Summary:
BulkAddRecord when finishing imports of nested JSON was throwing
an error due to unchecked access of referencing tables. This adds
a guard to prepare_new_values to handle such cases.
Imports happened to cause this to occur because the order that
imported tables are created/populated isn't aware of references
between tables, so it's possible for a reference column to
exist (momentarily) without a valid reference to another table.
These references are currently fixed after all imported tables are
created/populated.
Test Plan: Browser test.
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3144
Summary: Expanding search input field to full available height, to make the clickable area bigger.
Test Plan: Manual tests on browserstack
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3145
Summary:
The 'Open Access Rules' link is now hidden unless the
UserManager is opened inside a document, and the resource
that users are being managed for is a document.
Test Plan: Browser tests.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3142
Summary:
Currently when reloading a document, we may have two sqlite connections
to the document for a small period of time. This diff removes that
overlap.
Test Plan: added test
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3140
Summary: Added a helper to include lots of metadata in every logging call, added and converted many logging calls.
Test Plan: Existing tests pass
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3136
Summary:
Uses the new alwaysPreserveColIds option for action summaries in Triggers.ts.
Triggers.ts is now responsible for generating the summary to make it easy to pass this option. The value of the option is just all colIds mentioned in triggers configured in this document.
Test Plan: Tested adding 200 rows to a subscribed table to ensure the events are not truncated. Also tests batching nicely.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3135
Summary:
Recent python3 changes perturbed timing again, and a few more tests started failing.
Contains an unrelated correction for gvisor running under docker (a useful configuration on macs for debugging gvisor problems, but not supported by throttling code).
Test Plan: updated tests
Reviewers: dsagal, alexmojaki
Reviewed By: dsagal, alexmojaki
Differential Revision: https://phab.getgrist.com/D3129
Summary:
The Importer dialog is now maximized, showing additional column
matching options and information on the left, with the preview
table shown on the right. Columns can be mapped via a select menu
listing all source columns, or by clicking a formula field next to
the menu and directly editing the transform formula.
Test Plan: Browser tests.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3096
Summary: title
Test Plan: Added python unit test. Seems like the first test of sort_by in the whole codebase.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3124
Summary:
Added a new object type code `l` (for lookup) which can be used in user actions as a temporary cell value in ref[list] columns and is immediately converted to a row ID in the data engine. The value contains the original raw string (to be used as alt text), the column ID to lookup (typically the visible column) and one or more values to lookup.
For reflists, valueParser now tries parsing the string first as JSON, then as a CSV row, and applies the visible column parsed to each item.
Both ref and reflists columns no longer format the parsed value when there's no matching reference, the original unparsed string is used as alttext instead.
Test Plan: Added another table "Multi-References" to CopyPaste test. Made that table and the References table test with and without table data loaded in the browser.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3118
Summary:
For Windows and Ubuntu (on chrome) texteditor was showing scrollbars for
very long words. Adding more space to fix this issue.
Test Plan: manual
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3122
Summary: title
Test Plan: Tested manually, I don't think this needs an automated test. Made a text column with a value `[1, 2]` and converted the column to choice list. Previously this threw a JS error that `tag.trim` wasn't a function. Works now, suggests `1` and `2` as choices in the configuration.
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3128
Summary:
This enables ptrace and modifies a nofile limit, to be consistent with staging. The nofile change isn't really needed, it was for gvisor's vfs2 which we are no longer using, it is just for consistency.
This diff also documents the PYTHON_VERSION_ON_CREATION=3 flag active in staging and ~~soon to be active in~~ production.
Test Plan: settings work in staging
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3125
Summary:
Adds links to manage team and go to billing account in
the org menu (opened by clicking the dropdown in the
top-left corner of Grist). Tweaks some wording of items
in both AppHeader and AccountWidget, and adds a link
to create a new team site to the Site Switcher in both
menus.
Also tweaks the UI of UserManager by adding
an animation when the manager is opened from the
doc access dialog.
Test Plan: Browser tests.
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3121
Test Plan: Added check for these values in a relevant test case.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3117
Summary:
If PYTHON_VERSION_ON_CREATION is set in the environment, new documents will be created with a specific desired python version (2 or 3).
