echocat / adam Goto Github PK

Could you please implement new field types single select (radio button array) and multi select (check box array), where admin could specify values and user could pick them? Values would support wiki mark-up.

Example of single select: Employee office - only one value from given options.
Example of multi select: Employee projects - employee can participate on multiple projects at the time.

Thank you!

The manager field in the directory stores the DN of the users manager. The system could resolve this to a record using ".findEntityByDN()" and then allow the attributes of the user to be mapped to a field.

As an example use case:
LDAP Entry:
manager: cn=jsmith,ou=users,dc=example,dc=com
Mapped Entry:
Manager: John Smith (with link to /display/~jsmith in Confluence)

If this were a generic field type it would work for other attributes like "secretary" or possible custom DN based attributes.

When the directory has a large number of users, rendering the report can produce a very large document and take longer to generate.

Optional/configurable pagination would allow the report to be delivered in more manageable chunks.

After installing the initial config appears in one line.
Confluence version 5.4.3

2014-03-25 19:21:15,692 WARN [http-8090-1] [echocat.adam.synchronization.LdapDirectorySynchronizationRequiredEventListener] handleLogin The user 'xxxxxxx could not be synchronized. This means that this user could be out of date.
org.springframework.beans.factory.UnsatisfiedDependencyException: Error creating bean with name 'com.atlassian.crowd.directory.OpenLDAP$$Extension$$2': Unsatisfied dependency expressed through constructor argument with index 0 of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater]: : No unique bean of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater] is defined: expected single matching bean but found 2: [crowdLdapQueryTranslator, activeDirectoryQueryTranslater]; nested exception is org.springframework.beans.factory.NoSuchBeanDefinitionException: No unique bean of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater] is defined: expected single matching bean but found 2: [crowdLdapQueryTranslator, activeDirectoryQueryTranslater]

Full Stack Trace:

org.springframework.beans.factory.UnsatisfiedDependencyException: Error creating bean with name 'com.atlassian.crowd.directory.OpenLDAP$$Extension$$2': Unsatisfied dependency expressed through constructor argument with index 0 of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater]: : No unique bean of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater] is defined: expected single matching bean but found 2: [crowdLdapQueryTranslator, activeDirectoryQueryTranslater]; nested exception is org.springframework.beans.factory.NoSuchBeanDefinitionException: No unique bean of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater] is defined: expected single matching bean but found 2: [crowdLdapQueryTranslator, activeDirectoryQueryTranslater]
    at org.springframework.beans.factory.support.ConstructorResolver.createArgumentArray(ConstructorResolver.java:591)
    at org.springframework.beans.factory.support.ConstructorResolver.autowireConstructor(ConstructorResolver.java:193)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory.autowireConstructor(AbstractAutowireCapableBeanFactory.java:925)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory.createBeanInstance(AbstractAutowireCapableBeanFactory.java:835)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory.doCreateBean(AbstractAutowireCapableBeanFactory.java:440)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory$1.run(AbstractAutowireCapableBeanFactory.java:409)
    at java.security.AccessController.doPrivileged(Native Method)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory.createBean(AbstractAutowireCapableBeanFactory.java:380)
    at org.springframework.beans.factory.support.AbstractAutowireCapableBeanFactory.createBean(AbstractAutowireCapableBeanFactory.java:300)
    at com.atlassian.confluence.user.crowd.ConfluenceSpringContextInstanceFactory.getInstance(ConfluenceSpringContextInstanceFactory.java:52)
    at com.atlassian.confluence.user.crowd.ConfluenceSpringContextInstanceFactory.getInstance(ConfluenceSpringContextInstanceFactory.java:43)
    at org.echocat.adam.directory.ExtendingLDAPDirectoryInstanceLoader.newRemoteDirectory(ExtendingLDAPDirectoryInstanceLoader.java:90)
    at org.echocat.adam.directory.ExtendingLDAPDirectoryInstanceLoader.getRawDirectory(ExtendingLDAPDirectoryInstanceLoader.java:77)
    at org.echocat.adam.directory.ExtendingLDAPDirectoryInstanceLoader.getDirectory(ExtendingLDAPDirectoryInstanceLoader.java:83)
    at org.echocat.adam.synchronization.LdapDirectorySynchronizer.findRemoteDirectoryOf(LdapDirectorySynchronizer.java:267)
    at org.echocat.adam.synchronization.LdapDirectorySynchronizer$3.nextIterator(LdapDirectorySynchronizer.java:170)
    at org.echocat.adam.synchronization.LdapDirectorySynchronizer$3.nextIterator(LdapDirectorySynchronizer.java:166)
    at org.echocat.jomon.runtime.iterators.ChainedIterator.hasNext(ChainedIterator.java:51)
    at org.echocat.jomon.runtime.iterators.ChainedIterator.hasNext(ChainedIterator.java:48)
    at org.echocat.jomon.runtime.iterators.IteratorUtils$4.computeNext(IteratorUtils.java:141)
    at com.google.common.collect.AbstractIterator.tryToComputeNext(AbstractIterator.java:141)
    at com.google.common.collect.AbstractIterator.hasNext(AbstractIterator.java:136)
    at org.echocat.jomon.runtime.iterators.IteratorUtils$3.hasNext(IteratorUtils.java:109)
    at org.echocat.adam.synchronization.LdapDirectorySynchronizer.synchronize(LdapDirectorySynchronizer.java:215)
    at org.echocat.adam.synchronization.LdapDirectorySynchronizer.synchronize(LdapDirectorySynchronizer.java:199)
    at org.echocat.adam.administration.AdministrationAction$1.run(AdministrationAction.java:81)
Caused by: org.springframework.beans.factory.NoSuchBeanDefinitionException: No unique bean of type [com.atlassian.crowd.search.ldap.LDAPQueryTranslater] is defined: expected single matching bean but found 2: [crowdLdapQueryTranslator, activeDirectoryQueryTranslater]
    at org.springframework.beans.factory.support.DefaultListableBeanFactory.resolveDependency(DefaultListableBeanFactory.java:621)
    at org.springframework.beans.factory.support.ConstructorResolver.resolveAutowiredArgument(ConstructorResolver.java:622)
    at org.springframework.beans.factory.support.ConstructorResolver.createArgumentArray(ConstructorResolver.java:584)
    ... 25 more

When viewing a profile on the desktop, the fields shown are based on the view that is configured. However, when viewing on a mobile device, all fields are shown.

On the "All Users" report when you hover over an entry you see a popup like:

This gives you access to easily follow the user or jump to their personal space, etc.

It would be nice to be able to have this popup when using the custom reports as well

In the new views that were introduced in #20 the order of the elements/fields is dictated by the <profile> rather than the <view>

This presents a slight challenge: on one hand the <profile> allows for groups of elements to separate information logically. In this case it makes sense for <profile> to drive the ordering. However, for something like the "compact" view used for the hover profiles, it might be useful to be able to set the order in that view.

For example, in the sample configuration if we look at the fields Name, Email and Position. For the profile the order is Name, Email, grouped in Personal and Position grouped in Business. When we create a view for the hover, if we want to use the order: Name, Position, Email, this is not possible currently as the order is defined globally at <profile> and the <view> only controls visibility.

It may be worth considering changing <profile> to simply be the data definition only and <view> become the presentation definition. Thus we would move groups out of <profile> and into <view>

Is this real required? It is a hard work to do this and keep it in sync with the features of Adam. Could we expect that an administrator is able to configure XML (once #16 documentation is done) or it is really required?

As mentioned in #18 the changes made to allow configuration of the user popup doesn't look as clean as the original Confluence implementation.

Two suggestions:

1. Remove the line break and wrap the fields around the profile image to avoid the large block of white space
2. Remove the field titles as this is a small popup badge and the information presented within it should be obvious without labels (i.e. Name, Email, Phone, Title)

A marco that provides the same content like the people directory.

