Summary:
Implements the latest design of the Formula AI Assistant.
Also switches out brace to the latest build of ace.
Test Plan: Browser tests.
Reviewers: jarek
Reviewed By: jarek
Subscribers: jarek
Differential Revision: https://phab.getgrist.com/D3949
Test Plan: Added a check to the emoji test.
Reviewers: georgegevoian
Reviewed By: georgegevoian
Subscribers: georgegevoian
Differential Revision: https://phab.getgrist.com/D3951
Grist by default uses node-sqlite3 to manipulate data in an
SQLite database. If a single parameter is passed to `run`
and it is a list, the list is unpacked and its contents treated
as the actual parameters. In grist-static, we use other SQLite
interfaces that don't have that automatic unpacking. Most
calls like this have been removed from Grist, but at least one
was missed, and was causing symptoms such as
https://github.com/gristlabs/grist-static/issues/5
This change should make no difference to regular Grist, but
resolves the grist-static problems.
Summary:
add back config call for external storage
in the right order.
Test Plan: existing saas tests catch this
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3946
Summary:
Adding limits for AI calls and connecting those limits with a Stripe Account.
- New table in homedb called `limits`
- All calls to the AI are not routed through DocApi and measured.
- All products now contain a special key `assistantLimit`, with a default value 0
- Limit is reset every time the subscription has changed its period
- The billing page is updated with two new options that describe the AI plan
- There is a new popup that allows the user to upgrade to a higher plan
- Tiers are read directly from the Stripe product with a volume pricing model
Test Plan: Updated and added
Reviewers: georgegevoian, paulfitz
Reviewed By: georgegevoian
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3907
Summary:
In a selector table, when a selected row is filtered out of view, linked widgets should update based on the newly selected row.
There were a few bugs that contributed to this wrong behavior:
- Gridview wasn't subscribing to the current row id, and the row with id 'new' was being converted to the first row
- Cursor was keeping track of the currently selected row id, it was hiding a problem behind the proper rowIndex
- Undo/redo somehow leveraged the wrong rowId from the cursor during the position restore.
The `No data` text was also changed to be more meaningful.
Test Plan: Added and updated.
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3937
Summary:
A recent change broke the ability to open the account page if the current
org was inaccessible or invalid. Now the account page, and a few additional
pages, are excluded from needing an org to load.
Test Plan: Browser test.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3944
Summary:
It was impossible to focus the search input in the filter menu if either
of the range filter inputs were focused.
Test Plan: Browser test.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3942
Summary: Text in Document History > Activity tab is now selectable and one can copy it.
Test Plan: Manual
Reviewers: georgegevoian
Reviewed By: georgegevoian
Subscribers: georgegevoian
Differential Revision: https://phab.getgrist.com/D3939
Summary:
Adds a new Support Grist page (accessible only in grist-core), containing
options to opt in to telemetry and sponsor Grist Labs on GitHub.
A nudge is also shown in the doc menu, which can be collapsed or permanently
dismissed.
Test Plan: Browser and server tests.
Reviewers: paulfitz, dsagal
Reviewed By: paulfitz
Subscribers: jarek, dsagal
Differential Revision: https://phab.getgrist.com/D3926
Summary:
Here's a series of badness that easily leads to a crash, in reverse order:
- Lodash's map() function interprets an object with a .length property as an array.
- Some very old code generated human-friendly descriptions of user actions,
applying map() to parts of them. It so happens that this generated description
isn't even used.
- If a user action is encountered with a sufficiently large length propery,
map() would exhaust the server memory.
Fixed by removing old unneeded code, and replacing some other occurrences of
lodash's map() with native equivalents.
Test Plan: Tested manually on a local reproduction of the issue.
Reviewers: paulfitz
Reviewed By: paulfitz
Subscribers: paulfitz
Differential Revision: https://phab.getgrist.com/D3938
Summary:
- Detecting emoji is surprisingly tricky; we use a fancy regex as a decent heuristic.
