The root cause of #1218 is that `invokePartial` creates a stack of data frames
for nested partial blocks, but `resolvePartial` always uses the value at top of
the stack without "popping" it. The result is an infinite recursive loop, as
references to `@partial-block` in the partial at the top of the stack resolve to
itself.
So, walk up the stack of data frames when evaluating. This is accomplished by
1) setting the `partial-block` property to `noop` after use and
2) using `_parent['partial-block']` if `partial-block` is `noop`
Fix#1218
Users can utilize the UMD library if they are still using require.js and if they need to have specific modules, they can consume the cjs or es6 modules via webpack or similar.
Fix for #1096
Allows for ] literal characters to be used within [] IDs by prefixing them with the \ character. `\` literal at the end of the may be referenced by the `\\` sequence if conflicting. Under most circumstances the `\\` sequence will continue to work.
Potentially breaking change for users of [] ids that have `\\` anywhere in the id or `\` at the end of the id.
Fixes#1092
If a decorator is used within a partial but not in the calling template, the hash is not passed in. For now error on the side of always including as just assigning values has minimal overhead.
Fixes#1089
Allows for partials to be defined within the current template to allow for localized code reuse as well as for conditional behavior within nested partials.
Fixes#1018
These allow for a given block to be wrapped in helper methods or metadata and allow for more control over the current container and method before the code is run.
This allows for failover for missing partials as well as limited templating ability through the `{{> @partial-block }}` partial special case.
Partial fix for #1018
There is no real need for us to do `.call(container` other than for backwards compatibility with legacy versions. Using the 4.x release as a chance to optimize this behavior.