Re: [xwiki-users] Getting user name with a Java Component
On 28 Nov 2014 at 07:00:26, Marius Dumitru Florea (mariusdumitru.flo...@xwiki.com(mailto:mariusdumitru.flo...@xwiki.com)) wrote: On Nov 28, 2014 12:33 AM, Bryn Jeffries wrote: I wrote: What's the right way to get the current user from the execution context within a Java component? Thomas Mortagne replied: For what you need the simplest in a component is usually to use org.xwiki.bridge.DocumentAccessBridge component (from xwiki-platform-bridge module) until a proper user manager api is introduced. OK, so I should not use ExecutionContext at all? You’re right, our goal is to have all context information (including the user) be put in the Execution Context. However, currently, we’re in a transition phase, trying to move from the old XWikiContext to the new ExecutionContext and thus a lot of information is currently still held in the XWikiContext. See the details here: http://platform.xwiki.org/xwiki/bin/view/DevGuide/WritingComponents#HTheXWikicontext Thanks -Vincent It looks the the approach you're suggesting is: 1) Inject a DocumentAccessBridge instead of an Execution, e.g.: @Inject private DocumentAccessBridge bridge; 2) Get a reference to the current user (According to http://maven.xwiki.org/site/docs/xwiki-javadoc-5.0.x/org/xwiki/bridge/DocumentAccessBridge.html which might be out of date but I can't find a more recent API doc) DocumentReference userDoc = bridge.getCurrentUserReference(); 3) Presumably the document reference captures the full location of the user's profile page. The DocumentReference API doesn't appear to have a page name accessor, so presumably to extract the user name I need to pull it out of userDoc.toString(). What information do you need precisely? The user alias (what the user uses to log in)? Or the user pretty name? The user alias is the name of the user profile document. bridge.getCurrentUserReference().getName() For the pretty name you need to get the value of the first_name and last_name properties from the user profile document for which you have the reference: the user reference. bridge.getProperty... Hope this helps, Marius Is this right? Conceptually I find this less intuitive than using a context object, so it's a pity if that's been deprecated. Thanks, Bryn ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Handling exceptions from components
On 28 Nov 2014 at 01:00:37, Bryn Jeffries
(bryn.jeffr...@sydney.edu.au(mailto:bryn.jeffr...@sydney.edu.au)) wrote:
What's the preferred way to deal with errors occuring in Java components? Is
it OK to throw an exception from within the component,
yes
and can this be caught within the Velocity or Groovy code that calls it?
errors are automatically caught in the script macros ({{velocity}}, {{groovy}},
etc).
Thanks
-Vincent
___
users mailing list
users@xwiki.org
http://lists.xwiki.org/mailman/listinfo/users
[xwiki-users] Mocca Calendar Application 2.2.1 released - with a glitch
To folks who have the Mocca Calendar[1] application installed in their XWiki instances: I have just released a new version 2.2.1 to which you might want to update to, as it contains quite a few bugfixes and improvements, thanks to patches send in by several contributors.[2] However the new release does not show up when you try to update via the Extension Updater in your wiki admin; instead the latest version found there will be 2.1.9. This happened due to a glitch when reorganizing the code which changed the Extension-Id for that application. Instead please go to Add extensions, search for Mocca Calendar and when the search result comes up, the extension manager should propose you to update to the correct version 2.2.1 Also while upgrading at least in my test instances I got a few spurious conflicts. If you get them too, even while you have made no modifications to the code part of the calendar, please choose to keep the new version to get all updates installed property. Best regards and sorry for the hiccup, Clemens [1] http://extensions.xwiki.org/xwiki/bin/view/Extension/MoccaCalendar [2] http://jira.xwiki.org/browse/MOCCACAL-32?filter=13491 ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
[xwiki-users] Migration from 5.4.5 Standalone to 6.3 Tomcat
Hi, in our company we used the standalone version of XWiki for testing and are very satisfied with it so we want to migrate to the new version and a more robust setup. The setup is Windows Server 2003 R2 (sadly only 32 bits) with Apache 2.4, MySQL 5 and Tomcat 7. XWiki 6.3 is setup now including a Wiki Template so I'd like to migrate the data I exported as XAR from the 5.4.5 standalone instance. My approach was exporting data from 5.4.5, creating the Wikis using the template and importing the spaces and pages from the XAR export. The issue I'm facing is Java heap size which constantly gets in my way. As we're on a 32 bits system I can only assign ~1.2 GB of memory (the server has 4GB). This limitation makes it impossible to import the data into the new setup. I tried general export/import via administration page and the various options using the Admin Tools Application but on one of both ends I always get the out of memory error. - Using normal export/import: Export works but import throws the error - Admin Tools - Export pages and spaces: Export works but import throws the error - Admin Tools - Large Export on Disk: the heap size error is thrown Do you have any suggestions/hints on how I can migrate the spaces and pages I have in the 5.4.5 setup to the 6.3 one? Thanks in advance, Dennis ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] SOLR and object search