- Icons are a little larger than before.
- Styling tweaked for light and dark modes
- In case the OS doesn't render the emoji as one character, truncate what's
shown in the icon box.
Test Plan: Added a test case.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3904
Test Plan: Checked new looks manually, behavior should not be affected.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3934
A small test harness bundle was recently added that is breaking the docker image build. It could be added to the docker image, but that would introduce a bunch of extraneous test file dependencies. So this tweaks the build to simply skip the test bundle if its primary source file is not found.
Also added some other test fixes along the way:
* make a custom widget test more reliable
* update a localization test now that `pt` exists
* store more log info in artifact on error
There's a little nest of SelectBy tests that sometimes fail.
They are the only tests with an import of a helper function from
an other file that contains tests. Such imports have caused trouble
with mocha in the past. I'm not sure if that is the case now, but
I'd like to eliminate it as a possibility.
Summary:
This tweaks the prompting so that the user's message is given on its own instead of as a docstring within Python. This is so that the prompt makes sense when:
- the user asks a question such as "Can you write me a formula which does ...?" rather than describing their formula as a docstring would, or
- the user sends a message that doesn't ask for a formula at all (https://grist.slack.com/archives/C0234CPPXPA/p1687699944315069?thread_ts=1687698078.832209&cid=C0234CPPXPA)
Also added wording for the model to refuse when the user asks for something that the model cannot do.
Because the code (and maybe in some cases the model) for non-ChatGPT models relies on the prompt consisting entirely of Python code produced by the data engine (which no longer contains the user's message) those code paths have been disabled for now. Updating them now seems like undesirable drag, I think it'd be better to revisit this when iteration/experimentation has slowed down and stabilised.
Test Plan:
Added entries to the formula dataset where the response shouldn't contain a formula, indicated by the value `1` for the new column `no_formula`.
This is somewhat successful, as the model does refuse to help in some of the new test cases, but not all. Performance on existing entries also seems a bit worse, but it's hard to distinguish this from random noise. Hopefully this can be remedied in the future with more work, e.g. automatic followup messages containing example inputs and outputs.
Reviewers: paulfitz
Reviewed By: paulfitz
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3936
Summary:
This uses a newer version of mocha in grist-core so that tests can be run in parallel. That allows more tests to be moved without slowing things down overall. Tests moved are venerable browser tests; only the ones that "just work" or worked without too much trouble to are moved, in order to keep the diff from growing too large. Will wrestle with more in follow up.
Parallelism is at the file level, rather than the individual test.
The newer version of mocha isn't needed for grist-saas repo; tests are parallelized in our internal CI by other means. I've chosen to allocate files to workers in a cruder way than our internal CI, based on initial characters rather than an automated process. The automated process would need some reworking to be compatible with mocha running in parallel mode.
Test Plan: this diff was tested first on grist-core, then ported to grist-saas so saas repo history will correctly track history of moved files.
Reviewers: jarek
Reviewed By: jarek
Subscribers: jarek
Differential Revision: https://phab.getgrist.com/D3927
Summary:
Some edits to virtual tables (such as webhook lists) happen
via a route that was not yet handled. Actually Cyprien (the
original author) had handled this case but it got removed
because I didn't know what it was for :-). This brings back
support for edits by this route.
Test Plan: added a test
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3924
Summary: I looked through the template documents mentioned in `formula-dataset-index.csv` and selected formulas involving lookups to add to the CSV, particularly nontrivial formulas.
Test Plan: Running the test script on the new dataset gives a score of 47/61 compared to the previous 45/47, i.e. it scores 2/14 on the new entries. Lookups are clearly challenging and we'll need to add more information to the prompt, maybe even consider a more complicated strategy than a single prompt. This diff is purely for expanding the dataset, improving performance will come later.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3931
Summary: Converts `rec.` to `$` in AI generated formulas, and removes redundant `return` at the end.
