tamaya-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From anat...@apache.org
Subject [2/4] incubator-tamaya git commit: TAMAYA-19: Streamlined API and impl.
Date Fri, 26 Dec 2014 00:56:09 GMT
http://git-wip-us.apache.org/repos/asf/incubator-tamaya/blob/a60570e8/docs/src/main/asciidoc/design/3_Core.adoc
----------------------------------------------------------------------
diff --git a/docs/src/main/asciidoc/design/3_Core.adoc b/docs/src/main/asciidoc/design/3_Core.adoc
new file mode 100644
index 0000000..90794be
--- /dev/null
+++ b/docs/src/main/asciidoc/design/3_Core.adoc
@@ -0,0 +1,356 @@
+// Licensed to the Apache Software Foundation (ASF) under one
+// or more contributor license agreements.  See the NOTICE file
+// distributed with this work for additional information
+// regarding copyright ownership.  The ASF licenses this file
+// to you under the Apache License, Version 2.0 (the
+// "License"); you may not use this file except in compliance
+// with the License.  You may obtain a copy of the License at
+//
+//   http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+<<<
+[[Core]]
+== Tamaya Core
+=== Overview
+
+Tamaya Core provides an implementation of the Tamaya Configuration API and adds additional functionality and
+building blocks for supporting SPI implementations.
+
+Tamaya Core contains the following artifacts:
+
+* +DefaultConfigurationSpi, DefaultPropertyAdapterSpi+ implement the main API interfaces, backing up +Configuration+ and
+  +PropertyAdapter+
+* tbd
+
+The SPI contains the following core concepts/artifacts:
+
+* tbd
+
+* +org.apache.tamaya.core+ contains the main abstractions +Configuration, ConfigQuery, PropertyAdapter,
+  PropertySource+
+* +org.apache.tamaya.core.spi+ contains the SPI interfaces to be implemented by implementations and the +ServiceContext+ mechanism.
+
+
+[[CorePropertyAdapterSpi]]
+=== PropertyAdapter (PropertyAdapterSpi) Implementation
+
+tbd
+
+[[CoreConfigurationSpi]]
+=== Configuration (ConfigurationSpi) Implementation
+
+tbd
+
+
+[[Core ConfigurationBuilder]]
+=== Building Simple Configuration
+
+Looking at the structures of configuration system used by large companies we typically encounter some kind of configuration
+hierarchies that are combined in arbitrary ways. Users of the systems are typically not aware of the complexities in this
+area, since they simply know the possible locations, formats and the overriding policies. Framework providers on the other
+side must face the complexities and it would be very useful if Tamaya can support here by providing prebuilt functionality
+that helps implementing these aspects. All this leads to the feature set of combining property sources. Hereby the following
+strategies are useful:
+
+* aggregating configurations, hereby later configurations added
+  ** override any existing entries from earlier configurations
+  ** combine conflicting entries from earlier configurations, e.g. into a comma-separated structure.
+  ** may throw a ConfigException ig entries are conflicting
+  ** may only add entries not yet defined by former providers, preventing entries that are already present to be overwrite
+  ** any custom aggregation strategy, which may be a mix of above
+* intersecting configurations
+* subtracting configurations
+* filtering configurations
+
+These common functionality is provided by +ConfigurationBuilder+ instances. Additionally to the base strategies above a
++MetaInfo+ instance can be passed optionally as well to define the meta information for the newly created configuration.
+Let's assume we have two configurations with the following data:
+
+[source,properties]
+.Configuration 1
+--------------------------------------------
+a=a
+b=b
+c=c
+g=g
+h=h
+i=i
+--------------------------------------------
+
+[source,properties]
+.Configuration 2
+--------------------------------------------
+a=A
+b=B
+c=C
+d=D
+e=E
+f=F
+--------------------------------------------
+
+Looking in detail you see that the entries +a,b,c+ are present in both configurations, whereas +d,e,f+ are only present in Configuration 1,
+and +g,h,i+ only in Configuration 2.
+
+[source,java]
+.Example Combining Configurations
+--------------------------------------------
+Configuration cfg1 = ...
+Configuration cfg2 = ...
+
+// aggregate, hereby values from Configuration 2 override values from Configuration 1
+Configuration unionOverriding = ConfigurationBuilder.of().aggregate(cfg1, cfg2).build();
+System.out.println("unionOverriding: " + unionOverriding);
+
+// ignore duplicates, values present in Configuration 1 are not overriden by Configuration 2
+Configuration unionIgnoringDuplicates = ConfigurationBuilder.of()
+                       .withAggregationPolicy(AggregationPolicy.IGNORE_DUPLICATES).aggregate(cfg1, cfg2).build();
+System.out.println("unionIgnoringDuplicates: " + unionIgnoringDuplicates);
+
+// this variant combines/maps duplicate values into a new value
+Configuration unionCombined = ConfigurationBuilder.of().withAggregationPolicy(AggregationPolicy.COMBINE)
+                       .aggregate(cfg1, cfg2);
+System.out.println("unionCombined: " + unionCombined);
+
+// This variant throws an exception since there are key/value paris in both providers, but with different values
+try{
+    ConfigurationBuilder.of().withAggregationPolicy(AggregationPolicy.EXCEPTION).aggregate(provider1, provider2)
+                       .build();
+}
+catch(ConfigException e){
+    // expected!
