* support for creating linux arm64 artifacts

1) NOTE: workaround for arm64 with pkg via "--no-bytecode --public"

the packaging utilility (pkg) has following issues:

* bytecode generation fails for arm64:
  Warning Failed to make bytecode node16-arm64 for file /snapshot/zap/node_modules/...

  "--no-bytecode" flag is used to skip the bytecode generation

* --no-bytecode flag is broken:
  "Error! --no-bytecode and no source breaks final executable"

  "--public" flag is used as a workaround
  ref: https://github.com/vercel/pkg/issues/1139

2) update mac artifacts to be platform/architecture specific

3) pkg: version bump to pick up codesign error for macOS arm64

* updating code and dependencies to work with node 18

* updating documentation and other items from node 16 to 18

* keeping code consistent with startGeneration which uses the same function successfully

* adding --in so the correct files are selected

- Add zcl_struct_items_by_struct_and_cluster_name helper for retrieving struct items based on struct and cluster name in helper-zcl.js
- Updating selectAllStructItemsByStructName to handle cluster name as well
- Updating the xml for TargetStruct such that it is now present in binding and access control clusters
- Adding a test to make sure the right struct items are generated based on cluster names
- Github: ZAP#916

* Add isSupported helper for Darwin.

This supports marking this as "provisional", which unlike "removed" can be
overridden by a later "introduced".

* Make handling of introduced things inside a provisional thing better.

* Also include description in zcl_clusters, to match all_user_clusters.

* Add fileFormat to the file.
* Add the cli logic.
* Add both builtin metafiles as default and refresh matter test file.
* Wrap up the type 1 format, keep default to 0.
* Add unit test and trigger proper conversion.
* Allow for injected key/value pairs.
* Fix the reportable.
* Fix the reportable change.
* Fix a unit test.
* Add name ordering.
* Upgrade one matter file.
* Convert an all-cluster matter file.
* Convert some more files.
* Add some more unit tests.

Adding the type to bitmap fields. This change is specific to the dotdot loader and is a backwards compatible fix which was broken during the ZCL data type schema update. (#1001)

JIRA: ZAPP-1122

* Fixing error message when loading a .zap file meant for a newer version of ZAP

* 1.) Adding button which allows the user to close the notification. 2.)Editing the code so the notification only shows for the relevant session. 3.) Taking out Notification button color changing because it needs to be redesigned

Github: ZAP#682

* Gathering the files before checking them

* Adding check to make sure the user entered required arguments defined in the spec

Github: ZAP#682

Github: ZAP#971

- Deprecating chip_client_clusters, chip_cluster_commands, chip_cluster_responses and chip_cluster_command_arguments with all_user_clusters, zcl_commands, zcl_command_responses and zcl_command_arguments respectively
- Adding more attributes to the command in db-mapping along with the extension to the zcl_command query
- Updating selectAllClustersDetailsFromEndpointTypes such that all_user_clusters can be filtered based on side of cluster
- Adding zcl_command_responses to retrieve all commands which are also command responses.a
- Extending asJavaType to handle asJavaBoxedType properly under all block helpers and not just chipType block helpers
- Updating the local map for command query
- Minor cleanup
- Github: ZAP#971

Ensure db mapping aliases commandName for commands (template compatibility) and that commands by cluster contain response names (#981)

Github: ZAP#898

Github: ZAP#898

Minor test work, no functional changes.

improve doc: direct user to release page on landing page / document release management process. (#972)

* direct user to release page on Github landing page / document release management process.

Adding to enabled_attributes_for_cluster_and_side and zcl_attributes_server helper to replace chip_server_cluster_attributes (#899)

- chip_server_cluster_attributes is returning server attributes which are not actually enabled because the server side cluster is disabled.
- The reason chip_server_cluster_attributes is picking the server side attributes which are not enabled is because those are enabled in the .zap file(Which is the case for saving user selections and ease of use). However the helper is not actually checking if the cluster is enabled as well.
- enabled_attributes_for_cluster_and_side is solving this issue correctly and actually only showing attributes which are truly enabled.
- Extending endpointTypeAttributeExtended such that there is a common place for an attribute map
- Updating endpointTypeAttributeExtended with everything that attributeExportMapping had such that it can act as a common place for all other temporary maps that have been created across our code. Also sorting endpointTypeAttributeExtended for easier readability and avoiding duplicate keys
- Updating zcl_attributes_server block helper such that it can behave like chip_server_cluster_attributes.
- Updating attribute map such that it can be used more widely
- Adding additional helpers to remove the stateful helpers
- Adding removeKeys option for removing certain columns which are not needed in a block helper such as enabled_attributes_for_cluster_and_side and zcl_attributes_server
- Adding if_unsupported_attribute_type and if_attribute_complex to check for unsupported and complex zcl data types
- Deprecating old stateful helpers with new stateless helpers
- Adding select data type using type name and cluster id to data_types, enums, bitmap,  numbers and structs
- Adding if_unsupported_attribute_callback helper and if_basic_attribute if helpers for java code generation
- Deprecating java and python zcl type helpers with new zcl type helpers such that they do not need any stateful information coming from parent block helpers. For eg chipType and chipCallback.type
- Adding tests for all the newly introduced helpers.
- Updating Api documentation
- Adding if_is_data_type_signed and as_zcl_data_type_size to determine sign and size of zcl data types
- if_is_data_type_signed and as_zcl_data_type_size take type name and cluster id to return size and sign of the zcl data types
- Github: ZAP#898

when a user changes configurations, such change might have invalidated some configuration specific setting that has been cached already. By flushing the cache, it enables us to cache SQL results without disregarding user inputs.

