When computing i18n messages for templates there are two passes.
This is because messages must be computed before any whitespace
is removed. Then on a second pass, the messages must be recreated
but reusing the message ids from the first pass.
Previously ICUs were losing their legacy ids that had been computed
via the first pass. This commit fixes that by keeping track of the
message from the first pass (`previousMessage`) for ICU placeholder
nodes.
// FW-1637
PR Close#33318
Prior to this commit, the absolute spans (relative to template source
file rather than the start of an expression) of expressions in a
template attribute like `*ngIf` were generated incorrectly, equating to
the relative spans.
This fixes the bug by passing an `absoluteOffset` parameter when parsing
template bindings.
Through some levels of indirection, this is required for the Language
Service to support text replacement in
https://github.com/angular/angular/pull/33091.
PR Close#33189
Currently, the spans of expressions are recorded only relative to the
template node that they reside in, not their source file.
Introduce a `sourceSpan` property on expression ASTs that records the
location of an expression relative to the entire source code file that
it is in. This may allow for reducing duplication of effort in
ngtsc/typecheck/src/diagnostics later on as well.
Child of #31898
PR Close#31897
In an attempt to be compatible with previous translation files
the Angular compiler was generating instructions that always
included the message id. This was because it was not possible
to accurately re-generate the id from the calls to `$localize()` alone.
In line with https://hackmd.io/EQF4_-atSXK4XWg8eAha2g this
commit changes the compiler so that it only renders ids if they are
"custom" ones provided by the template author.
NOTE:
When translating messages generated by the Angular compiler
from i18n tags in templates, the `$localize.translate()` function
will compute message ids, if no custom id is provided, using a
common digest function that only relies upon the information
available in the `$localize()` calls.
This computed message id will not be the same as the message
ids stored in legacy translation files. Such files will need to be
migrated to use the new common digest function.
This only affects developers who have been trialling `$localize`, have
been calling `loadTranslations()`, and are not exclusively using custom
ids in their templates.
PR Close#32867
Metadata blocks are delimited by colons. Previously the code naively just
looked for the next colon in the string as the end marker.
This commit supports escaping colons within the metadata content.
The Angular compiler has been updated to add escaping as required.
PR Close#32867
Previously the metadata and placeholder blocks were serialized in
a variety of places. Moreover the code for creating the `LocalizedString`
AST node was doing serialization, which break the separation of concerns.
Now this is all done by the code that renders the AST and is refactored into
helper functions to avoid repeating the behaviour.
PR Close#32867
Now that the `$localize` translations are `MessageId` based the
compiler must render `MessageId`s in its generated `$localize` code.
This is because the `MessageId` used by the compiler is computed
from information that does not get passed through to the `$localize`
tagged string.
For example, the generated code for the following template
```html
<div id="static" i18n-title="m|d" title="introduction"></div>
```
will contain these localization statements
```ts
if (ngI18nClosureMode) {
/**
* @desc d
* @meaning m
*/
const MSG_EXTERNAL_8809028065680254561$$APP_SPEC_TS_1 = goog.getMsg("introduction");
I18N_1 = MSG_EXTERNAL_8809028065680254561$$APP_SPEC_TS_1;
}
else {
I18N_1 = $localize \`:m|d@@8809028065680254561:introduction\`;
}
```
Since `$localize` is not able to accurately regenerate the source-message
(and so the `MessageId`) from the generated code, it must rely upon the
`MessageId` being provided explicitly in the generated code.
The compiler now prepends all localized messages with a "metadata block"
containing the id (and the meaning and description if defined).
Note that this metadata block will also allow translation file extraction
from the compiled code - rather than relying on the legacy ViewEngine
extraction code. (This will be implemented post-v9).
Although these metadata blocks add to the initial code size, compile-time
inlining will completely remove these strings and so will not impact on
production bundle size.
PR Close#32594
The `goog.getMsg()` function requires placeholder names to be camelCased.
This is not the case for `$localize`. Here placeholder names need
match what is serialized to translation files.
Specifically such placeholder names keep their casing but have all characters
that are not in `a-z`, `A-Z`, `0-9` and `_` converted to `_`.
PR Close#32509
This commit changes the Angular compiler (ivy-only) to generate `$localize`
tagged strings for component templates that use `i18n` attributes.
