User clients
In this chapter, we provide an overview of the clients that are available for IBM Content Manager OnDemand (Content Manager OnDemand), including the various web client offerings that are based on the Content Manager OnDemand Web Enablement Kit (ODWEK). We describe the differences between web and Windows clients and their viewing options.
In the later sections, we focus on the integration and application programming interface (API) client options of Content Manager OnDemand, such as the ODWEK API, the Content Management Interoperability Services (CMIS) web services, the mid-server SAPI, and integration with other IBM Enterprise Content Manager products, such as IBM Information Integrator and IBM FileNet P8. We describe how to use the existing API to build your own web client interface for Content Manager OnDemand.
In this chapter, we cover the following topics:
8.1 Choosing the correct client for your implementation
Customers are faced with challenges in choosing the interface to Content Manager OnDemand that makes the most sense for their implementation. Content Manager OnDemand has many different user interfaces. Many aspects come into play when you consider the best design for access to Content Manager OnDemand to meet all of your requirements in the most cost-effective manner. Licensing costs, hardware costs, performance, and maintainability are just a few considerations, but the most important requirement is meeting the business needs for many different user types.
The Content Manager OnDemand Client choices enable the product to meet the ever-changing world of information technology and the way content is delivered. For example, delivering documents that are stored in Content Manager OnDemand to a mobile device was not relevant a few years ago. However, it is an important consideration for enterprise content delivery today. Technology drives change with current Content Manager OnDemand customers, and IBM delivers options to meet current and future business requirements. A customer’s goal is to use a single user interface for access to all of its Enterprise Content Management content. IBM met that goal with the IBM Content Navigator user interface, but IBM continues to retain multiple Content Manager OnDemand Client interfaces to meet the various needs of its customers.
When you choose the correct client for your implementation of Content Manager OnDemand, two primary considerations are the client functionality and the client architecture.
Concerning the client functionality, the most powerful client is the Microsoft Windows client. All other clients contain only a subset of the features of the Windows client. The most prominent difference is the viewer capability.
Determine whether your users require functionality that is specific to the Windows client only. If not, see the range of viewer options that are described in 8.1.1, “Viewer options” on page 186, which compares the different viewers across the various client options.
8.1.1 Viewer options
Different viewer options for the data that is stored within a Content Manager OnDemand system exist. The following general types of viewers are available:
•The viewing capabilities that are provided by the Windows client.
•The web viewers that are shipped with ODWEK.
•Generic web viewers that are available in Content Navigator or other third-party web viewers. The built-in viewers of Content Navigator are described in 8.2.1, “IBM Content Navigator” on page 193.
•Conversion and transformation services that are started by ODWEK.
•External applications that are opened according to their associated document types (for example, Microsoft Word for .doc or .docx files).
•Special client applications, such as the CICS client, the Structured APIs, or Java API access.
The content that is displayed by certain viewers can be changed by either transforms (ODWEK) or exits. For more information about exits, see Chapter 11, “Exits” on page 241.
Windows client viewers
The Content Manager OnDemand Windows client contains native capabilities for viewing typical archive data types:
•Line Data and SCS
•AFP
•Images
The Windows client reflects the richest set of capabilities in terms of viewing these data types. Because it directly communicates with the Content Manager OnDemand server, we reference the Windows client for all of its features that relate to document display.
The Line Data viewer of the Windows client is the most sophisticated viewer that is available for Content Manager OnDemand from the selection of readily available viewers.
The viewing of these primary data types happens within the same application. The Windows client provides other features, such as thumbnails, and configurable and saveable views.
The Content Manager OnDemand Windows client also contains other capabilities for viewing archive data types, such as Portable Document Format (PDF) and User-Defined.
Starting with Content Manager OnDemand version 9.5, for both DocType=PDF and user-defined PDF, the Windows Client will attempt to view a PDF document with Adobe Acrobat, if it is installed. If Adobe Acrobat is not installed, for DocType=PDF, Adobe Acrobat Reader will be used instead when the PDF document is viewed.