7x speed up on zap-config.h generation speed

machine: mac x86 2.6 GHz 6-Core Intel Core i7
Before:
PASS  test/gen-zigbee-9.test.js (91.223 s)
  ✓ Basic gen template parsing and generation (18 ms)
  ✓ Validate package loading (1 ms)
  ✓ Load the zap file: /Users/jiteng/repo/zap/test/resource/full-th.zap (1149 ms)
  ✓ Generate "zap-cluster-command-parser.h" (196 ms)
  ✓ Generate "zap-print.h" (14 ms)
  ✓ Generate "zap-id.h" (402 ms)
  ✓ Generate "zap-event.c" (111 ms)
  ✓ Generate "zap-enabled-incoming-commands.h" (41 ms)
  ✓ Generate "zap-event.h" (746 ms)
  ✓ Generate "zap-config.h" (72763 ms)
  ✓ Generate "zap-type.h" (223 ms)
  ✓ Generate "zap-tokens.h" (574 ms)
  ✓ Generate "zap-cli.c" (3542 ms)
  ✓ Generate "zap-command.h" (1161 ms)
  ✓ Generate "zap-command-structs.h" (448 ms)
  ✓ Generate "zap-cluster-command-parser.c" (275 ms)

After:

 PASS  test/gen-zigbee-9.test.js (24.919 s)
  ✓ Basic gen template parsing and generation (21 ms)
  ✓ Validate package loading
  ✓ Load the zap file: /Users/jiteng/repo/zap/test/resource/full-th.zap (1667 ms)
  ✓ Generate "zap-cluster-command-parser.h" (156 ms)
  ✓ Generate "zap-print.h" (39 ms)
  ✓ Generate "zap-id.h" (439 ms)
  ✓ Generate "zap-event.c" (127 ms)
  ✓ Generate "zap-enabled-incoming-commands.h" (49 ms)
  ✓ Generate "zap-event.h" (905 ms)
  ✓ Generate "zap-config.h" (10471 ms)
  ✓ Generate "zap-type.h" (241 ms)
  ✓ Generate "zap-tokens.h" (627 ms)
  ✓ Generate "zap-cli.c" (3802 ms)
  ✓ Generate "zap-command.h" (1238 ms)
  ✓ Generate "zap-command-structs.h" (555 ms)
  ✓ Generate "zap-cluster-command-parser.c" (251 ms)

See project-chip/connectedhomeip#24994

* node-cache: return refs instead of clone for speed.
* add '--disableDbCaching' flag
* reduce heap usage by 50% (118MB -> 58.8MB via Chrome Heap Profiler) via disabling source map during build
* improve generation speed by 75% via SQL query caching

Using full-th sample app as reference, there are 108,871 SQL query calls
spread over the following query functions

* selectAtomicType()
* selectBitmapByName()
* selectDataTypeByName()
* selectEnumByName()
* selectNumberByName()
* selectStructByName()

Performance metric:
  test runner: ./src-script/gen-test-runner.js
  machine: macbook 2.6 Ghz 6-core Intel Core i7 / 32GB DDR4

  Before:
    average generation time: 42.851s

  After:
    initial cache warm up generation time: 17.699s
    average generation time: 11.766s

* Cleanup some helper functions.
* Some further cleanup.
* Remove reference to xvfb.

* Add some more content-based matter tests.
* Check the content to further depth of details.
* Add a cautionary tale that nobody will read... :(
* Get coverage reports.
* Clean up some typos.
* Use proper English.
* Remove xvfb-run, we do not need it.
* Clean up coverage for cypress, which is not being produced.
* Add the differently-ordered list to the generation.
* Fix the code coverage job.

The token attributes are being set in the following order:
For Standard attributes:
- All singleton attributes
- All non singleton attributes
- Both of the above are followed by subsequent ordering of cluster mfg code, attribute mfg code, endpoint identifier(smallest endpoint identifier for singleton), cluster code, attribute code
Configuring the tests according to the above order

* Cut down unnecessary repeat generations from same source.
* Break up zigbee tests into individual generations.
* Split database files.
* Fix the problem with the result.
* Clean up some custom package stuff.
* Consistently name tests.
* Add another generation for performance analysis.
* Add a full-th test for performance analysis.
* Sort out some nuances of the test.
* Separate each individual file for generation.
* There is no need to run Zigbee cypress tests twice.
* Support array for generateOnly option, and limit unit test generation to just what we test.
* Exclude more generations of things not needed for tests.
* Restore Matter test data to reasonably recent, and mandate the Matter tests to pass without errors.
* Add the all-clusters test for Matter.
* Bring back old matter XML files to keep Cypress fixtures happy.
* Comment out a code coverage. It was not doing much useful now, so we are commenting it out for now, until we restore it to correct functioning later, using the cache.

* "clusterSize:" had a stray ':' that was causing generation to produce invalid
  output.
* The old code had things indented sanely, and this is just restoring that.
* There was no detection of unknown tokens, which would have quickly caught the
  "clusterSize" problem.

- Fixing Token Attribute helpers to include cluster defines
- Sorting token endpoints by endpoint id
- Bumping the feature level pointer
- JIRA: ZAPP-1085

Device types are 32-bit in the spec.

* Restore one matter unit test to being operational.
* Allow for sorting and omitting certain items in the struct population.
* Increase feature level.