+}
+--------------------------------------------
+
+The example above produces the following outpout:
+
+[source,listing]
+.Example Combining Configurations
+--------------------------------------------
+AggregatedConfiguration{
+  (name = dynamicAggregationTests)
+  a = "[a][A]"
+  b = "[b][B]"
+  c = "[c][C]"
+  d = "[D]"
+  e = "[E]"
+  f = "[F]"
+  g = "[g]"
+  h = "[h]"
+  i = "[i]"
+}
+unionOverriding: AggregatedConfigurations{
+  (name = <noname>)
+  a = "A"
+  b = "B"
+  c = "C"
+  d = "D"
+  e = "E"
+  f = "F"
+  g = "g"
+  h = "h"
+  i = "i"
+}
+unionIgnoringDuplicates: AggregatedConfigurations{
+  (name = <noname>)
+  a = "a"
+  b = "b"
+  c = "c"
+  d = "D"
+  e = "E"
+  f = "F"
+  g = "g"
+  h = "h"
+  i = "i"
+}
+unionCombined: AggregatedConfigurations{
+  (name = <noname>)
+  a = "a,A"
+  b = "b,B"
+  c = "c,C"
+  d = "D"
+  e = "E"
+  f = "F"
+  g = "g"
+  h = "h"
+  i = "i"
+}
+--------------------------------------------
+
+No +AggregationPolicy+ is also a functional interface that can be implemented:
+
+[source,java]
+.AggregationPolicy Interface
+--------------------------------------------
+@FunctionalInterface
+public interface AggregationPolicy {
+    String aggregate(String key, String value1, String value2);
+}
+--------------------------------------------
+
+So we can also define our own aggregation strategy using a Lambda expression:
+
+[source,java]
+.Use a Custom AggregationPolicy
+--------------------------------------------
+Configuration cfg1 = ...;
+Configuration cfg2 = ...;
+Configuration config = ConfigurationBuilder.of("dynamicAggregationTests")
+      .withAggregationPolicy((k, v1, v2) -> (v1 != null ? v1 : "") + '[' + v2 + "]")
+      .aggregate(cfg1, cfg2).build();
+System.out.println(config);
+--------------------------------------------
+
+The output of this code snippet is as follows:
+
+[source,listing]
+.Listing of dynamic aggregation policy
+--------------------------------------------
+AggregatedConfiguration{
+  (name = dynamicAggregationTests)
+  a = "[a][A]"
+  b = "[b][B]"
+  c = "[c][C]"
+  d = "[D]"
+  e = "[E]"
+  f = "[F]"
+  g = "[g]"
+  h = "[h]"
+  i = "[i]"
+}
+--------------------------------------------
+
+Summarizing the +ConfigurationBuilder+ allows to combine providers in various forms:
+
+[source,listing]
+.Methods provided on PropertySources
+--------------------------------------------
+public final class ConfigurationBuilder {
+
+    private ConfigurationBuilder() {}
+
+    public static ConfigurationBuilder of();
+    public static ConfigurationBuilder of(PropertySource config);
+    public static ConfigurationBuilder of(String name);
+
+    public ConfigurationBuilder withAggregationPolicy(AggregationPolicy aggregationPolicy);
+    public ConfigurationBuilder withName(String name);
+
+    public ConfigurationBuilder addArgs(String... args);
+    public ConfigurationBuilder addPaths(List<String> paths);
+    public ConfigurationBuilder addUrls(URL... urls);
+    public ConfigurationBuilder addUrls(List<URL> urls);
+    public ConfigurationBuilder addMap(Map<String, String> map);
+
+    public Configuration empty();
+    public Configuration empty(String name);
+    public ConfigurationBuilder addEnvironmentProperties();
+    public ConfigurationBuilder addSystemProperties();
+    public ConfigurationBuilder aggregate(AggregationPolicy policy, Configuration... configs){
+    public ConfigurationBuilder aggregate(AggregationPolicy policy, List<Configuration> configs) {
+    public ConfigurationBuilder intersected(Configuration... providers) {
+    public ConfigurationBuilder subtracted(Configuration target, Configuration... providers) {
+    public ConfigurationBuilder filtered(Predicate<String> filter, Configuration config) {
+    public ConfigurationBuilder contextual(Supplier<Configuration> mapSupplier,
+                                              Supplier<String> isolationKeySupplier) {
+    public ConfigurationBuilder delegating(Configuration mainMap, Map<String, String> parentMap) {
+    public ConfigurationBuilder replacing(Configuration mainMap, Map<String, String> replacementMap);
+
+    public Configuration build();
+    public Configuration buildFrozen();
+}
+--------------------------------------------
+
+
+
+=== Environment
+
+The environment basically is also a kind of property/value provider similar to +System
+.getenv()+ in the JDK. Nevertheless it provides additional functionality:
+
+[source,java]
+.Interface Environment
+--------------------------------------------
+public interface Environment {
+
+    Optional<String> get(String key);
+    boolean containsKey(String key);
+    Set<String> keySet();
+    Map<String,String> toMap();
+
+    public static Environment current();
+    public static Environment root();
+--------------------------------------------
+
+* Basically an environment can contain any properties. The root environment
+  hereby must contain at least
+  ** all JDK's environment properties.
+  ** additional root properties are allowed as well.
+* the root environment is always directly accessible by calling +Environment.root()+
+* the current environment can be accessed by calling +Environment.current()+.
+
+Summarizing the Environment can be seen as a runtime context. This also implies, that this context changes
+depending on the current runtime context. Developers implementing an environment mechanism should be aware that
+an environment can be accessed very frequently, so evaluation and access of an +Environment+ must be fast. For
+further details we recommend the SPI details section of the core implementation.
+
+
+== SPI
+
+[[API PropertySourceBuilder]]
+==== Building Property Sources
+
+In [[PropertSource]] we have outlines that the essence of a property key store for configuration can be modelled by
+the +PropertySource+ interface. Similarly to the +ConfigurationBuilder+ you can also combine several +PropertySource+
+instances to assemble more complex configuration scenarios. Typically assembling is done within a +ConfigProvider+,
+which is responsible for providing correct Configuration corresponding to the current environment.
+
+Summarizing you can
+* aggregate providers, hereby later providers added
+  ** override any existing entries from earlier providers
+  ** combine conflicting entries from earlier providers, e.g. into a comma-separated structure.
+  ** may throw a ConfigException ig entries are conflicting
+  ** may only add entries not yet defined by former providers, preventing entries that are already present to be overwritten
+  ** any custom aggregation strategy, which may be a mix of above
+* intersecting providers
+* subtracting providers
+* filtering providers
+
+The following code snippet gives a couple of examples:
+
+[source,java]
+.Example Combining PropertySources
+--------------------------------------------
+PropertySource provider1 = ...
+PropertySource provider2 = ...
