Indexing

Before artifacts can be searched in Archiva, they need to be indexed first. Aritfacts are indexed during these instances:

• when a repository is scanned
• when an artifact is deployed or uploaded in the repository
• when an artifact is proxied or downloaded from a remote repository
• when remote index is available and downloaded

During repository scanning, the index-content consumer (see Consumers) creates an indexing task and queues it in an indexing task queue. This indexing task is in turn executed by a task executor and that's where the actual indexing happens. A similar indexing task is also created and queued as artifacts are deployed in the Archiva repository or proxied from a remote repository. The reason behind this is to avoid multiple or simultaneous update of the index (a repository scan might be executing while artifacts are being deployed in the repository). Note: This behavior applies for versions starting 1.2.1.

What Can Be Searched?

The repository files that are indexed during repository scan can be configured in the Repository Scanning page, in the Repository Scanning - File Types section (indexable-content). By default, the indexable-content file types are:

**/*.txt
**/*.TXT
**/*.block
**/*.config
**/*.pom
**/*.xml
**/*.xsd
**/*.dtd
**/*.tld

Quick Search

The contents of all the files in the repository that matches the above configuration will be indexed during repository scan. So when you type a word in the search box, Archiva will search the index for that query string and return all the search hits in the form of a paginated list of artifacts. For example you typed in 'ant', the search result would contain the list of artifacts that use or reference 'ant'.

For query strings containing whitespaces, Archiva searches for the query string including the whitespaces. So if you typed in 'commons logging', the search result would contain the list of artifacts that reference 'commons logging' which is still separated by whitespace. In this case, 'commons-logging' (hyphen separated) is not included in the search results.

Archiva can also search the index for java classes, packages, and methods and return a list of artifacts containing the query string. In the previous versions, we need to append the keyword 'bytecode:' before the query string in the search box (like 'bytecode:myMethod'). But since version 1.2, we no longer need to type the keyword. Just type the java class or package name in the search box.. and voila!

You can use various criterias to search artifacts:

Simple Query string

Advanced Paramters

OSGI Parameters

Search results

You can apply various filters on the search results.

Below is a sample screenshot of the search results:

When the user clicks a specific artifact in the search results list, Archiva will direct the user in the Artifact Info page (the same page from the Repository Browse). Otherwise if it is not a pom file (ex. metadata.xml), the file will appear as a file to be downloaded from the browser.

Security

The Search functionality respects the repository permissions. If the user is not logged in to Archiva and does a search, only those repositories with guest permissions will be searched for the query string. And if the user is logged in to Archiva and does a search, only those repositories in which that user has permission to access will be searched for the query string.