refs https://github.com/TryGhost/Team/issues/597
- To be able to transpile the library for different runtimes (make it polymorphic) had to get rid of dependencies that were not compatible with ES Modules
- By making errors an injectable constructor option it removes the depencency and allows to transpile the library for multiple targets
- The `errors` option is now a required parameter for `loadLimits` method. It errors if it's missing (error message copy inspired by content api error 69fcea0582/packages/content-api/lib/index.js (L21))
refs https://github.com/TryGhost/Team/issues/597
- Before adding more parameters documented existing ones
- Created LimitConfig type definition to have easier look into the structure of limit conifiguration
refs https://github.com/TryGhost/Team/issues/587
- Documented common usecases such as:
1. initialization and configuration of limit service
2. usage of "max" types of limits
refs https://github.com/TryGhost/Team/issues/587
- There was no test coverage for MaxLimit's errorIfIsOverLimit check. Added basic test to make sure upcoming modifications don't break existing functionality
refs https://github.com/TryGhost/Team/issues/587
- The optional {max} passed as an option allows to override currently configured limit and do a theoretical new limit check. For example: check if the max limit will be exceeded if the limit changes (user changes plans)
refs https://github.com/TryGhost/Team/issues/587
- Having a JSDoc gives better intellisense when the class is instantiated and provides clues about what each parameter might be used for
refs https://github.com/TryGhost/Team/issues/587
- When the 'max' configuration is missing the instance of the class breaks when used unexpectedly. Followed similar approach to currentCountQuery check by failing fast in the constructor
refs https://github.com/TryGhost/Team/issues/587
- Test were missing for class initialization and around how the limit currently works.
- Before extending it's behavior throught its valuable to cover current functionality to not accidentally break anything
refs: https://github.com/TryGhost/Team/issues/510
- Ghost config always uses camelcase. This was incorrectly implemented with snake case originally
- Swap to use camelCase by default, which is desirable, but support both
- It's really easy to support both in the loader and isLimited check, so we do this to stop ourselves tripping on this later
refs: https://github.com/TryGhost/Team/issues/510
- we need to make sure we take into account any invites that could be accepted at any time
- this counts all invites for non-contributor roles as well as all users who aren't contributors
- this should stop there being loop holes to inviting staff users
refs: https://github.com/lodash/lodash/issues/705
- Was seeing unexpected token = errors when using lodash templates in Ghost
- This is because we're setting template settings globally in this dependency and it affects every other user of lodash
- Using runInContext keeps this templateSettings change local to this lib
- Test proves that after requiring limits we can require lodash and have the default values again
refs https://github.com/TryGhost/Ghost/issues/12496
- `workerMessageHandler` option allows for custom worker message handling and allows to eliminate a need for loggers of any type inside of jobs.
- removing loggers from jobs solves file hanle leak which used to cause Ghost process to crash (see referenced issue)
refs #122
- In future changes there's a plan to add "inline" scheduled jobs, which would conflict current method naming.
- The amount of parameters in the methods was more than 3, so it made sense to transform them into an options object
- Scheduling still doesn't work for "inline" jobs. This should be solved as a part of upstream library (https://github.com/breejs/bree/issues/68)
closes https://github.com/TryGhost/Ghost-Utils/issues/118
- Custom error handling is needed to be able to override default bree
error handling logic.
- bree bump to 4.1.0 also fixed logging errors (object Object fix in
tests)
- The handler function receives two parameters. First contains an error
that has been thrown by the job. Second, job and worker metadata
no issue
- Provided more context about which type of job does what and introduced naming to be able to distinguish them
- The naming is still to be reviewed by peers
closes#117
- Having immediately executable offloaded jobs is necessary to be able to run usecases like: send batched emails now, or any other job that does not need to be scheduled
- Changed "simple" job timeout to make tests run faster
closes#119
- A future use-case which this feature caters for is allowing to migrate "post scheduler" to use job manager instead of managing scheduling itself
- removeJob method will be needed to allow "rescheduling" of the post
- this helps simplify the code and gets rid of Promise chaining
- apparently I can't easily use an async function within filter, so I've
left it for now
- this helps bring all the code together so we can extract it in the
future
- turning it into a class also lets us easily inject the i18n instance
and store it locally
refs https://github.com/TryGhost/Ghost/issues/12402
- Describes different types of jobs that could be executed depending on the nature of the jobs.
