When a search is started, it is sent to the
archive server for processing. The results are returned as a list in
the web client.
The search results can be set in various ways,
usually presented in 3 columns: thumbnail image of the page with the
article or pages in case of a spread, abstract of the article text,
and the photos. The search word is highlighted in yellow.
archive search usually generates many hits. By clicking on the
article name, the text is displayed, so a particular article can be
found quickly. Search result lists can be sorted by date, page,
author, etc. Text content can easily be copied and pasted into Scoop
articles or other programs.
When an article is activated from a
search result, you can see on which page it has been placed. A
simple command can display the whole page.
• PDF Image of entire page displayed where the searched article
is highlighted in red
• The searched article displayed as text,
and PDF with full graphics
• Images are linked with captions and
• Search results can be limited to just headers,
etc at your choice
• Search result always displays date,
edition, product and page
We tested search in
a database of 30,000+ articles, where the search took less than one
second on a server with a 1.6 GHz P4 processor.
When the page editor has finished a Quark or
InDesign page daily for production, eLibrary automatically generates
page and article images, indexes everything accordingly, and adds to
the exiting archive.
The process is automatic, and is normally
executed by clicking a button. But the page editor has full control
over which parts of the page to archive, keeping relevant elements
together. The page editor does this, having just created the page,
since nobody else knows that page better.
When the page editor
sends the page for archiving, our eXporter add-on converts the
articles to XML tagged documents, based on the typographic style
sheet formatting used in the page document. This module is the same
as in Scoop ePaper, making it very easy to run eLibrary and ePaper
The XML documents are finally indexed to the SQL
database. Indexing takes place at a speed of about 300,000 articles
per hour, which is more than enough. The index process can be
started manually, or automatically, a few hours after the latest
production. During indexing, searches can still be performed.
• Smart information retrieval. High
quality tagging with minimal manual effort.
de-selection of unwanted material.
• Articles are archived for
each edition, creating multiple versions if they are changed between
• Generating structured standard XML, useful for web
publishing or open-source configurations.
• The generated XML
documents are IPTC compatible, enabling export to external archive
systems, if needed.
Only an authorised editor or
system administrator may change or delete articles in the archive.
This can be done by deleting complete time intervals, deleting
individual articles, or editing article content or the articles
category and search tags.
Back-up and security
automatic backup system in eLibrary is designed for maximum
security. A special routine for reconstructing the indexes helps