housekeeping: Update dependency esbuild to ^0.18.0 #2732
Merged
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Outdated suggestions cannot be applied.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.
This PR contains the following updates:
^0.17.0
->^0.18.0
Release Notes
evanw/esbuild (esbuild)
v0.18.11
Compare Source
Fix a TypeScript code generation edge case (#3199)
This release fixes a regression in version 0.18.4 where using a TypeScript
namespace
that exports aclass
declaration combined with--keep-names
and a--target
ofes2021
or earlier could cause esbuild to export the class from the namespace using an incorrect name (notice the assignment toX2._Y
vs.X2.Y
):v0.18.10
Compare Source
Fix a tree-shaking bug that removed side effects (#3195)
This fixes a regression in version 0.18.4 where combining
--minify-syntax
with--keep-names
could cause expressions with side effects after a function declaration to be considered side-effect free for tree shaking purposes. The reason was because--keep-names
generates an expression statement containing a call to a helper function after the function declaration with a special flag that makes the function call able to be tree shaken, and then--minify-syntax
could potentially merge that expression statement with following expressions without clearing the flag. This release fixes the bug by clearing the flag when merging expression statements together.Fix an incorrect warning about CSS nesting (#3197)
A warning is currently generated when transforming nested CSS to a browser that doesn't support
:is()
because transformed nested CSS may need to use that feature to represent nesting. This was previously always triggered when an at-rule was encountered in a declaration context. Typically the only case you would encounter this is when using CSS nesting within a selector rule. However, there is a case where that's not true: when using a margin at-rule such as@top-left
within@page
. This release avoids incorrectly generating a warning in this case by checking that the at-rule is within a selector rule before generating a warning.v0.18.9
Compare Source
Fix
await using
declarations insideasync
generator functionsI forgot about the new
await using
declarations when implementing lowering forasync
generator functions in the previous release. This change fixes the transformation ofawait using
declarations when they are inside loweredasync
generator functions:Insert some prefixed CSS properties when appropriate (#3122)
With this release, esbuild will now insert prefixed CSS properties in certain cases when the
target
setting includes browsers that require a certain prefix. This is currently done for the following properties:appearance: *;
=>-webkit-appearance: *; -moz-appearance: *;
backdrop-filter: *;
=>-webkit-backdrop-filter: *;
background-clip: text
=>-webkit-background-clip: text;
box-decoration-break: *;
=>-webkit-box-decoration-break: *;
clip-path: *;
=>-webkit-clip-path: *;
font-kerning: *;
=>-webkit-font-kerning: *;
hyphens: *;
=>-webkit-hyphens: *;
initial-letter: *;
=>-webkit-initial-letter: *;
mask-image: *;
=>-webkit-mask-image: *;
mask-origin: *;
=>-webkit-mask-origin: *;
mask-position: *;
=>-webkit-mask-position: *;
mask-repeat: *;
=>-webkit-mask-repeat: *;
mask-size: *;
=>-webkit-mask-size: *;
position: sticky;
=>position: -webkit-sticky;
print-color-adjust: *;
=>-webkit-print-color-adjust: *;
tab-size: *;
=>-moz-tab-size: *; -o-tab-size: *;
text-decoration-color: *;
=>-webkit-text-decoration-color: *; -moz-text-decoration-color: *;
text-decoration-line: *;
=>-webkit-text-decoration-line: *; -moz-text-decoration-line: *;
text-decoration-skip: *;
=>-webkit-text-decoration-skip: *;
text-emphasis-color: *;
=>-webkit-text-emphasis-color: *;
text-emphasis-position: *;
=>-webkit-text-emphasis-position: *;
text-emphasis-style: *;
=>-webkit-text-emphasis-style: *;
text-orientation: *;
=>-webkit-text-orientation: *;
text-size-adjust: *;
=>-webkit-text-size-adjust: *; -ms-text-size-adjust: *;
user-select: *;
=>-webkit-user-select: *; -moz-user-select: *; -ms-user-select: *;
Here is an example:
Browser compatibility data was sourced from the tables on https://caniuse.com. Support for more CSS properties can be added in the future as appropriate.