- Lays down ground rules on how to design scheduled jobs
no issue
- When providing a crontab schedule expression it should always contain 6 elements first one of them being a "seconds" schedule description . For example: '0/5 * * * * *' - meaning to run every 5 seconds
no issue
- When jobs are performing CPU intensive tasks they block main process'
event loop. They also can cause memory leaks or unexpected crashes
effectively crashing the parent proccess. To address these issues jobs need to be performed off of main
process. Worker Threads (https://nodejs.org/dist/latest-v12.x/docs/api/worker_threads.html)
are the best candidate for such work.
- These changes introduce an integration on top of bree
(https://github.com/breejs/bree/) which allows to run recurring
jobs in worker thereads. It falls back to child process execution for
Node v10 running without `--experimental-worker` flag.
- bree was chosen not only because it gives a polyfill for older Node
versions. It has support for some of the future use-cases Ghost is looking to
implement, like scheduled jobs.
- This changeset also includes a complete example of job running on an
interval with a possibility for graceful shutdown
no issue
- This is a quick implementation change to prevent from queue stalling and never becoming idle in case there's an error thrown from within the job function/module
- More robust error handling should be designed soon!
no issue
- Jobs should not always be functions, one of the standard practices is having a job defined in a module as a self contained executable function
- First parameter of `addJob` function now also handles path to modules which it imports and executes
no issue
- Accepting the schedule or a data when scheduled job should be run would follow a signature used in other established frameworks: (1) https://github.com/mperham/sidekiq/wiki/Ent-Periodic-Jobs#definition
- Another reason to put scheduling parameter first is this would allow leaving an optional "data" parameter as last
no issue
- CRON format is the most common one used for job scheduling and is well known to most developers
- This will become one of supported formats for job scheduling
refs https://github.com/TryGhost/Ghost/issues/11878
- To be able to identify the reason behind comparison failure on more granular level (like token expiration) had to provide additional information in return result for falsy token comparisons
refs https://github.com/TryGhost/Ghost/issues/11878
- There are multiple reasons why the token can be invalid. This coverage is meant cover these reasons and pave the way for introduction of more rganular errors causing the invlid token
* extracted promise lib files and tests from Ghost
* includes commits:
Updated var declarations to const/let and no lists
Move tests from core to root (#11700)
Extended sequence utility
Moved pipeline/sequence to lib/promise
* included commits:
Updated var declarations to const/let and no lists
Move tests from core to root (#11700)
Updated to use slugify method from SDK for safe string
Added Node v10 Support (#10058)
Dynamic Routing: Added migration for routes.yaml file (#9692)
Fixed missing Bluebird require in `security/password.js` (#9624)
🔥 Drop Node v4 Support
Added unit tests for models.Invite.add
Added lib.security.password lib
Moved unique identifier generation to lib/security
Moved tokens, url safe and safe string utility to lib/security
- History merged from Ghost, setup package structure and added a test
- This is exactly the kind of code that we should keep separate to Ghost, as it's rarely touched
- It's much easier to reason about and test when it's on its own
- `shellSync` was removed in execa v2
- this commit switches the usage to the recommended replacement; `sync`
- also adds a test for changelog, the part of code which uses execa
- the symlink test has been failing for some time because recent minors
of Node 12 produce different size zip files
- we were checking that the produced zip is a specific size, but we
basically just want to check it's smaller than the original folder
- this commit updates the test with the size of the original folder,
found by running `du -b test/fixtures/test-theme/`
refs https://github.com/TryGhost/Ghost/issues/11944
- it's useful for consuming applications to know if images can be transformed outside of actually performing the transform (eg, when outputting "resized" image URLs)
- added `canTransformFiles()` that returns a boolean for whether or not `sharp` is installed
refs TryGhost/Ghost#11835
- this error was added after this package was created. Once this gets released, we will be able to remove the errors file in the core 🥳
no-issue
This is a major change and will require that we explicitly define the
versions of the services which we depend on. This will make sure that
updated services will always be sure of what actions their dependents
will have.
