposing raw card/revlog data to frontend 2020-06-29 15:48:00 +10:00
Damien Elmes
fd258d15c5 some initial work on updating the graphs 2020-06-29 15:48:00 +10:00
Damien Elmes
710e5ecce2 formatting fix 2020-06-22 11:58:31 +10:00
Damien Elmes
cec1fdece5 move revlog out of sync code, and add enums 
and remove type=4, which does not appear to be used
 2020-06-16 16:31:36 +10:00
Damien Elmes
f5694a7e4a migrate card stats to backend 
Currently this renders the HTML directly like the previous Python
implementation - doing it in JS would probably make more sense in the
future.
 2020-06-15 17:22:16 +10:00
Damien Elmes
965fd848d1 on upgrade, mark decks modified that are renamed 2020-06-14 14:48:58 +10:00
Damien Elmes
d5cd9af53b fix review cards with high intervals not being reset properly 
would lead to a "fixed .. with invalid card property" message that
didn't go away
 2020-06-09 20:11:26 +10:00
Damien Elmes
75471224fa add progress to db check 2020-06-08 21:07:36 +10:00
Damien Elmes
e74e5c604b add daily count updating to backend 2020-06-05 19:49:53 +10:00
Damien Elmes
c5e65d02e1 add note/card removal to backend 2020-06-04 18:21:04 +10:00
Damien Elmes
d7a2e78c0b change sync label to indicate sync state 
- blue for normal sync, red for full sync required
- refactor status fetching code so we don't hold a collection lock
during the network request, which slows things down
- fix sync spinner restarting when returning to deck list
 2020-06-02 13:23:01 +10:00
Damien Elmes
1c52d11133 rework usn handling in sync 2020-06-01 13:57:10 +10:00
Damien Elmes
0787381a99 login/full up/full down plugged in 2020-06-01 13:57:10 +10:00
Damien Elmes
c4032c0176 another snapshot of sync work 
progress handler is going to need to be reworked to function across
threads
 2020-06-01 13:57:10 +10:00
Damien Elmes
2114fdf6f5 more syncing work 
no checks yet
 2020-06-01 13:57:10 +10:00
Damien Elmes
91e3256e6a move dupe check to backend 2020-05-24 19:48:56 +10:00
Damien Elmes
01ebe964e9 bytes->blob in sql 2020-05-18 13:36:45 +10:00
Damien Elmes
336349d8d1 update before_upload() 2020-05-17 14:13:21 +10:00
Damien Elmes
a682865169 move deck config to protobuf 2020-05-17 13:28:41 +10:00
Damien Elmes
e85911c4e4 add option to limit deck tree counts to a particular node 2020-05-17 08:38:49 +10:00
Damien Elmes
17efa85d75 update rusqlite 2020-05-16 19:49:08 +10:00
Damien Elmes
cbfc9d5a87 handle default deck and filtered deck suppression in the backend 2020-05-15 21:21:10 +10:00
Damien Elmes
14c0a5891b fetch template and note fields in backend during normal card render 
Saves having to serialize the note fields and q/a templates, which
is particularly a win when rendering question/answer in the browse
screen.