+
+// aggregate, hereby values from provider 2 override values from provider 1
+PropertySource unionOverriding = PropertySourceBuilder.of()
+             .aggregate(provider1, provider2).build(); // OVERRIDE policy is default
+System.out.println("unionOverriding: " + unionOverriding);
+
+// ignore duplicates, values present in provider 1 are not overriden by provider 2
+PropertySource unionIgnoringDuplicates = PropertySources
+             .aggregate(AggregationPolicy.IGNORE_DUPLICATES(), provider1, provider2).build();
+System.out.println("unionIgnoringDuplicates: " + unionIgnoringDuplicates);
+
+// this variant combines/maps duplicate values into a new value
+PropertySource unionCombined = PropertySourceBuilder.of().withAggregationPolicy(AggregationPolicy.COMBINE))
+            .aggregate(provider1, provider2).build();
+System.out.println("unionCombined: " + unionCombined);
+
+// This variant throws an exception since there are key/value paris in both providers, but with different values
+try{
+    PropertySourceBuilder.of().withAggregationPolicy(AggregationPolicy.EXCEPTION).aggregate(provider1, provider2);
+}
+catch(ConfigException e){
+    // expected!
+}
+--------------------------------------------
+
+

http://git-wip-us.apache.org/repos/asf/incubator-tamaya/blob/a60570e8/docs/src/main/asciidoc/design/3_Extensions.adoc
----------------------------------------------------------------------
diff --git a/docs/src/main/asciidoc/design/3_Extensions.adoc b/docs/src/main/asciidoc/design/3_Extensions.adoc
deleted file mode 100644
index db949ac..0000000
--- a/docs/src/main/asciidoc/design/3_Extensions.adoc
+++ /dev/null
@@ -1,841 +0,0 @@
-// Licensed to the Apache Software Foundation (ASF) under one
-// or more contributor license agreements.  See the NOTICE file
-// distributed with this work for additional information
-// regarding copyright ownership.  The ASF licenses this file
-// to you under the Apache License, Version 2.0 (the
-// "License"); you may not use this file except in compliance
-// with the License.  You may obtain a copy of the License at
-//
-//   http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing,
-// software distributed under the License is distributed on an
-// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
-// KIND, either express or implied.  See the License for the
-// specific language governing permissions and limitations
-// under the License.
-<<<
-[[CoreConcepts]]
-== {name} Core Concepts
-Though {name} is a very powerful and flexible solution there are basically only a few simple core concepts required that build
-the base of all the other mechanisms:
-
-The API contains the following core concepts/artifacts:
-
-* Literal Key/Value Pairs
-* _PropertyProvider:_ is the the SPI for a source that provides configuration data. A +PropertyProvider+
-     hereby defines
-     ** a minimalistic SPI to be implemented by the config data source
-     ** provides data key/value pairs in raw format as String key/values only
-     ** providers should not have any dependencies other than to the datasource
-     ** providers may read context dependent data, but basically providers themselves are not contextual.
-        Context management should be done by the ConfigurationProvider implementation that also is responsible
-        for combining a set of property providers to a Configuration.
-  _Configuration_ is the API that users of Tamaya will see, when they access configuration in raw format. Hereby +Configuration+
-     ** adds type support for non String types
-     ** provides functional extension points (+with,query+)
-     ** allows registering/deregistering of change listeners
-     ** is the entry point for evaluating the current +Configuration+
-     ** each +PropertyProvider+ can be easily converted into a +Configuration+
-     ** allows configuration entries to be injected
-     ** to access configuration _templates_ (annotated interfaces).
-     ** Configuration may support mutability by allowing instances of +ConfigChangeSet+ to be passed.
-* _PropertyProviders_ allows to aggregate different property providers. Hereby property providers are
-  seen as sets, which can be combined to new providers using set styled operations (aggregate, intersect, subtract).
-  This allows to model and create composite container providers, to build up more complex configuration models
-  step by step.
-* _MetaInfo_ is provided by each +Configuration, PropertyProvider+ and describes the configuration/provider and its entries.
-* _Environment_ is the base model for modelling the environment and the accessor for getting the current +Environment+ instance.
-* _Annotations_ a set of annotations allows to configure configuration injection on classes or interface (aka config templates).
-
-The SPI contains the following core concepts/artifacts:
-
-* _Bootstrap_ is the delegate singleton that is used by the framework to resolve components. The effective component
-  loading can be accessed by implementing and registering an instance of +ServiceProvider+ using +java.util.ServiceLoader+.
-* All the singleton used explicitly (+PropertyAdapters,PropertyProviders+ are backed up corresponding API interfaces.
-  To override a singleton's behaviour the corresponding SPI has to be implemented and registered, so it can be loaded
-  by the current +Bootstrap+ setup (by default ServiceLoader based).
-* Also the singleton used implicitly by +Configuration, Environment, Stage+ are backed up corresponding SPI interfaces.
-  To override a singleton's behaviour the corresponding SPI has to be implemented and registered, so it can be loaded
-  by the current +Bootstrap+ setup (by default ServiceLoader based).
-
-This is also reflected in the main parts of the API, which is quite small:
-
-* +org.apache.tamaya+ contains the main abstractions +Configuration, ConfigOperator, ConfigQuery, PropertyAdapter, Stage,
-  Environment, PropertyProvider, MetaInfo+
-* +org.apache.tamaya.spi+ contains the SPI interfaces to be implemented by implementations and the +Bootstrap+ mechanism.
-+ +org.apache.tamaya.annot+ contains the annotations defined.
-
-In the implementation are there are additional projects:
-
-* +org.apache.tamaya.core+ contains the core implementation of the API. Deploying it together with the API results in a
-  flexible framework that can be easily used for configuration of different complexity. But out of the box this framework
-  will not do much more than exposing system and environment properties, its power comes when an additional meta-model
-  is defined and deployed. Hereby you can write your own, or use on e of the provided ones (see later).
-* the core part is extended by multiple additional modules
-  ** CDI integration
-  ** Default configuration meta-models and providers for the most common usage scenarios
-    *** standalone applications
-    *** Java EE
-    *** ...
-
-These parts are explained in the following sections. It is recommended that user's of the API read through this part.
-All subsequent parts are building upon this concepts and may be more difficult to understand without having read
-this section.
