[rush-lib] Fix parallelism percentage parsing

…3-reference-alias

[api-extractor] Fix #3593 Incorrect canonical reference to aliased class in .api.json

Update the .gitignore template to ignore files generated by IntelliJ IDEA

…ngConsumableParent()" method because it does nontrivial work

#3622

…IDEA

…g-test

[localization utilities] Add a test for a RESX file with a duplicate entry.

…extractor-fix-3593-reference-alias

# Conflicts:
#	apps/api-extractor/src/collector/CollectorEntity.ts
#	build-tests/api-extractor-scenarios/etc/exportImportStarAs2/api-extractor-scenarios.api.json
#	build-tests/api-extractor-scenarios/etc/namespaceImports/api-extractor-scenarios.api.json
#	build-tests/api-extractor-scenarios/etc/namespaceImports2/api-extractor-scenarios.api.json

[api-extractor] Fix a bug where an item is exported from both the entry point and an exported `AstNamespaceImport`.

[rush-lib] Add "beforeInstall" hook

As suggested by @zelliott, the logic when to fall back
to Navigation.Exports (`.`) must use `consumable`, not
`exported`.
Updated the improved build-tests/api-extractor-scenarios
results.

Some proposed changes to #3602

[rush-lib] Add RUSH_PNPM_VERIFY_STORE_INTEGRITY environment variable

"Incorrect canonical reference to aliased class in .api.json"
#3593

This fixes the other aliasing, namely through (nested)
namespaces.
Added a test scenario "docReferencesNamespaceAlias", which
consists of nested namespaces that have their own index.d.ts,
re-exporting the default export of each sibling file and
their respective sub-namespace (top-level -> 'renamed'
-> 'sub').
The *.api.json canonicalReferences now correctly use the
nested namespace "fully-qualified name".

Again, some other test result changed unexpectedly, maybe
revealing an existing bug / wrong expected result.
The scenarios are "exportImportStarAs2", "importEquals"
and "includeForgottenExports". The pattern is that
before, namespace and class were separate tokens, but
are now aggregated to one dot-separated expression.
I hope that these are improvements, but I am not sure,
so this should be reviewed.

Implementation:
---------------
When DeclarationReferenceGenerator#_getParentReference()
looks for a parent symbol, it must also resolve
namespace parent symbols the Collector derived from
'import * as ___'.
To that end, Collector now has a method
getExportingNamespace() that pulls namespace imports from
the CollectorEntities of a namespace symbol (if present).
For that, namespace imports are now also registered in the
_entitiesBySymbol Map.
In DeclarationReferenceGenerator, the handling of
namespaces had to be improved. _getNavigationToSymbol()
now detects namespaces, even behind an alias, and
then always returns "." as separator. To find the
parent, it also considers namespaces now.
_symbolToDeclarationReference() takes care not to
follow an alias that leads to a *-imported namespace,
because the target is the imported source file that
no longer allows to retrieve the namespace name.

"Incorrect canonical reference to aliased class in .api.json"
#3593

Fix first scenario from #3593.

Added test scenario "docReferencesAlias", inspired by
the test repo referenced in the GitHub issue:

`Item` and `Options` are default exports of two
separate source files, and `index` re-exports both,
renaming `Options` to `renamed_Options` (preserving
the "namespacing" information of the directory).

`DeclarationReferenceGenerator` now receives a function
to resolve a Symbol to its emit name.
`ApiModelGenerator` fetches this function from `Collector`,
which now maintains another map from symbol to
`CollectorEntity`. If for a given symbol, there is
a `CollectorEntity`, then its `emitName` is returned,
the symbol's name otherwise.

While this change fixes the `export { __ as __ }`
scenario, for _some_ reason, it changes the output
of one other test scenario, namely "ambientNameConflict".
In its `*.api.json` file, some reference to the
class re-exported as `MyPromise`, the name clash
of localFile's `Promise` and the global `Promise`
now leads to `Promise_2` being used instead of
`Promise`, the old expected result that was also
wrong, because it should have been `MyPromise`,
and `MyPromise` should appear in `*.api.json`, but
that is something completely different, so I'll
ignore the problem for now and check in the changed
expected `*.api.json` of "ambientNameConflict".