On Mon, Nov 24, 2014 at 9:46 AM, Gerritjan Koekkoek gerritjankoekk...@gmail.com wrote: Hi, I'm trying to understand the new SOLR search (i did not understand the old Lucene very well) My use-case is the following. We have a special FAQ application where the object has, amongst others, the following attributes: - Subject - Topic-group - Language - Question - Answer he default search returns page-title/name, but this is in our case a non-informational, generated by the system code. So instead of Page name we would like to show: Subject You have 3 options: (1) Use Velocity to output the value of the subject property in the FAQ title. See https://github.com/xwiki/xwiki-platform/blob/master/xwiki-platform-core/xwiki-platform-blog/xwiki-platform-blog-ui/src/main/resources/Blog/BlogPostSheet.xml#L36 Gerritjan: Thanks this is helpful but see below (2) I doesn't make sense to have a separate String property to store the subject when you can store it in the document title. The designer of our FAQ has opted to store multiple objects of same class in one page. (because XWiki does not support multiple locales (language-country) on a object level. So the title must pick the subject name of the object of the current language-country combination. When we just display the subject we do not need to solve this challenge. (3) Customize the entire search page just to change the way the search results are displayed Yes I fear there is no way around this? As facets we would like to show Topic-group and Language, Default (on entering the page with search box) we would like to set the context-language as a search filter... So when reader is reading french the result only shows french FAQ (with a french subject title) if english only english. By checking and unchecking languages in the facets the user could extend or reduce the search. This is already the case with the default search. The context language is checked by default in the locale facet. See https://github.com/xwiki/xwiki-platform/blob/master/xwiki-platform-core/xwiki-platform-search/xwiki-platform-search-solr/xwiki-platform-search-solr-ui/src/main/resources/Main/SolrSearchConfig.xml#L74 . It looks the context language is only looking at language attribute of page. We need to compare with attribute in object. I think it would be useful if some explanation is how the object-based facets can be defined. A challenge is the topic field, this is a list with translation key. So if user is french-language the list will show french topics, but if he/she check english as well in facets things get complicated. We have a business-rule that the english collection of FAQ's is the baseline, and most comprehensive. The other languages are only translations of the same. So another languages can not have a question not translated. So my question is: How to define the search http://extensions.xwiki.org/xwiki/bin/view/Extension/Solr+Search+Application#HForDevelopers http://design.xwiki.org/xwiki/bin/view/Design/SolrSchema http://extensions.xwiki.org/xwiki/bin/view/Extension/Solr+Search+Query+API http://lucene.apache.org/solr/ How to modify the output so page-title is no longer showing How to modify facets so only the two fields can be set See https://github.com/xwiki/xwiki-platform/blob/master/xwiki-platform-core/xwiki-platform-faq/xwiki-platform-faq-ui/src/main/resources/FAQCode/FAQSearch.xml#L49 .This helps... Hope this helps, Marius ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Migration from 5.4.5 Standalone to 6.3 Tomcat
On Fri, Nov 28, 2014 at 3:02 PM, D R rir@gmail.com wrote: Hi, in our company we used the standalone version of XWiki for testing and are very satisfied with it so we want to migrate to the new version and a more robust setup. The setup is Windows Server 2003 R2 (sadly only 32 bits) with Apache 2.4, MySQL 5 and Tomcat 7. XWiki 6.3 is setup now including a Wiki Template so I'd like to migrate the data I exported as XAR from the 5.4.5 standalone instance. My approach was exporting data from 5.4.5, creating the Wikis using the template and importing the spaces and pages from the XAR export. The issue I'm facing is Java heap size which constantly gets in my way. As we're on a 32 bits system I can only assign ~1.2 GB of memory (the server has 4GB). This limitation makes it impossible to import the data into the new setup. I tried general export/import via administration page and the various options using the Admin Tools Application but on one of both ends I always get the out of memory error. - Using normal export/import: Export