-
-
-[[APIKeyValues]]
-=== Key/Value Pairs
-
-Basically configuration is a very generic concept. Therefore it should be modelled in a generic way. The most simple
-and similarly most commonly used are simple literal key/value pairs. So the core building block of {name} are key/value pairs.
-You can think of a common +.properties+ file, e.g.
-
-[source,properties]
-.A simple properties file
---------------------------------------------
-a.b.c=cVal
-a.b.c.1=cVal1
-a.b.c.2=cVal2
-a=aVal
-a.b=abVal
-a.b2=abVal
---------------------------------------------
-
-Now you can use +java.util.Properties+ to read this file and access the corresponding properties, e.g.
-
-[source,properties]
-.Accessing some properties
---------------------------------------------
-Properties props = new Properties();
-props.readProperties(...);
-String val = props.getProperty("a.b.c");
-val = props.getProperty("a.b.c.1");
-...
---------------------------------------------
-
-This looks familiar to most of you. Nevertheless when looking closer to the above key/value pairs,
-there are more concepts in place: looking at the keys +a.b.c+, +a.b.c.1+, +a.b.c.2+, +a+, +a.b+ we
-see that the key names build up a flattened tree structure. So we can define the following:
-
-Given a key +p1.p2.p3.k=value+:
-
-* +p1.p2.p3.k+ is called the _qualified key_
-* +p1.p2.p3+ is the key's _area_
-* the child areas +p1.p2", "p1+ are called _areas_ as well
-* +k+ is the _(unqualified) key_
-
-Given that you can perform some very useful actions:
-
-* you can filter the keys with an area. E.g. in the example before you can query for all keys within the area +a.b.c+
-  and map them to new properties set as follows:
-
-[source,properties]
-.Accessing an area
---------------------------------------------
-1=cVal1
-2=cVal2
---------------------------------------------
-
-Similarly accessing the area +a+ results in the following properties:
-
-[source,properties]
-.Accessing the area +a+
---------------------------------------------
-b=abVal
-b2=abVal
---------------------------------------------
-
-Additionally you can access all values of an area recursively, so accessing +a+ recursively results in
-the following properties:
-
-[source,properties]
-.Accessing area +a+ recursively
---------------------------------------------
-b.c=cVal
-b.c.1=cVal1
-b.c.2=cVal2
-b=abVal
-b2=abVal
---------------------------------------------
-
-Why this is useful? Well there are different use cases:
-
-* you can segregate your configuration properties, e.g. a module can access its module configuration by
-  querying all properties under the area +config.modules.myModule+ (or whatever would be appropriate).
-* you can use this mechanism to configure maps (or more generally: collections).
-* you can easily filter parts of configuration
-* ...and more.
-
-==== Why Using Strings Only
-
-Using Strings as base representation of configuration comes with several huge advantages:
-
-* Strings are simple to understand
-* Strings are human readable and therefore easy to prove for correctness
-* Strings can easily be used within different language, different VMs, files or network communications.
-* Strings can easily be compared and manipulated
-* Strings can easily be searched, indexed and cached
-* It is very easy to provide Strings as configuration, which gives much flexibility for providing configuration in
-  production as well in testing.
-* and more
-
-On the other side there are also disadvantages:
-
-* Strings are inherently not type safe, they do not provide validation out of the box for special types, such as
-numbers,
-  dates etc.
-* Often you want not to work with Strings, but with according types.
-* Strings are not hierarchical, so mapping hierarchical structures requires some extra efforts.
-
-Nevertheless most of these advantages can be mitigated easily, hereby still keeping all the benefits from above:
-
-* Adding type safe converters on top of String allow to add any type easily, that can be directly mapped out of Strings.
-  This includes all common base types such as numbers, dates, time, but also timezones, formatting patterns and more.
-* Even more complex mappings can be easily realized, by using String not as a direct representation of configuration,
-  but a reference that defines where the more complex configuration artifact is available. This mechanism is similarly
-  easy to understand as parsing Strings to numbers, but is powerful enough to provide e.g. all kind of deployment
-  descriptors in Java EE.
-* Hierarchical and collection types can be mapped in different ways:
-** The keys of configuration can have additional syntax/semantics. E.g. when adding dor-separating path semantics
-*** trees/maps can also simply be mapped.
-
-[APIPropertyProviders]
-=== Property Providers
-==== Basic Model
-
-We have seen that constrain configuration aspects to simple literal key/value pairs provides us with an easy to
-understand, generic, flexible, yet expendable mechanism. Looking at the Java language features a +vava.util.Map<String,
-String>+ and +java.util.Properties+ basically model these quite well out of the box.
-So it makes sense to build configuration on top of the JDK's +Map+ interface. This creates immediately additional
-benefits:
-
-* we inherit full Lambda and collection support
-* Maps are widely known and well understood
-
-Nevertheless there are some things to be considered:
-
-* Configuration also requires meta-data, such as
-** the origin of a certain configuration entry and how it was derived from other values
-** the sensitivity of some data
-** the provider that have read the data
-** the time, when the data was read
-** the timestamp, when some data may be outdated
-** ...
-
-Basically the same is also the not related to some single configuration key, but also to a whole map.
-The +PropertyProvider+ interface models exact these aspects and looks as illustrated below:
-
-[source,java]
-.Interface PropertyProvider
---------------------------------------------
-public interface PropertyProvider{
-
-      Optional<String> get(String key);
-      boolean containsKey(String key);
-      Map<String, String> toMap();
-      MetaInfo getMetaInfo();
-
-      default Set<String> keySet();
-      default ConfigChangeSet load();
-      default boolean isMutable();
-      default void apply(ConfigChangeSet change);
-}
---------------------------------------------
-
-Hereby
-
-* +getMetaInfo()+ return the meta information for the property provider, as well as for individual property key/value pairs.
-* +get, containsKey, keySet+ look similar to the methods on +Map+, though +get+ uses the +Optional+ type introduced
-  with Java 8. This avoids returning +null+ or throwing exceptions in case no such entry is available and also
-  reduced the API's footprint, since default values can be easily implemented by calling +Optional.orElse+.
