The European Library architecture is designed around a structure that allows users to obtain "search results" from european national libraries' online collections. The European Library portal manages and handles requests in different ways in order to deliver the "search result" and to adapt its structure to national libraries ones. Here are the different components of this architecture.

The European Library portal is the "dynamic" part of the website and includes also the libraries section. It is a software application written in Java. The static part of the website includes the organisation sections and the online exhibitions space. These different components of technical infrastructure are currently migrating to the University of London Computer Centre (ULCC), who will become The European Library' hosting partner.

The following diagram illustrates this technical structure:

The European Library's priority is to provide the most relevant results as quickly as possible to end-users. There are two key methods of providing integrated search:

• collect all metadata and store it in a central database(via OAI-PMH)
• send queries to individual off-site databases and collect the search result also in federated search (SRU) and Z39.50

Currently 3 ways of accessing the collections are available:

-via the OAI-PMH Central Index using the SRU protocol (central access)

-directly via the SRU protocol (federated search)

-via a central gateway using the SRU protocol (federated search)

The protocols OAI-PMH, SRU and Z39.50 provide access to metadata. Access to objects and related services is provided via the URL in the identifier field or via an OpenURL service. Searching is always via SRU in SRU enabled targets. The Central Index is accessible by SRU. For Z39.50 targets searching is done via the SRU to Z39.50 gateway.

With the move to ULCC, the current Central Index (Verity) will be replaced by the open source search server: SOLR (Apache Software Foundation) based on the Lucene Java search library. SOLR is a fast, full-featured and a full-text search library, which also permits the indexation.

SRU is a standard search and retrieval protocol that uses the Internet to carry queries between a user and a target database. This protocol uses CQL (Common Query Language), a standard query syntax for representing queries. CQL can hardle simple searching for one or more words, to complex Boolean expressions using different indexes. An SRU request consists of a SRU base-URL and associated parameters, such as the query, the start record, the maximum number of returned records and the record schema. The European Library currently uses version 1.1 of SRU.

The OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting) environment is a mechanism for data providers to expose their metadata and facilitate the efficient dissemination of content. It makes it possible to retrieve and centralise partners' metadata records and to reduce the average time of searching. OAI-PMH supports any metadata format encoded in XML, however Dublin Core is the minimal format specified for interoperability. Already, over 70% of The European Library collections are OAI-PMH compatible within The European Library environment. It is the favoured access protocol :

The OAI-PMH protocol defines two different entities:
• Data providers are entities who possess metadata and are willing to share this with others
• Service Providers, such as The European Library, are entities who harvest data from Data Providers in order to provide higher-level services to users (e.g. searching, browsing, etc.).

OAI-PMH harvesting is done through a set of commands, known as "verbs" (e.g. ListIdentifiers to get the identifiers of the last modified records and GetRecord to retrieve a metadata record from a repository) and parameters (e.g. set to specify a set and metadataPrefix to define a metadata namespace. Harvesting results are delivered in XML format.

The European Library intends to use a standalone OAI-PMH server called REPOX to manage incremental harvesting of the collections in the Central Index. Repox is developed by the Technical University of Lisbon. The main objective is to increase the regularity of collections' harvesting. Repox will allow The European Library to schedule the automatic harvesting of collections. Repox also provides a central metadata repository service that will be used by other functionalities of The European Library portal.

Metadata from OAI-PMH enabled collections is harvested and indexed in The European Library central index. The access to the full metadata records will be provided from the central metadata repository. These two components will be migrated to a new central index and a central metadata repository using SOLR.

The Z39.50 Gateway is a standard communication protocol for online search and retrieval of bibliographic data from remote computer databases. A SRU/Z39.50 Gateway converts a SRU search to a search in the Z39.50 protocol. Then the Z39.50 session will find and retrieve the requested records from a pre-specified Z39.50 target. The European Library runs an operational SRU/Z39.50-gateway, which The European Library refers to as the "central gateway" or "The European Library gateway". The Z39.50-SRU gateway allows the adoption of the SRU protocol by TEL partners whose library systems can only support Z39.50. This central gateway allows Z39.50 servers based in partner's libraries to communicate with The European Library portal.

In addition to this protocol conversion, the gateway can perform other conversions:

1. Character set conversions

2. Conversion of the record format provided by the Z39.50 target to The European Library Application profile

3. Addition of other data not included in the catalogue records, for example links to the digital objects

The gateway is needed by The European Library portal (the SEARCH part on http://search.theeuropeanLibrary.org) to access Z39.50 targets, because the portal only speaks SRU, so it needs a translator to be able to talk to Z39.50 databases, hence the translator = the gateway.