as a partner we are interested also in new Captiva 7.0 solution. I saw a few videos but firstly, before qualitative discussion, can someone provide here a quick link to new Captiva manuals? Unfortunatelly we didn't visit last EMC Momentum for some reasons. (i also searched on powerlink - but nothing)
Thank you in advance!
here are the questions:
1. Dispatcher for InputAccel is major part of pre 7.0 versions of InputAccel; however very "loosely coupled" (read "not very well integrated") with InputAccel. I haven't seen it mentioned in any of Captiva 7.0 announcements. Is it:
a) completely integrated into Captiva 7.0, or
b) is it still a separate subsystem inside Captiva system, or
c) something else...?
2. What is the upgrade procedure for existing Dispatcher projects (*.dpp) to version 7.0?
Is it possible to get detailed product documentation for Captiva 7.0?
Captiva 7 definitely improves the integration between Captiva Capture (formerly InputAccel) and Advanced Recognition (formerly Dispatcher).
Captiva 7 is the first phase for unifying the data models and recognition frameworks (zonal, free-form, full-page, etc.) The practical impacts are: 1) single place to configure options, 2) full exposure of features (e.g. no more of features being available in NuanceOCR module, but not in zonal OCR with Nuance), 3) re-use in all modules, whether used for full-page OCR, or rubberband OCR, and 4) one way for OCR extraction data to be attached to metadata. It is all inside a new rich object called the Document Type, which includes the index data, and is export ready. So now it's easier to pass data between Capture and Advanced Recognition modules. There is no value to value mapping. Only one IA value is passed from Extraction to Desktop to Export. Data can be flattened for use in legacy exporters.
Dispatcher Manager (known as Recognition Designer in Captiva 7) is an integrated component within Captiva Designer and is launched from there. If a customer chooses to continue using the Validation and Recognition modules, then they will continue to use Dispatcher Manager separately from Captiva Designer. If they want to migrate to use Captiva Desktop, then they can import their old DPPs into Captiva Designer, and work with Recognition projects instead.
Yes, you understand it correctly. In Captiva 7, Validation Edit is considered a deprecated module. Deprecated modules are fully supported, & available from installer. They include roll-up of bug fixes, they are fully tested, but don’t include enhancements. Deprecated modules will not be shipped in next major release, and documentation for deprecated modules are not be shipped or installed.
Other deprecated modules in Captiva 7 include Auto Annotate, File System Export, Image, Image Enhancement, Image Export, Image Quality Assurance, IndexPlus, Index Export, PDF Export, Spawn, Values to XML, Dispatcher Recognition, Dispatcher Validation, & Dispatcher Statistics Tool.
We recommend new users do not use deprecated modules, & existing users begin thinking of a migration path away from deprecated modules.
At the moment, some advanced IA projects that include advanced recognition consist of:
1. Captureflow designer project .xpp
2. Dispatcher project .dpp with VBA script libraries and OCR templates
3. Visual studio c# project for IA advanced scripting
4. IA process settings files
Now, I have two questions:
1. Is it possible to run this kind of project on Captiva 7 without modifications (i.e. with minimal effort)?
2. What does it take to migrate this kind of project to "Captiva 7 route" (i.e. use non-deprecated modules only). Is it possible/feasible to do, or is it better/easier to start from scratch (e.g. what does .dpp project convert to)?
In addition, is it possible to extend functionality of Captiva Desktop client using .net scripting like it is currently possible with IndexPlus and other quickmodulehost.exe modules?
With C7, there are essentially two types of DPP files. A Dispatcher Project is a DPP that works w/ Validation and Recognition, and a Recognition Project which is a DPP that works with the new Desktop and Extraction and uses the new data and document model. The new way uses Document Types. At design-time, a Doc Type is a document-level data entry form containing Field and control definitions, Layout, List of validation rules, Doc properties, etc. At run-time, the doc type object is a Level 1 IA value “string”, but is encoded with data.
When upgrading from a 6.x XPP (and not an IPP), and you simply want to use the existing XPP, DPP, and deprecated modules, you’d do the following:
- Check if xpp filename meets new name restrictions
- Copy the xpp into the target capture system folder\GlobalData\XPP.
- Open the CaptureFlow within Captiva Designer to make sure it is correct. Note any warnings and save it.
- Create installed processes for each instance used in the 6.x server
- Connect to server that has the 6.x process settings.
- For each installed process, download settings
- Save the CaptureFlow.
And that’s it. If you want to migrate to the new modules, in addition to the above steps, you’d open the xpp and replace deprecated modules within your captureflow with the new modules (e.g. replace Validation with Desktop, replace Recognition with Extraction, etc.). The new Captiva Designer supports importing of a Dispatcher Project, and will automatically convert the index families into document types, and link those document types to the new Recognition project. You’d then point Extraction to the new project. Lastly, you’d want to go configure the modules (in C7 you do that in Designer, and not Admin Console) to account for the new document object and any new functionality you wish to utilize.
With regard to scripting, Captiva 7 introduces Document Type scripting for different types of events (e.g. Global, Document, Validation Rule, UI, Form, and Table). It's used for mid-task data manipulation in Desktop and Extraction. You can use C# or VB.NET. The scripts are separate from the Captureflow for reuse. One DLL contains all doc types, not one DLL per doc type. This simplifies deployment and runtime complexity.
If you want to use the existing IPP w/out changes to the IPP or to modules or settings:
- Use Process Designer to update IPP, compile and install to the Captiva Server
- New module types can added to IPP
- Use Captiva Designer to prepare profiles & deploy to Server. If profiles are not used, this step can be skipped
- Use Captiva Administrator to configure process steps
- Install needed files onto module servers, if not already there:
- .NET code module assemblies
- Install CSS assemblies using step setup.
- Create batch
- Use Process Designer to attach to a batch and debug process flow
If you wish to migrate your IPP to an XPP to use w/ Captiva 7 Designer:
- Benefits of migrating to XPP:
- More maintainable
- Easier to understand
- Easier make changes
- Deployment support
- Configure within the same Designer tool
- Source files ready for source control
- Re-design process completely with Captiva Captureflow Designer
- In many cases, complex IPP routing code is no longer necessary (longer term: will save you maintenance costs)
- Freedom to use new modules and features
- Keep existing modules and duplicate routing logic in XPP
- Re-use existing steps and configurations
- Done by constructing a new XPP with same step names and downloading the settings
Migrating IPP scripts to Captiva Designer scripts:
- CFD scripts:
- Prepare and Finish routines
- Uses VB.net or C#
- .NET 3.5
- Unsupported routines:
- Error: CFD provides ErrorCode that can be used in Finish routine
- Tree PostNodeAdd, PostNodeMove: IA values SubTreeModified and TreeNodeModified can be used in Finish routine
- If other routines were used, process may need to be redesigned, or continue to use IPP
Regarding ODBC Export, there were no changes in C7.
- Use Process Designer to update IPP, compile and install to the Captiva Server
Check out this data sheet and let me know if you have additional questions:
eInput 2.2 is compatible w/ Captiva 7.