The Aggregator is based on and part of the Eclipse b3 project. b3 provides a versatile and adaptable framework supporting build, assembly and deployment processes. It supports a rich set of use cases. One of those - the aggregation of repositories - is the focus of the b3 Aggregator tool.
The Aggregator combines repositories from various sources into a new aggregated p2 repository. It can also be configured to produce a hybrid p2/Maven2 repository. There are many situations where using aggregated repositories is a good solution. The reasons vary from licensing issues to organizational requirements:
- Owners of a p2 repo for a given project may not be in position to host all required or recommended components due to licensing issues - Buckminster's SVN support can serve as an example here, as it requires components available through the main Eclipse p2 repo as well as third-party components. Hence users have to visit several repos for a complete install.
- Projects want to provide convenient access to their products - Installation instructions requiring the user to visit several repos for a complete install are not uncommon. An aggregated repo for all those locations provides a convenient one-stop strategy. The aggregation may support mirroring all consumed p2 repos or simply providing an indirection via a composite repo.
- Organizations or teams want control over internally used components - It may be necessary to have gated access to relevant p2 repos and do an organizational "healthcheck" of those before internal distribution. Furthermore, internally used aggregated repos can provide a common basis for all organizational users.
- Increase repository availability - by aggregating and mirroring what is used from multiple update sites into an internally controlled server (or several).
- Distributed Development Support - an overall product repository is produced by aggregating contributions from multiple teams.
The Aggregator tool is focused on supporting these specific requirements, and it plays an important role in the full scope of the b3 project. The Aggregator is however used in scenarios outside of the traditional "build domain" and this has been reflected in the user interface which does not delve into the details of "building" and should therefore be easy to use by non build experts.
Furthermore, it is worth noting that:
- the Aggregator is based on EMF models
- headless execution of aggregation definitions (once they have been created) is possible using a command line packaging of the Aggregator
The b3 Aggregator performs aggregation and validation of repositories. The input to the aggregator engine (that tells it what to do) is a b3aggr EMF model. Such a model is most conveniently created by using the b3 Aggregator editor. This editor provides both editing and interactive execution of aggregation commands. The editor is based on a standard EMF "tree and properties view" style editor where nodes are added and removed to from a tree, and the details of nodes are edited in a separate properties view. Once a b3aggr model has been created it is possible to use the command line / headless aggregator to perform aggregation (and other related commands). (Note that since the b3aggr is "just an EMF model", it can be produced via EMF APIs, transformation tools, etc., and thus support advanced use cases).
The model mainly consists of Contributions - specifications of what to include from different repositories, and Validation Repositories - repositories that are used when validating, but which are not included in the produced aggregation (i.e., they are not copied). The model also contains specification of various processing rules (exclusions, transformation of names, etc.), and specification of Contacts - individuals/mailing-lists to inform when processing fails.
Here are some of the important features supported by the b3 Aggregator:
- p2 and maven2 support — the aggregator can aggregate from and to both p2 and maven2 repositories.
- Maven2 name mapping support — names in the p2 domain are automatically mapped to maven2 names using built-in rules. Custom rules are also supported.
- Mirroring — artifacts from repositories are mirrored/downloaded/copied to a single location.
- Selective mirroring — an aggregation can produce an aggregate consisting of a mix of references to repositories and mirrored repositories.
- Cherry picking — it is possible to pick individual items when the entire content of a repository is not wanted. Detailed picking is supported as well as picking transitive closures like a product, or a category to get everything it contains/requires.
- Pruning — it is possible to specify mirroring based on version ranges. This can be used to reduce the size of the produced result when historical versions are not needed in the aggregated result.
- Categorization — categorization of installable units is important to the consumers of the aggregated repository. Categories are often choosen by repository publishers in a fashion that makes sense when looking at a particular repository in isolation, but when they are combined with others it can be very difficult for the user to understand what they relate to. An important task for the constructor of an aggregation is to be able to organize the aggregated material in an easily consumable fashion. The b3 aggregator has support for category prefixing, category renaming, addition of custom categories, as well as adding and removing features in categories.