refs https://github.com/TryGhost/Ghost/issues/11794
- archiver has an undocumented dot option, that allows you to include dotfiles
- our existing tests had a dotfile, but we didn't check properly that everything exists
- Swapped to using folder-hash to compare if the original, and compressed-then-decompressed folders are identical
- Added an example of a dotfolder with nested dotfile
- Updated compress to use the dot option, make it optional but on by default
- Use array destructuring
- Use @tryghost/errors
- Part of the big move towards decoupling, this gives visibility on what's being used where
- Biting off manageable chunks / fixing bits of code I'm refactoring for other reasons
- All var declarations are now const or let as per ES6
- All comma-separated lists / chained declarations are now one declaration per line
- This is for clarity/readability but also made running the var-to-const/let switch smoother
- ESLint rules updated to match
How this was done:
- npm install -g jscodeshift
- git clone https://github.com/cpojer/js-codemod.git
- git clone git@github.com:TryGhost/Ghost.git shallow-ghost
- cd shallow-ghost
- jscodeshift -t ../js-codemod/transforms/unchain-variables.js . -v=2
- jscodeshift -t ../js-codemod/transforms/no-vars.js . -v=2
- yarn
- yarn test
- yarn lint / fix various lint errors (almost all indent) by opening files and saving in vscode
- grunt test-regression
- sorted!
- All var declarations are now const or let as per ES6
- All comma-separated lists / chained declarations are now one declaration per line
- This is for clarity/readability but also made running the var-to-const/let switch smoother
- ESLint rules updated to match
How this was done:
- npm install -g jscodeshift
- git clone https://github.com/cpojer/js-codemod.git
- git clone git@github.com:TryGhost/Ghost.git shallow-ghost
- cd shallow-ghost
- jscodeshift -t ../js-codemod/transforms/unchain-variables.js . -v=2
- jscodeshift -t ../js-codemod/transforms/no-vars.js . -v=2
- yarn
- yarn test
- yarn lint / fix various lint errors (almost all indent) by opening files and saving in vscode
- grunt test-regression
- sorted!
- All var declarations are now const or let as per ES6
- All comma-separated lists / chained declarations are now one declaration per line
- This is for clarity/readability but also made running the var-to-const/let switch smoother
- ESLint rules updated to match
How this was done:
- npm install -g jscodeshift
- git clone https://github.com/cpojer/js-codemod.git
- git clone git@github.com:TryGhost/Ghost.git shallow-ghost
- cd shallow-ghost
- jscodeshift -t ../js-codemod/transforms/unchain-variables.js . -v=2
- jscodeshift -t ../js-codemod/transforms/no-vars.js . -v=2
- yarn
- yarn test
- yarn lint / fix various lint errors (almost all indent) by opening files and saving in vscode
- grunt test-regression
- sorted!
- Creating a better, more long-term API here
- compress and extract are opposite, neat terms
- Use new Promise to get rid of callback argument when using archiver to compress a folder
- Add options argument, and make a couple of key details configurable
- Make the response intelligable
- Ensure both functions are consistent
- Updated tests to match
- switch back from @tryghost/extract-zip to extract-zip now that it has been fixed (and is much better maintained)
- switch the internal interface to be fully promise-based and test promise-based too
- move all test files from core/test to test/
- updated all imports and other references
- all code inside of core/ is then application code
- tests are correctly at the root level
- consistent with other repos/projects
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
- move all test files from core/test to test/
- updated all imports and other references
- all code inside of core/ is then application code
- tests are correctly at the root level
- consistent with other repos/projects
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
- move all test files from core/test to test/
- updated all imports and other references
- all code inside of core/ is then application code
- tests are correctly at the root level
- consistent with other repos/projects
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
- mock non existant module util was defined twice
- split it out properly from the rest of the utils, update all references
- this allows us to move this util out of the codebase along with other code, e.g. the image manipulation code
no issue
- Updated Test & linting packages
- Updated use of hasOwnProperty
- Using Object.prototype.hasOwnProperty instead (ref. eslint.org/docs/rules/no-prototype-builtins)
- Removed already defined built-in global variable Intl
- Applied `--fix` with lint command on `core/test` folder
- The rules were broken because some of them were made stricter for `eslint: recommended` ruleset (ref. https://eslint.org/docs/user-guide/migrating-to-6.0.0#eslint-recommended-changes)
- Removed redundant global variable declarations to pass linting
refs #9389
- https://github.com/sinonjs/sinon/blob/master/CHANGELOG.md
Breaking changes for Ghost:
- no need to create a sandbox anymore, each file get's it's own sandbox
- just require sinon and use this sandbox
- you can still create separate sandboxes with .createSandbox
- reset single stubs: use .resetHistory instead of .reset
This is a global replace for any sandbox creation.