Fix an obscure identifier minification bug (#2809)
Function declarations in nested scopes behave differently depending on whether or not
"use strict"
is present. To avoid generating code that behaves differently depending on whether strict mode is enabled or not, esbuild transforms nested function declarations into variable declarations. However, there was a bug where the generated variable name was not being recorded as declared internally, which meant that it wasn't being renamed correctly by the minifier and could cause a name collision. This bug has been fixed:Fix a bug in esbuild's compatibility table script (#3179)
Setting esbuild's
target
to a specific JavaScript engine tells esbuild to use the JavaScript syntax feature compatibility data from https://kangax.github.io/compat-table/es6/ for that engine to determine which syntax features to allow. However, esbuild's script that builds this internal compatibility table had a bug that incorrectly ignores tests for engines that still have outstanding implementation bugs which were never fixed. This change fixes this bug with the script.The only case where this changed the information in esbuild's internal compatibility table is that the
hermes
target is marked as no longer supporting destructuring. This is because there is a failing destructuring-related test for Hermes on https://kangax.github.io/compat-table/es6/. If you want to use destructuring with Hermes anyway, you can pass--supported:destructuring=true
to esbuild to override thehermes
target and force esbuild to accept this syntax.This fix was contributed by @ArrayZoneYour.
v0.18.8
Compare Source
Implement transforming
async
generator functions (#2780)With this release, esbuild will now transform
async
generator functions into normal generator functions when the configured target environment doesn't support them. These functions behave similar to normal generator functions except that they use theSymbol.asyncIterator
interface instead of theSymbol.iterator
interface and the iteration methods return promises. Here's an example (helper functions are omitted):This is an older feature that was added to JavaScript in ES2018 but I didn't implement the transformation then because it's a rarely-used feature. Note that esbuild already added support for transforming
for await
loops (the other part of the asynchronous iteration proposal) a year ago, so support for asynchronous iteration should now be complete.I have never used this feature myself and code that uses this feature is hard to come by, so this transformation has not yet been tested on real-world code. If you do write code that uses this feature, please let me know if esbuild's
async
generator transformation doesn't work with your code.v0.18.7
Compare Source
Add support for
using
declarations in TypeScript 5.2+ (#3191)TypeScript 5.2 (due to be released in August of 2023) will introduce
using
declarations, which will allow you to automatically dispose of the declared resources when leaving the current scope. You can read the TypeScript PR for this feature for more information. This release of esbuild adds support for transforming this syntax to target environments without support forusing
declarations (which is currently all targets other thanesnext
). Here's an example (helper functions are omitted):The injected helper functions ensure that the method named
Symbol.dispose
is called onnew Foo
when control exits the scope. Note that as with all new JavaScript APIs, you'll need to polyfillSymbol.dispose
if it's not present before you use it. This is not something that esbuild does for you because esbuild only handles syntax, not APIs. Polyfilling it can be done with something like this:This feature also introduces
await using
declarations which are likeusing
declarations but they callawait
on the disposal method (not on the initializer). Here's an example (helper functions are omitted):The injected helper functions ensure that the method named
Symbol.asyncDispose
is called onnew Foo
when control exits the scope, and that the returned promise is awaited. Similarly toSymbol.dispose
, you'll also need to polyfillSymbol.asyncDispose
before you use it.Add a
--line-limit=
flag to limit line length (#3170)Long lines are common in minified code. However, many tools and text editors can't handle long lines. This release introduces the
--line-limit=
flag to tell esbuild to wrap lines longer than the provided number of bytes. For example,--line-limit=80
tells esbuild to insert a newline soon after a given line reaches 80 bytes in length. This setting applies to both JavaScript and CSS, and works even when minification is disabled. Note that turning this setting on will make your files bigger, as the extra newlines take up additional space in the file (even after gzip compression).v0.18.6
Compare Source
Fix tree-shaking of classes with decorators (#3164)
This release fixes a bug where esbuild incorrectly allowed tree-shaking on classes with decorators. Each decorator is a function call, so classes with decorators must never be tree-shaken. This bug was a regression that was unintentionally introduced in version 0.18.2 by the change that enabled tree-shaking of lowered private fields. Previously decorators were always lowered, and esbuild always considered the automatically-generated decorator code to be a side effect. But this is no longer the case now that esbuild analyzes side effects using the AST before lowering takes place. This bug was fixed by considering any decorator a side effect.