- Validation — the b3 aggregator validates the aggregated result to ensure that everything in the repository is installable.
- Blame Email — when issues are found during validation, the aggregator supports sending emails describing the issue. This is very useful when aggregating the result of many different projects. Advanced features include specifying contacts for parts of the aggregation, which is useful in large multi-layer project structures where issues may be related to the combination of a group of projects rather than one individual project - someone responsible for the aggregation itself should be informed about these cross-project issues. The aggregator supports detailed control over email generation, including handling of mock emails when testing aggregation scripts.
Start by installing a fresh Eclipse 3.7 SDK from http://download.eclipse.org/eclipse/downloads
The b3 aggregator can either be integrated in your Eclipse SDK or it can be installed as a standalone headless product (i.e. pure command line, without any graphical UI).
The instructions below show the current URLs (when this document was written). Always check the latest information on the b3 download page before installing.
Eclipse SDK installation
Installation of the headless version of the Aggregator is similar to a typical headless b3 installation. The following steps focus on the installation of the headless Aggregator feature.
- Start by Downloading the (headless) director which can be found here.
- Next unpack the director to a location of your choice. You may want to set the
PATHto include the install location and make the director accessible for additional use.
- Install b3 with the following command:
director -r <HEADLESS_REPO> -d <INSTALL_DIR> -p b3 -i org.eclipse.b3.cli.product -i org.eclipse.b3.aggregator.engine.feature.feature.groupwhere
- -r <HEADLESS_REPO> is the headless p2 update site: Current stable version is: http://download.eclipse.org/modeling/emft/b3/headless-3.7, and our latest build can be found at https://hudson.eclipse.org/hudson/job/emft-b3-build/lastSuccessfulBuild/artifact/b3.headless.p2.repository
- -d <INSTALL_DIR> is the chosen install location of the headless b3
- -p b3 is the name of the p2 profile
- -i org.eclipse.b3.cli.product is the name of the headless b3 base
- -i org.eclipse.b3.aggregator.engine.feature.feature.group is the name of the aggregator feature
Getting started with standard examples
In the following sections we provide two simple examples that are easy to replicate and should highlight the most important features of the Aggregator. The first example deals with the creation of two variations of a p2 repo. The second shows the Aggregator's Maven support.
Aggregating a p2 repo
The first sample aggregation is build around Buckminster and its support for Subversive. The objective of this aggregated repo is to:
- provide a "one-stop shop" experience
- conveniently pull in third-party components that are not hosted at Eclipse
- provide this repo as an indirection mechanism if required
(This example aggregation can be downloaded via the b3 project SVN and opened in an appropriately set up workbench: buckminster_indigo.b3aggr).
The background is that Buckminster provides support for Subclipse. In addition to all components hosted at Eclipse, a complete installation will also require Subclipse components from Tigris.org (http://subclipse.tigris.org/update_1.6.x). We want to create a repo that combines these components and makes them accessible from one location. We want to make several platform configurations available.
This example already includes some of the more advanced aggregation features. Stepping through the model view from the top node the following components can be identified:
- The top node Aggregation groups all model definitions regarding ValidationSets and Contributions. Looking at the properties section at the bottom we see that:
- the top node has been provided with a mandatory Label with the value "Indigo + Buckminster for Subclipse"; this is also the label that is shown to users when accessing the aggregated repo via the p2 update manager
- the Build Root identifies the location to which the artifacts of the aggregated repo will be deployed
- The aggregation is defined for three configurations (i.e. os=win32, ws=win32, arch=x86; etc)
- any number of configurations can be defined
- during the aggregation process all dependencies of the contributed components will be verified for all provided configurations, unless exceptions are defined (see below)
- We have one ValidationSet labeled "main". A ValidationSet constitutes everything that will be validated as one unit by the p2 planner.