-* +isMutable()+ allows to easy check, if a property provider is mutable, which is more elegant than catching
-  +NonSupportedOperation+ exception thrown on the according methods of +Map+.
-* +load()+ finally allows to (re)load a property map. It depends on the implementing source, if this operation
-  has any effect. If the map changes an according +ConfigChange+ must be returned, describing the
-  changes applied.
-* +hasSameProperties+ allows to perform a comparison with another provider.
-* +toMap+ allows to extract thing to a +Map+.
-
-This simple model will be used within the spi, where configuration can be injected/provided from external resources.
-But we have seen, that we have to consider additional aspects, such as extendability and type safety. Therefore we
-extend +PropertyMap+ and hereby also apply the 'composite pattern', which results in the following key abstraction.
-
-==== Meta Information
-
-Each instance also provides an instance of +MetaInfo+, which provides meta information for the providers and its properties:
-
-[source,java]
-.Accessing Meta Information
---------------------------------------------
-PropertyProvider prov = ...;
-MetaInfo metaInfo = prov.getMetaInfo();
-Set<String> keys = metaInfo.keySet();  // returns the attribute keys, for which meta-information is accessible.
-String metaData = metaInfo.get("a.b.c.value"); // access meta information
-String itemName = metaInfo.getName(); // access meta information for the provider
---------------------------------------------
-
-As we have seen above there is as well a +MetaInfoBuilder+, which must be used to create instances of
-+MetaInfo+.
-
-==== Mutability
-
-Property providers optionally may be mutable. This can be checked by calling +boolean isMutable()+. If a provider
-is mutable a +ConfigChangeSet+ can be passed. This change set can then be applied by the provider. On creation
-of the +ConfigChangeSetBuilder+ a provider can pass version information, so _optimistic locking_ can be implemented
-easily:
-
-[source,java]
-.Creating and applying a +ConfigChangeSet+ to a provider
---------------------------------------------
-PropertyProvider prov = ...;
-ConfigChangeSet changeSet = ConfigChangeSetBuilder.of(provider)  // creating a default version
-   .remove("key1ToBeRemoved", +key2ToBeRemoved")
-   .put("key2", "key2Value")
-   .put("key3", 12345)
-   .put("key4", 123.45)
-   .build();
-provider.apply(changeSet);
---------------------------------------------
-
-[[API CombineProviders]]
-==== Combining Property Providers
-
-Looking at the structures of configuration system used by large companies we typically encounter some kind of configuration
-hierarchies that are combined in arbitrary ways. Users of the systems are typically not aware of the complexities in this
-area, since they simply know the possible locations, formats and the overriding policies. Framework providers on the other
-side must face the complexities and it would be very useful if Tamaya can support here by providing prebuilt functionality
-that helps implementing these aspects. All this leads to the feature set of combining property providers. Hereby the following
-strategies are useful:
-
-* aggregating providers, hereby later providers added
-  ** override any existing entries from earlier providers
-  ** combine conflicting entries from earlier providers, e.g. into a comma-separated structure.
-  ** may throw a ConfigExcepotion ig entries are conflicting
-  ** may only add entries not yet defined by former providers, preventing entries that are already present to be overwritte
-  ** any custom aggregation strategy, which may be a mix of above
-* intersecting providers
-* subtracting providers
-* filtering providers
-
-These common functionality is provided by the +PropertyProviders+ singleton. Additionally to the base strategies above a +MetaInfo+
-instance can be passed optionally as well to define the meta information for the newly created provider instances.
-Let's assume we have two property providers with the following data:
-
-[source,properties]
-.Provider 1
---------------------------------------------
-a=a
-b=b
-c=c
-g=g
-h=h
-i=i
---------------------------------------------
-
-[source,properties]
-.Provider 2
---------------------------------------------
-a=A
-b=B
-c=C
-d=D
-e=E
-f=F
---------------------------------------------
-
-Looking in detail you see that the entries +a,b,c+ are present in both providers, whereas +d,e,f+ are only present in provider 1,
-and +g,h,i+ only in provider 2.
-
-[source,java]
-.Example Combining PropertyProviders
---------------------------------------------
-PropertyProvider provider1 = ...
-PropertyProvider provider2 = ...
-
-// aggregate, hereby values from provider 2 override values from provider 1
-PropertyProvider unionOverriding = PropertyProviders.aggregate(AggregationPolicy.OVERRIDE(), provider1, provider2);
-System.out.println("unionOverriding: " + unionOverriding);
-
-// ignore duplicates, values present in provider 1 are not overriden by provider 2
-PropertyProvider unionIgnoringDuplicates = PropertyProviders.aggregate(AggregationPolicy.IGNORE_DUPLICATES(), provider1, provider2);
-System.out.println("unionIgnoringDuplicates: " + unionIgnoringDuplicates);
-
-// this variant combines/maps duplicate values into a new value
-PropertyProvider unionCombined = PropertyProviders.aggregate(AggregationPolicy.COMBINE(), provider1, provider2);
-System.out.println("unionCombined: " + unionCombined);
-
-// This variant throws an exception since there are key/value paris in both providers, but with different values
-try{
-    PropertyProviders.aggregate(AggregationPolicy.EXCEPTION(), provider1, provider2);
-}
-catch(ConfigException e){
-    // expected!