Test Plan: Expanded unit test.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3933
Summary:
The GristDocTutorial table is now always visible to users with edit
access to the trunk, and the Share menu is now available within
tutorial forks, making it easier for editors to replace the original
tutorial trunk with changes made in the fork, and for viewers to export
their copy of the tutorial.
Also, changes to the GristDocTutorial table are now immediately reflected
in the tutorial popup.
Test Plan: Browser tests.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3930
Summary:
- Cleaning stripe data after billing tests
- Better stripe webhook test integration, that should fix test interference
- Not importing why-is-node-running when its not needed, which improves dev experience.
Test Plan: Modified
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3932
Summary:
- Fly flyctl cleanup which has been failing; deleting volumes before app no longer seems needed
- Move fly-deploy and add workflow scripts to make fly preview work in grist-core too
Test Plan: Tested that previews still work in this repo; and tested with a copy in grist-core.
Reviewers: jarek
Reviewed By: jarek
Subscribers: jarek
Differential Revision: https://phab.getgrist.com/D3928
Summary:
1. Introduces another highlight for link-selector rows, with the same color as
regular selection, and allowing to overlap with regular selection.
2. Don't show "secondary" cursors (those in inactive sections), to keep a single
cursor on the screen, since having multiple (which different in color) could
cause confusion.
3. An unrelated improvement (prompted by a new fixture doc) is to default the
active section to the top-left one (rather than the one with smallest rowId).
4. Another unrelated improvement (prompted by a test affected by the previous unrelated improvement) is to skip chart widgets when searching (previously search would step through those with an invisible "cursor").
Includes also tweaks for better testing on Arm-based Macs:
- Add support for TEST_CHROME_BINARY_PATH environment variable (helpful for a Mac arm64 architecture workaround)
- Remove unsetting of SELENIUM_REMOTE_URL when running headless (unlikely to affect anyone, and can be done outside the script, but interferes with the Mac workaround)
Test Plan: Added a new test case that cursor and linking-selector CSS classes are present or absent appropriately. Fixed test affected by the fix to default active section.
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3891
Summary:
It became hard to detect aborted connections in node 16.
In node 14, req.on('close', ...) did the job. Thid diff adds a
work-around, until a better way is discovered or added.
Aborting a req will typically lead to 'close' being called
on the response, without writableFinished being set.
- https://github.com/nodejs/node/issues/38924
- https://github.com/nodejs/node/issues/40775
Test Plan:
existing DocApiForwarder test passes; manually
checking on various node versions.
Reviewers: JakubSerafin
Reviewed By: JakubSerafin
Differential Revision: https://phab.getgrist.com/D3923
Summary:
The previous code for extracting a Python formula from the LLM completion involved some shaky string manipulation which this improves on.
Overall the 'test results' from `runCompletion` went from 37/47 to 45/47 for `gpt-3.5-turbo-0613`.
The biggest problem that motivated these changes was that it assumed that code was always inside a markdown code block
(i.e. triple backticks) and so if there was no block there was no code. But the completion often consists of *only* code
with no accompanying explanation or markdown. By parsing the completion in Python instead of JS,
we can easily check if the entire completion is valid Python syntax and accept it if it is.
I also noticed one failure resulting from the completion containing the full function (instead of just the body)
and necessary imports before that function instead of inside. The new parsing moves import inside.
Test Plan: Added a Python unit test
Reviewers: paulfitz
Reviewed By: paulfitz
Subscribers: paulfitz
Differential Revision: https://phab.getgrist.com/D3922
Not only GREP_TESTS can be assigned a single word like:
GREP_TESTS=DocApi yarn test
But also can be assigned a whole sentence part:
GREP_TESTS="supports ascending sort" yarn test
That's especially useful to run a single test (and not a whole test
suit)
Co-authored-by: Florent FAYOLLE <florent.fayolle@beta.gouv.fr>
