Now that ngtsc performs type checking using a dedicated `__ng_typecheck__.ts`
file, `NgtscProgram` always wraps its `ts.CompilerHost` in a shim host. This
shim fails to delegate `resolveModuleNames` so no custom module resolution
logic is considered. This introduces a problem for the CLI, as the compiler
host it passes kicks of ngcc for any imported module such that Ivy's
compatibility compiler runs automatically behind the scenes.
This commit adds delegation of the `resolveModuleNames` to fix the issue.
Fixes#30064
PR Close#30068
Previously, an instance of HttpParams would retain its list of mutations
after they have been materialized as a result of a read operation. Not
only does this unnecessarily hold onto memory, more importantly does it
introduce a bug where branching of off a materialized instance would
reconsider the set of mutations that had already been applied, resulting
in repeated application of mutations.
This commit fixes the bug by clearing the list of pending mutations
after they have been materialized, such that they will not be considered
once again for branched off instances.
Fixes#20430
PR Close#29045
This minimises the risk of unexpected failures due to breaking changes,
when building a new image (e.g. as a result of an unrelated config
change in Dockerfile).
PR Close#29976
Previously, the preview server docker image was based on Debian 8
(jessie). Recently, `jessie-updates` and `jessie-backborts` were removed
from the Debian mirrors ([more info][1]), thus breaking new builds of
the image.
Instead of updating `/etc/apt/sources.list` to remove the obsolete
sources, this commit upgrades to Debian 9 (stretch).
(The GCE VM running the preview server docker container was also
upgraded from Debian 8 to 9 this morning.)
---
Other changes:
- Removed dependency on `chkconfig`, which is not supported on Debian 9.
- Installing `nginx` from the regular repositories (instead of
`*-backports).
- Upgraded to `pm2` v3, which can handle hooking itself up to system
startup better (without `chkconfig` - see above).
- Updated tests to reflect the fact that `nginx` has dropped the reason
phrase in response status lines for HTTP/2 (in compliance with
[the spec][2]). (HTTP/1.1: `HTTP/1.1 200 OK` | HTTP/2: `HTTP/2 200`)
[1]: https://www.lucas-nussbaum.net/blog/?p=947
[2]: https://http2.github.io/http2-spec/#rfc.section.8.1.2.4
PR Close#29976
In #29926, the size of the build artifacts has increased due to turning
on differential loading (which generates an es2015/es5 pair for each JS
resource).
To avoid the preview server's rejecting the build artifacts (as in
[288181][1]), this commit increases the max allowed artifact size from
20MB to 25MB (current artifact size after #29926 is ~22MB).
[1]: https://circleci.com/gh/angular/angular/288181
PR Close#29976
Currently when someone runs `ng update` with the static-query migration,
the migration can fail with an error saying that the `AOT` compiler could not
be created. This can happen if the CLI project contains a test `tsconfig.json`
that is picked up by the schematic.
Due to the fact that spec tsconfig files cannot be ran with NGC (e.g. test
components are not part of a module; not all source files are guaranteed to
be included), test `tsconfig` projects will now use a new `test` migration
strategy where all queries within tests are left untouched and a TODO is added.
PR Close#30034
This commit unifies the way auxillary RootScopeModule and DynamicTestModule are compiled in R3TestBed by calling `compileNgModuleDefs` explicitly for RootScopeModule. This change also resolves the problem where TestBed's code was used from the @angular/core NPM package: due to the "jit" flag, the @NgModule decorator on the RootScopeModule was transformed to RootScopeModule.decorators = [...], but actual ngModuleDef was never defined.
PR Close#30037
The compiler uses metadata to represent what it statically knows about
various expressions in a program. Occasionally, expressions in the program
for which metadata is extracted may contain sub-expressions which are not
representable in metadata. One such construct is an arrow function.
The compiler does not always need to understand such expressions completely.
