(for upgrading users, please see the notes at the bottom)
Bazel brought a lot of nice things to the table, such as rebuilds based on
content changes instead of modification times, caching of build products,
detection of incorrect build rules via a sandbox, and so on. Rewriting the build
in Bazel was also an opportunity to improve on the Makefile-based build we had
prior, which was pretty poor: most dependencies were external or not pinned, and
the build graph was poorly defined and mostly serialized. It was not uncommon
for fresh checkouts to fail due to floating dependencies, or for things to break
when trying to switch to an older commit.
For day-to-day development, I think Bazel served us reasonably well - we could
generally switch between branches while being confident that builds would be
correct and reasonably fast, and not require full rebuilds (except on Windows,
where the lack of a sandbox and the TS rules would cause build breakages when TS
files were renamed/removed).
Bazel achieves that reliability by defining rules for each programming language
that define how source files should be turned into outputs. For the rules to
work with Bazel's sandboxing approach, they often have to reimplement or
partially bypass the standard tools that each programming language provides. The
Rust rules call Rust's compiler directly for example, instead of using Cargo,
and the Python rules extract each PyPi package into a separate folder that gets
added to sys.path.
These separate language rules allow proper declaration of inputs and outputs,
and offer some advantages such as caching of build products and fine-grained
dependency installation. But they also bring some downsides:
- The rules don't always support use-cases/platforms that the standard language
tools do, meaning they need to be patched to be used. I've had to contribute a
number of patches to the Rust, Python and JS rules to unblock various issues.
- The dependencies we use with each language sometimes make assumptions that do
not hold in Bazel, meaning they either need to be pinned or patched, or the
language rules need to be adjusted to accommodate them.
I was hopeful that after the initial setup work, things would be relatively
smooth-sailing. Unfortunately, that has not proved to be the case. Things
frequently broke when dependencies or the language rules were updated, and I
began to get frustrated at the amount of Anki development time I was instead
spending on build system upkeep. It's now about 2 years since switching to
Bazel, and I think it's time to cut losses, and switch to something else that's
a better fit.
The new build system is based on a small build tool called Ninja, and some
custom Rust code in build/. This means that to build Anki, Bazel is no longer
required, but Ninja and Rust need to be installed on your system. Python and
Node toolchains are automatically downloaded like in Bazel.
This new build system should result in faster builds in some cases:
- Because we're using cargo to build now, Rust builds are able to take advantage
of pipelining and incremental debug builds, which we didn't have with Bazel.
It's also easier to override the default linker on Linux/macOS, which can
further improve speeds.
- External Rust crates are now built with opt=1, which improves performance
of debug builds.
- Esbuild is now used to transpile TypeScript, instead of invoking the TypeScript
compiler. This results in faster builds, by deferring typechecking to test/check
time, and by allowing more work to happen in parallel.
As an example of the differences, when testing with the mold linker on Linux,
adding a new message to tags.proto (which triggers a recompile of the bulk of
the Rust and TypeScript code) results in a compile that goes from about 22s on
Bazel to about 7s in the new system. With the standard linker, it's about 9s.
Some other changes of note:
- Our Rust workspace now uses cargo-hakari to ensure all packages agree on
available features, preventing unnecessary rebuilds.
- pylib/anki is now a PEP420 implicit namespace, avoiding the need to merge
source files and generated files into a single folder for running. By telling
VSCode about the extra search path, code completion now works with generated
files without needing to symlink them into the source folder.
- qt/aqt can't use PEP420 as it's difficult to get rid of aqt/__init__.py.
Instead, the generated files are now placed in a separate _aqt package that's
added to the path.
- ts/lib is now exposed as @tslib, so the source code and generated code can be
provided under the same namespace without a merging step.
- MyPy and PyLint are now invoked once for the entire codebase.
- dprint will be used to format TypeScript/json files in the future instead of
the slower prettier (currently turned off to avoid causing conflicts). It can
automatically defer to prettier when formatting Svelte files.
- svelte-check is now used for typechecking our Svelte code, which revealed a
few typing issues that went undetected with the old system.
- The Jest unit tests now work on Windows as well.