This diff commits to offering a choice of engine, so the engine for a document no longer starts to initialize until the document has been fetched and read. Staging (and dev, and testing) has been like this for a while.
Test Plan: added test; manual testing of forks/copies etc
Reviewers: dsagal, alexmojaki
Reviewed By: dsagal, alexmojaki
Differential Revision: https://phab.getgrist.com/D3119
Summary:
Redesigning column type section to make it more user-friendly. Introducing column behavior concept.
Column can be either:
- Empty Formula Column: initial state (user can convert to Formula/Data Column)
- Data Column: non formula column with or without trigger (with option to add trigger, or convert to formula)
- Formula Column: pure formula column, with an option to convert to data column with a trigger.
Test Plan: Existing tests.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3092
Summary:
This marks the engine setting in document settings as experimental,
with a skull and cross-bones.
It also makes sure the setting is shown if PYTHON_VERSION_ON_CREATION
is set (this relates to a separate change to set the default python
version to 3).
Test Plan: manual
Reviewers: alexmojaki, dsagal
Reviewed By: alexmojaki, dsagal
Subscribers: anaisconce
Differential Revision: https://phab.getgrist.com/D3120
Summary:
Grist has, up to now, used a throttling mechanism that allows a sandbox free rein until it starts using above some threshold percentage of a cpu for some time - at that point, we start sending STOP and CONT signals on a duty cycle, with longer and longer STOPped periods until cpu usage is at a threshold. The general idea is to do short jobs quickly, while throttling long jobs (thus unfortunately making them even longer) in order to continue doing other short jobs quickly.
The runsc sandbox is not a single process, there are in fact 5 per sandbox in our setup. Runsc can work with kvm or ptrace. Kvm is not available to us, so we use ptrace. With ptrace, there is one process that is the appropriate one to duty cycle, and another that needs to receive a signal in order to yield. This diff adds the necessary machinery.
This is a conservative change, where I stick with our existing throttling mechanism and adapt it to the new sandbox. It would be reasonable to consider switching throttling. There's a lot the OS allows. We can set a quota for how much cpu a process can use within a given period, for example. However the overall behavior with that would be quite different to what we have, so feels like this would need more discussion.
The implementation contains use of a linux utility `pgrep` since portability is not important (runsc is only available on linux) and there's no node api for enumerating children of a process.
The diff contains some tweaks to `buildtools/contain.sh` to streamline experimenting with Grist and runsc on a mac. It is important for throttling that node and the sandbox processes are in the same process name space, if docker is in between them then some extra machinery is needed (a proxy throttler and a way to communicate with it) which I chose not to implement.
Test Plan: added test; a lot of manual testing
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3113
Summary:
The /assign endpoint checks if a document is on the desired worker
and moves it if not. This is never done under regular operation, but
is useful when quarantining a misbehaving document.
The endpoint was failing to operate correctly if the requester did
not have access to the document. This diff makes the endpoint
accessible through a /housekeeping route, using the same pattern as
the /force-reload endpoint.
Test Plan: added test
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3109
Summary:
- Change "Continue" button to "Review" (we don't charge immediately,
first show a review screen)
- Show more informative messages for certain failures with discount
coupons.
- Focus form elements with error, or at least the part of the form
containing an error.
- Auto-focus discount input box when it gets toggled on.
- Show warning about URL changes only when subdomain is changed.
Test Plan: Updated tests; tested focus and changed error messages manually.
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3115
Summary:
Pasting data like `A\nB\tC` was failing because the first row, used for
checking column type (to handle rich data), was failing on an undefined value.
Discovered while trying out the fix in https://phab.getgrist.com/D3110.
Test Plan: Tested manually. The case mentioned now works as expected.
Reviewers: alexmojaki
Reviewed By: alexmojaki
Differential Revision: https://phab.getgrist.com/D3111
Summary: title
Test Plan:
Tested manually. Blank rows at the end are no longer pasted. Pasting multiple columns separated by tabs can still have blank cells in some cells of the final rows.
I don't think this needs an automated test.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3110
Summary:
Push webhook events to redis queue with key based on docId.