For example, for a provider defined with `useValue`, the compiler does not
need to understand the value at all, only the outer provider definition. In
this case, the compiler employs a technique known as "expression lowering",
where it rewrites the provider expression into one that can be represented
in metadata. Chiefly, this involves extracting out the dynamic part (the
`useValue` expression) into an exported constant.
Lowering is applied through a heuristic, which considers the containing
statement as well as the field name of the expression.
Previously, this heuristic was not completely accurate in the case of
route definitions and the `loadChildren` field, which is lowered. If the
route definition using `loadChildren` existed inside a decorator invocation,
lowering was performed correctly. However, if it existed inside a standalone
variable declaration with an export keyword, the heuristic would conclude
that lowering was unnecessary. For ordinary providers this is true; however
the compiler attempts to fully understand the ROUTES token and thus even if
an array of routes is declared in an exported variable, any `loadChildren`
expressions within still need to be lowered.
This commit enables lowering of already exported variables under a limited
set of conditions (where the initializer expression is of a specific form).
This should enable the use of `loadChildren` in route definitions.
PR Close#30038
At the moment `ts_api_guardian_test` rule depends on the entire node_modules tree, this is causing an input file tree of over 71473 in the material repo which is failing RBE
With this change we add only the required dependencies to as data attribute instead of the `@npm//:node_modules`
With this change we should also reduce the `Too many open files in system` error that sometimes crops up
PR Close#30051
Previously, a template's context name would only be included in an embedded
template function if the element that the template was declared on has a
tag name. This is generally true for elements, except for `ng-content`
that does not have a tag name. By omitting the context name the compiler
could introduce duplicate template function names, which would fail at runtime.
This commit fixes the behavior by always including the context name in the
template function's name, regardless of tag name.
Resolves FW-1272
PR Close#30025
Previously, ngcc would insert new imports at the beginning of the file, for
convenience. This is problematic for imports that have side-effects, as the
side-effects imposed by such imports may affect the behavior of subsequent
imports.
This commit teaches ngcc to insert imports after any existing imports. Special
care has been taken to ensure inserted constants will still follow after the
inserted imports.
Resolves FW-1271
PR Close#30029
Fixes Ivy throwing an error because it tries to generate styling instructions for empty `style` and `class` bindings.
This PR resolves FW-1274.
PR Close#30024
Currently the `template-var-assignment` migration incorrectly warns if
the template writes to a property in the component that has the same
`ast.PropertyWrite´ name as a template input variable but different
receiver. e.g.
```html
<!-- "someProp.element" will be incorrectly reported as template variable assignment -->
<button *ngFor="let element of list" (click)="someProp.element = null">Reset</button>
```
Similarly if an output writes to a component property with the same name as a
template input variable, but the expression is within a different template scope,
the schematic currently incorrectly warns. e.g.
```html
<button *ngFor="let element of list">{{element}}</button>
<!-- The "element = null" expression does not refer to the "element" template input variable -->
<button (click)="element = null"></button>
```
PR Close#30026
Previously, we had a bug where directive matching could fail if the directive
attribute was bound and followed a certain number of classes. This is because
in the matching logic, we were treating classes like normal attributes. We
should instead be skipping classes in the attribute matching logic. Otherwise
classes will match for directives with attribute selectors, and as we are
iterating through them in twos (when they are stored as name-only, not in
name-value pairs), it may throw off directive matching for any bound attributes
that come after. This commit changes the directive matching logic to skip
classes altogether.
PR Close#30002
`ng build` by default builds the prodapp because there is not
a dev build in Bazel. This PR restores the `--prod` to do the
same to prevent achitect from showing missing config error.
PR Close#30005
We recently added the "uname -a" command to the CircleCI
integration test in order to identify Linux kernels that
broke Chrome sandboxing.
Since this issue seems to be claimed as fixed by the CircleCI
support and we didn't see any sandboxing issues the last month,
we are removing the debugging command we added.