-}
---------------------------------------------
-
-The example above produces the following outpout:
-
-[source,listing]
-.Example Combining PropertyProviders
---------------------------------------------
-AggregatedPropertyProvider{
-  (name = dynamicAggregationTests)
-  a = "[a][A]"
-  b = "[b][B]"
-  c = "[c][C]"
-  d = "[D]"
-  e = "[E]"
-  f = "[F]"
-  g = "[g]"
-  h = "[h]"
-  i = "[i]"
-}
-unionOverriding: AggregatedPropertyProvider{
-  (name = <noname>)
-  a = "A"
-  b = "B"
-  c = "C"
-  d = "D"
-  e = "E"
-  f = "F"
-  g = "g"
-  h = "h"
-  i = "i"
-}
-unionIgnoringDuplicates: AggregatedPropertyProvider{
-  (name = <noname>)
-  a = "a"
-  b = "b"
-  c = "c"
-  d = "D"
-  e = "E"
-  f = "F"
-  g = "g"
-  h = "h"
-  i = "i"
-}
-unionCombined: AggregatedPropertyProvider{
-  (name = <noname>)
-  a = "a,A"
-  b = "b,B"
-  c = "c,C"
-  d = "D"
-  e = "E"
-  f = "F"
-  g = "g"
-  h = "h"
-  i = "i"
-}
---------------------------------------------
-
-No +AggregationPolicy+ is also an interface that can be implemented:
-
-[source,java]
-.AggregationPolicy Interface
---------------------------------------------
-@FunctionalInterface
-public interface AggregationPolicy {
-    String aggregate(String key, String value1, String value2);
-}
---------------------------------------------
-
-So we can also define our own aggregation strategy using a Lambda expression:
-
-[source,java]
-.Use a Custom AggregationPolicy
---------------------------------------------
-PropertyProvider provider1 = ...;
-PropertyProvider provider2 = ...;
-PropertyProvider props = PropertyProviders.aggregate(
-      (k, v1, v2) -> (v1 != null ? v1 : "") + '[' + v2 + "]",
-      MetaInfo.of("dynamicAggregationTests"),
-      props1, props2);
-System.out.println(props);
---------------------------------------------
-
-Additionally we also pass here an instance of +MetaInfo+. The output of this code snippet is as follows:
-
-[source,listing]
-.Listing of dynamic aggregation policy
---------------------------------------------
-AggregatedPropertyProvider{
-  (name = dynamicAggregationTests)
-  a = "[a][A]"
-  b = "[b][B]"
-  c = "[c][C]"
-  d = "[D]"
-  e = "[E]"
-  f = "[F]"
-  g = "[g]"
-  h = "[h]"
-  i = "[i]"
-}
---------------------------------------------
-
-Summarizing the +PropertyProviders+ singleton allows to combine providers in various forms:
-
-[source,listing]
-.Methods provided on PropertyProviders
---------------------------------------------
-public final class PropertyProviders {
-
-    private PropertyProviders() {}
-
-    public static PropertyProvider fromArgs(String... args) {
-    public static PropertyProvider fromArgs(MetaInfo metaInfo, String... args) {
-    public static PropertyProvider fromPaths(AggregationPolicy aggregationPolicy, String... paths) {
-    public static PropertyProvider fromPaths(String... paths) {
-    public static PropertyProvider fromPaths(List<String> paths) {
-    public static PropertyProvider fromPaths(AggregationPolicy aggregationPolicy, List<String> paths) {
-    public static PropertyProvider fromPaths(MetaInfo metaInfo, List<String> paths) {
-    public static PropertyProvider fromPaths(AggregationPolicy aggregationPolicy, MetaInfo metaInfo, List<String> paths) {
-    public static PropertyProvider fromUris(URI... uris) {
-    public static PropertyProvider fromUris(AggregationPolicy aggregationPolicy, URI... uris) {
-    public static PropertyProvider fromUris(List<URI> uris) {
-    public static PropertyProvider fromUris(AggregationPolicy aggregationPolicy, List<URI> uris) {
-    public static PropertyProvider fromUris(MetaInfo metaInfo, URI... uris) {
-    public static PropertyProvider fromUris(AggregationPolicy aggregationPolicy, MetaInfo metaInfo, URI... uris) {
-    public static PropertyProvider fromUris(MetaInfo metaInfo, List<URI> uris) {
-    public static PropertyProvider fromUris(AggregationPolicy aggregationPolicy, MetaInfo metaInfo, List<URI> uris) {
-    public static PropertyProvider fromMap(Map<String, String> map) {
-    public static PropertyProvider fromMap(MetaInfo metaInfo, Map<String, String> map) {
-    public static PropertyProvider empty() {
-    public static PropertyProvider emptyMutable() {
-    public static PropertyProvider empty(MetaInfo metaInfo) {
-    public static PropertyProvider emptyMutable(MetaInfo metaInfo) {
-    public static PropertyProvider fromEnvironmentProperties() {
-    public static PropertyProvider fromSystemProperties() {
-    public static PropertyProvider freezed(PropertyProvider provider) {
-    public static PropertyProvider aggregate(AggregationPolicy mapping, MetaInfo metaInfo, PropertyProvider... providers){
-    public static PropertyProvider aggregate(PropertyProvider... providers) {
-    public static PropertyProvider aggregate(List<PropertyProvider> providers) {
-    public static PropertyProvider aggregate(AggregationPolicy mapping, PropertyProvider... propertyMaps) {
-    public static PropertyProvider aggregate(AggregationPolicy mapping, List<PropertyProvider> providers) {
-    public static PropertyProvider mutable(PropertyProvider provider) {
-    public static PropertyProvider intersected(AggregationPolicy aggregationPolicy, PropertyProvider... providers) {
-    public static PropertyProvider intersected(PropertyProvider... providers) {
-    public static PropertyProvider subtracted(PropertyProvider target, PropertyProvider... providers) {
-    public static PropertyProvider filtered(Predicate<String> filter, PropertyProvider provider) {
-    public static PropertyProvider contextual(Supplier<PropertyProvider> mapSupplier,
-                                              Supplier<String> isolationKeySupplier) {
-    public static PropertyProvider delegating(PropertyProvider mainMap, Map<String, String> parentMap) {
-    public static PropertyProvider replacing(PropertyProvider mainMap, Map<String, String> replacementMap) {
-}
---------------------------------------------
-
-
-[[API Configuration]]
-=== Configuration
-==== Basic Model
-
-Configuration inherits all basic features from +PropertyProvider+, but additionally adds functionality for
-type safety and extension mechanisms:
-
-[source,java]
-.Interface Configuration
---------------------------------------------
-public interface Configuration extends PropertyProvider{
-
-    default OptionalBoolean getBoolean(String key);
-    default OptionalInt getInteger(String key);
-    default OptionalLong getLong(String key);
-    default OptionalDouble getDouble(String key);
-    default <T> Optional<T> getAdapted(String key, PropertyAdapter<T> adapter);
-    <T> Optional<T> get(String key, Class<T> type);
-
-    // accessing areas
-    default Set<String> getAreas();
-    default Set<String> getTransitiveAreas();
-    default Set<String> getAreas(final Predicate<String> predicate);
-    default Set<String> getTransitiveAreas(Predicate<String> predicate);
-    default boolean containsArea(String key);
-
-    // extension points
-    default Configuration with(ConfigOperator operator);
-    default <T> T query(ConfigQuery<T> query);
-
-    // versioning
-    default String getVersion(){return "N/A";}
-    void addPropertyChangeListener(PropertyChangeListener l);
-    void removePropertyChangeListener(PropertyChangeListener l);
-
-    // singleton accessors
-    public static boolean isDefined(String name);
-    public static <T> T current(String name, Class<T> template);
-    public static Configuration current(String name);
-    public static Configuration current();
-    public static <T> T current(Class<T> type){
-    public static void configure(Object instance);
-    public static String evaluateValue(String expression);
-    public static String evaluateValue(Configuration config, String expression);
-    public static void addGlobalPropertyChangeListener(PropertyChangeListener listener);
-    public static void removeGlobalPropertyChangeListener(PropertyChangeListener listener);
-}
---------------------------------------------
-
-Hereby
-
-* +XXX getXXX(String)+ provide type safe accessors for all basic wrapper types of the JDK.