Fix a minification bug involving function expressions (#3125)
When minification is enabled, esbuild does limited inlining of
const
symbols at the top of a scope. This release fixes a bug where inlineable symbols were incorrectly removed assuming that they were inlined. They may not be inlined in cases where they were referenced by earlier constants in the body of a function expression. The declarations involved in these edge cases are now kept instead of being removed:v0.18.5
Compare Source
Implement auto accessors (#3009)
This release implements the new auto-accessor syntax from the upcoming JavaScript decorators proposal. The auto-accessor syntax looks like this:
This syntax is not yet a part of JavaScript but it was added to TypeScript in version 4.9. More information about this feature can be found in microsoft/TypeScript#49705. Auto-accessors will be transformed if the target is set to something other than
esnext
:You can also now use auto-accessors with esbuild's TypeScript experimental decorator transformation, which should behave the same as decorating the underlying getter/setter pair.
Please keep in mind that this syntax is not yet part of JavaScript. This release enables auto-accessors in
.js
files with the expectation that it will be a part of JavaScript soon. However, esbuild may change or remove this feature in the future if JavaScript ends up changing or removing this feature. Use this feature with caution for now.Pass through JavaScript decorators (#104)
In this release, esbuild now parses decorators from the upcoming JavaScript decorators proposal and passes them through to the output unmodified (as long as the language target is set to
esnext
). Transforming JavaScript decorators to environments that don't support them has not been implemented yet. The only decorator transform that esbuild currently implements is still the TypeScript experimental decorator transform, which only works in.ts
files and which requires"experimentalDecorators": true
in yourtsconfig.json
file.Static fields with assign semantics now use static blocks if possible
Setting
useDefineForClassFields
to false in TypeScript requires rewriting class fields to assignment statements. Previously this was done by removing the field from the class body and adding an assignment statement after the class declaration. However, this also caused any private fields to also be lowered by necessity (in case a field initializer uses a private symbol, either directly or indirectly). This release changes this transform to use an inline static block if it's supported, which avoids needing to lower private fields in this scenario:Fix TypeScript experimental decorators combined with
--mangle-props
(#3177)Previously using TypeScript experimental decorators combined with the
--mangle-props
setting could result in a crash, as the experimental decorator transform was not expecting a mangled property as a class member. This release fixes the crash so you can now combine both of these features together safely.v0.18.4
Compare Source
Bundling no longer unnecessarily transforms class syntax (#1360, #1328, #1524, #2416)
When bundling, esbuild automatically converts top-level class statements to class expressions. Previously this conversion had the unfortunate side-effect of also transforming certain other class-related syntax features to avoid correctness issues when the references to the class name within the class body. This conversion has been reworked to avoid doing this:
This conversion process is very complicated and has many edge cases (including interactions with static fields, static blocks, private class properties, and TypeScript experimental decorators). It should already be pretty robust but a change like this may introduce new unintentional behavior. Please report any issues with this upgrade on the esbuild bug tracker.
You may be wondering why esbuild needs to do this at all. One reason to do this is that esbuild's bundler sometimes needs to lazily-evaluate a module. For example, a module may end up being both the target of a dynamic
import()
call and a staticimport
statement. Lazy module evaluation is done by wrapping the top-level module code in a closure. To avoid a performance hit for staticimport
statements, esbuild stores top-level exported symbols outside of the closure and references them directly instead of indirectly.Another reason to do this is that multiple JavaScript VMs have had and continue to have performance issues with TDZ (i.e. "temporal dead zone") checks. These checks validate that a let, or const, or class symbol isn't used before it's initialized. Here are two issues with well-known VMs:
JavaScriptCore had a severe performance issue as their TDZ implementation had time complexity that was quadratic in the number of variables needing TDZ checks in the same scope (with the top-level scope typically being the worst offender). V8 has ongoing issues with TDZ checks being present throughout the code their JIT generates even when they have already been checked earlier in the same function or when the function in question has already been run (so the checks have already happened).
Due to esbuild's parallel architecture, esbuild both a) needs to convert class statements into class expressions during parsing and b) doesn't yet know whether this module will need to be lazily-evaluated or not in the parser. So esbuild always does this conversion during bundling in case it's needed for correctness (and also to avoid potentially catastrophic performance issues due to bundling creating a large scope with many TDZ variables).