If you're upgrading from Bazel, updated usage instructions are in docs/development.md and docs/build.md. A summary of the changes:
- please remove node_modules and .bazel
- install rustup (https://rustup.rs/)
- install rsync if not already installed (on windows, use pacman - see docs/windows.md)
- install Ninja (unzip from https://github.com/ninja-build/ninja/releases/tag/v1.11.1 and
place on your path, or from your distro/homebrew if it's 1.10+)
- update .vscode/settings.json from .vscode.dist
* Fix reparented_name not correctly handling tags that are prefixes of the
new parent
To reproduce the issue:
1. Add two tags: `a` and `ab`.
2. From the browser's sidebar, drag & drop `a` into `ab`.
Result: panic
* Fix reparent_tags panicking if new parent is a child of source tag
This is the "foo, foo::bar" case that should be a no-op.
* Add more tests for tag reparenting
* Add crate snafu
* Replace all inline structs in AnkiError
* Derive Snafu on AnkiError
* Use snafu for card type errors
* Use snafu whatever error for InvalidInput
* Use snafu for NotFoundError and improve message
* Use snafu for FileIoError to attach context
Remove IoError.
Add some context-attaching helpers to replace code returning bare
io::Errors.
* Add more context-attaching io helpers
* Add message, context and backtrace to new snafus
* Utilize error context and backtrace on frontend
* Rename LocalizedError -> BackendError.
* Remove DocumentedError.
* Have all backend exceptions inherit BackendError.
* Rename localized(_description) -> message
* Remove accidentally committed experimental trait
* invalid_input_context -> ok_or_invalid
* ensure_valid_input! -> require!
* Always return `Err` from `invalid_input!`
Instead of a Result to unwrap, the macro accepts a source error now.
* new_tempfile_in_parent -> new_tempfile_in_parent_of
* ok_or_not_found -> or_not_found
* ok_or_invalid -> or_invalid
* Add crate convert_case
* Use unqualified lowercase type name
* Remove uses of snafu::ensure
* Allow public construction of InvalidInputErrors (dae)
Needed to port the AnkiDroid changes.
* Make into_protobuf() public (dae)
Also required for AnkiDroid. Not sure why it worked previously - possible
bug in older Rust version?
* Show warning if multiple type boxes are used
* Report templates referencing media in Media Check
* Apply suggestions from code review
* Fix media-check.ftl
* Only report media references with fields
Like `<img src={{Front}}>`.
Also report Anki sound tags and latex.
* Loop existing media regexes
* Enable state-dependent custom scheduling data
* Next(Card)States -> SchedulingStates
The fact that `current` was included in `next` always bothered me,
and custom data is part of the card state, so that was a bit confusing
too.
* Store custom_data in SchedulingState
* Make custom_data optional when answering
Avoids having to send it 4 extra times to the frontend, and avoids the
legacy answerCard() API clobbering the stored data.
Co-authored-by: Damien Elmes <gpg@ankiweb.net>
* Keep filtered decks when importing apkg
If all original decks exist and scheduling is included.
* Create missing decks from csv
* Export original decks if with_scheduling
* Also remap original deck ids on import
* Update imported filtered decks
* Fix meta column being mapped to tags
* Fix ids in csv deck and notetype columns
Note: This implies names which parse to an i64 will be seen as ids,
likely resulting in the intended deck/notetype not being found.
* Check for scheduling with revlog and deck configs
Might help with cases in which scheduling was included, but all cards
are new. In such a case, filtered deck should not be converted.
* Fix duplicate with same GUID being created
* Remove redundant `distinct`s from sql query
* Match notes by _either_ guid _or_ first field
* Refactor to emphasise GUID/first field distinction
* Export default deck and config if with scheduling
* Fix default deck being exported if it's a parent
* Add card meta for persisting custom scheduling state
* Rename meta -> custom_data
* Enforce limits on size of custom data
Large values will slow down table scans of the cards table, and it's
easier to be strict now and possibly relax things in the future than
the opposite.
* Pack card states and customData into a single message
+ default customData to empty if it can't be parsed
Co-authored-by: Damien Elmes <gpg@ankiweb.net>
* Introduce setting to collapse field by default
* Fix schema order
* Change wording from adjective to imperative
sounds a bit less clunky
* Update rslib/src/notetype/schema11.rs (dae)
* Keep settings in single column
* Add back Toggle Visual Editor string
* Add RichTextBadge component and show it conditionally
* Reverse input order depending on default setting
* Make PlainTextInput border-radius responsive to toggle states
* Prevent first Collapsible transition differently
* Focus inputs after Collapsible transition
The double tick calls are just a temporary solution until I find the exact moment an input is focusable again.