-* +getAdapted+ allow accessing any type, hereby also passing a +PropertyAdapter+ that converts
-  the configured literal value to the type required.
-* +getAreas()+, +getTransitiveAreas()+ allow to examine the hierarchical tree modeled by the configuration tree.
-  Optionally also predicates can be passed to select only part of the tree to be returned.
-* +containsArea+ allows to check, if an area is defined.
-* +with, query+ provide the extension points for adding additional functionality.
-
-* the static accessor methods define:
-  ** +current(), current(Class), current(String), current(String, Class)+ return the configuration valid for the current runtime environment.
-  ** +addPropertyChangeListener, removePropertyChangeListener+ allow to register or unregister
-     global config change listener instances.
-  ** evaluateValue allows to evaluate a configuration expression based on a given configuration.
-  ** +configure+ performs injection of configured values.
-
-[[TypeConversion]]
-==== Type Conversion
-
-Configuration extend +PropertyProvider+ and add additional support for non String types. This is achieved
-with the help of +PropertyAdapter+ instances:
-
-[source,java]
-.PropertyAdapter
---------------------------------------------
-@FunctionalInterface
-public interface PropertyAdapter<T>{
-    T adapt(String value);
-}
---------------------------------------------
-
-PropertyAdapter instances can be implemented manually or registered and accessed from the
-+PropertyAdapers+ singleton. Hereby the exact mechanism is determined by the API backing up the singleton.
-By default corresponding +PropertyAdapter+ instances can be registered using the Java +ServiceLoader+
-mechanism, or programmatically ba calling the +register(Class, PropertyAdapter)+ method.
-
-[source,java]
---------------------------------------------
-public final class PropertyAdapters{
-    public static <T> PropertyAdapter<T> register(Class<T> targetType, PropertyAdapter<T> adapter);
-    public static boolean isTargetTypeSupported(Class<?> targetType);
-    public static  <T> PropertyAdapter<T> getAdapter(Class<T> targetType);
-    public static  <T> PropertyAdapter<T> getAdapter(Class<T> targetType, WithPropertyAdapter annotation);
-}
---------------------------------------------
-
-Whereas this mechanism per se looks not very useful it's power shows up when combining it with the annotations
-API provided, e.g. look at the following annotated class:
-
-[source,java]
-.Annotated Example Class
---------------------------------------------
-public class ConfiguredClass{
-
-    @ConfiguredProperty
-    private String testProperty;
-
-    @ConfiguredProperty("a.b.c.key1")
-    @DefaultValue("The current \\${JAVA_HOME} env property is ${env:JAVA_HOME}.")
-    String value1;
-
-    @ConfiguredProperty("a.b.c.key2")
-    private int value2;
-
-    @ConfiguredProperty
-    @DefaultValue("http://127.0.0.1:8080/res/api/v1/info.json")
-    private URL accessUrl;
-
-    @ConfiguredProperty
-    @DefaultValue("5")
-    private Integer int1;
-
-    @ConfiguredProperty("a.b.customType")
-    private MyCustomType myCustomType;
-
-    @ConfiguredProperty("BD")
-    private BigDecimal bigNumber;
-
-    ...
-}
---------------------------------------------
-
-The class does not show all the possibilities that are provided, but it shows that arbitrary types can be supported easily.
-This applied similarly to collection types, whereas collections are more advanced and therefore described in a separate section
-later.
-
-Given the class above and the current configuration can provide the values required, configuring an instance of the
-class is simple:
-
-[source,java]
-.Configuring the Example Class
---------------------------------------------
-ConfiguredClass classInstance = new ConfiguredClass();
-Configuration.configure(configuredClass);
---------------------------------------------
-
-Additional types can transparently be supported by implementing and registering corresponding SPI instances. This is explained
-in the SPI documentation of {name}.
-
-==== Extension Points
-
-We are well aware of the fact that this library will not be able to cover all kinds of use cases. Therefore
-we have added similar functional extension mechanisms that were used in other areas of the Java eco-system as well:
-
-* +ConfigOperator+ define unary operations on +Configuration+. They can be used for filtering, implementing
-  configuration views, security interception etc.
-* +ConfigQuery+ defines a function returning any kind of result based on a configuration instance. Typical
-  use cases of queries could be the implementation of configuration SPI instances that are required
-  by other libraries or frameworks.
-
-Both interfaces hereby are defined as functional interfaces:
-
-[source,java]
-.ConfigOperator and ConfigQuery
---------------------------------------------
-@FunctionalInterface
-public interface ConfigOperator{
-    Configuration operate(Configuration config);
-}
-
-@FunctionalInterface
-public interface ConfigQuery<T>{
-    T query(Configuration config);
-}
---------------------------------------------
-
-Both interfaces can be applied on a +Configuration+ instance:
-
-[source,java]
-.Applying Config operators and queries
---------------------------------------------
-Configuration secured = Configuration.of().apply(ConfigSecurity::secure);
-ConfigSecurity securityContext = Configuration.of().query(ConfigSecurity::targetSecurityContext);
---------------------------------------------
-
-NOTE: +ConfigSecurity+ is an arbitrary class.