BREAKING CHANGE
Since `$localize` is a global function, it must be included in any applications
that use i18n. This is achieved by importing the `@angular/localize` package
into an appropriate bundle, where it will be executed before the renderer
needs to call `$localize`. For CLI based projects, this is best done in
the `polyfills.ts` file.
```ts
import '@angular/localize';
```
For non-CLI applications this could be added as a script to the index.html
file or another suitable script file.
PR Close#31609
Similar to interpolation, we do not want to completely remove whitespace
nodes that are siblings of an expansion.
For example, the following template
```html
<div>
<strong>items left<strong> {count, plural, =1 {item} other {items}}
</div>
```
was being collapsed to
```html
<div><strong>items left<strong>{count, plural, =1 {item} other {items}}</div>
```
which results in the text looking like
```
items left4
```
instead it should be collapsed to
```html
<div><strong>items left<strong> {count, plural, =1 {item} other {items}}</div>
```
which results in the text looking like
```
items left 4
```
---
**Analysis of the code and manual testing has shown that this does not cause
the generated ids to change, so there is no breaking change here.**
PR Close#31962
Template AST nodes for (bound) attributes, variables and references will
now retain a reference to the source span of their value, which allows
for more accurate type check diagnostics.
PR Close#30181
Currently, template expressions and statements have their location
recorded relative to the HTML element they are in, with no handle to
absolute location in a source file except for a line/column location.
However, the line/column location is also not entirely accurate, as it
points an entire semantic expression, and not necessarily the start of
an expression recorded by the expression parser.
To support record of the source code expressions originate from, add a
new `sourceSpan` field to `ASTWithSource` that records the absolute byte
offset of an expression within a source code.
Implement part 2 of [refactoring template parsing for
stability](https://hackmd.io/@X3ECPVy-RCuVfba-pnvIpw/BkDUxaW84/%2FMA1oxh6jRXqSmZBcLfYdyw?type=book).
PR Close#31391
Fixes all TypeScript failures caused by enabling the `--strict`
flag for test source files. We also want to enable the strict
options for tests as the strictness enforcement improves the
overall codehealth, unveiled common issues and additionally it
allows us to enable `strict` in the `tsconfig.json` that is picked
up by IDE's.
PR Close#30993
Since `goog.getMsg` does not process ICUs (post-processing is required via goog.i18n.MessageFormat, https://google.github.io/closure-library/api/goog.i18n.MessageFormat.html) and placeholder format used for ICUs and regular messages inside `goog.getMsg` are different, the current implementation (that assumed the same placeholder format) needs to be updated. This commit updates placeholder format used inside ICUs from `{$placeholder}` to `{PLACEHOLDER}` to better align with Closure. ICU placeholders (that were left as is prior to this commit) are now replaced with actual values in post-processing step (inside `i18nPostprocess`).
PR Close#31459
There's no build time dependency from @angular/core to @angular/compiler,
so core can't directly refer to compiler types. To overcome this limitation,
there's a facade layer defined in the compiler and duplicated in core,
such that during runtime all types will correctly align.
There's a testcase in the compiler that verifies that all such facade types
are compatible across core and compiler, such that the core types can't get
misaligned with the actual definitions in the compiler. This suite of
tests were missing the `R3BaseMetadataFacade` facade type, so it was possible
for this type to get out of sync.
PR Close#31210
Change the Element constructor in r3_ast to create a new ParseSourceSpan when regenerating it rather than extending an object, which does not contain the overloaded toString().
PR Close#31190
This is a tentative fix for the error `Cannot write file '/node_modules/@angular/core/core.ngfactory.d.ts' because it would overwrite input file.` that is showing in codefresh windows ci.
PR Close#30482
It's unnecessary for a jasmine_node_test rule to depend on a TypeScript library. This dependency is already satisfied via the 'data' and also having it in 'deps' causes CI flakiness on Windows
PR Close#30482
Leading trivia, such as whitespace or comments, is
confusing for developers looking at source-mapped
templates, since they expect the source-map segment
to start after the trivia.
This commit adds skipping trivial characters to the lexer;
and then implements that in the template parser.
PR Close#30095
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
In some cases ivy expects projectable nodes to be passed in a different order
to ViewEngine. Specifically, ivy expects the catch-all ("*") to be at index
0, whereas ViewEngine expects it to be at its position at which it was parsed
in the template.