---
From https://sinonjs.org/releases/v7.2.3/sandbox/
> Default sandbox
> Since sinon@5.0.0, the sinon object is a default sandbox. Unless you have a very advanced setup or need a special configuration, you probably want to just use that one.
closes#10301
* Redirected to original image for gifs & svgs
* Created canTransformFileExtension method
* Updated image middlewares to use canTransformFileExtension
no-issue
* Added InternalServerError to resizeImage
* Added a redirect to original image if sharp is missing
* Improved naming - safeMethod -> method
* Updated process method to follow same sharp check pattern
* Refactor safety wrapper into makeSafe function
* Moved generic manipulation error to makeSafe function
* Refactored unsafeProcess to use unsafeResizeImage
* Removed CRAZY catch
refs #10181
* Added initial handleImageSizes middleware
* Implemented saveRaw method on local file storage
* Wired up handleImageSizes middleware
* Implemented delete for LocalFileStorage
* Removed delete method from theme Storage class
* Deleted sizes directory when theme is activated
* Ensured that smaller images are not enlarged
* Renamed sizes -> size
* Exited middleware as early as possible
* Called getStorage as late as possible
* Updated image sizes middleware to handle dimension paths
* Revert "Deleted sizes directory when theme is activated"
This reverts commit 9204dfcc73a6a79d597dbf23651817bcbfc59991.
* Revert "Removed delete method from theme Storage class"
This reverts commit b45fdb405a05faeaf4bd87e977c4ac64ff96b057.
* Revert "Implemented delete for LocalFileStorage"
This reverts commit a587cd6bae45b68a293b2d5cfd9b7705a29e7bfa.
* Fixed typo
Co-Authored-By: allouis <fabien@allou.is>
* Redirected to original image if no image_sizes config
* Refactored redirection because rule of three
* Updated comments
* Added rubbish tests
* Added @TODO comment for handleImageSizes tests
* Added safeResizeImage method to image manipulator
* Used image manipulator lib in image_size middleware
closes#10144
- When the input image is well optimized and has smaller byte size than the processed one it's still being used
- Bumped sharp version to have access to `size` property
refs #4453
* On by default
* Added config to disable resizing
* Added basic image optimization processing
* Added dep: sharp (optional dep)
* Added resize middleware
* Take care of rotation based on EXIF information
* Removed all meta data from optimised image
* Added handling if sharp could not get installed
* Do not read ext twice - optimisation
* Do not call sharp if config is disabled
* Do not remove the original image which was uploaded (store 2 images)
* Support of `req.files` for internal logic
* Disabled cache to enable file removal on Windows
no issue
- `{{#get}}` can slow down requests a lot if not used carefully, typically by using `limit="all"` or similar which can force a lot of data to be fetched and processed
- adds a warning log if we detect any `{{#get}}` helper call which takes longer than a certain threshold (default 200ms)
- allow log level and threshold to be configured via config to allow for different environments behaviours and requirements
New config options:
```
{
"logging": {
"slowHelper": {
"level": "warn",
"threshold": 200
}
}
}
```
Example output for `{{#get "tags" limit="all" order="name asc"}}` with a lot of tags:
```
[2019-06-07 10:35:52] WARN {{#get}} helper took 453ms to complete
{{#get}} helper took 453ms to complete
Error ID:
062daed0-8910-11e9-8185-3b615ad8677d
Error Code:
SLOW_GET_HELPER
Details:
api: v2.tagsPublic.browse
apiOptions:
order: name asc
limit: all
returnedRows: 1698
```
refs #9178
- first iteration of tidying up the unit tests
- this is useful in the current stage, because if i move files in the server folder, i need a clean folder/file structure to detect which tests needs to move
- this is a simple cleanup to reflect the current server folder structure
- Apps are marked as removed in 3.0, never officially launched and have been deprecated for at least 2 years.
