t-call-context is a feature that's supposed to mask the rendering
context completely, but currently the component remains available
through `this`.
This commit stops treating `this` as a reserved word, so that it's
compiled to a lookup in the rendering context, and adds `this` to the
rendering context when binding the component's rendering function. With
these changes, `this` behaves the same as before when outside a
t-call-context, but when the rendering context is overriden, the
template can no longer access `this`. `this` still represents the
instance of the component inside of the rendering function since it's
needed by owl internally.
A side-effect of this change is that now the rendering context is no
longer the instance of the component by default, but is always an object
with the component in its prototype chain. This was already the case
before in some contexts (eg inside t-foreach, or inside components with
a t-set/t-call anywhere in its template). This can cause issues in rare
cases when a component method was called directly on the rendering
context, as before this change, the method's bound this would be the
component instance (except in a t-foreach, component with a
t-set/t-call, etc), while after this change it is now never the
component instance. When the method only reads on `this` there is no
issue as all the components properties are available on the rendering
contexts, but setting a value on `this` will write on the rendering
context and not the component which is likely a mistake.
While this is a breaking change, simply adding a t-set/t-call to any
template would break components that would be broken by this change,
with this in mind we decided to make this change anyway so that
developers get the error as early as possible in the development cycle
rather than having a seemingly inocuous change break code under them.
Previously, when a template contained backslashes, they were not escaped
when creating the blockstring, meaning they would be interpreted as an
escape sequence within the JS string. This means that backslashes
preceding most characters were completely ignored and didn't end up in
the final block, double backslashes were collapsed to a single one, and
backslashes that constituted valid escape sequences in JS would be
treated as those (eg, \n would be a newline).
When creating a JS string expression from a string value, all characters
with special meaning in JS should be escaped, we were correctly escaping
backticks as these would unexpectedly close the string if not escaped,
but forgot to escape backslashes. This commit fixes that.
closes#1300
With the introduction of t-call-context, there is now no guarantee that
the ComponentNode can be found in the rendering context, any attempt to
do so can crash when combined with t-call-context. This commit fixes
that by using `this` instead, which in compiled templates refers to the
component that is being rendered.
The t-out directive is compiled internally into a LazyValue, which
represents a value that may or may not be created sometimes in the
future. It can also be reused more than once, and this is where there
may be an issue: if a component is contained in the lazyvalue, it needs
a unique key (coming from the t-foreach) to be properly indexed in the
parent children map. However, the LazyValue does not keep the key
information, so it is not able to provide it to its content.
The fix is then quite clear: the LazyValue class should store the key
information, and provides it to its content. This allows the LazyValue
to be used multiple times, in any place in a template.
closes#1270
Before this commit, the template compiler would guess the next block id
that will be generated when compiling the body of a tcall. This is
correct IF there are not nested t-call, but otherwise wrong, because the
next block id could be mixed up: the first t-call would save the next
block id (let's say n), then the inner t-call would also save the same
block id (so, n), will then generate its own block (n+1), then the outer
t-call would use the block n index instead of n+1
The best fix, in my opinion, is to make sure we get the next block var
name, so we do not have to guess. To do that, each compile block type
function needs to properly return the information.
closes#1267
Recently, we made it so that when a component is rendered, it always
updates the property values for computed properties. This was done by
wrapping the value in a String or Boolean object. One issue with this is
that wrapping a falsy value in a String doesn't yield an empty string,
but a string containing the value as text (eg new String(undefined) ->
"undefined"), which causes the value to not remain empty as per the
spec. This commit fixes that by adding a fallback to the empty string
for falsy values before converting to a String object.
closes: #1236
Before this commit, owl would incorrectly skip patching html properties
when the new value is the same as the value used in the previous render.
However, this is incorrect, since the user may have changed the value by
clicking on a checkbox, or changing some text in an input.
So, to be more correct, owl has to force the update whenever it
encounters a prop. This is however hard to do without impacting the
main update loop, so we kind of work around the problem by using String
and Boolean instance, instead of primitive values.
Before this commit, whenever Owl encounter a t-if, it generates an
anchor (a "hole") in the current block being compiled. However, in some
cases, the content of the t-if may not have any content at all, and the
anchor is then useless. Worse, the code generating the anchor generates
an index based on the number of sub blocks, but if there is no content,
the next anchor being created will have the same index, which then may
cause weird bugs.
A possible way to fix this could be to make sure we increment properly
the anchor index, but we could even do better: not having an anchor at
all.
Before this commit, the generated code for the component directive was
using the current context as the place to look for static informations
(such as the sub components). However, it is not entirely correct, since
the current context may be different than the current component (which
is easily accessed by using the this variable).
Also, while doing this, we fix some issues in the t-set directive, which
as calling lazy values with the wrong this.
Previously, having a leading or trailing space caused
HTMLElement.classList.add to be called with an empty string (because we
are splitting on whitespace), which is not allowed and caused a crash.
This commit fixes that by trimming the keys of the class object in the
same way that we already do it for class strings.
Before this commit, the generated code crashed if an attribute in
the template contained backticks. The compiler output something
like
```js
let block1 = createBlock(`<div foo="`bar`"/>`);
```
The backticks are now properly escaped.
This means that unrelated ids (eg the id of a template, variable or key)
not longer share the same incrementing counter, meaning that you no
longer see a variable named "v2" unless another variable "v1" was
generated previously, this is also true for block data.
The problem was that the compiler is based on the assumption that the
multi block received by the parser only occurs in some cases
where the structure of the template require a multi block, and it does
not work when we have random multiblock elsewhere.
We could fix the issue by modifying the code generator code to support
these usecases, or by simply removing these cases in the parser. Since
this seems more efficient, this is the approach taken by this commit.
Note that it was a good opportunity to simplify the parser.
t-out automatically escaped content when it is a string not marked
with the `markup` function
t-out renders the raw content if it is a Block, or if it has been marked
with the `markup` funtion.
t-esc has been kept since it is safe and is optimized to render text nodes.
all t-raw calls are in fact the same as t-out.
Samuel Degueldre