* Use requestAnimationFrame to await focusable state
Note: Svelte tick doesn't seem to work in this scenario.
* Keep content of unmapped fields when importing
* Test new behaviour
* Fix typo in `canonify_tags_without_resgistering`
* Log updated note instead of original one
* Revert merging imported tags
But keep old note tags if no new ones are provided.
* Introduce field setting to use plain text editor by default
* Remove leftover function from #1476
* Use boolean instead of string
* Simplify clear_other_field_duplicates
* Convert plain text key to camelCase
* Move HTML item below the existing checkbox, instead of to the right (dae)
Showing it on the right is more space efficient, but feels a bit
cluttered IMHO.
Partially completes #1068, and will allow mobile clients to drop
their separate diff-match-patch imports. Does not yet try to handle
case folding or combining-char stripping, and leaves some of the outer
HTML wrapping up to the frontend for now.
The logic for rendering the provided string has changed: missing chars
are now only inserted if they follow a correct section, and the original
text is shown instead of hyphens. This is an experiment, and can be
changed if it's not well received.
* Refactor MediaManager transactions
* Add media entries when importing
* Fix legacy apkg import not setting csum
Also test for this in the roundtrip test.
* Avoid reallocating MediaCopier's buffer
* Make sha1 optional (dae)
* Add deck-specific limits to DeckNormal
* Add deck-specific limits to schema11
* Add DeckLimitsDialog
* deck_limits_qt6.py needs to be a symlink
* Clear duplicate deck setting keys on downgrade
* Export deck limits when exporting with scheduling
* Revert "deck_limits_qt6.py needs to be a symlink"
This reverts commit 4ee7be1e10c4e8c49bb20de3bf45ac18b5e2d4f6.
* Revert "Add DeckLimitsDialog"
This reverts commit eb0e2a62d33df0b518d9204a27b09e97966ce82a.
* Add day limits to DeckNormal
* Add deck and day limits mock to deck options
* Revert "Add deck and day limits mock to deck options"
This reverts commit 0775814989e8cb486483d06727b1af266bb4513a.
* Add Tabs component for daily limits
* Add borders to tabs component
* Revert "Add borders to tabs component"
This reverts commit aaaf5538932540f944d92725c63bb04cfe97ea14.
* Implement tabbed limits properly
* Add comment to translations
* Update rslib/src/decks/limits.rs
Co-authored-by: Damien Elmes <dae@users.noreply.github.com>
* Fix camel case in clear_other_duplicates()
* day_limit → current_limit
* Also import day limits
* Remember last used day limits
* Add day limits to schema 11
* Tweak comment (dae)
* Exclude day limit in export (dae)
* Tweak tab wording (dae)
* Update preset limits on preset change
* Explain tabs in tooltip (dae)
* Omit deck and today limits if v2 is enabled
* Preserve deck limit when switching to today limit
* Expose cloze text as attribute on front side
* Update test_models.py
* Update template_filters.rs
* Escape HTML for data-attribute
* Use minimal HTML encoding in Rust
to match Python's html.escape and pass tests.
* Rename attribute to data-cloze
to make it more generic.
* Run formatter
* Revert to using Rust encode_attribute and add helper function for tests
* Adjust remaining steps after config update
* Handle relearning steps separately
Also refactor a lot.
* Also adjust remaining steps after deck change
* Test step adjustment after config update
* Fix `SearchBuilder::(re)learning_cards()`
* Fix step adjustment after deck change
* Test step adjustment after deck change
* Fix test name
* Readjust remaining steps according to last delay
Also atomize tests and add some tooling.
* Throw error for unknown condition fields as well
So if 'foo' is not a field, refuse to save a template containing
`{{#foo}}bar{{/foo}}`. Previously, only `{{foo}}` would be checked.
As a side effect, templates which *only* contain fields as conditions
may be saved. Meh.
* Display template errors in q/a columns only
So the affected browser row remains active and the user can fix the
template more easily.
* Specify if error occured in a browser template
* Minor wording tweak (dae)
There's an argument for using the exact wording as well, but this just
reads a little more naturally to me.