Enforce TDZ errors in computed class property keys (#2045)
JavaScript allows class property keys to be generated at run-time using code, like this:
Previously esbuild treated references to the containing class name within computed property keys as a reference to the partially-initialized class object. That meant code that attempted to reference properties of the class object (such as the code above) would get back
undefined
instead of throwing an error.This release rewrites references to the containing class name within computed property keys into code that always throws an error at run-time, which is how this JavaScript code is supposed to work. Code that does this will now also generate a warning. You should never write code like this, but it now should be more obvious when incorrect code like this is written.
Fix an issue with experimental decorators and static fields (#2629)
This release also fixes a bug regarding TypeScript experimental decorators and static class fields which reference the enclosing class name in their initializer. This affected top-level classes when bundling was enabled. Previously code that does this could crash because the class name wasn't initialized yet. This case should now be handled correctly:
Fix a minification regression with negative numeric properties (#3169)
Version 0.18.0 introduced a regression where computed properties with negative numbers were incorrectly shortened into a non-computed property when minification was enabled. This regression has been fixed:
v0.18.3
Compare Source
Fix a panic due to empty static class blocks (#3161)
This release fixes a bug where an internal invariant that was introduced in the previous release was sometimes violated, which then caused a panic. It happened when bundling code containing an empty static class block with both minification and bundling enabled.
v0.18.2
Compare Source
Lower static blocks when static fields are lowered (#2800, #2950, #3025)
This release fixes a bug where esbuild incorrectly did not lower static class blocks when static class fields needed to be lowered. For example, the following code should print
1 2 3
but previously printed2 1 3
instead due to this bug:Use static blocks to implement
--keep-names
on classes (#2389)This change fixes a bug where the
name
property could previously be incorrect within a class static context when using--keep-names
. The problem was that thename
property was being initialized after static blocks were run instead of before. This has been fixed by moving thename
property initializer into a static block at the top of the class body:This change was somewhat involved, especially regarding what esbuild considers to be side-effect free. Some unused classes that weren't removed by tree shaking in previous versions of esbuild may now be tree-shaken. One example is classes with static private fields that are transformed by esbuild into code that doesn't use JavaScript's private field syntax. Previously esbuild's tree shaking analysis ran on the class after syntax lowering, but with this release it will run on the class before syntax lowering, meaning it should no longer be confused by class mutations resulting from automatically-generated syntax lowering code.
v0.18.1
Compare Source
Add the
--drop-labels=
option (#2398)If you want to conditionally disable some development-only code and have it not be present in the final production bundle, right now the most straightforward way of doing this is to use the
--define:
flag along with a specially-named global variable. For example, consider the following code:You can build this for development and production like this:
esbuild --define:DEV=true
esbuild --define:DEV=false
One drawback of this approach is that the resulting code crashes if you don't provide a value for
DEV
with--define:
. In practice this isn't that big of a problem, and there are also various ways to work around this.However, another approach that avoids this drawback is to use JavaScript label statements instead. That's what the
--drop-labels=
flag implements. For example, consider the following code:With this release, you can now build this for development and production like this:
esbuild
esbuild --drop-labels=DEV
This means that code containing optional development-only checks can now be written such that it's safe to run without any additional configuration. The
--drop-labels=
flag takes comma-separated list of multiple label names to drop.Avoid causing
unhandledRejection
during shutdown (#3219)All pending esbuild JavaScript API calls are supposed to fail if esbuild's underlying child process is unexpectedly terminated. This can happen if
SIGINT
is sent to the parentnode
process with Ctrl+C, for example. Previously doing this could also cause an unhandled promise rejection when esbuild attempted to communicate this failure to its own child process that no longer exists. This release now swallows this communication failure, which should prevent this internal unhandled promise rejection. This change means that you can now use esbuild's JavaScript API with a customSIGINT
handler that extends the lifetime of thenode
process without esbuild's internals causing an early exit due to an unhandled promise rejection.Update browser compatibility table scripts
The scripts that esbuild uses to compile its internal browser compatibility table have been overhauled. Briefly:
caniuse-lite
and@mdn/browser-compat-data
as new data sources (replacing manually-copied information)This change means it's now much easier to keep esbuild's internal compatibility tables up to date. You can review the table changes here if you need to debug something about this change:
v0.18.0
Compare Source
This release deliberately contains backwards-incompatible changes. To avoid automatically picking up releases like this, you should either be pinning the exact version of
esbuild
in yourpackage.json
file (recommended) or be using a version range syntax that only accepts patch upgrades such as^0.17.0
or~0.17.0
. See npm's documentation about semver for more information.The breaking changes in this release mainly focus on fixing some long-standing issues with esbuild's handling of
tsconfig.json
files. Here are all the changes in this release, in detail:Add a way to try esbuild online (#797)
There is now a way to try esbuild live on esbuild's website without installing it: https://esbuild.github.io/try/. In addition to being able to more easily evaluate esbuild, this should also make it more efficient to generate esbuild bug reports. For example, you can use it to compare the behavior of different versions of esbuild on the same input. The state of the page is stored in the URL for easy sharing. Many thanks to @hyrious for creating https://hyrious.me/esbuild-repl/, which was the main inspiration for this addition to esbuild's website.