Remove events from redis after sending using LTRIM.
Put failed events back on the end of the queue under normal circumstances.
When the event queue gets too long:
- Wait until it gets consumed before continuing.
- Drop failed events (i.e. don't put them back on the end of the queue)
- Limit webhook retries to 5
Test Plan: Tested that interactions with redis are as expected using redis MONITOR command.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3100
Summary: Allows any timezone abbreviation associated with the given timezone, and simply ignores it. Previously only certain abbreviations worked and they were not unique so using them outside the US was broken.
Test Plan: Added parseDate tests
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3106
Summary:
Adding sort options for columns.
- Sort menu has a new option "More sort options" that opens up Sort left menu
- Each sort entry has an additional menu with 3 options
-- Order by choice index (for the Choice column, orders by choice position)
-- Empty last (puts empty values last in ascending order, first in descending order)
-- Natural sort (for Text column, compares strings with numbers as numbers)
Updated also CSV/Excel export and api sorting.
Most of the changes in this diff is a sort expression refactoring. Pulling out all the methods
that works on sortExpression array into a single namespace.
Test Plan: Browser tests
Reviewers: alexmojaki
Reviewed By: alexmojaki
Subscribers: dsagal, alexmojaki
Differential Revision: https://phab.getgrist.com/D3077
Summary: Handle reflist columns in ViewFieldRec.parseValue
Test Plan: Reused section of test of reference columns
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3105
Summary:
- Grouping series may result in series with inconsistent number of values. This can result in inconsistent ordering of the bars displayed by plotly.
- This diff fixes it by consolidating grouped series by adding unll values for each missing xvalues in the series.
Here a is a minimal example of that bug:
{F36639}
Test Plan: Includes new nbrowser test.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3085
Summary:
Handle reference columns in ViewFieldRec.valueParser.
Extracted code for reuse from ReferenceEditor to look up values in the visible column. While I was at it, also extracted a bit of common code from ReferenceEditor and ReferenceListEditor into a new class ReferenceUtils. More refactoring could be done in this area but it's out of scope.
Changed NTextEditor to use field.valueParser, which affects numeric and reference fields. In particular this means numbers are parsed on data entry, it doesn't change anything for references.
Test Plan:
Added more CopyPaste testing to test references.
Tested entering slightly formatted numbers in NumberFormatting.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3094
Summary:
Searching with the keyboard shortcut selected the previous search text in the
search box, but using the Search icon did not. A user reported it as an
inconvenience: having to manually delete the value before searching for a new
one.
Test Plan: Verified manually
Reviewers: georgegevoian
Reviewed By: georgegevoian
Subscribers: georgegevoian
Differential Revision: https://phab.getgrist.com/D3102
Summary:
The document collecting new user info
(https://docs.getgrist.com/doc/GristNewUserInfo) got very slow, taking 40+
seconds for cold open. Sign-up submissions had to wait this time to proceed to
next step, because they waited for the write to this doc, which was blocked on
the Calculate action to complete.
Two changes were made: one to remove all expensive columns and summaries in the
actual doc, so the doc is back to opening in single seconds, and times should
be acceptable now.
The second change is this diff: to avoid waiting for the write step, so that it
doesn't affect users even if it gets slow again.
Test Plan: Existing test continues to work with a minor reliability tweak.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3103
Summary:
Action summaries by default will drop rows in bulk changes, keeping only a few of them as examples. This diff allows overriding that, or selectively preserving some columns in their entirety.
This is intended for use with webhooks.
Test Plan: added test
Reviewers: alexmojaki
Reviewed By: alexmojaki
Differential Revision: https://phab.getgrist.com/D3098
Summary:
Shares of the same role (e.g. viewer) at different levels could interact for a resource (e.g. a doc) shared with everyone@, potentially blocking the listing of that resource. This diff removes the interaction.
The permission of a user on a resource is calculated by finding all acl rules that link that resource to a group to which the user belongs, or to a group that has a subgroup to which the user belongs, etc, and then bitwise-or-ing the permissions on the acl rules. A later wrinkle was to allow public sharing via special users. A still later wrinkle was to avoid listing resources if they were only shared with the special everyone@ user, while allowing access to them if user has their full link. That wrinkle had a bug, where if e.g. a doc were shared with everyone@ as a viewer, and the org the doc was in was shared with someone@ as a viewer, and the doc inherited the org permissions via a workspace, then that doc would end up not being listed.