Also some work towards being able to preview notes without having to
commit them to the database.
 2020-05-13 10:38:49 +10:00
Damien Elmes
21ac2c14a7 handle scheduling preferences in the backend 2020-05-12 21:13:34 +10:00
Damien Elmes
befa66fb27 fix default rollover 2020-05-12 21:13:34 +10:00
Damien Elmes
dc9200e639 add tests for the other checks, and fix new card due limiting 2020-05-12 21:13:34 +10:00
Damien Elmes
bdf5daa422 add tests for card fixes; implement deck recovery 2020-05-12 21:13:34 +10:00
Damien Elmes
c227298565 stop using memory for sqlite temp store 
On large collections this can end up consuming hundreds of megabytes
of RAM on operations like a vacuum.
 2020-05-12 21:13:34 +10:00
Damien Elmes
d7f9bccd1e rewrite DB check 
- notes with wrong field count are now recovered instead of
being deleted
- notes with missing note types are now recovered
- notes with missing cards are now recovered
- recover_missing_deck() still needs implementing
- checks required
 2020-05-12 21:13:34 +10:00
Damien Elmes
b03dcd2643 fetch notes in nid order 
May reduce the amount of seeking on conventional disks. Chunking and
fetching batches of notes at once would improve things further, at the
cost of more memory use.
 2020-05-12 21:13:34 +10:00
Damien Elmes
667f9e6554 bulk tag add/remove/update; canonify on note save 
also remove the tag list updated hook - we'll need a better solution in
the future than having the library code call back into the GUI code
 2020-05-12 21:13:34 +10:00
Damien Elmes
ad1c519d1c move find&replace to backend 2020-05-12 21:13:34 +10:00
Damien Elmes
2d0a6283e4 add an index on notes(mid) so we can avoid the tablescan 2020-05-12 21:13:34 +10:00
Damien Elmes
7ef72aba04 use the backend for the deck due tree 
- approx 3x faster on a large test deck
- counts are no longer capped to 1000 in the tree
 2020-05-12 21:13:33 +10:00
Damien Elmes
97e857a41c add browserCollapsed as part of tree build 2020-05-12 21:13:33 +10:00
Damien Elmes
611959e5a8 switch primary key to deck id 2020-05-12 21:13:33 +10:00
Damien Elmes
87551574c0 add partial index on odid, and search odid as well 
the odid != 0 check in cards_for_deck.sql is necessary for sqlite
to be able to take advantage of the index
 2020-05-12 21:13:33 +10:00
Damien Elmes
24ef0cc6f1 add separate decks table, and start on moving deck handling to Rust 
The Python tests are passing, but there are still a number of issues
to work through, and the table/protobuf schema is not yet finalized.
 2020-05-12 21:13:33 +10:00
Damien Elmes
cbfe20712e add (currently disabled) check for missing/empty decks on card add 2020-05-12 21:13:33 +10:00
Damien Elmes
c989e23012 handle negative due numbers when gathering existing cards 2020-05-12 21:13:33 +10:00
Damien Elmes
4edeca6ed9 fix use counts not including unused note types 2020-05-12 21:13:33 +10:00
Damien Elmes
4cc070a240 note type removal 2020-05-12 21:13:33 +10:00
Damien Elmes
c8cbc13bba start on exposing notes and individual note type methods 
changes to note:

- add_note() now takes a provided deck id instead of looking it up
in the notetype
- note type use counts fetched using a single table scan
- make sure note type changes are persisted
- expose optionalness of ords in templates and fields json
 2020-05-12 21:13:33 +10:00
Damien Elmes
f95fe72795 mod schema if field/templates changed 
Instead of throwing an error if schema not marked as changed,
just mark it changed, as that way it can be included as part of
the same transaction.
 2020-05-12 21:13:33 +10:00
Damien Elmes
d810ba5db9 template changes and card generation 
Cloze cards are not yet supported, missing decks are not handled,
and more testing is still required.
 2020-05-12 21:13:33 +10:00
Damien Elmes
0c8763edf3 cache notetypes 2020-05-12 21:13:33 +10:00
Damien Elmes
7ff6226a0e always fetch full notetype 2020-05-12 21:13:33 +10:00
Damien Elmes
6b8a45bcb6 more work towards note adding 
Still a prototype at this stage - we'll likely want a caching layer
for note types, and I'm not sure of the merit of having fields in
a separate table, since they're almost always required.
 2020-05-12 21:13:33 +10:00
Damien Elmes
4a2256aecb the start of note adding, and note type changes 2020-05-12 21:13:33 +10:00
Damien Elmes
cbb76290c2 start work on field changes, and add search_cards to col 2020-05-12 21:13:33 +10:00
Damien Elmes
638e69af91 move note db code to storage 2020-05-12 21:13:33 +10:00
Damien Elmes
3807dcacd5 native struct for CardTemplate 2020-05-12 21:13:33 +10:00
Damien Elmes
91767fc87b native struct for NoteField 
ord is made optional so we can keep track of changes in the future
 2020-05-12 21:13:33 +10:00
Damien Elmes
9c41210cf4 use native struct for NoteType 2020-05-12 21:13:33 +10:00
Damien Elmes
65f12eeda2 add stock note types in backend 2020-05-12 21:13:33 +10:00
Damien Elmes
96300f5858 migrate the schema11 usages to new structs/sql queries 2020-05-12 21:13:33 +10:00
Damien Elmes
4b39a8b727 move schema11 note type code into separate file 2020-05-12 21:13:33 +10:00
Damien Elmes
475478cec5 split note types into separate tables 
- store the config in protobuf instead of json
- still loading+saving in bulk for now
- code using the schema11 structs needs to be migrated
 2020-05-12 21:13:33 +10:00
Damien Elmes
e2c57b54a2 reset due on load if outside i32 bounds 
Some third party clients that create apkg files appear to be saving
a millisecond timestamp into this column. While this could also be
solved by increasing this field to a 64 bit number, some older clients
expect this column to be a 32 bit number.
 2020-04-30 11:57:35 +10:00
Damien Elmes
b7696e6381 merge the 12-14 upgrade code 2020-04-10 17:46:57 +10:00
Damien Elmes
0538d9b764 (de)serialize decks in backend 2020-04-09 12:41:59 +10:00
Damien Elmes
bca49bd054 load/save note types in backend 
This allows us to normalize bad data, and is the first step towards
splitting note types into separate tables.
 2020-04-08 10:05:07 +10:00
Damien Elmes
d05722aa82 use blobs for config 2020-04-06 15:39:47 +10:00
Damien Elmes
b5022ad354 store config in separate DB table 
- mtime is tracked on each key individually, which will allow
merging of config changes when syncing in the future
- added col.(get|set|remove)_config()
- in order to support existing code that was mutating returned
values (eg col.conf["something"]["another"] = 5), the returned list/dict
will be automatically wrapped so that when the value is dropped, it
will save the mutated item back to the DB if it's changed. Code that
is fetching lists/dicts from the config like so:

col.conf["foo"]["bar"] = baz
col.setMod()

will continue to work in most case, but should be gradually updated to:

conf = col.get_config("foo")
conf["bar"] = baz
col.set_config("foo", conf)
 2020-04-06 15:39:47 +10:00
Damien Elmes
3c7eabc51e add flag to skip downgrade on collection close 
Disabled for now; when enabled it will allow faster collection
open and close in the normal case, while continuing to downgrade
when exporting or doing a full sync.

Also, when downgrading is disabled, the journal mode is no longer
changed back to delete.
 2020-04-04 17:21:45 +10:00
Damien Elmes
ea2a3d7586 use separate schema version for tag changes 
Will prevent issues if user upgrades after an unclean shutdown
 2020-04-04 10:14:51 +10:00
Damien Elmes
ac54f85840 update tag handling 
- tag list stored in a separate DB table
- non-wildcard searches now do full unicode case folding
(eg tag:masse matches 'Maße')
- wildcard matches do simple unicode case folding
- some functions haven't been updated yet, so ascii folding will
continue to be used in some operations
 2020-04-03 19:34:46 +10:00
Damien Elmes
8f427dbbd7 FString -> TR 2020-04-01 18:16:43 +10:00
Damien Elmes
15ba839003 translate default deck config name 2020-04-01 17:49:25 +10:00
Damien Elmes
cffe146397 split deck config into separate SQL table 
- on collection load, the schema is upgraded to 12
- on collection close, the changes are reversed so older clients
can continue to open the collection
- in the future, we could potentially skip the reversal except
when exporting/doing a full sync
- the same approach should work for decks, note types and tags in the
future too
- the deck list code needs updating to cache the deck confs for the
life of the call
 2020-03-30 20:01:16 +10:00
Damien Elmes
089a68eb43 move deck conf handling to backend 2020-03-30 14:39:46 +10:00
Damien Elmes
d41124343d initial work on undo support 2020-03-29 17:52:16 +10:00
Damien Elmes
3924dedbb7 remove usn cache 
SqliteStorage no longer contains any mutable state
 2020-03-29 12:50:38 +10:00
Damien Elmes
0451a4d4e0 drop separate RequestContext/StorageContext 2020-03-29 12:12:35 +10:00
Damien Elmes
c4a78ffd82 roll back the vec cache changes 
prepare_cached() is sufficiently fast, and allowing the vec cache
to persist across calls is complicated due to lifetime restrictions
 2020-03-29 09:26:09 +10:00
Damien Elmes
cb03c87f8b Revert "add benchmark for vec cache" 
This reverts commit 0247d65ccd.
 2020-03-29 09:16:11 +10:00
Damien Elmes
1112fc8e8b add (unused) add_card, and move sql to separate files 2020-03-27 14:48:25 +10:00
Damien Elmes
0247d65ccd add benchmark for vec cache 
test storage::sqlite::bench::bench_hash_cache ... bench:         399 ns/iter (+/- 27)
test storage::sqlite::bench::bench_no_cache   ... bench:       4,854 ns/iter (+/- 499)
test storage::sqlite::bench::bench_vec_cache  ... bench:           0 ns/iter (+/- 0)
 2020-03-27 09:59:48 +10:00
Damien Elmes
d1c44ec06d store cached statements in a vec instead of separate optionals 2020-03-27 09:51:19 +10:00
Damien Elmes
62be1b2688 update card in transaction 2020-03-27 07:58:25 +10:00
Damien Elmes
e28fdde984 handle mtime/usn bump in backend, and tweak integer sizes 2020-03-26 20:55:04 +10:00
Damien Elmes
0510ab7c9c update_card() 2020-03-26 18:54:20 +10:00
Damien Elmes