Two forms of build options are supported: either CLI-style (example) or JS-style (example). Both are converted into a JS object that's passed to esbuild's WebAssembly API. The CLI-style argument parser is a custom one that simulates shell quoting rules, and the JS-style argument parser is also custom and parses a superset of JSON (basically JSON5 + regular expressions). So argument parsing is an approximate simulation of what happens for real but hopefully it should be close enough.
Changes to esbuild's
tsconfig.json
support (#3019):This release makes the following changes to esbuild's
tsconfig.json
support:Using experimental decorators now requires
"experimentalDecorators": true
(#104)Previously esbuild would always compile decorators in TypeScript code using TypeScript's experimental decorator transform. Now that standard JavaScript decorators are close to being finalized, esbuild will now require you to use
"experimentalDecorators": true
to do this. This new requirement makes it possible for esbuild to introduce a transform for standard JavaScript decorators in TypeScript code in the future. Such a transform has not been implemented yet, however.TypeScript's
target
no longer affects esbuild'starget
(#2628)Some people requested that esbuild support TypeScript's
target
setting, so support for it was added (in version 0.12.4). However, esbuild supports reading from multipletsconfig.json
files within a single build, which opens up the possibility that different files in the build have different language targets configured. There isn't really any reason to do this and it can lead to unexpected results. So with this release, thetarget
setting intsconfig.json
will no longer affect esbuild's owntarget
setting. You will have to use esbuild's own target setting instead (which is a single, global value).TypeScript's
jsx
setting no longer causes esbuild to preserve JSX syntax (#2634)TypeScript has a setting called
jsx
that controls how to transform JSX into JS. The tool-agnostic transform is calledreact
, and the React-specific transform is calledreact-jsx
(orreact-jsxdev
). There is also a setting calledpreserve
which indicates JSX should be passed through untransformed. Previously people would run esbuild with"jsx": "preserve"
in theirtsconfig.json
files and then be surprised when esbuild preserved their JSX. So with this release, esbuild will now ignore"jsx": "preserve"
intsconfig.json
files. If you want to preserve JSX syntax with esbuild, you now have to use--jsx=preserve
.Note: Some people have suggested that esbuild's equivalent
jsx
setting override the one intsconfig.json
. However, some projects need to legitimately have different files within the same build use different transforms (i.e.react
vs.react-jsx
) and having esbuild's globaljsx
setting overridetsconfig.json
would prevent this from working. This release ignores"jsx": "preserve"
but still allows otherjsx
values intsconfig.json
files to override esbuild's globaljsx
setting to keep the ability for multiple files within the same build to use different transforms.useDefineForClassFields
behavior has changed (#2584, #2993)Class fields in TypeScript look like this (
x
is a class field):TypeScript has legacy behavior that uses assignment semantics instead of define semantics for class fields when
useDefineForClassFields
is enabled (in which case class fields in TypeScript behave differently than they do in JavaScript, which is arguably "wrong").This legacy behavior exists because TypeScript added class fields to TypeScript before they were added to JavaScript. The TypeScript team decided to go with assignment semantics and shipped their implementation. Much later on TC39 decided to go with define semantics for class fields in JavaScript instead. This behaves differently if the base class has a setter with the same name:
When you run
tsc
, the value ofuseDefineForClassFields
defaults tofalse
when it's not specified and thetarget
intsconfig.json
is present but earlier thanES2022
. This sort of makes sense because the class field language feature was added in ES2022, so before ES2022 class fields didn't exist (and thus TypeScript's legacy behavior is active). However, TypeScript'starget
setting currently defaults toES3
which unfortunately means that theuseDefineForClassFields
setting currently defaults to false (i.e. to "wrong"). In other words if you runtsc
with all default settings, class fields will behave incorrectly.Previously esbuild tried to do what
tsc
did. That meant esbuild's version ofuseDefineForClassFields
wasfalse
by default, and was alsofalse
if esbuild's--target=
was present but earlier thanes2022