The fix is straightforward enough, but needs different code for postgres and sqlite, and is a bit verbose because we unwrap subgroups to a few levels rather than doing recursion (which looks cleaner but was slower in benchmarks).
Test Plan: added test that fails without this fix
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3095
Summary: Adds parseDateStrict function based on parseDate, uses it in DateParser subclass of ValueParser.
Test Plan:
Tweaked parseDate test to check parseDateStrict.
Extended test in CopyPaste to test parsing dates as well as numbers.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3088
Summary:
- This should make these easier to work with and make changes to.
- Removes one unused method.
Test Plan: No changes of behavior, existing tests should pass.
Reviewers: alexmojaki
Reviewed By: alexmojaki
Differential Revision: https://phab.getgrist.com/D3091
Summary:
This gives more guidance to users when editing document shares in the UserManager dialog.
* For a document on a team site, any shares with team members are marked `Team member`.
* Shares that count as external collaborators are marked for documents on a team or personal site as `collaborator` (personal site) or `outside collaborator` (team site).
* Collaborators are marked `1 of 2`, `2 of 2`, and then `limit exceeded`.
* On a team site, links are offered for each collaborator to add them to the team. The links lead to a prefilled dialog for managing team membership which can be confirmed immediately, allowing the user to continue without interruption.
* On a personal site, for the last collaborator and beyond, a link is added for creating a team. This isn't seamless since creating a team involves billing etc.
There's a small unrelated tweak in tests to remove a confusing import from `test/browser` in `test/server`.
One thing I didn't get to is checking if owner of doc is owner of site. If they aren't, they may try to add a member and be denied at that point - it would be more polite to change messaging earlier for them.
Test Plan: added and updated tests
Reviewers: georgegevoian
Reviewed By: georgegevoian
Differential Revision: https://phab.getgrist.com/D3083
Summary:
- Sharing, Client, DocClients, HostingStorageManager all include available info.
- In HostingStorageManager, log numSteps and maxStepTimeMs, in case that helps
debug SQLITE_BUSY problem.
- Replace some action-bundle logging with a JSON version aggregating some info.
- Skip logging detailed list of actions in production.
Test Plan: Tested manually by eyeballing log output in dev environment.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3086
Summary:
When a value $B.A is a ReferenceList, returning it in an Any column can cause
an "unmarshallable object" error, if the RecordSet happens to be storing
row_ids in the form of a nested RecordList object. This happened consistently
when $B.A started off as another type and got converted to ReferenceList.
A specific manifestation was when a reference column $B uses "A" as a display
column, and this column gets converted from Text to ReferenceList.
Test Plan: Added a test that reproduces the problem.
Reviewers: alexmojaki
Reviewed By: alexmojaki
Differential Revision: https://phab.getgrist.com/D3089
Summary:
Add ValueParser file, base class, and subclasses for column types. Only NumericParser is used for now.
Add valueParser field to ViewFieldRec.
Use valueParser when parsing pasted text data in Grid and Detail views.
Test Plan: Add test to nbrowser CopyPaste suite, copying into a numeric column with different currency and locale settings.
Reviewers: dsagal
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3082
Summary: Upgrade chardet version from 2.3.0 to 4.0.0 to improve encoding detection when importing files with non-ascii characters.
Test Plan: the tests
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3080
Summary: return NOW(tz=tz).date()
Test Plan: None, curious to see if this fixes test_time_defaults when run near midnight.
Reviewers: jarek
Reviewed By: jarek
Differential Revision: https://phab.getgrist.com/D3079
Summary:
Record numbers of rows, columns, cells, and bytes of marshalled data for most calls to table_data_from_db
Export new function get_table_stats in the sandbox, which gives the raw numbers and totals.
Get and log these stats in ActiveDoc right after loading tables, before Calculate, so they are logged even in case of errors.
Tweak logging about number of tables, especially number of on-demand tables, to not only show in debug logging.
