-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix gridded yxt-Card styling for 2 columns #784
Closed
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The navigation param in the base config is called verticalPages now, and is an array rather than an object. TEST=none
* Always use CustomEvent polyfill This commit changes DOM.trigger to always use our custom event polyfill, instead of looking in the window first to see if CustomEvent exists as a function. Consulting was running into issues with ie11 saying that it doesn't know how to use the Event constructor. This likely could have also been fixed by changing `return new Event` to `return new CustomEvent`, however it felt more robust to me to always rely on our own polyfill, instead of something somebody using the sdk adds. TEST=manual Went on ie11, checked that adding a CustomEvent function onto the window manually through the dev tools causes the error consulting was seeing. This was checked by typing into the searchbar, since the searchcomponent uses DOM.trigger. Then, removed code and checked that there were no errors, and autocomplete was displaying values that I could click on to run a search. Also checked on chrome.
This PR adds the infrastructure needed to write TestCafe acceptance tests for the SDK. It also creates the first such test. This acceptance test is relatively simple and verifies a basic flow on a universal search page. Essentially, it types a manual query, clears it, picks an autocomplete option and verifies that the expected VerticalResults sections are there. This test is a jumping off point and is by no means exhaustive. We should look to add more to this test, test other flows, and make sure we have Block models for every component. Right now, tests are only run in Safari and Chrome. Firefox support and IE11 support we come in later PRs. J=SPR-1963 TEST=auto Made sure the new acceptance test could pass. Watched as the test ran in both Safari and Chrome to make sure the test matched what I thought it should be doing.
The correct Selector was not being used to select the various autocomplete options in the DOM. This PR corrects that. TEST=auto Reran acceptance test and made sure it still worked.
This PR ensures that the SearchBar component has a border when viewed on Mobile. The border is the same as the one that already exists on the Desktop view. Without a border, the SearchBar's input was impossible to find against a white background on Mobile. J=SPR-2001 TEST=manual Verified that the border appears the same accross Desktop and Mobile in both IE11 and Chrome.
Update the legacy card to properly reference big date, fixes this issue: #679. TEST=manual Locally, create event vertical page and display big date.
Checked for other instances of data-component="IconComponent" inside #ifs for similar issues, did not find any. TEST=manual Tested that can specify a custom iconUrl (using accordion card in vertical results) with just an iconUrl and no icon option specified
Admins were adding this css by default when migrating to v0.13 TEST=manual see that vertical on univeral results with useAccordion = true has this border-top styling
This was causing a breaking change and being overridden manually TEST=manual open vertical search page, check that on hover and using the dev tools hover button the cta does not change color
_config._parentOpts.target is a valid fallback option for standard and legacy cards now, and the places where the target attribute defaulted to _blank have been reverted to default to _self. J=SPR-2025 TEST=manual Manually verified that setting target in the VerticalResults config would be reflected in the card links. Also verified that not setting a target would default to _self instead of _blank.
…700) For this issue #677 (https://github.com/yext/answers/issues/677), add documentation for the existing UniversalResults config option useAccordion.
When displayAllResults is true, before a page would display applied filters in the results header, like it does for normal results. It's still the case that only nlp filters are displayed. J=SPR-2003 TEST=manual Checked that page was displaying applied filters for no results Made change checked that filters no longer display for no results checked that filters still display for regular results
Adds a noResults configuration which alters the behavior of the map component when a vertical search returns no results. J=SPR-2002 TEST=manual Verified that inputting a no-results query while displayAllResults was set to false would cause the map to be empty (and hidden unless otherwise set to display an empty map).