It also have to include the ability to provide a search filter.

For the first version we do not want to hide same columns from the reports. We only want to provide to select a report to be displayed.

People search does not return any results.

1. Go people
2. Click Search box
3. Try to filter some user - nothing is returned

With disabled plugin search works as expected.

When there is not enough space for User Profile macro, then macro start behaving weird

• overlapping itself
• overflows into wrong area

The Confluence UserDetailsManager that stores the rendered profile fields is limited to 255 characters.

This results in truncated data if the directory contains more (or if the template results in more) than 255 characters.

Our groups are formed using colons to join the group path. Example:

employees:all
employees:us:all
employees:us:california

When attempting to use these groups in the filter an error is raised indicating a colon is not a valid character.
Example:

<includingGroup id="employees:all" />

Currently it is not possible to set order of elements for each view independently. Sort order reflect order in which are elements defined. Ideally order of elements within the View in XML.

Thanks!

Currently the synchronized attributes are stored in Bandana which does not perform well as the amount of data stored grows. Atlassian's recommendation is to use Active Objects as it will perform better with larger datasets.

Unless I'm missing something, I can't find anything to configure inside the xml to even connect to my OpenLDAP server.

Is there a way to display the total number of employees at the bottom of the page? Or add a column that starts with 1 and autoincrements?

Add a HTML element type so that the velocity template can return any HTML content without escaping. This will allow more customized formatting without the constraints of wiki markup. Examples include, image tags, javascript, and custom links.

Given the templates can only be edited by admins, this should not present a security concern.

When renaming a Internal User, we always run in to NullPointerException

Below's the error from the log file:

2015-02-12 08:46:32,722 DEBUG [ajp-bio-8009-exec-7] [net.sf.hibernate.SQL] log select confluence0_.user_key as user_key, confluence0_.username as username, confluence0_.lower_username as lower_us3_ from user_mapping confluence0_ where (confluence0_.lower_username=? )
2015-02-12 08:46:32,725 TRACE [ajp-bio-8009-exec-7] [sf.hibernate.type.StringType] nullSafeSet binding 'jrajchert' to parameter: 1
2015-02-12 08:46:32,731 ERROR [ajp-bio-8009-exec-7] [atlassian.confluence.servlet.ConfluenceServletDispatcher] sendError Could not execute action
 -- referer: https://confluence.artin.cz/admin/users/edituser.action?username=jrajchert | url: /admin/users/doedituser.action | userName: sysadmin
java.lang.NullPointerException

This issue cannot be replicated if we disabled the echocat Adam plugin

We noticed that the user profile's UI/field was different from the original design, could it be the cause?

Thank you

Regards,

Wayne Wong

Currently the reports allow free text search via the UI. It would also be useful to allow field specific filters for dynamic reports.

A few examples:

browsepeople.action?report=employees&department=QA

browsepeople.action?report=employees&position=Engineer

browsepeople.action?report=employees&position=*Manager

which would match "Product Manager", "Engineering Manager" etc...

reading and poking around brought me to a conclusion that Adam does not yet support Confluence instances hooked into Crowd or with multiple directories (internal + Crowd).

please consider adding Crowd support

I found the problem.

• User is not find by user search
• View / edit permissions based on group do not work
• User is not within the group

until his profile is resaved. For example when I will make report showing all employees, based on group employees (default settings), such report is empty even if there are users within the group. When I will resave the profile (no need to make change, just start edit and then save) then the user is finally shown in that report. It have to be made after each change. When there is user appearing corrently on report Employees and i will add him to admin group, he will not be shown on report based on admin group until his profile is resaved again.

Tested on Confluence 5.5.6 and 5.6.1 with internal directory or Crowd directory.

Could you please add new type of element for easy input if users. I would like to use it local Manager field. Currently i can use WikiMarkup field with syntax [~user], but it is not something what I would like to give to end-users.

It is possible to add disabled users to Reports? They are excluded from that by default. For example as an parameter of the Report?