- The main ValidationSet contains three different contributions.
- The first Contribution to the aggregation is labeled "Indigo". This contribution includes binary configuration-specific artifacts which are only available for linux. If a simple contribution would be defined the aggregation would fail for all non-linux configurations, and hence the aggregation would fail as a whole.
- this requires a definition of Valid Configurations Rules that state exceptions
- the rules defined for the the three components in question essentially state that the verification process for those components should only be performed for linux-based configurations
- one Mapped Repository is defined for this contribution (it can have multiple); all that is needed is a user-defined label and the URL of the repository that should be included
- the result of this definition is that all categories, products, and features from Indigo p2 repo will be included in the aggregated repo.
- The second Contribution is labeled "Subclipse" and deals with the inclusion of bundles provided from the Subclipse project. #*this contribution represents the simplest example of a contribution
- The third Contribution is labeled "Buckminster (latest)". It shows another advanced feature - an Exclusion Rule.
- remember that the objective of the sample repo is to provide convenient setup of Buckminster with Subclipse support, and since Buckminster's Subclipse and Subversive support are mutually exclusive, we want to exclude the features for Subversive from the aggregated repo to make it easier for the user.
- this is done using an Exclusion Rule defined for each Installable Unit that should be excluded
- A list of all included repos is displayed at the bottom of the model editor view
- this list allows browsing the contents of all repos
- this part of the model is not editable
The aggregation can be run by right-clicking any node in the model and selecting Build Aggregation. This example was setup to use a mirroring approach for all contributed repos. Hence, the complete contents of all included can be found in the aggregated repos target location specified under Build Root.
Check the next section for a slightly different approach.
Providing a repo indirection
Mirroring all repo artifacts of your aggregated contributions is a very valuable and important feature when performing aggregation, but there are also many cases where this is not necessary. It is possible to turn off artifact mirroring/copying by changing one property for a defined contribution.
Each Mapped Repository has a boolean property called Mirror Artifacts which can be set to
The following buckminster_indigo_redirect.b3aggr is a variation of the first example with the Mirror Artifacts property set to
Creating a Maven-conformant p2 repo
You will need a headless installation of b3 with the Aggregator feature installed.
Running from the command lineJust type:
b3 aggregate <options>
For a detailed listing of the available options consult the next section.
Command line options
Currently there is no support for Hudson.
Aggregator model components and specific actions
This section provides an in-depth description and reference of the Aggregator model, listing all model components, properties and available actions.
The following aggregator-specific actions are available via the context menu that can be invoked on any node in the Aggregator model editor:
The root node of any aggregation model is the Aggregator node. It specifies a number of global properties including the Build Root (the target location of the aggregated repository) as well as the repo structure (maven-conformant or classic p2 setup). There are several child components some of which can be reference in other parts of the model: Configuration, Contribution, Contact, Custom Category, Validation Repository, and Maven Mapping.
An Aggregation may have one or more Configuration definitions. The aggregated repo will be verified for all added configurations. If dependencies for any of the given configurations fails the aggregation as a whole fails. It is however possible to specify exceptions for individual Contributions.
A Configuration is a combination of the following properties:
Contributions are the key element of any aggregation. Contributions specify which repositories (or parts thereof (category, feature, product, IU)) to include, and the constraints controlling their inclusion in the aggregated repository. A contribution definition may consist of several Mapped Repository and Maven Mapping components.
A Contribution may define several Mapped Repositories defining the actual content of the contribution. The Aggregator provides fine-grained control over the contribution from each Mapped Repository through references to Products, Bundles, Features, Categories, Exclusion Rules, Valid Configuration Rules.
Defining Product components allows the addition of individual Eclipse products to the aggregation to be specified (as opposed to mapping the complete contents of a given Mapped Repository. This naturally requires that products are present in the repositories being mapped.