-
-=== Configuration Injection
-
-The +Configuration+ interface provides static methods that allow to anykind of instances be configured
-ny just passing the instances calling +Configuration.configure(instance);+. The classes passed hereby must
-be annotated with +@ConfiguredProperty+ to define the configured properties. Hereby this annotation can be
-used in multiple ways and combined with other annotations such as +@DefaultValue+,
-+@WithLoadPolicy+, +@WithConfig+, +@WithConfigOperator+, +@WithPropertyAdapter+.
-
-To illustrate the mechanism below the most simple variant of a configured class is given:
-
-[source,java]
-.Most simple configured class
---------------------------------------------
-pubic class ConfiguredItem{
-  @ConfiguredProperty
-  private String aValue;
-}
---------------------------------------------
-
-When this class is configured, e.g. by passing it to +Configuration.configure(Object)+,
-the following is happening:
-
-* The current valid +Configuration+ is evaluated by calling +Configuration cfg = Configuration.of();+
-* The current property value (String) is evaluated by calling +cfg.get("aValue");+
-* if not successful, an error is thrown (+ConfigException+)
-* On success, since no type conversion is involved, the value is injected.
-* The configured bean is registered as a weak change listener in the config system's underlying
-  configuration, so future config changes can be propagated (controllable by applying the
-  +@WithLoadPolicy+ annotation).
-
-In the next example we explicitly define the property value:
-[source,java]
---------------------------------------------
-pubic class ConfiguredItem{
-
-  @ConfiguredProperty
-  @ConfiguredProperty("a.b.value")
-  @configuredProperty("a.b.deprecated.value")
-  @DefaultValue("${env:java.version}")
-  private String aValue;
-}
---------------------------------------------
-
-Within this example we evaluate multiple possible keys. Evaluation is aborted if a key could be successfully
-resolved. Hereby the ordering of the annotations define the ordering of resolution, so in the example above
-resolution equals to +"aValue", "a.b.value", "a.b.deprecated.value"+. If no value could be read
-from the configuration, it uses the value from the +@DefaultValue+ annotation. Interesting here
-is that this value is not static, it is evaluated by calling +Configuration.evaluateValue(Configuration, String)+.
-
-=== Environment
-
-The environment basically is also a kind of property/value provider similar to +System.getProperties()+ and +System
-.getenv()+ in the JDK. Nevertheless it provides additional functionality:
-
-[source,java]
-.Interface Environment
---------------------------------------------
-public interface Environments {
-
-    String getEnvironmentType();
-    String getEnvironmentId();
-    Environment getParentEnvironment();
-
-    Optional<String> get(String key);
-    boolean containsKey(String key);
-    Set<String> keySet();
-    Map<String,String> toMap();
-
-    public static Environment current(){
-    public static Environment getRootEnvironment(){
-    public static List<String> getEnvironmentTypeOrder(){
-    public static List<String> getEnvironmentHierarchy(){
-    public static Optional<Environment> getInstance(String environmentType, String contextId){
-    public static Set<String> getEnvironmentContexts(String environmentType){
-    public static boolean isEnvironmentActive(String environmentType){
---------------------------------------------
-
-* environments are hierarchical. Hereby all environments inherit from the root environment. The root environment
-  hereby must contain
-  ** all JDK's system properties, with same keys, values
-  ** all JDK's environment properties, prefixed with +env:+.
-  ** additional root properties are allowed as well.
-* the root environment is always directly accessible by calling +Environment.getRootEnvironment()+
-* the current environment can be accessed by calling +Environment.of()+.
-* each environment also defines a +Stage+ (implementing +StageSupplier+). Hereby, if not set explicitly the +Stage+ is inherited from the root
-  environment. Consequently the root environment must provide a +Stage+, which by default is +Stage.development()+.
-
-Additionally each environment instance is uniquely identified by the environment type (accessible from
-+getEnvironmentType()+ and the environment id (accessible from +getEnvironmentId()+). So it is possible to access
-an +Environment+ by calling +of(String environmentType, String environmentId)+. Implementations may restrict access
-to environments depending on the current runtime environment (runtime context) active. The API does
-not require further aspects.
-
-The call to +getEnvironmentIds(String)+ returns all context ids of the known +Environment+ instances
-of a given type. E.g. assuming there is an environment type +war+ calling +Environment.getEnvironmentIds("war")+
-may return +"/web/app1", "/web/app2"+ (assuming the war context ids equal the web applications root contexts).
-
-All environments are basically ordered. The ordering can be accessed by calling +getEnvironmentTypeOrder()+. Hereby
-not every environment type in a hierarchy must necessarily present. This is reflected by +getEnvironmentHierarchy()+
-which returns the environment type ids in order, but only containing the types of the environments
-currently present and accessible in the hierarchy. As an example an environment type order in an advanced
-use case could be something like +"root","ear","war","saas","user"+, whereas the concrete environment type hierarchy
-may be +"root","war","saas"+, because the application was not included
-in an additional ear archive and no user is currently active (anonymous). The call to +isEnvironmentActive(String)+
-allows to determine if an environment of the given type is currently active.
-Finally the environment hierarchy is of course similarly reflected by the relationship (+getParentEnvironment()+).
-The following code should illustrate some of these concepts:
-
-[source,java]
-.Interface Environment
---------------------------------------------
-List<String> envHierarchy = Environment.getEnvironmentHierarchy();
-  // -> "root","war","saas"
-Environment env = Environment.of();
-System.out.println(env.getEnvironmentContext()); // saas
-System.out.println(env.getEnvironmentId());      // mysolution_pro
-env = env.getParentEnvironment();
-System.out.println(env.getEnvironmentContext()); // war
-System.out.println(env.getEnvironmentId());      // pro
-env = env.getParentEnvironment();
-System.out.println(env.getEnvironmentContext()); // root
-System.out.println(env.getEnvironmentId());      // system
-env = env.getParentEnvironment();
-// env is null now!
---------------------------------------------
-
-


Mime
View raw message