This commit adds one test that breaks under ivy and others that just describe
more accurately what happens in corner cases.
PR Close#27791
This PR alligns markup language lexer with the previous behaviour in version 7.x:
https://stackblitz.com/edit/angular-iancj2
While this behaviour is not perfect (we should be giving users an error message
here about invalid HTML instead of assuming text node) this is probably best we
can do without more substential re-write of lexing / parsing infrastructure.
This PR just fixes#29231 and restores VE behaviour - a more elaborate fix will
be done in a separate PR as it requries non-trivial rewrites.
PR Close#29328
This PR alligns markup language lexer with the previous behaviour in version 7.x:
https://stackblitz.com/edit/angular-iancj2
While this behaviour is not perfect (we should be giving users an error message
here about invalid HTML instead of assuming text node) this is probably best we
can do without more substential re-write of lexing / parsing infrastructure.
This PR just fixes#29231 and restores VE behaviour - a more elaborate fix will
be done in a separate PR as it requries non-trivial rewrites.
PR Close#29328
BREAKING CHANGE:
Certain elements (like `<tr>` or `<col>`) require parent elements to be of a certain type by the HTML specification
(ex. <tr> can only be inside <tbody> / <thead>). Before this change Angular template parser was auto-correcting
"invalid" HTML using the following rules:
- `<tr>` would be wrapped in `<tbody>` if not inside `<tbody>`, `<tfoot>` or `<thead>`;
- `<col>` would be wrapped in `<colgroup>` if not inside `<colgroup>`.
This meachanism of automatic wrapping / auto-correcting was problematic for several reasons:
- it is non-obvious and arbitrary (ex. there are more HTML elements that has rules for parent type);
- it is incorrect for cases where `<tr>` / `<col>` are at the root of a component's content, ex.:
```html
<projecting-tr-inside-tbody>
<tr>...</tr>
</projecting-tr-inside-tbody>
```
In the above example the `<projecting-tr-inside-tbody>` component culd be "surprised" to see additional
`<tbody>` elements inserted by Angular HTML parser.
PR Close#29219
Currently with ViewEngine, if someone runs the platform's
`bootstrapModule` method in order to boostrap a module in
JIT mode, external component resources are properly resolved
*automatically*.
Currently with Ivy, the developer would need to manually call
`resolveComponentResources` in order to asynchronously fetch
the determined external component resources. In order to make
this backwards compatible with ViewEngine, and also since
platforms can already specify a `ResourceLoader` compiler
provider, we need to automatically resolve all external
component resources on module bootstrap.
--
Since the `ResourceLoader` is part of the `@angular/compiler`,
because ViewEngine performed the factory creation in the compiler,
we can't access the `ResourceLoader` token from within core.
In order to workaround this without introducing a breaking change,
we just proxy the `ResourceLoader` token to `core` through the
compiler facade. In the future, we should be able to move the
`ResourceLoader` to core when ViewEngine code no longer exists in
the `@angular/compiler`.
PR Close#29083
At the moment, certain tests relies on resolving the module with an index.d.ts, this root cause might be some implementations are missing from the mocks.
Similar to: 58b4045359
PR Close#28884
The content projection mechanism is static, in that it only looks at the static
template nodes before directives are matched and change detection is run.
When you have a selector-based content projection the selection is based
on nodes that are available in the template.
For example:
```
<ng-content selector="[some-attr]"></ng-content>
```
would match
```
<div some-attr="..."></div>
```
If you have an inline-template in your projected nodes. For example:
```
<div *ngIf="..." some-attr="..."></div>
```
This gets pre-parsed and converted to a canonical form.
For example:
```
<ng-template [ngIf]="...">
<div some-attr=".."></div>
</ng-template>
```
Note that only structural attributes (e.g. `*ngIf`) stay with the `<ng-template>`
node. The other attributes move to the contained element inside the template.
When this happens in ivy, the ng-template content is removed
from the component template function and is compiled into its own
template function. But this means that the information about the
attributes that were on the content are lost and the projection
selection mechanism is unable to match the original
`<div *ngIf="..." some-attr>`.
This commit adds support for this in ivy. Attributes are separated into three
groups (Bindings, Templates and "other"). For inline-templates the Bindings
and "other" types are hoisted back from the contained node to the `template()`
instruction, so that they can be used in content projection matching.