Test Plan: Updated doc regression tests, that shows what the data looks like nicely.
Reviewers: dsagal, paulfitz
Reviewed By: dsagal
Differential Revision: https://phab.getgrist.com/D3081
Summary:
New plan signups now include a discount code field in
the signup form. If a valid discount code is entered, a
discount will be applied on the confirmation page.
Test Plan: Browser and server tests.
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: jarek
Differential Revision: https://phab.getgrist.com/D3076
Summary:
Added NumberParse.ts, counterpart of NumberFormat.ts.
Contains generic functionality for parsing numbers formatted by Intl.NumberFormat, not tied to documents or anything.
This doesn't change any actual behaviour, applying this parsing when pasting/typing in numeric columns will be a separate diff.
Test Plan: New file with extensive unit tests.
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: jarek
Differential Revision: https://phab.getgrist.com/D3078
Summary:
This verifies that all existing tests are capable of running under python3/gvisor, and fixes the small issues that came up. It does not yet activate python3 tests on all diffs, only diffs that specifically request them.
* Adds a suffix in test names and output directories for tests run with PYTHON_VERSION=3, so that results of the same test run with and without the flag can be aggregated cleanly.
* Adds support for checkpointing to the gvisor sandbox adapter.
* Prepares a checkpoint made after grist python code has loaded in the gvisor sandbox.
* Changes how `DOC_URL` is passed to the sandbox, since it can no longer be passed in as an environment variable when using checkpoints.
* Uses the checkpoint to speed up tests using the gvisor sandbox, otherwise a lot of tests need more time (especially on mac under docker).
* Directs jenkins to run all tests with python2 and python3 when a new file `buildtools/changelogs/python.txt` is touched (this diff counts as touching that file).
* Tweaks miscellaneous tests
- some needed fixes exposed by slightly different timing
- a small number actually give different results in py3 (removal of `u` prefixes).
- some needed a little more time
The DOC_URL change is not the ultimate solution we want for DOC_URL. Eventually it should be a variable that gets updated, like the date perhaps. This is just a small pragmatic change to preserve existing behavior.
Tests are run mindlessly as py3, and for some tests it won't change anything (e.g. if they do not use NSandbox). Tests are not run in parallel, doubling overall test time.
Checkpoints could be useful in deployment, though this diff doesn't use them there.
The application of checkpoints doesn't check for other configuration like 3-versus-5-pipe that we don't actually use.
Python2 tests run using pynbox as always for now.
The diff got sufficiently bulky that I didn't tackle running py3 on "regular" diffs in it. My preference, given that most tests don't appear to stress the python side of things, would be to make a selection of the tests that do and a few wild cards, and run those tests on both pythons rather then all of them. For diffs making a significant python change, I'd propose touching buildtools/changelogs/python.txt for full tests. But this is a conversation in progress.
A total of 6886 tests ran on this diff.
Test Plan: this is a step in preparing tests for py3 transition
Reviewers: dsagal
Reviewed By: dsagal
Subscribers: dsagal
Differential Revision: https://phab.getgrist.com/D3066
Summary:
Adding validation for api /records endpoint, that checks if the json payload is valid.
Modifying POST /records endpoint to allow creating blank or partial records.
Test Plan: Updated tests
Reviewers: alexmojaki
Reviewed By: alexmojaki
Differential Revision: https://phab.getgrist.com/D3061
Summary:
- Puts events on a queue in memory and ensures they are sent in the order they were generated.
- Makes the caller (Sharing.ts) wait until changed records have been fetched from the DB, but allows it to continue after while remaining work happens asynchronously.
- Gathers all new webhook events into an array so they can be backed up to the queue on redis in a single command (in a future diff).
- Uses changes in isReady to determine event type, no more 'existed before'
The structure of the code has changed a lot, so I think the scope of the diff needs to stop here. Lots of work is still deferred in TODOs.
Test Plan: Updated existing test. Actually dropped testing of retry on failures and slowness because it no longer made sense to keep that as part of the current test, so a new test for that will be added in a future diff.
Reviewers: paulfitz
Reviewed By: paulfitz
Differential Revision: https://phab.getgrist.com/D3074
