Before sqlite 3.15, the parameter to vacuum was ignored. Since sqlite
3.15, it became a database name parameter. The "vacuum analyze" syntax
was never supported by sqlite, seems to be just a psql thing.
The error for this is "OperationalError: unknown database analyze", and
happens when doing a media sync.
we need to switch to autocommit mode before executing vacuums &
certain pragmas
also removed the pysqlite version workaround; we just use the
python-provided versions in all cases now as it is new enough
- wrap request in AnkiRequestsClient so we can keep track of
upload/download bytes without having to monkey patch anything
- force a 64kB buffer size instead of the default 8kB
- show one decimal point in up/down so small requests still give
visual feedback
- update add-on downloading and update check to use requests
- remove the update throttling in aqt/sync.py, as it's not really
necessary anymore
- fix an extra \r\n being included at the end of the last form var
- req()s in the middle of a sync should throw an exception if they
receive a 403, as the calling code does not check for an empty return
universal_newlines uses system locale which is ascii on osx unless
LC_CTYPE/LANG is set to utf8, so we need to be explicit about the
encoding we want instead
we also lose line buffering, so we have to explicitly flush
Previously, y axis ticks could be fractional. Fractional ticks were always rounded to the nearest whole number for display. This leads to confusing graphs where the bar and ticks do not match up. For example, if the bar is 3 and the tick is 2.5, then the bar renders just above the tick but the tick is rendered as 3. They both then appear to have value 3 but don't line up.
To fix this behavior, we now indicate to flot that for the y axis we don't want fractional tick values (by setting tickDecimals to 0). flot will pick tick values to accommodate this setting. If for some reason the ticks are fractional, which shouldn't happen, we will render to one decimal place. Otherwise we render whole numbers without the decimal. Since we are counting reviews, this behavior makes more sense, because reviews are always whole numbers and never fractional.
When adding a new cloze tag while a card is in a filtered deck, anki currently adds the new card to the Default deck.
My proposed fix checks if the card has the odid field set and then sets the did to use when generating the new cards as the odid.
- we now allow exports over 2gb/64k files - AnkiMobile and AnkiDroid
will need to be updated to support this
- avoid compressing media files in export, as in the common case of
jpg/mp3 it's much faster with no increase in size
- exports and imports now show # of files processed
- mw.progress.update() now limits # of updates per sec
When the override deck is not valid, the more intuitive behavior for users is to not override. Advanced users that really want to use this feature will immediately open the deck override settings and see that it's referencing an invalid deck. Advanced users and developers have all been stumped by the old behavior, it's simply not intuitive.
fixes https://anki.tenderapp.com/discussions/ankidesktop/15795-export
-error-doubling-br-tags
This code dates back a few years, and was probably a naive solution
for files breaking when exported with newlines. Ideally we should be
preserving the newlines and wrapping the field in quotes, but since
some people may be relying on exported files not to be quoted, we'll
wait to change this until the next major release. For now, we'll use
a space instead, which should not alter the appearance of the
rendered HTML.
Currently, field searches are confirmed by a regex search with the single option of case-insensitive (?i), and with the beginning and ending markers ^ and $. Since multi-line is not enabled, and re.DOTALL (option s) is not enabled, the field search will fail for any field with a new line
fixes the following:
- create a filtered deck and sync it
- review cards in the filtered deck and delete it
- sync again
The filtered deck deletion was bumping the mod time on cards at the
start of the sync, preventing the reviews from being synced from the
other side, leading to lost reviews and sanity check errors.
When a user renames or deletes a field, Anki tries to go through the
user's templates to substitute occurrences of that field name with its
new name or remove the field.
However, if the user has multiple Mustache placeholders on a single
line, this does not work as intended because the leading (.*) capture
group grabs the text in-between the two (or more) Mustache placeholders.
Examples:
{{#Flag}}
...
{{/Flag}}{{^Flag}}
...
{{/Flag}}
or
{{Front}} {{#Flag}}...{{/Flag}}
or
{{Front}} Hello. {{Victim}}
Previously they were allowed to be added, but all searches and
operations treated them as case-sensitive, creating an inconsistency.
The new implementation will change new tags with different case than
existing ones to the version currently in the database.
This may cause some slowdown in collections with a very large number of
tags since the only way to do this at the moment is to scan through
every one of them. Changing the format tags are stored in in the future
may be useful.
deletes in particular take some time for the server to process, but
don't require much bandwidth, leading to the progress appearing to
have pause when content is actually being processed
this also gives the user an idea of how long the process will take to
complete
- media syncing no longer locks the account, so it can be done
in the background in the future, and multiple clients can safely
sync media at the same time
- all operations are now idempotent, so they can be repeatedly safely
in the event of a connection error
- whether it's a normal incremental sync, an initial sync,
or the media database has been deleted, no files will be uploaded
or downloaded if they already exist on the other side
- file removals are now chunked like additions & updates, preventing
timeouts due to large requests
- if the server can't process a chunk in time, it will return a count
of what it did process, so the client can retry the rest
Notes for AnkiDroid:
- when porting this, recommend you pick a different name for the
.media.db2 file, so users don't accidentally copy the AD version to
the desktop or vice versa
- please make sure filenames are added to the zip in NFC form
If duplicate notes are already in the collection and, based on the first
field of all of them, they should be updated, update all of them instead
of just the first one we come across.
When "allow HTML" was turned off, Anki was replacing newlines with <br>s
in the text of the new notes before it escaped HTML characters, so the
line breaks were becoming <br>.
Even if there are several (duplicate) notes in the collection that
match, only send the message once to avoid making it look like multiple
dupes have been added.
This change keeps drag-and-drop behavior the same except for displaying
an error if an otherwise acceptable move would nest anything underneath
a filtered deck.
Dvi isn't compatible with quite a few Latex packages. I have been trying to include chemical formulas with the chemfig package. Exposing the tool chain to plugins would be very useful. For the record, I'm currently using:
´´´python
latexCmds = [
["pdflatex", "-interaction=nonstopmode", "tmp.tex"],
["pdflatex", "-interaction=nonstopmode", "tmp.tex"],
["pdfcrop", "tmp.pdf"],
["convert", "-density", "300", "tmp-crop.pdf", "tmp.png"]
]
´´´
The 64 bit built of portaudio on OSX seems to generate a wav
file that says it's 44100 but is actually the default rate, leading
to samples playing too fast or slow.
This allows {{kana:cloze:Text}} to work. May allow users to
shoot themselves it the foot - will see how it goes in the wild
and reconsider if necessary