Removed extra bordering on filter component so it looks less bad TEST=manual Eye test
Across all browser but internet explorer, searchbar placeholder text was gray, but on ie11 was black. This was changed to gray. TEST=manual check ie11 in browserstack, and that on chrome things look the same
Add documentation for Map Component. Resolves this issue: #394 (https://github.com/yext/answers/issues/394)
* Make CustomEvent Polyfill match previous non-polyfilled behavior instead of first trying to use the event constructor, if it exists on the window, before resorting to the polyfill. However, this polyfill did not match the behavior of the event constructor it was replacing, and had cancelable and bubbles set to false by default instead of true. Also untrack .DS_Store and add it to the gitignore TEST=manual Tested on ie11, firefox, safari, chrome. Test that typing in search bar and submitting a search still works. Check that submitting a near me query, then resubmitting another query does not prompt for the geolocation a second time.
This new option can be set in the top-level Answers.init configuration. It can be used by both VerticalResults and the MapComponent. Both of these components have logic that will favor an individual noResults config to the one in globalStorage. Additionally, both components will apply the same sane defaults to any noResults attributes not specified. A change was also made to ensure that the VerticalResults component passes along its noResults config to any child MapComponent, to ensure consistency. Note that Rose requested the displayAllResults default be false for the MapComponent TEST=manual Ensured correct behavior in all the following situations: VerticalResults with included MapComponent - Specified noResults config for VerticalResults. - Specified noResults config at global level. - Specified noResults config at both levels. - Specified noResults config at neither level. VerticalResults and separate MapComponent - Specified noResults config for both components. - Specified noResults config at global level. - Specified noResults config at global and component levels. - Specified noResults config at neither levels.
* update universal results borders margin on top and bottom of yxt-Results-filters was not receiving border-left and border-right (because margin does not receive border in css) causing gaps to appear Also fix double border TEST=manual check that vertical results on vertical search, vertical results on universal searchh, accordion results do not have gaps in the border, and otherwise appearance is unchanged
This change caused an undesired breaking styling change to the dignity funerals site. The margin-top: 0 and margin-bottom: 0 can be kept, since that was done to override user-agent styling. TEST=manual check that padding has same value as before
UniversalResults is dumb and this accommodates the dumbness. TEST=manual UniversalResults shows a map now. Same difference for Vertical Results.
If a user enters a cardType that is built in, no changes. If they enter a cardType that is not built in, the library interprets it as a component name. TEST=manual Update the config to use built in vs custom cardTypes. Observe changes in browser.
If vertical key is falsy because it is null or empty we should fall back to the url. Fixes issue where universal's vertical key was null and it was not showing in nav TEST=manual,auto Run local experience and see that universal tab shows up and component renders without errors.
#535) If verticalPages config is present in the component config, prefer it over the one in the top-level config. TEST=manual Run answers experience locally putting config in top level config, component config, then both.
Pagination component now prefers verticalKey from its component config. Falls back to the one from answers top-level config. TEST=manual Run answers experience locally putting config in top level config, component config, and both.
#578) If an implementer passed in a verticalKey that was empty or undefined, its entry would be ignored in the navigation component. Changing the null check so any falsy verticalKey is treated the same. TEST=manual Test on local answers instance that universal verticalPages config is used when omitting vertical key.
…ds compatibility (#712) TEST=manual Replace analytics config option with eventType and notice that analytics event type is updated.
Legacy card styling was overflowing horizontally. Cards generally were adding extra white space vertically. This fixes both so cards in IE11 look the same as other browsers. TEST=manual Tested in Chrome, IE11 and Firefox.
TEST=manual Visually test using dev site and client site on Universal Results page.
) Change so placeholders are not added when the maxNumberOfColumns is 1. Update placeholder to be an empty element rather than a repeated result card. The placeholder has the same horizontal margins and border as a card so it has the same width; add number of placeholders needed - 1 to account for the requirement that column count changes on screen re-size. Addresses this issue: #664 TEST=manual On a Universal page, visually inspect on dev site locally and client site locally. Test on Vertical page with 4, 3, 2, and 1 columns, adjusting screen size and noting that the cards stay the same size.
Added for backwards compatibility. The legacy card is the only one that uses this variable. TEST=manual Test locally on Universal and Vertical pages that the config is passed through and shows in the DOM.