PR Close#30006
Previously, all of a program's files would be copied into the __ivy_ngcc__
folder where ngcc then writes its modifications into. The set of source files
in a program however is much larger than the source files contained within
the entry-point of interest, so many more files were copied than necessary.
Even worse, it may occur that an unrelated file in the program would collide
with an already existing source file, resulting in incorrectly overwriting
a file with unrelated content. This behavior has actually been observed
with @angular/animations and @angular/platform-browser/animations, where
the former package would overwrite declaration files of the latter package.
This commit fixes the issue by only copying relevant source files when cloning
a bundle's content into __ivy_ngcc__.
Fixes#29960
PR Close#30020
Previously, during the evaluation of a function call where no argument
was provided for a parameter that has a default value, the default value
would be taken from the context of the caller, instead of the callee.
This commit fixes the behavior by resolving the default value of a
parameter in the context of the callee.
PR Close#29888
Previously, ngtsc's static evaluator did not take spread operators into
account when evaluating function calls, nor did it handle rest arguments
correctly. This commit adds support for static evaluation of these
language features.
PR Close#29888
We only set ng-reflect properties on directive input bindings.
This PR ensures that we also add ng-reflect properties on unbound inputs for backwards compatibility.
FW-1266 #resolve
PR Close#29973
Template type-checking is enabled by default in the View Engine compiler.
The feature in Ivy is not quite ready for this yet, so this flag will
temporarily control whether templates are type-checked in ngtsc.
The goal is to remove this flag after rolling out template type-checking in
google3 in Ivy mode, and making sure the feature is as compatible with the
View Engine implementation as possible.
Initially, the default value of the flag will leave checking disabled.
PR Close#29698
Previously, Template.templateAttrs was introduced to capture attribute
bindings which originated from microsyntax (e.g. bindings in *ngFor="...").
This means that a Template node can have two different structures, depending
on whether it originated from microsyntax or from a literal <ng-template>.
In the literal case, the node behaves much like an Element node, it has
attributes, inputs, and outputs which determine which directives apply.
In the microsyntax case, though, only the templateAttrs should be used
to determine which directives apply.
Previously, both the t2_binder and the TemplateDefinitionBuilder were using
the wrong set of attributes to match directives - combining the attributes,
inputs, outputs, and templateAttrs of the Template node regardless of its
origin. In the TDB's case this wasn't a problem, since the TDB collects a
global Set of directives used in the template, so it didn't matter whether
the directive was also recognized on the <ng-template>. t2_binder's API
distinguishes between directives on specific nodes, though, so it's more
sensitive to mismatching.
In particular, this showed up as an assertion failure in template type-
checking in certain cases, when a directive was accidentally matched on
a microsyntax template element and also had a binding which referenced a
variable declared in the microsyntax. This resulted in the type-checker
attempting to generate a reference to a variable that didn't exist in that
scope.
The fix is to distinguish between the two cases and select the appropriate
set of attributes to match on accordingly.
Testing strategy: tested in the t2_binder tests.
PR Close#29698
This commit adds support for template type-checking a pipe binding which
previously was not handled by the type-checking engine. In compatibility
mode, the arguments to transform() are not checked and the type returned
by a pipe is 'any'. In full type-checking mode, the transform() method's
type signature is used to check the pipe usage and infer the return type
of the pipe.
Testing strategy: TCB tests included.
PR Close#29698
The template type-checking engine previously would assemble a type-checking
program by inserting Type Check Blocks (TCBs) into existing user files. This
approach proved expensive, as TypeScript has to re-parse and re-type-check
those files when processing the type-checking program.
Instead, a far more performant approach is to augment the program with a
single type-checking file, into which all TCBs are generated. Additionally,
type constructors are also inlined into this file.
This is not always possible - both TCBs and type constructors can sometimes
require inlining into user code, particularly if bound generic type
parameters are present, so the approach taken is actually a hybrid. These
operations are inlined if necessary, but are otherwise generated in a single
file.
