Symfony bundle that integrates gdbots/ncr library.
Follow the standard bundle install using gdbots/ncr-bundle as the composer package name.
The examples below assume you're running the DynamoDb Ncr and Elastica NcrSearch.
config/packages/ncr.yml:
# many of the configurations below are defaults so you can remove them
# from your configuration, added here for reference
parameters:
cloud_region: 'us-west-2'
es_clusters:
default:
debug: '%kernel.debug%'
timeout: 300 # default
persistent: true # default
round_robin: true # default
servers:
- {host: '127.0.0.1', port: 9200} # default
gdbots_ncr:
ncr:
provider: dynamodb
memoizing:
enabled: true # default
#
# IMPORTANT read_through notes for the memoizer
#
# If true, the NcrCache will be updated when a cache miss occurs.
# When the Pbjx request bus is in memory, then you'd want this
# to be false so there aren't two processes updating the NcrCache.
# One is the memoizer and the other are event listeners which are
# updating cache after successful get node requests.
read_through: false # default
psr6:
enabled: true # default
# by default this cache is the same cache as your application
# it is recommend to configure this separately because Ncr items
# can be in cache much longer than app cache, for example, they
# don't need to be cleared on app deployments.
provider: cache.app # default
read_through: true # default
# to customize the cache keys or tweak cache times
# you must provide your own class and override those methods.
# see Psr6Ncr::getCacheKey and Psr6Ncr::beforeSaveCacheItem
#class: Acme\Ncr\Repository\Psr6Ncr
ncr_cache:
max_items: 500 # default
dynamodb:
config:
# these apply to batch get operations
batch_size: 100 # default
concurrency: 25 # default
table_manager:
# multi-tenant applications will likely need to provide a custom
# table manager so node tables can be derived at runtime.
# class: Acme\Ncr\Repository\DynamoDb\TableManager
table_name_prefix: my-ncr
node_tables:
# any SchemaCurie not defined here will end up using the default
# the entire default key is not needed unless you're changing it
default:
class: Gdbots\Ncr\Repository\DynamoDb\NodeTable # default
table_name: multi # default
'acme:article':
table_name: article
'acme:user':
# class is optional but needed if you want to have custom
# global secondary indexes or customize items before they
# are put into the table.
class: Acme\Ncr\Repository\DynamoDb\UserTable
table_name: user
ncr_search:
provider: elastica
elastica:
# your app will at some point need to customize the queries
# override the class so you can provide these customizations.
class: Acme\Ncr\Search\Elastica\ElasticaNcrSearch
query_timeout: '500ms' # default
clusters: '%es_clusters%'
index_manager:
# multi-tenant apps will probably need to use a custom class
#class: Acme\Ncr\Search\Elastica\IndexManager
index_prefix: my-ncr
indexes:
default:
number_of_shards: 5 # default
number_of_replicas: 1 # default
types:
# types do not need to be configured unless there is a custom
# mapping or they must be put into different indexes or types
# than the automatic handling would provide
'acme:user':
index_name: members # default is "default"
mapper_class: Acme\Ncr\Search\Elastica\UserMapper
# typically these would be in services.yml file.
services:
# If you are using AWS ElasticSearch service, use AwsAuthV4ClientManager
gdbots_ncr.ncr_search.elastica.client_manager:
class: Gdbots\Ncr\Search\Elastica\AwsAuthV4ClientManager
arguments:
- '@aws_credentials'
- '%cloud_region%'
- '%es_clusters%'
- '@logger'
tags:
- {name: monolog.logger, channel: ncr}
# recommended, create your own lazy loading handler
# to optimize batch requests.
acme_ncr.ncr_lazy_loading_handler:
class: Acme\Ncr\NcrLazyLoadingHandler
public: false
arguments: ['@ncr_lazy_loader']
tags:
- {name: pbjx.event_subscriber}
It is recommended to have data retrieval be the responsibility of Pbjx requests, however, that strategy doesn't work for all uses cases. Use Symfony autowiring and typehint the interface in your constructor or setter methods to get key Ncr services.
Autowiring supported for these interfaces:
Gdbots\Ncr\NcrGdbots\Ncr\NcrCacheGdbots\Ncr\NcrLazyLoaderGdbots\Ncr\NcrSearch
The NcrExtension provides a function called ncr_get_node. It is important to note that this does NOT make a query to get a node, instead it pulls from NcrCache.
This might change in the future, but this strategy eliminates horribly performing twig templates that make Ncr queries.
The function will accept a MessageRef, NodeRef or a string version of a NodeRef.
Example use of ncr_get_node:
{% set created_by = ncr_get_node(pbj.get('created_by_ref')) %}
Here is the creator:
{{ created_by }}Other twig functions (documentation wip):
- ncr_deref_nodes
- ncr_get_preloaded_nodes
- ncr_get_preloaded_published_nodes
- ncr_is_node_published
- ncr_preload_node
- ncr_preload_nodes
- ncr_preload_embedded_nodes
- ncr_to_node_ref
This library provides the basics for creating and extracting data from the Ncr services. Run the Symfony console and look for ncr commands.
ncr:create-search-storage Creates the NcrSearch storage.
ncr:create-storage Creates the Ncr storage.
ncr:describe-search-storage Describes the NcrSearch storage.
ncr:describe-storage Describes the Ncr storage.
ncr:export-nodes Pipes nodes from the Ncr to STDOUT.
ncr:get-node Fetches a single node by its NodeRef and writes to STDOUT.
ncr:reindex-nodes Pipes nodes from the Ncr and reindexes them.
ncr:sync-node Syncs a single node from the Ncr with the EventStore.
ncr:sync-nodes Syncs nodes from the Ncr with the EventStore.Review the --help on the ncr commands for more details.