works but import throws the error - Admin Tools - Export pages and spaces: Export works but import throws the error - Admin Tools - Large Export on Disk: the heap size error is thrown Quite frankly the one in Admin Tools should change it's name since it's far from supporting anything close to large... Do you have any suggestions/hints on how I can migrate the spaces and pages I have in the 5.4.5 setup to the 6.3 one? The extensions that have the less memory footprint (probably by far) I can think of are http://extensions.xwiki.org/xwiki/bin/view/Extension/Large+Wiki+Export and http://extensions.xwiki.org/xwiki/bin/view/Extension/Large+XAR+Import so you might want to try them. Note that at least for export it's fixed in 6.2 but for import even if it's been improved a lot the way the UI work you still have to be able to attach your attachment. Thanks in advance, Dennis ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users -- Thomas Mortagne ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Getting user name with a Java Component
From: vinc...@massol.net [vinc...@massol.net] OK, so I should not use ExecutionContext at all? You’re right, our goal is to have all context information (including the user) be put in the Execution Context. However, currently, we’re in a transition phase, trying to move from the old XWikiContext to the new ExecutionContext and thus a lot of information is currently still held in the XWikiContext. See the details here: http://platform.xwiki.org/xwiki/bin/view/DevGuide/WritingComponents#HTheXWikicontext Thanks. I've actually read through that page several times, but it's only now making sense to me after having given things a try and getting the help on this mailling list. FWIW the problem for a new user is that there is a mix of documentation applying to several versions of XWiki (especially when searching via Google). This is partly the curse of wikis, since they don't go through structured revision control. It's so good that the mailling list is so responsive to put us straight. Thanks, Bryn ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Getting user name with a Java Component
From Marius: 3) Presumably the document reference captures the full location of the user's profile page. The DocumentReference API doesn't appear to have a page name accessor, so presumably to extract the user name I need to pull it out of userDoc.toString(). What information do you need precisely? The user alias (what the user uses to log in)? Or the user pretty name? The user alias is the name of the user profile document. bridge.getCurrentUserReference().getName() I think this is what I need to do. I have a separate database which uses proxy authentication so that queries give different results based on the user (as a form of ACL). So I need the uid for the current user, which is mapped to the wiki alias. Many thanks, Bryn ___ users mailing list users@xwiki.org http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Handling exceptions from components
Said Vincent:
and can this be caught within the Velocity or Groovy code that calls it?
errors are automatically caught in the script macros ({{velocity}},
{{groovy}}, etc).
OK, looks like Groovy has explicit try/catch
(http://groovy.codehaus.org/JN3035-Exceptions). I couldn't see anything about
Velocity catching exceptions (only about Velocity throwing exceptions -
http://velocity.apache.org/engine/devel/developer-guide.html#exceptions).
So if I have a component that I call from velocity with $mycomponent.foo() and
the method throws an exception (IOException, for instance), can I control what
the Velocity script does about this?
___
users mailing list
users@xwiki.org
http://lists.xwiki.org/mailman/listinfo/users
Re: [xwiki-users] Handling exceptions from components
On Fri, Nov 28, 2014 at 9:17 PM, Bryn Jeffries
bryn.jeffr...@sydney.edu.au wrote:
Said Vincent:
and can this be caught within the Velocity or Groovy code that calls it?
errors are automatically caught in the script macros ({{velocity}},
{{groovy}}, etc).
OK, looks like Groovy has explicit try/catch
(http://groovy.codehaus.org/JN3035-Exceptions). I couldn't see anything about
Velocity catching exceptions (only about Velocity throwing exceptions -
http://velocity.apache.org/engine/devel/developer-guide.html#exceptions).
Yes there is nothing in standard Velocity but XWiki add some tools to
Velocity and among other things some try/catching support, see
http://extensions.xwiki.org/xwiki/bin/view/Extension/Velocity+Module#HVelocityDirectives.
What Vincent meant by automatically caught is that you get to see the
error in the error where you have put the {{velocity}} macro, it does
not break the whole page.
So if I have a component that I call from velocity with $mycomponent.foo()
and the method throws an exception (IOException, for instance), can I control
what the Velocity script does about this?
___
users mailing list
users@xwiki.org
http://lists.xwiki.org/mailman/listinfo/users
--
Thomas Mortagne
___
users mailing list
users@xwiki.org
http://lists.xwiki.org/mailman/listinfo/users