. However, TypeScript's legacy class field behavior is becoming increasingly irrelevant and people who expect class fields in TypeScript to work like they do in JavaScript are confused when they use esbuild with default settings. It's also confusing that the behavior of class fields would change if you changed the language target (even though that's exactly how TypeScript works).So with this release, esbuild will now only use the information in
tsconfig.json
to determine whetheruseDefineForClassFields
is true or not. SpecificallyuseDefineForClassFields
will be respected if present, otherwise it will befalse
iftarget
is present intsconfig.json
and isES2021
or earlier, otherwise it will betrue
. Targets passed to esbuild's--target=
setting will no longer affectuseDefineForClassFields
.Note that this means different directories in your build can have different values for this setting since esbuild allows different directories to have different
tsconfig.json
files within the same build. This should let you migrate your code one directory at a time without esbuild's--target=
setting affecting the semantics of your code.Add support for
verbatimModuleSyntax
from TypeScript 5.0TypeScript 5.0 added a new option called
verbatimModuleSyntax
that deprecates and replaces two older options,preserveValueImports
andimportsNotUsedAsValues
. SettingverbatimModuleSyntax
to true intsconfig.json
tells esbuild to not drop unused import statements. Specifically esbuild now treats"verbatimModuleSyntax": true
as if you had specified both"preserveValueImports": true
and"importsNotUsedAsValues": "preserve"
.Add multiple inheritance for
tsconfig.json
from TypeScript 5.0TypeScript 5.0 now allows multiple inheritance for
tsconfig.json
files. You can now pass an array of filenames via theextends
parameter and yourtsconfig.json
will start off containing properties from all of those configuration files, in order. This release of esbuild adds support for this new TypeScript feature.Remove support for
moduleSuffixes
(#2395)The community has requested that esbuild remove support for TypeScript's
moduleSuffixes
feature, so it has been removed in this release. Instead you can use esbuild's--resolve-extensions=
feature to select which module suffix you want to build with.Apply
--tsconfig=
overrides tostdin
and virtual files (#385, #2543)When you override esbuild's automatic
tsconfig.json
file detection with--tsconfig=
to pass a specifictsconfig.json
file, esbuild previously didn't apply these settings to source code passed via thestdin
API option or to TypeScript files from plugins that weren't in thefile
namespace. This release changes esbuild's behavior so that settings fromtsconfig.json
also apply to these source code files as well.Support
--tsconfig-raw=
in build API calls (#943, #2440)Previously if you wanted to override esbuild's automatic
tsconfig.json
file detection, you had to create a newtsconfig.json
file and pass the file name to esbuild via the--tsconfig=
flag. With this release, you can now optionally use--tsconfig-raw=
instead to pass the contents oftsconfig.json
to esbuild directly instead of passing the file name. For example, you can now use--tsconfig-raw={"compilerOptions":{"experimentalDecorators":true}}
to enable TypeScript experimental decorators directly using a command-line flag (assuming you escape the quotes correctly using your current shell's quoting rules). The--tsconfig-raw=
flag previously only worked with transform API calls but with this release, it now works with build API calls too.Ignore all
tsconfig.json
files innode_modules
(#276, #2386)This changes esbuild's behavior that applies
tsconfig.json
to all files in the subtree of the directory containingtsconfig.json
. In version 0.12.7, esbuild started ignoringtsconfig.json
files insidenode_modules
folders. The rationale is that people typically do this by mistake and that doing this intentionally is a rare use case that doesn't need to be supported. However, this change only applied to certain syntax-specific settings (e.g.jsxFactory
) but did not apply to path resolution settings (e.g.paths
). With this release, esbuild will now ignore alltsconfig.json
files innode_modules
instead of only ignoring certain settings.Ignore
tsconfig.json
when resolving paths withinnode_modules
(#2481)Previously fields in
tsconfig.json
related to path resolution (e.g.paths
) were respected for all files in the subtree containing thattsconfig.json
file, even within a nestednode_modules
subdirectory. This meant that a project'spaths
settings could potentially affect any bundled packages. With this release, esbuild will no longer usetsconfig.json
settings during path resolution inside nestednode_modules
subdirectories.Prefer
.js
over.ts
withinnode_modules
(#3019)The default list of implicit extensions that esbuild will try appending to import paths contains
.ts
before.js
. This makes it possible to bundle TypeScript projects that reference other files in the project using extension-less imports (e.g../some-file