Defining Category components allows the addition of the content in specific categories (provided by the contributed repository) rather than the complete contents of a given Mapped Repository.
Defining Bundle components allows addition of individual Eclipse bundles to the aggregation to be specified (rather than the complete contents of a given Mapped Repository).
Defining Feature components allows the addition of individual Eclipse features to the aggregation to be specified (rather than the complete contents of a given Mapped Repository). The features to include must be present in the mapped repository.
Furthermore, this component provides the means to group features implicitly into Custom Categories.
The Exclusion Rules provides an alternative way to control the content of the aggregated repository. An exclusion rule may specify exclusion of any bundle, feature or product. The excluded IU will not be considered in the aggregation and verification process. Each exclusion rule can only reference one IU id.
Valid Configuration Rule
By default all contributed contents of a Mapped Repository will be verified for all Configurations defined for the aggregation. A Valid Configuration Rule provides more control over validation. When using a Valid Configuration Rule, the referenced IUs (product, feature, or bundle) will only be verified and aggregated for the configurations specified in the rule. The rule only applies if the whole repository is mapped (i.e. when no explicit features, products, bundles or categories are mapped, regardless if they are enabled or disabled).
Maven Mapping (Contribution)
See Maven Mapping for a detailed description and list of properties.
Defines a resuseable contact element which can be referenced in other parts of the model and may be used to send notifications about the aggregation process.
A Custom Category provides a grouping mechanism for features in the aggregated repository. A custom category can be referenced by Features defined for inclusion from a Mapped Repository. The relationship to between Custom Category and a Feature is bi-directional. Thus, adding the feature to a custom category will update this property automatically in the Feature definition, and vice versa.
A Validation Repository is used to define that a repository should be used when validating dependencies for the aggregation but whose contents should not be included in the aggregated repository. It supports the cases where the objective is to create a repository that is not self sufficient, but rather a complement to something else (typical use case is an aggregation of everything produced by an organization with validation against the main Eclipse repository).
The Aggregator supports the creation of Maven-conformant repositories. A Maven repository requires a structure and use of naming conventions that may have to be achieved by a transformation of the Bundle-SymbolicName (BSN) when working with Eclipse bundles. There is a default translation from BSN naming standard to Maven naming. If that is not satisfactory, custom transformations are supported by the definition of one or more Maven Mappings which can be defined at the Aggregator and the Contribution level.
This only applies when the Maven Result property of the Aggregator model is set to true. In that case all defined Maven Mappings are applied in the order in which they appear in the model starting from the most specific to the most generic. That means for each artifact that a Contribution adds to the aggregated repository:
The most generic mapping is a default pattern that is applied whenever a Maven is created. It does not need to be added explicitly to the model. A mapping is specified using a regular expression that is applied to each BSN. The regular expression must specify two replacements; one for the maven groupId, and one for the maven artifactId. The group and artifact ids have an effect on the resulting Maven repo structure. The default pattern is:
The default maven mapping use the replacement
org | eclipse | b3 | org.eclipse.b3.aggregator | <folder name after version> | org.eclipse.b3.aggregator-<version>.jar ...
What else should be documented
When launched from IDE, only --buildModel and --action options are set (all other options have default values). Perhaps it's worth adding a launching dialog which would enable setting all options that are available in headless mode.
You can, for example, redirect a copy of the console output to a publicly available resource (a text file served by a http server). The public URL should be passed in the --logURL option so that a link to it would appear in the informational email.
The Hudson Support article was copied from the old buckminster documentation. It can be removed.
Only listed configurations are supported (they are a part of the model). Adding an option means changing the model and creation of a new aggregator build.
Yes, it is possible to load p2 and maven2 repositories by default (if someone adds a custom loader then he can load any type of repository). The output is always p2 compatible, optionally combined with maven2 (with no extensibility - at least now). However, if a maven2 repo is loaded and the result is also maven2 compatible, it is not identical to the original (not all attributes loaded from maven are stored in the p2 model).