|
JRA Software, Publisher
SearchPDF Product Architecture
Introduction
SearchPDF 2.0 is an advanced ASP Application that requires Microsoft IIS on a Windows web server, and the dtSearch Text Retrieval Engine, available from dtSearch Corp. It runs on Windows NT, Windows 2000 and Windows XP servers. Programmers and developers who are familar with ASP coding can easily create custom implementations of SearchPDF to suit any requirements. We give you three application models for starters as you create your own tailored interfaces. JRA can also create custom interfaces for you (see Services), so even without any ASP experience we can quickly get you up and running.
Most configuration work for your custom search implementation can be accomplished with parameter management in the global.asa file. The high-level portion of the ASP code is left open for further customizing if desired. The low-level (generic) portion of the ASP code for this application is compiled in a DLL file and this results in a big performance gain.
SearchPDF 2.0 also features a search navigation toolbar written in Javascript, and this code can be customized if desired. For example, if you decide to integrate with ePublishing, you can add a "Buy Now" button on the toolbar.
The open design of this application gives you the ability to rapidly deploy searchable collections using standard interface models, and to extend these interfaces for your custom requirements. You get a perfect fit every time, for every document collection.
dtSearch Text Retrieval Engine
|
|
The SearchPDF 2.0 product works with a copy of the dtSearch Text Retrieval Engine. SearchPDF is an advanced application that has been developed "on top of" the dtSearch engine.
|
dtSearch Corp. has a decade of experience in text search and retrieval. Hundreds of commercial applications currently incorporate the dtSearch Text Retrieval Engine. Refer to the dtSearch case studies page here. Read the specific case study review of SearchPDF (PDF WebSearch) integration with dtSearch here.
SearchPDF uses both the dtSearch Text Retrieval Engine and the dtSearch Indexer application. This dtSearch Indexer application allows you to generate a search-index for your document collection that is then opened and used by SearchPDF 2.0.
The dtSearch Engine is written in C++ and is implemented as a DLL file, dten600.dll, and supporting library files.
SearchPDF Application Code
The SearchPDF application is implemented as a set of ASP scripts, with an integrated navigation toolbar written entirely in Javascript. The benefit of Javascript (as opposed to Java Applets) is that no special plugin or code download is required on the client computer. SearchPDF also features its own fast-performing recordset for management of search results.
Much of the SearchPDF ASP application code has been compiled (binarized) into a DLL file (searchpdf.dll) that is called by a set of 9 high level ASP configuration files. This greatly simplifies customization, as there is then far less customizable code to work with.
SearchPDF Building Block Diagram:
SearchPDF configurable ASP application scripts, and Javascript
|
SearchPDF binarized, generic application scripts (searchpdf.dll)
|
dtSearch Text Retrieval Engine (dten600.dll)
|
Microsoft Internet Information Server (IIS)
|
Windows Operating System (NT, 2000, XP)
|
Why Not Just use dtSearch Web?
dtSearch Web is a component product offering of dtSearch Corp., that generates a sample web interface for dtSearch indexes. The dtSearch Engine includes source code to the sample web interface, allowing it to be customized or extended.
As a sample or quick interface (because it can be generated), dtSearch Web is fine, but it is not a professional interface such as is offered by bigger (and much more expensive) search product vendors. It is not intended to be.
This is where SearchPDF steps in. SearchPDF not only provides a professional interface that can be customized, it also nearly doubles the features of dtSearch Engine, with features like search refinement, search results re-sorting, advanced date field management and metadata display.
Significantly, the SearchPDF design allows for almost the entire page (browser viewing area) to be used for document viewing, instead of "wasting" screen real-estate for the side-by-side display of search results. Similarly, instead of squeezing search results into a left-frame, it can be displayed in SearchPDF in a full-screen recordset.
As a final consideration, SearchPDF is the ONLY commercial search product to support the DjVu file format. You won't find this support in dtSearch Web or in other dtSearch products.
High-Level Customizing Interface
We hide the complexity of a sophisticated web application, giving you high-level configuration ASP scripts with which to quickly customize your interface.
Looking ahead to SearchPDF 3.0
In the planed upgrade to SearchPDF, we will be adding two new layers to make it easier to customize:
SearchPDF 3.0 Building Block Diagram:
Web-Browser-Based Application Adminstration Interface
|
XML Configuration File (web.config) for .NET Framework
|
SearchPDF configurable ASP application scripts, and Javascript
|
SearchPDF binarized, generic application scripts (searchpdf.dll)
|
dtSearch Text Retrieval Engine (dten600.dll)
|
Microsoft Internet Information Server (IIS)
|
Windows Operating System (NT, 2000, XP)
|
With the upgrading of ASP script to ASP.NET files, we will also migrate all settings and variables to the web.config file, where they will be stored in XML format. A set of web pages will provide the interface to update the XML-based web.config file, making it simple to modify and customize the interface.
Comparison of dtSearch Web and SearchPDF interfaces
dtSearch Web Generated Search Interface (fixed)
Sample SearchPDF Interface (including fields) (customized for each project)
dtSearch Web Search Results Interface (left frame)
SearchPDF Search Results Interface (full page) (project fields displayed)
|