| RE: [corona-dev] ProjectContainer Definition |
|
If you are interested in CVS
repository itself, you define one repository and one access method. The content
of CVS is always files. It is the matter of the class that accesses and
displays the repository. You know exactly that it is CVS type and you know how
to connect to it. We had a short discussion here
about a related problem: take for example that you want to access a team
repository located in an xml file in CVS repository. In this case, you define
accessMethod as xml and contents (kind) as teamrepository, parameters define that
the xml file is not located locally, but in CVS. Pawel From:
corona-dev-bounces@xxxxxxxxxxx [mailto:corona-dev-bounces@xxxxxxxxxxx] On Behalf Of Everitt, Glenn Seems to me that the accessMethod (what
you were calling type) belongs in the repository-connection I am still not sure what to do when I have
a repository that contains resources that we may want to process/display differently
for example CVS holds java source files and XML files and Schema Files.
Do we just define 3 repositories with the same repository-connection (as well
as same accessMethod) and then specify a different content type? Glenn Everitt From: We need the information about type
and kind to identify how to process a repository. E.g. a class (say
MyViewClass) implements a tab to be displayed on the multi-page project
browser. Repository kind is used to determine, which class should be used
(whether it is team members or icons repository or bugs repository). Repository
type is used by MyViewClass to actually connect to the repository. All three
repositories can use the same "type", i.e. they all access XML or a
database, but the business logic behind is very different. Classes on the client side access
repositories and require their types and connection parameters. Let's change the word kind to
"contents" and the word type to "accessMethod" and the word
uri to "identifier" in repository definition. Pawel From: What software components will use the ‘kind’ attribute http://www.eclipse.org/corona/repositoryKinds/.... What will the ‘kind’ attribute
be used for? I think you are using the type
incorrectly. I think the type tells you it is a TeamMember Repository and
I think the Repository connection information tells you how to connect to
it. So, once we are connected to the repository do we care whether
it is a database or XML? I think we only care about the API to get / set
resources to in the repository. What uses the API for interacting with
the repository? Glenn Everitt From: One more comment: for uri used inside
repository kind attribute should start with http://www.eclipse.org/corona/repositoryKinds/.... From: On the basis of our current project
container model I’ve defined some project container having 2 team members
repositories defined to show why repository kind and type attributes
could/should(?) coexist: <?xml
version="1.0" encoding="UTF-8"?> <sdo:datagraph
xmlns:project="http://www.eclipse.org/corona/container/project"
xmlns:sdo="commonj.sdo"> <changeSummary
xmlns=""/> <project:project-context-container
name="SampleEclipseProject">
<container-uri>http://www.eclipse.org/corona/container/sampleEclipseProject</container-uri>
<container-path>
<container-path-element>http://www.eclipse.org/corona/container/sampleEclipseProject</container-path-element>
</container-path>
<container-type>http://www.eclipse.org/corona/containerTypes/eclipseProject</container-type>
<container-description>Sample eclipse
project</container-description>
<repositories> <repository
kind="http://www.eclipse.org/corona/repositoryTypes/teamMemberRepository"
name="TeamMemberRepository"
type="http://www.eclipse.org/corona/repositoryTypes/xmlRepository"
uri="http://www.eclipse.org/corona/repository/TeamMemberRepository">
<description>Team member xml repository definition</description>
<repository-configurations
default-configuration-name="members_1">
<repository-configuration name="members_1">
<repository-connection-parameters>
<repository-connection-parameter name="fileURI"
value="file:///model/data/teamMembers.xml"/>
</repository-connection-parameters>
<resource-subset selection-criteria="//team/members" type="http://www.eclipse.org/corona/resourceSubsetTypes/XPathResultSet"/>
</repository-configuration>
</repository-configurations> </repository>
<repository
kind="http://www.eclipse.org/corona/repositoryTypes/teamMemberRepository"
name="TeamMemberRepository2"
type="http://www.eclipse.org/corona/repositoryTypes/jdbcRepository"
uri="http://www.eclipse.org/corona/repository/TeamMemberRepository2">
<description>Team member jdbc repository definition </description>
<repository-configurations
default-configuration-name="members_2">
<repository-configuration name="members_2">
<repository-connection-parameters>
<repository-connection-parameter name="url" value="jdbc:postgresql://corona:5432/teamMembers2"/>
<repository-connection-parameter name="driver"
value="org.postgresql.Driver"/>
<repository-connection-parameter name="user"
value="anonymous"/>
</repository-connection-parameters>
<resource-subset selection-criteria="select * from users"
type="http://www.eclipse.org/corona/resourceSubsetTypes/JdbcResultSet"/>
</repository-configuration>
</repository-configurations>
</repository>
</repositories>
</project:project-context-container> </sdo:datagraph> Is such situation possible
– in other words can we have two separate team member repositories
defined within the same project container (in such situation to receive list of
all team members we would have to sum results parsed from xml and tuples from
database)???? From: Thanks Edyta, I had forgotten that the
selection criteria were specified as an attribute the resource-subset. I
was looking for it on repository I couldn’t decide where to put when I
first defined this schema. I mis-typed one of the comments below: . I am sure what is meant by
persistence method regarding repository. Should have read . I am not sure what is meant by persistence method regarding
repository. Regarding this comment We defenitely need a "kind" attribute in the repository
definition. Currently there is a "type" attribute which determins what is
a persistance method for the respository (eg. XML, JDBC, file, etc.). But
we still need to know what kind of information it provides - if it is team member, artifact or web page repository. The repository definition points to a
repository configuration. The repository configuration points at the
repository implementation. The same repository implementation can be
pointed at by more than one repository definition. For example: - <repository name="CoronaBugRepository" type="Corona Defects" version="01.00.00" repository-configuration="CoronaBugRepositoryConfiguration" /> - <repository-configuration
name="CoronaBugRepositoryConfiguration"> - <repository-connection> <repository-uri>https://bugs.eclipse.org/bugs/enter_bug.cgi?product=Corona </repository-uri>
</repository-connection> </repository-configuration> repository
name="MylarBugRepository" type="Mylar Defects"
version="01.00.00"
repository-cofiguration="MylarBugRepositoryConfiguration"> - <repository-configuration
name="MylarBugRepositoryConfiguration"> - <repository-connection> <repository-uri>https://bugs.eclipse.org/bugs/enter_bug.cgi?product=Mylar</repository-uri>
</repository-connection> Or possibly using URI’s to point to
the definition, but then the user interface would have to look up type
definition before displaying. type="http://org.eclipse.corona.repository.type.defect/CoronaDefects" Glenn Everitt The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. The contents of this e-mail are intended for the named addressee only.
It contains information that may be confidential. Unless you are the named
addressee or an authorized designee, you may not copy or use it, or disclose it
to anyone else. If you received it in error please notify us immediately and
then destroy it. |