- We've slowly removed bits that got in our way or were insecure over time meaning they mostly didn't work
- This cleans up the remainder of the logic
- The tables should be cleaned up in a future major
- Apps are marked as removed in 3.0, never officially launched and have been deprecated for at least 2 years.
- We've slowly removed bits that got in our way or were insecure over time meaning they mostly didn't work
- This cleans up the remainder of the logic
- The tables should be cleaned up in a future major
no issue
- extracted from https://github.com/expressjs/vhost/pull/21
- uses `req.hostname` if it's available. `req.hostname` is set to the `x-forwarded-host` value when `'trust proxy'` setting is enabled in express
no issue
Direct copy of current master from https://github.com/expressjs/vhost/ with automatic eslint fixes to match our styleguide.
Forked because we need a reverse proxy fix and the original module appears abandoned.
no issue
Ghost's relative->absolute handling is a little strange when the rootUrl includes a subdirectory. Root-relative paths such as /content/image.jpg are actually treated as subdirectory-relative. This means that it's possible to migrate from a root config to a subdirectory config without migrating data in the database, _however_ that means that the database will now have a mix of path styles (/content/image.png and /subdir/content/image.png). To handle this when all root-relative paths are treated as subdir-relative we have to rely on subdirectory deduplication.
- updates tests to reflect correct subdirectory handling according to the above rules
- fixes missing subdirectories when root urls contain subdirectories but relative paths do not
- fixes subdirectory deduplication when the supplied url/path does not have a trailing slash but matches the root url's subdirectory
no issue
- Removed v1 'author' leftover in include statement for preview controller
- Removed v1 'author' leftover in include statement for preview controller
- Removed v1 'author' leftover in include statement in entry lookup routing helper
- Migrated related test to use v2 API controller
- Removed v0.1 routing confif
- Removed v0.1 url config
- Fixed tests that had to do with url's in resources after removing v0.1 resources from URL cache
- Removed v1 'author' leftover in include statement in static routing helper
- Modified the test to use v2 API
- Removed v1 specific condition with 'page' in context helper
- Fixed dynamic routing spec after theme switch to v2. All tested users have to have at least one published post to be shown as an author
- Fixed URL Service spec to use theme engine v2
no issue
- extracted most standalone util functions into separate files+tests
- switched the main `UrlUtils` function into a class (we had mixed uses of the returned object previously as both `urlUtils()` and `new UrlUtils()` - the class makes it obvious that the `new` keyword is required)
- removed some unnecessary use of lodash and switched to using `URL` in place of `url` so that the package can have parity across node+browser without needing an additional package bundled in
- changelogPath can be a single path as before, or an array of paths with optional content which will be pre-pended
- Removed non-generic casper option from releases.create options
no issue
- Added option to allow appending casper changelog to main release draft
- Uses capser changelog path and new version info to append casper changes
The emoji *must* be at index=2 as each line looks like:
"* <commit message>"
We use index=2 to make test whether emoji is first character(s) in the
commit message
- Added the first version of release-utils
- Will release a v0.0.1
- Be able to create changelog.md based on previous version.
- Be able to merge & clean changelog.md.
- Be able to create a gist
- Be able to filter commits by emojis (user-facing changes)
- Be able to create a draft release
- Be able to upload a release zip to Github
The goal is to use our release-utils in multiple places (e.g. in Casper).
We will discuss if we want to replace some of our manual written utilities by existing npm packages. But to avoid breaking everything, we do it iteratively. Our release utils works and was designed based on our needs. We've tested some release utility in the past, they were all not really satisfying. TBC...
- this was originally created in the slimer repo, as that's where it was first used
- it's also used in migrate, should be used in gscan, and hopefully eventually, will also be part of Ghost-CLI
- therefore, it belongs somewhere more general