PR Close#29041
Previously the start of a character indicated by an escape sequence
was being incorrectly computed by the lexer, which caused tokens
to include the start of the escaped character sequence in the
preceding token. In particular this affected the name extracted
from opening tags if the name was terminated by an escape sequence.
For example, `<t\n>` would have the name `t\` rather than `t`.
This fix refactors the lexer to use a "cursor" object to iterate over
the characters in the template source. There are two cursor implementations,
one expects a simple string, the other expects a string that contains
JavaScript escape sequences that need to be unescaped.
PR Close#28978
The parts of a token are supposed to be an array of not-null strings,
but we were using `null` for tags that had no prefix. This has been
fixed to use the empty string in such cases, which allows the `null !`
hack to be removed.
PR Close#28978
Prior to this change i18n block bindings were converted to Expressions right away (once we first access them), when in non-i18n cases we processed them differently: the actual conversion happens at instructions generation. Because of this discrepancy, the output for bindings in i18n blocks was generated incorrectly (with invalid indicies in pipeBindN fns and invalid references to non-existent local variables). Now the bindings processing is unified and i18nExp instructions should contain right bind expressions.
PR Close#28969
During build time we remap particular property bindings, because their names don't match their attribute equivalents (e.g. the property for the `for` attribute is called `htmlFor`). This breaks down if the particular element has an input that has the same name, because the property gets mapped to something invalid.
The following changes address the issue by mapping the name during runtime, because that's when directives are resolved and we know all of the inputs that are associated with a particular element.
PR Close#28765
With #28594 we refactored the `@angular/compiler` slightly to
allow opting out from external symbol re-exports which are
enabled by default.
Since symbol re-exports only benefit projects which have a
very strict dependency enforcement, external symbols should
not be re-exported by default as this could grow the size of
factory files and cause unexpected behavior with Angular's
AOT symbol resolving (e.g. see: #25644).
Note that the common strict dependency enforcement for source
files does still work with external symbol re-exports disabled,
but there are also strict dependency checks that enforce strict
module dependencies also for _generated files_ (such as the
ngfactory files). This is how Google3 manages it's dependencies
and therefore external symbol re-exports need to be enabled within
Google3.
Also "ngtsc" also does not provide any way of using external symbol
re-exports, so this means that with this change, NGC can partially
match the behavior of "ngtsc" then (unless explicitly opted-out).
As mentioned before, internally at Google symbol re-exports need to
be still enabled, so the `ng_module` Bazel rule will enable the symbol
re-exports by default when running within Blaze.
Fixes#25644.
PR Close#28633
When template bindings are being parsed the event handlers
were receiving a source span that included the whole attribute.
Now they get a span that is focussed on the handler itself.
PR Close#28055
When tokenizing markup (e.g. HTML) element attributes
can have quoted or unquoted values (e.g. `a=b` or `a="b"`).
The `ATTR_VALUE` tokens were capturing the quotes, which
was inconsistent and also affected source-mapping.
Now the tokenizer captures additional `ATTR_QUOTE` tokens,
which the HTML related parsers understand and factor into their
token parsing.
PR Close#28055
In order to support source mapping of templates, we need
to be able to tokenize the template in its original context.
When the template is defined inline as a JavaScript string
in a TS/JS source file, the tokenizer must be able to handle
string escape sequences, such as `\n` and `\"` as they
appear in the original source file.
This commit teaches the lexer how to unescape these
sequences, but only when the `escapedString` option is
set to true. Otherwise there is no change to the tokenizing
behaviour.
PR Close#28055
The lexer that does the tokenizing can now process only a part the source
string, by passing a `range` property in the `options` argument. The
locations of the nodes that are tokenized will now take into account the
position of the span in the context of the original source string.
This `range` option is, in turn, exposed from the template parser as well.
Being able to process parts of files helps to enable SourceMap support
when compiling inline component templates.
PR Close#28055
Pete Bacon Darwin
ayazhafiz
crisbeto
Kara Erickson
Misko Hevery
JoostK
Paul Gschwendtner
Andrew Kushnir
Alan Agius
Ben Lesh
Keen Yee Liau
Alex Rickabaugh
Pawel Kozlowski
Matias Niemelä
Greg Magolan