to load./some-file.ts
instead of./some-file.js
). However, this behavior is undesirable withinnode_modules
directories. Some package authors publish both their original TypeScript code and their compiled JavaScript code side-by-side. In these cases, esbuild should arguably be using the compiled JavaScript files instead of the original TypeScript files because the TypeScript compilation settings for files within the package should be determined by the package author, not the user of esbuild. So with this release, esbuild will now prefer implicit.js
extensions over.ts
when searching for import paths withinnode_modules
.These changes are intended to improve esbuild's compatibility with
tsc
and reduce the number of unfortunate behaviors regardingtsconfig.json
and esbuild.Add a workaround for bugs in Safari 16.2 and earlier (#3072)
Safari's JavaScript parser had a bug (which has now been fixed) where at least something about unary/binary operators nested inside default arguments nested inside either a function or class expression was incorrectly considered a syntax error if that expression was the target of a property assignment. Here are some examples that trigger this Safari bug:
It's not clear what the exact conditions are that trigger this bug. However, a workaround for this bug appears to be to post-process your JavaScript to wrap any in function and class declarations that are the direct target of a property access expression in parentheses. That's the workaround that UglifyJS applies for this issue: mishoo/UglifyJS#2056. So that's what esbuild now does starting with this release:
This fix is not enabled by default. It's only enabled when
--target=
contains Safari 16.2 or earlier, such as with--target=safari16.2
. You can also explicitly enable or disable this specific transform (calledfunction-or-class-property-access
) with--supported:function-or-class-property-access=false
.Fix esbuild's TypeScript type declarations to forbid unknown properties (#3089)
Version 0.17.0 of esbuild introduced a specific form of function overloads in the TypeScript type definitions for esbuild's API calls that looks like this:
This more accurately reflects how esbuild's JavaScript API behaves. The result object returned by
transformSync
only has thelegalComments
property if you passlegalComments: 'external'
:However, this form of function overloads unfortunately allows typos (e.g.
egalComments
) to pass the type checker without generating an error as TypeScript allows all objects with unknown properties to extendTransformOptions
. These typos result in esbuild's API throwing an error at run-time.To prevent typos during type checking, esbuild's TypeScript type definitions will now use a different form that looks like this:
This change should hopefully not affect correct code. It should hopefully introduce type errors only for incorrect code.
Fix CSS nesting transform for pseudo-elements (#3119)
This release fixes esbuild's CSS nesting transform for pseudo-elements (e.g.
::before
and::after
). The CSS nesting specification says that the nesting selector does not work with pseudo-elements. This can be seen in the example below: esbuild does not carry the parent pseudo-element::before
through the nesting selector&
. However, that doesn't apply to pseudo-elements that are within the same selector. Previously esbuild had a bug where it considered pseudo-elements in both locations as invalid. This release changes esbuild to only consider those from the parent selector invalid, which should align with the specification:Forbid
&
before a type selector in nested CSSThe people behind the work-in-progress CSS nesting specification have very recently decided to forbid nested CSS that looks like
&div
. You will have to use eitherdiv&
or&:is(div)
instead. This release of esbuild has been updated to take this new change into consideration. Doing this now generates a warning. The suggested fix is slightly different depending on where in the overall selector it happened:v0.17.19
Compare Source
Fix CSS transform bugs with nested selectors that start with a combinator (#3096)
This release fixes several bugs regarding transforming nested CSS into non-nested CSS for older browsers. The bugs were due to lack of test coverage for nested selectors with more than one compound selector where they all start with the same combinator. Here's what some problematic cases look like before and after these fixes:
Configuration
📅 Schedule: Branch creation - "every 2 week on monday" (UTC), Automerge - At any time (no schedule defined).
🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.
♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.
🔕 Ignore: Close this PR and you won't be reminded about this update again.
This PR has been generated by Mend Renovate. View repository job log here.