Before Content Manager OnDemand version 9.5, PDF documents can be viewed by the Windows client in two ways:
•If they are configured in the application as data type “PDF”, the rich feature set of the AFP and Line Data viewer applies, but Adobe Acrobat Professional is required.
•If the data type is configured as “User Defined” and “.pdf” as the extension, the documents are started externally. Therefore, you can view the documents with the no-charge Adobe Acrobat viewer or any other installed PDF viewer.
Any data type can be specified as “User Defined”, for example, Word documents (.docx). User-defined data is viewed by invoking its associated application.
Web-based viewing options
The web-based viewing options for Content Manager OnDemand are provided primarily by ODWEK. ODWEK includes different viewers that are dedicated to Content Manager OnDemand documents that can use Content Manager OnDemand functions, such as the segment-wise retrieval of large objects or annotations. These viewers are used in web applications, such as Content Navigator or any other custom-developed web client:
•Line Data applet
•Browser plug-in for image viewing
•AFP browser plug-in
•AFP Transforms
•Generic Transforms
Detailed information about ODWEK’s viewers and transforms is in IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646. Only a brief overview is provided in this chapter.
The line data applet is a Java applet that is provided by ODWEK. It is similar to the line data viewing capabilities of the Windows client, but it does not contain all of the parallel functionality for viewing line data within the Windows client. For example, the applet does not support saving and selecting custom views.
The plug-ins for AFP and images are shipped as setup packages, which must be installed on the user’s computer. The plug-ins integrate themselves with Mozilla Firefox browsers and Microsoft Internet Explorer. The AFP plug-in provides similar viewing capabilities to the Windows client.
The image plug-in can view image files, with the added benefit of displaying TIFF images (which current web browsers usually cannot display).
Conversions and transforms
In addition to the viewers, ODWEK uses conversion or transformation engines, which convert the document into another data type. ODWEK allows the integration of AFP Transform components for converting AFP into HTML or PDF documents, and it provides a generic transform interface, which can be used to plug in any conversion or transformation engine.
The transforms apply only to documents that are served by ODWEK. They are available to web clients that are based on ODWEK (such as Content Navigator) and to any other application that is written by using the ODWEK Java API. They are not available on the Windows client.
Web viewing considerations
When you choose a viewer strategy in web clients, it is important to know the differences among the viewer architectures:
•Java applet viewers, such as the line data applet or Content Navigator’s generic applet viewer, are downloaded automatically to the user’s computer and run within the browser. No deployment is needed, but a Java installation must be present on the PC. They are effectively cached on the user computers, and they can provide sophisticated functionality. On the downside, each Java applet requires a Java virtual machine (JVM) to run. On terminal servers that serve multiple users at once, this requirement might lead to larger memory consumption.
•Plug-in viewers are native applications that must be installed through a setup routine on the user’s computer. They integrate with the browser and provide their own viewing logic, which can be sophisticated (for example, with the AFP plug-in).
•The generic and Ajax viewers that are provided by Content Navigator provide limited rendering and viewing capabilities. They do not require any rollout or JVM.
•Transforms, such as the Ricoh AFP2PDF or other vendor-provided transforms, result in a PDF document that is viewed in the Acrobat viewer. Although this viewer is deployed on most user PCs, the rendering consumes processing power on the mid-tier system. Also, large documents cannot be rendered into PDFs. Because the PDF is displayed by an external application, it cannot communicate with the Content Manager OnDemand server like the line data applet.
Depending on the data that you are working with, consider these options:
•For Line Data:
– The line data applet supports annotations. It can work with large object (LOB) reports if the large object functionality is employed at load time.
– The Ajax viewer and direct rendering capabilities of Content Navigator work only on shorter reports. Additionally, the viewing of annotations and large object documents is not supported.
•For AFP data:
– The AFP plug-in is the best choice, because it is almost identical to the client. However, it does not support annotations.
The only viewers that use this functionality are the line data applet, the AFP plug-in viewer, and the Content Manager OnDemand Windows client.