It is critically important that the original program also include an empty
version of the type-checking file, otherwise the shape of the two programs
will be different and TypeScript will throw away all the old program
information. This leads to a painfully slow type checking pass, on the same
order as the original program creation. A shim to generate this file in the
original program is therefore added.
Testing strategy: this commit is largely a refactor with no externally
observable behavioral differences, and thus no tests are needed.
PR Close#29698
This commit adds support in the template type-checking engine for handling
the logical not operation and the safe navigation operation.
Safe navigation in particular is tricky, as the View Engine implementation
has a rather inconvenient flaw. View Engine checks a safe navigation
operation `a?.b` as:
```typescript
(a != null ? a!.b : null as any)
```
The type of this expression is always 'any', as the false branch of the
ternary has type 'any'. Thus, using null-safe navigation throws away the
type of the result, and breaks type-checking for the rest of the expression.
A flag is introduced in the type-checking configuration to allow Ivy to
mimic this behavior when needed.
Testing strategy: TCB tests included.
PR Close#29698
View Engine's implementation of naive template type-checking is less
advanced than the current Ivy implementation. As a result, Ivy catches lots
of typing bugs which VE does not. As a result, it's necessary to tone down
the Ivy template type-checker in the default case.
This commit introduces a mechanism for doing that, by passing a config to
the template type-checking engine. Through this configuration, particular
checks can be loosened or disabled entirely.
Testing strategy: TCB tests included.
PR Close#29698
Previously the template type-checking code only considered the metadata of
directive classes actually referenced in the template. If those directives
had base classes, any inputs/outputs/etc of the base classes were not
tracked when generating the TCB. This resulted in bindings to those inputs
being incorrectly attributed to the host component or element.
This commit uses the new metadata package to follow directive inheritance
chains and use the full metadata for a directive for TCB generation.
Testing strategy: Template type-checking tests included.
PR Close#29698
Previously, metadata registration (the recording of collected metadata
during analysis of directives, pipes, and NgModules) was only used to
produce the `LocalModuleScope`, and thus was handled by the
`LocalModuleScopeRegistry`.
However, the template type-checker also needs information about registered
directives, outside of the NgModule scope determinations. Rather than
reuse the scope registry for an unintended purpose, this commit introduces
new abstractions for metadata registration and lookups in a separate
'metadata' package, which the scope registry implements.
This paves the way for a future commit to make use of this metadata for the
template type-checking system.
Testing strategy: this commit is a refactoring which introduces no new
functionality, so existing tests are sufficient.
PR Close#29698
Previously, bindings to [class] and [style] were treated like any other
property binding. That is, they would result in type-checking code that
attempted to write directly to .class or .style on the element node.
This is incorrect, however - the mapping from Angular's [class] and [style]
onto the DOM properties is non-trivial.
For now, this commit avoids the issue by only checking the expressions
themselves and not the assignment to the element properties.
Testing strategy: TCB tests included.
PR Close#29698
Previously the template type-checking engine processed templates in a linear
manner, and could not handle '#' references within a template. One reason
for this is that '#' references are non-linear - a reference can be used
before its declaration. Consider the template:
```html
{{ref.value}}
<input #ref>
```
Accommodating this required refactoring the type-checking code generator to
be able to produce Type Check Block (TCB) code non-linearly. Now, each
template is processed and a list of TCB operations (`TcbOp`s) are created.
Non-linearity is modeled via dependencies between operations, with the
appropriate protection in place for circular dependencies.
Testing strategy: TCB tests included.
PR Close#29698
JoostK
George Kalpakas
Paul Gschwendtner
Andrew Kushnir
Alex Rickabaugh
Alan Agius
Trevor Karjanis
Ben Lesh
José I. Escudero
Brandon
Kenny Schank
iliesaithamouda
Kristiyan Kostadinov
Adam Plumer
Kara Erickson
Keen Yee Liau
Suguru Inatomi
Trotyl Yu
Olivier Combe