Sites
Each event belongs to a single site. A site does not have to be a web site, but can be any logical grouping of events that is suitable for the event data.
|
For the purposes of linguistic processing, content belonging to a site is assumed to be mono-lingual. |
Site groups
A site group is a combination of multiple sites. Each site can be a member of one or more site groups. Site groups are used in both the API and GUI to allow generation of reports based on events from all sites in the group.
Permissions
There are no permissions on a site itself, all permissions are inherited from the site groups the site is a member of. If the user has access to multiple site groups containing the site, the permissions will be combined. Site groups can be shared by giving permissions to other users.
In order to add an existing site to a site group, the user will need write access to both the site and site group before performing the operation. In practice this means the user must have write access to another site group containing the site. To remove a site form a site group, the user will only need write access to that site group. After a site is removed from a site group, the user removing the site might no longer have access to it if the user does not have access to another site group containing the site.
|
The site can not be removed from the site group it was added to when it was created. This limitation might be removed in the future. |
URL and domain ownership
During content processing ownership of each URL is found using the site ids in the Insight Javascript snippet included on the page. If such a script is found, the URL is deemed to belong to the given site. Otherwise we attempt to find the site by looking at site ids for documents further up in the path, including index.html and similar. Finally we will look for the site id of the top private domain directory index. The same procedure is used when verifying permissions to crawl settings and URL normalization parameters.