b2fba4e1ff use backend to get card 2020-03-26 17:47:43 +10:00
Damien Elmes
1af3de6f70 DeckID, CardID 2020-03-26 15:10:40 +10:00
Damien Elmes
4150480709 NoteTypeID 2020-03-26 15:00:24 +10:00
Damien Elmes
bb0b817d9c use a macro for newtype defs 2020-03-26 13:50:20 +10:00
Damien Elmes
d95cb93d7a usn newtype 2020-03-26 13:06:02 +10:00
Damien Elmes
bfc0287e51 use newtypes for distinguishing between second and millisecond stamps 2020-03-26 12:59:51 +10:00
Damien Elmes
6aced5f125 handle fields with embedded nuls 2020-03-24 14:53:33 +10:00
Damien Elmes
b7fdb8aeb4 fix setting of wal 2020-03-23 13:52:57 +10:00
Damien Elmes
e02f032a47 use backend for v1 and v2 cutoff calculations 
this should also fix the CI failures, which were happening because
the datetime module wasn't matching the shifted time.time()
 2020-03-22 14:28:26 +10:00
Damien Elmes
cf50821d3b add search that ignores combining chars 
On a test of a ~40k card collection, the 'ignore accents' add-on
takes about 1150ms, and this code takes about 70ms.
 2020-03-21 15:15:59 +10:00
Damien Elmes
982e1a811d add unicase collation 
sqlite's like is hard-coded to use ASCII comparisons, so we can't
take advantage of this yet
 2020-03-21 12:40:20 +10:00
Damien Elmes
e0e7560d2b add regexp() to sqlite 2020-03-21 10:23:11 +10:00
Damien Elmes
9810b584d5 support sorting on note type, card template and decks 2020-03-20 21:15:23 +10:00
Damien Elmes
c3c56d7cf2 fix decks 2020-03-20 21:15:23 +10:00
Damien Elmes
6c469a96eb finish the remaining searches 
Searches that require multiple deck or note type lookups won't perform
very well at the moment - it either needs caching or to be split up
at the DB level.

Nothing tested yet.
 2020-03-20 21:15:23 +10:00
Damien Elmes
4f1c85bf8a add v1 and v2 legacy timing code 2020-03-20 21:15:23 +10:00
Damien Elmes
1b2ebfcf6f deck searching 
A bit more complicated than it needs to be, as we don't have the
full deck manager infrastructure yet.
 2020-03-20 21:15:23 +10:00
Damien Elmes
eef158b20c add field_at_index() sql func 2020-03-20 21:15:23 +10:00
Damien Elmes
e72d21e09e move note code into notes.rs, add ability to rollback when unchanged 2020-03-20 21:15:23 +10:00
Damien Elmes
3f3a45e2ba add Collection struct, and get media check working again 
- media check no longer needs collection to be closed
- use savepoints for operations initiated by Rust, so they are
atomic without forcing a commit
 2020-03-20 21:15:23 +10:00
Damien Elmes
700d713f80 add begin/commit/rollback, and support creating collections 
all but one unit test is now passing
 2020-03-20 21:15:23 +10:00
Damien Elmes
38d4ea5311 use (or)json for DB bridge 
Some initial testing with orjson indicates performance varies from
slightly better than pysqlite to about 2x slower depending on the type
of query.

Performance could be improved by building the Python list in rspy
instead of sending back json that needs to be decoded, but it may make
more sense to rewrite the hotspots in Rust instead. More testing is
required in any case.
 2020-03-20 21:15:23 +10:00
Damien Elmes
a129b8f8cf initial work on DB handling in Rust 
committing the Protobuf implementation for posterity, but will replace
it with json, as Protobuf measures about 6x slower for some workloads
like 'select * from notes'
 2020-03-20 21:15:23 +10:00