Example

<report id="all" showActiveUsers="ture" showDisabledUsers="ture">
        <localization locale="en_US" title="All users" />
        <localization locale="de_DE" title="Alle Nutzer" />
        <access>
            <anonymous view="forbidden" />
            <default view="allowed" />
        </access>
 </report>

Then admin could define reports of Active, Disabled or both. Such overview might come handy when we want to someone re-hire or contact him from any other reason. Also brings some insight into fluctuation.

Then it would be great if such info could be present as an profile element. When someone is reviewing user profile it is not obvious if the user is active - another element could clarify it.

When i click user mentioned on Confluence page or shown as an author I can see this:

No idea if the user is active or not. Should I contact him?

The following exception occurs when Adam tries to sync:

Exception in thread "LDAP directory synchronization job" java.lang.LinkageError: loader constraint violation: when resolving field "LOG" the class loader (instance of org/apache/felix/framework/ModuleImpl$ModuleClassLoader) of the referring class, com/opensymphony/xwork/ActionSupport, and the class loader (instance of org/apache/catalina/loader/WebappClassLoader) for the field's resolved type, port, have different Class objects for that type
    at org.echocat.adam.administration.AdministrationAction.access$100(AdministrationAction.java:37)
    at org.echocat.adam.administration.AdministrationAction$1.run(AdministrationAction.java:79)

Provide the ability to sort the columns in the following ways:

1. Allow the default sort to be selected in the configuration
2. Allow the end user to dynamically re-sort the report by clicking on the column headers.

Currently when including the user profile macro on a page, all fields are shown. It may be desirable for the user to display only a subset of fields on the macro and the rest be accessible from the profile page itself.

The visibility could be configured by a property as with #18

My original intent with #18 was to show the popup when hovering over the links on the custom Adam report pages (i.e. "Employees"). Right now the popup is only visible on the default "All users" report.

The thumbnailPhoto field in AD has been used by Outlook 2010 onwards, and we already had it uploaded in AD.

It would really great to have this.

Gary

There are issues with the latest release.

Provide support for mapping fields to multi-valued attributes in LDAP.

For example if a user has two "telephoneNumber" attributes you could show the following on their profile page:

Phone: xxxxx
Phone: yyyyy

Provide the way to export a report as CSV

Provide a way to export the reports as PDF.

If this is a desirable effect, would it be possible to add an option to configure it as it removes the ability to get to certain Confluence features (Status Updates, Profile Picture, etc)

Before Enabling:

After Enabling:

The latest code from master (0.2.0) has been tested as working against 5.5.x. It should be released and marked as such so users can update with confidence.

In public places is compact view just fine. In administration it is deformed. Tester of confluence 5.5.6 and 5.6.1 (both unsupported I know)

When viewing a user profile on the Confluence mobile interface, only the original builtin fields are shown (Position, Department, Location, About Me).

Are there hooks into the mobile UI to allow rendering of custom fields as well?

It is somehow possible show (system) username in vews or reports? It might be handy because as username could be used in other systems (i.e. Version control system)

Give a user and admin the chance to edit fields in confluence (example mobile phone) and sync it back to the LDAP directory. In this case the administrators and users have a easy user interface to edit their data.

It is possible add column with (resized) avatars into table report?

When using the "Edit Profile" UI in Confluence, The "IM" field cannot be changed while Adam is enabled. You can modify the text but after clicking save it reverts to its previous value.

By comparison changing the "About Me" text works fine.

When installed on an instance that also contains Confluence Questions the profile overwrites the Questions "Expertise" panel shown at the top right side of the profile page.

Is it possible to inject the custom profile without removing the "Expertise" panel? There must be some way to add panels to the profile page given that the "Expertise" panel itself was injected by the Confluence Questions plugin.

When clicking next or on an explicit page number under "All users" the user is directed to the following url:
/browsepeople.action&startIndex=50

This produces a "Page Not Found" error:

We will start this issue if the feature set are stable.