– AFP to PDF is a choice that does not require a plug-in rollout at the users’ computers if the Acrobat plug-in is installed on their workstations. Font mappings must be configured at a central location. The additional workload on a rendering system and additional license costs must be considered. Large reports might not be able to be rendered or viewed.
|
Note: The AFP viewer plug-in, which is available with ODWEK and Content Manager OnDemand, is a version of the AFP viewer plug-in from the InfoPrint Solutions Company. Although the standard InfoPrint viewer can be used for viewing AFP, the ODWEK version uses direct communication with the Content Manager OnDemand server, enabling segmented document transfer for LOB documents.
|
Annotations
Only the native ODWEK viewers and the Windows client support annotations. These viewers and Windows clients support annotations in the following ways:
•Line data applet: Supports text. Starting with version 9, the viewer can work with graphical annotations, also.
•Windows Client: Supports maximum capabilities for all data types.
•Other viewers, for example, the AFP plug-in viewer: Do not support and are not aware of annotations.
Web clients, such as Content Navigator or the ODWEK Java API, can work with annotations and provide access to them through the hit list. Graphical annotations cannot be accessed that way because they are not exposed through the Java API.
Large object support
Large object (LOB) support is the methodology for working with large reports. For more information about how LOB affects your reports, see “Large object” on page 52.
From a viewer’s perspective, if a large document is transferred, it generates high network traffic, resource consumption, and long wait times for users. If the viewer supports LOB documents, the viewer communicates with the server to transfer only the chunk of data that the user is looking at (for example, a 200 page chunk out of a 10,000 page report). If the user scrolls to a different chunk of pages, the viewer downloads only that relevant portion of the document that the user scrolled to.
The ODWEK Java API provides line-of-business operations. For more information, see IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646.
8.1.2 Client infrastructure options
Several basic architectural options, Windows client, Content Navigator, or API-based client integration into your line-of-business application, are available.
Windows client
Consider the following items when you are planning a Windows client infrastructure:
•It is faster than the web clients and more powerful.
•It requires native installation on each user’s workstation or notebook. Server version upgrades might also require a new client installation.
•This client supports Citrix and Terminal services environments.
•It does not support the Transforms interface for transforming and converting data formats because the data formats are provided by ODWEK only.
Content Navigator
When you choose a ready-for-use web client, consider the IBM strategic client, IBM Content Navigator, because it is the most complete, most recent web client.
Special use cases might require the development of a custom client application for Content Manager OnDemand. For more information about development APIs, see 8.3, “Client API overview” on page 202.
With Content Navigator, you can run a cross-repository search to search for content across multiple types of repositories, including Content Manager OnDemand. For example, Content Manager OnDemand search results can be included in the same hit list as search results from other supported repositories to help provide a comprehensive view of content.
When you create a cross-repository search, you can specify the following information:
•Specify the scope of the search on each repository. You can specify the search or the classes that you want to include in the cross-repository search by using IBM Content Manager OnDemand. On IBM FileNet Content Manager and IBM Content Manager, you also can limit the search to a specific folder.
•Specify how properties from each repository are related to each other.
•Specify any default search criteria that you want displayed when users open the search.
For more information about how to configure a cross-repository search, see the IBM Content Navigator Knowledge Center at the following web address:
Consider the following items when you choose Content Navigator or other clients:
•The viewers that are provided by Content Navigator are limited compared to the Windows client.
•A Content Manager OnDemand client that focuses only on Content Manager OnDemand is probably the easiest to maintain.
•A general Enterprise Content Manager Client, such as Content Navigator, with setup specifications that support the Content Manager OnDemand model and capabilities, might increase the dependency footprint of the client tier while it provides access to other systems through the same user interface.
Developing your own client
When you develop your own applications (web client), you can use the ODWEK Java APIs. For more information about the ODWEK APIs, see 8.3.1, “Content Manager OnDemand Web Enablement Kit” on page 202 and IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646.
If you are developing a Windows application, you optionally can use the Object Linking and Embedding (OLE) (ActiveX Control) API, which is provided by the Windows client. This API requires a Windows client installation.