Since we added a border to yxt-Results-titleBar oobx, and since the titleBar sits on top of yxt-Accordion-list, yxt-Accordion-list probably needs a border-left and border-right as well so the results line up with the titlebar. TEST=manual test that dignity funerals site has lined up AccordionResults and yxt-Results-titleBar after updating
Previously the dependency was at version 3.6.4. Dependabot suggested that we upgrade. This upgrade should not cause any issues as the new version has the same major and minor revision as the current version. J=SPR-2069 TEST=manual Did a manual smoke test in IE11 since this dependency is used for the polyfills. Made sure I could navigate between pages, click autocomplete options, clear a search, submit QA, see a Map, etc. Ensured there were no console errors.
Dependabot suggested we update the versions of the following @babel plugins: - @babel/plugin-proposal-object-rest-spread - @babel/plugin-transform-arrow-functions - @babel/plugin-transform-runtime Even though these plugins are devDependencies, they should be kept up-to-date. We use these plugins to transform our source code when creating the various asset bundles. It's possible, then, that security issues in these plugins could make their way into the builds of the SDK. These were only minor version upgrades, so the risk should be low. TEST=manual These @babel plugins are mainly used to help polyfill the SDK's source code and make it IE11 friendly. To test, then, I ran a few smoke tests on a demo site in IE11. I made sure search, navigation, spell check, autocomplete, and qa submission functionality worked and no errors were present in the console. I also made sure to exercise some areas of the code where we use arrow functions or the spread operator, to ensure the polyfilling of those continued to work.
Allow an implementer to specify an iconURL in the verticalPages config. This allows them to use an icon that is not in the built-in set that ships with the template bundle. If both iconUrl and icon are passed in, they will both be passed through to the IconComponent which has logic to determine which takes priority. TEST=manual In dev site for a Vertical page, use in an iconUrl in the verticalPages config. View in browser to confirm custom icon displays. Also test to verify that built-in icons still work.
* Let url be null on Legacy card. Let url be null-ed by the user, so that it can be forced to default to a linkless title. * Add parens and update standard card
…ages (#737) On chrome, this lines up ctas perfectly whether there are images or not On ie11 they still don't line up properly, but I hope this can be a more short-term fix? TEST=manual test with maxNumberOfColumns = 4 and maxNumberOfColumns = 1, using element.getBoundingClientRect(); for cards with images and without images
* Fix tall mobile cards * Alexis did it!
…702) * Update circle ci deploy to also update major version When deploying a tagged commit, also create and override the major version for that commit. In other words previously we only stored v1.0 and v1.0.0, now also store v1 TEST=manual set $CIRCLE_TAG in local bash and tested that was able to export $MAJOR_VERSION
TEST=manual Proofread the README and check veracity.
TEST=none
J=SPR-2131 TEST=none
This PR allows the SearchBar to be configured to not use a form as the query submission context. By default, a form is used, unless a client sets the new 'useForm' config option to false. This was done to support clients that wanted to use a SearchBar on an ASP.NET page. J=SPR-2175 TEST=manual Tested the following scenarios for SearchBars set to useForm true and false: - Entering text and hitting enter. - Entering text and hitting the submit button. - Entering text and clearing. - Clicking autocomplete option. - Using keys to navigate to and select an autocomplete option. - Clear button appears in the correct situations. - Search animation works as expected. - Near me query works as expected. In the above scenarios, made sure Analytics worked correctly, if applicable.
#753) Add modifier option to the VerticalResults portion of the documentation. TEST=none
Remove hardcoded "DEVICE" accuracy which forces the device accuracy help text.
Add the BSD3 license
The 'limit' here does not accept strings, only numbers. Updating example in the documentation to have the correct type. TEST=none
…758) The limit param is not used and is not supported by the back end of Answers for universal searches. Removing the unused param from the README and the component because it is misleading.
Fixes #783 We should be applying width to the Card for 2 columns just as we are doing for 3 and 4 columns.
Closing to start off of v0.14.0 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes #783
We should be applying width to the Card for 2 columns just as we are doing for 3 and 4 columns.