This script is similar to the existing one that schedules
report results to to be sent to a PM on a regular basis,
but instead takes a topic ID and posts to that. This way
people can have report results sent to a public topic regularly
too and not have to deal with PM recipients and so on.
The old float validation had several bugs. It will recognize strings
like "a1.2" and "3.4b" as valid doubles, but will not recognize integers
like "1234" as doubles. Also, since an empty string is not falsy in Ruby,
it will recognize "Inf" as -Infinity.
This commit fixes these issues
Validation of `user_id` parameter will throw a 500 error because
`User.find_by_username_or_email` does not throw
`ActiveRecord::RecordNotFound`, but silently returns `nil`.
This results in a `NoMethodError` in `object.id` on the next line
What's the problem?
===================
TL;DR: When the user enters an incorrect default value, its
corresponding param-input will disappear
When creating a parameter from SQL, we perform cast_to_ruby, which means
that if the default value given by the user is invalid, the entire
parameter will not be added to the param_list of the query.
This behavior is very confusing, and users will not understand why an
incorrect initial value will cause the param-input to disappear.
What's the fix?
================
The cast_to_ruby process is canceled in create_from_sql, so that
param-input with incorrect default value will still be displayed.
We have a simple validation process on the front end, which is enough to
prompt whether some default inputs are incorrect.
We used `with_deleted` incorrectly in the code. This method does not exist
on `User`, `Badge`, and `Group`. This will cause an error when entering a
numeric id in these three parameter input, forcing the user to enter the
name/username of these inputs.
See https://github.com/discourse/discourse-data-explorer/pull/307#issuecomment-2291017256
The main change here is that we now send a single PM to a group rather than individual PMs to each group member.
This change also adds email recipients correctly as target_emails since support was added within Discourse Automation.
* Added can_see_bookmarkable? methods to BaseBookmarkable, implement
that in QueryGroupBookmarkable
* Update spec to check that Notification for reminder has
bookmarkable_id and bookmarkable_type
c.f. https://github.com/discourse/discourse/pull/25905
If a column is payload or contains _payload it will be assumed
it has JSON data in it, then we will show the truncated JSON in the
result column with a button to show the full-screen formatted
JSON using our full-screen code viewer. We also do the same if
the column is the `json` postgres data type.
This change includes the Discourse base url to allow for subfolder installs, it will also fix another potential issue where reports are sent via email (since we use the email_group_user component) that would require the base_url to know which domain to point to.
This PR fixes 2 issues that were picked up by users for the Scheduled Data Explorer Report automation script and a couple of small improvements to better match the format of manual data explorer query results.
The first issue is that within result_to_markdown the colrender contains null values, and we are currently checking length (previously treated as a packed array but it is actually a sparse array). Therefore we can check if the current index of the array is null rather than checking the size of the array.
The second issue addresses the blank query_params field. When the data explorer script does not require any params to be passed in via the automation script then it will have a nil value, however it should be defaulted to {} within the plugin.
To improve formatting the markdown table for PMs is now aligned to left and where values are substituted (for example user_id becomes username) we then include the id within brackets, for example:
user_id becomes username (user_id)
This feature enables admins to create reports automatically based on a recurring schedule.
It introduces a new automation script that includes the new email_group_user field added to discourse-automation, along with a query_id and query_params to pass in parameters to the existing data explorer query.
The output of the report will be sent via pm (as a markdown table) to the recipients entered within the automation script.
The automation (supports individual users, email addresses and groups).
This commit updates the plugin to the latest guidelines, as shown in
discourse-plugin-skeleton, which involves moving a lot of the code to
dedicated files, use proper namespaces, use the autoloader as much as
possible, etc.
Reintroduces the migration removed in `cdfc5d4`, and fixes it to work with `query_groups`. Since the code is fairly complex, I moved into a rake task so it can be tested and make sure it works.
* FEATURE: Add hide button (toggleable) for all queries (frontend only)
* Switches between hide/unhide on click
* Works almost like the delete button, but toggles between the query's
hidden attribute instead
* So far this is only a frontend feature, the backend implementation
still needs work
* Revert "FEATURE: Add hide button (toggleable) for all queries (frontend only)"
This reverts commit a8771d2ad5.
REVERT: Remove button that hides queries (frontend)
* Prepare for migration of old frontend logic to backend
* We are going to reuse the existing delete button, but change its
backend logic to enable soft deletion. From the user's perspective
nothing will change, but any deletion mistakes can be reverted.
