3.5

ABBYY Recognition Server 3.5 System Administrator’s Guide
38
4. If necessary, change the temporary folders for source images and resulting files. Images to be
recognized are copied to the Input folder. If crawling and recognition are not carried out
simultaneously, images will stack up in the Input folder. The Output folder contains the
recognition results and the images that couldn’t be recognized. Use folders on the hard drive
with enough disk space for Input and Output folders.
Important! For the integration of ABBYY Recognition Server with the Google Search Appliance to work correctly, the
following conditions must be met:
• indexing of graphical file formats must be allowed in the Google Search Appliance settings;
• reception of XML feeds from the IP address of the Server Manager computer must be allowed in the
Google Search Appliance settings;
• viewing of URL addresses assigned to indexed documents (smb://machine.domain.com/path) must be
allowed in the Google Search Appliance settings.
In case any of these conditions is not met, the XML feed submitted by the Connector will be ignored by the Google
Search Appliance without notification.
Notes:
• The Connector stores information about crawled files and folders in the FeedGeneratorState.db
database file in the <ABBYY Recognition Server 3.5 Installation
Folder>\Bin\RecognitionServer3Temp folder. Information about the date of the last modification and
indexing is available for each indexed file. When crawling the folders, the Connector checks this
information in the database for every found file of the specified format. If the file has been created or
modified since the last indexing, it will be sent to ABBYY Recognition Server for recognition. The files
that haven’t changed since the last indexing will not be re-recognized.
• Indexing of XML feeds by Google Search Appliance may be performed with delay. Thus, it may take
some time (up to a few hours) for the data from the XML feeds to appear in the index.
Information about functioning of the Google Search Appliance Connector is registered in a special log (the
FeedGeneratorLog.txt file in the Bin subfolder of the ABBYY Recognition Server 3.5 installation folder). The log contains
information about:
• the Connector start and stop occurrences;
• files sent to ABBYY Recognition Server for recognition;
• successfully recognized files with contents sent to the Google Search Appliance for indexing;
• errors of the Connector.
Information about critical errors of the Connector is also written to the Server Log.
Setting Up Microsoft Search IFilter
ABBYY Recognition Server 3.5 incorporates the Microsoft Search IFilter component which allows the program to
interact with the following search engines developed by Microsoft: Microsoft Search Server, Microsoft Office SharePoint
Server and Microsoft Windows Search.
The search engine by Microsoft indexes documents in specified folders. If these folders contain documents in graphical
formats to be indexed, they are sent to ABBYY Recognition Server through the IFilter. ABBYY Recognition Server
recognizes documents within a special hidden workflow and exports the results as text. When the recognition results
appear, IFilter returns them to the Microsoft engine for indexing. After that the documents become searchable.
The IFilter is based on the ABBYY Recognition Server 3.5 IFilter Backend service (Start > Control Panel >
Administrative Tools > Services > ABBYY Recognition Server 3.5 IFilter Backend) and performs two main
functions:
1. Receives image files from the Microsoft search system crawler and sends them to ABBYY Recognition Server for
OCR;
2. Returns the recognized text to the Microsoft search system for indexing.