It's misleading to imply that an .env doesn't exist when it's not what the actual check looks for.
It's also poor design to hardcode an unrelated error message in a "redirect to installer" function,
which only worked because this function was called from exactly one other place where this
error message was correct.
This method should typehint the incoming value once union types are
available, but for now this ensures that method_exists() is not called
on scalar values, which is unsupported in PHP 8.
It does not make sense to require a psr-container-implementation when the package is itself providing an implementation, as it fulfills its own requirement.
* DOCS: Add new graphql 4 docs
* Reorganise docs
* Docs done
* Basic graphql index page
* TOC for getting started
* show folders on graphql index page
* Add middleware note
* Docs update
* Update docs to reflect flushless schema
* Docs updates
* Docs for getByLink
* Query caching docs
* Docs on nested operations
* update docs for new graphql dev admin
* Docs for configurable operations
* Replace readSiteTrees with readPages
* Schema defaults docs
* Docs for inherited plugins
* Docs for customising *
* Docs for field whitelisting
* Change whitelist word
* New docs on modelConfig
* Document dev/build extension
* Document default/global plugins
* Document new input type fields config
* Apply suggestions from code review
Co-authored-by: Andre Kiste <bergice@users.noreply.github.com>
* Note about when procedural schema gets built
* Fix link
* Apply suggestions from code review
Co-authored-by: Andre Kiste <bergice@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Andre Kiste <bergice@users.noreply.github.com>
* DOCS Note about plugins in custom queries
* DOCS Note about filter and custom resolvers
* DOCS Note about canview paging
* DOCS Updated guidance on _extend
See https://github.com/silverstripe/silverstripe-graphql/issues/296
* Apply suggestions from code review
Co-authored-by: Andre Kiste <bergice@users.noreply.github.com>
* DOCS Pre-release warning
Co-authored-by: Ingo Schommer <ingo@silverstripe.com>
Co-authored-by: Andre Kiste <bergice@users.noreply.github.com>
Co-authored-by: Ingo Schommer <me@chillu.com>
Note that it's currently unclear whether Silverstripe Cloud or CWP support this,
but it shouldn't block us from recommend this in the open source project.
It's documented in the "server requirements", which should make it pretty
clear that this requires you to have control over server configuration (or check with those that have).
See https://github.com/silverstripe/silverstripe-framework/issues/7710
For whatever reason (different locale version) my local dev env uses
narrow-NBSPs (Unicode 8239) rather than regular NBSP in its localised
strings. This patch makes the tests robust against this difference.
Note that this occurred running the tests in Lando.
Co-authored-by: Robbie Averill <robbie@averill.co.nz>