* DEV: Hide user queries upon deletion, but keep them in store
* Creating a new query will set its hidden attribute to false by
default
* Deleting a user-made query will not delete it from the store, but
set its hidden attribute to true
* User queries will not be indexed if they are hidden
* Undeleting a query will unhide it, and will be indexed
* Updating a hidden query will unhide it, and will be indexed
* SPEC: Add spec for hidden/deleted queries
* Hidden queries should not be shown
* FEATURE: Add ability to delete/hide system queries
* System queries are now able to be deleted from view, but will remain
in the backend for retrieval, if necessary
* FEATURE/DEV: Add rake commands for query soft deletion
* query:list_hidden - Shows a list of hidden queries
* query:hide_all[only_default] - Hides all queries, w/ boolean arg to
hide only default ones
* query:unhide[id] - Unhides a query by id
* query:unhide_all[exclude_default] - Unhides all hidden queries,
w/ boolean arg to exclude default ones
* Remove rails loggers
* UX/DEV: Update query rake tasks to be more user friendly
* Split query:hide_all[only_default] into two tasks:
* query:hide_all - Hides all queries
* query:hide_all:only_default - Hide only default queries
* Split query:unhide_all[exclude_default] into two tasks:
* query:unhide_all - Unhides all hidden queries
* query:unhide_all:exclude_default - Unhides all non-default
queries
* Rename file to match task name
* UX: query:unhide can accept multiple arguments
* Example: rake query:unhide[-5,-6,-7,3,5,6,-13,-14,-10]
* UX: Update query rake tasks to output cleaner messages
* Remove unneeded comment
* DEV: Keep only necessary rake tasks, use more specific naming
* UX/DEV: Add rake task for hard deletion, better console logs
* User is able to hard delete a query only if it is hidden, otherwise
output a message stating so
* Add commented examples above each task
* Add rainbow support for more readable console logs
* Successful messages will display green, failures display red,
additional info displays yellow
* Separate multiple queries with spaces instead of lines
* DEV: Remove rainbow colorizing in console logs
* Rainbow is a dependency of rubocop and it may go away in the future
* Rainbow is only used for dev and test environments
* DEV: Add Rails engine to enable rake tasks to be loaded at runtime
* DEV: Favor require - load files only if they are not already loaded
* SPEC: Add tests for data_explorer[id] rake command
* Test if a single query is hidden correctly
* Expect length of query list to not be modified
* Expect array of hidden queries to have exactly 1 element
* Expect that one element to have the same ID as the one invoked to
be hidden
* Test if multiple queries are hidden correctly
* Expect length of query list to not be modified
* Expect array of hidden queries to have the number of elements
equal to the number invoked to be hidden
* Expect the elements to have the same ID as the ones invoked to be
hidden
* Test if a query exists in PluginStore
* Expect query list to be empty
* DEV: Clear pre-existing tasks before redefining
* This prevents double invocation when user invokes the task once
* SPEC: Add tests for unhide_query rake task
* Test if a single query unhides correctly
* Expect length of query list to not be modified
* Expect array of hidden queries to have exactly 1 element after
unhiding 1 of 2 queries
* Expect remaining element to be hidden
* Test if multiple queries unhide correctly
* Expect length of query list to not be modified
* Expect array of hidden queries to have exactly 1 element after
unhiding 3 of 4 queries
* Expect remaining element to be hidden
* Test if a query exists in PluginStore
* Expect query list to not be modified
* SPEC: Add tests for hard_delete rake task
* Test if a single query hard deletes correctly
* Expect length of query list to be shorter by 1
* Expect array of hidden queries to have exactly 1 element after
hard deleting 1 of 2 queries
* Expect 1 remaining hidden element
* Test if multiple queries hard delete correctly
* Expect length of query list to be shorter by 3 after hard deleting
3 of 4 queries
* Expect array of hidden queries to have exactly 1 element after
hard deleting 3 of 4 queries
* Expect 1 remaining hidden element
* Test if a query exists in PluginStore
* Expect hidden query list to not be modified
* Test if a query is not hidden
* Expect query list to not be modified
* UX: Favor newline char in place of puts for logs
* Condensed console logs to output newline char instead of another puts
statement (reduces number of lines used significantly)
This new feature means that if you name a column `xyz_url` then the return
value will be treated as a URL.
If you would like to name the link use the format `name,https://cnn.com`
Martin Brennan
Jarek Radosz
Linca
Robert
Loïc Guitaut
Natalie Tay
David Battersby
Penar Musaraj
TheJammiestDodger
Canapin
Bianca Nenciu
David Taylor
Frank
Krzysztof Kotlarek
Roman Rizzi
jjaffeux
Ricky Chon
Michelle Saquetim
Rafael dos Santos Silva
Sam Saffron
Guo Xiang Tan
Rishabh
Simon Cossar
Maja Komel
Rishabh