Another option is to use an intermediate API that is based on the ODWEK Java API for the Content Manager OnDemand access portion. Content Management Interoperability Services (CMIS) or other web services can be used as the intermediate API. The web service application uses ODWEK to access Content Manager OnDemand and relays this access through its own web services to any other application. In this case, the Windows application only needs to talk to the web service. For more information about CMIS and its limitations, see 8.3.2, “Content Management Interoperability Services” on page 204.
The use of an intermediate API increases complexity and potentially decreases performance, but it decouples a Windows application and Content Manager OnDemand in terms of API versioning and requiring a Content Manager OnDemand installation.
8.1.3 Client compatibility
During the development history of Content Manager OnDemand, features were added and internal API schemes were changed. Therefore, not every client level can work with every server level. When you choose a client infrastructure for your Content Manager OnDemand environment, you must consider version dependencies.
Client compatibility matrix
At the API level, all user clients share a common API core that is based on the Windows client and ODWEK. Almost all other client and API implementations are based on these common APIs. An up-to-date overview of the compatibility matrix that shows the client and ODWEK level that can work with each server level is available at the following website:
Determining version levels
Especially on IBM i and IBM z Systems, the server release level might not be obvious, because it is set by program temporary fixes (PTFs). The most convenient way to determine the server level of your Content Manager OnDemand system is to log on to either the Administrator Client or the user Windows client. After you are logged on, the clients show the server version in the status bar, as shown in Figure 8-1 on page 192.
Figure 8-1 Server version is displayed in the client
Every ars command on the server displays its current server software version, as well.
You can view the version of the Windows client by clicking Help → About.
To determine the version of ODWEK, you can either look for the readme file in the ODWEK application directory or use a client. If you are running a web client (for example, Content Navigator), open a line data report by using the line data applet viewer. Because this viewer is provided by ODWEK directly, the viewer shows the current ODWEK version level in the About dialog box under the Help menu.
Cross-server calls with server console commands
Several of the ars commands that are provided by the server software installation, for example, the ARSDOC and ARSLOAD commands, can work with remote servers. This capability applies to cross-platform calls, for example, loading data with the ARSLOAD command that is running on Linux to a Content Manager OnDemand server that is running on the mainframe.
For more information, see “Server commands” on page 205.
Multiple versions at the same time
Before version 9.5, only one installation of the Content Manager OnDemand Windows client (user and administrative) was installed on a workstation concurrently. Multiple different versions were not allowed to coexist.
Starting with version 9.5 and later, you can run multiple versions of the Content Manager OnDemand Windows client (at the release level only, not the PTF level) on a single workstation. The client code is now installed in the c:Program Files (x86)IBMOnDemand ClientsV9.5 directory.
For ODWEK, you can run multiple versions of ODWEK on a single system. Although this capability might not be a preferred scenario from a maintenance point of view, it can be helpful during upgrades and existing system access scenarios. Each application that uses the ODWEK API must point to the correct installation path and load the correct corresponding libraries.
For more information, see the technote at the following website:
8.2 Content Manager OnDemand Client options
In this section, we describe the common client options for Content Manager OnDemand, including web and non-web clients.
8.2.1 IBM Content Navigator
Content Navigator is the strategic client for IBM Content Manager, IBM FileNet P8, and Content Manager OnDemand. Access to Content Manager OnDemand servers is through the ODWEK Java API. Content Navigator is a Web 2.0 web client and requires a web application server, such as IBM WebSphere® Application Server.
Content Navigator can be used to access documents from multiple content repositories:
•IBM Content Manager Enterprise Edition repositories
•IBM Content Manager OnDemand repositories
•IBM FileNet P8 repositories
•Organization for the Advancement of Structured Information Standards (OASIS) CMIS repositories
With Content Navigator, users can perform these tasks:
•Search documents from any of the content repositories
•View documents side-by-side
•Edit document properties
•Add annotations to documents
•Send documents and document links through email
•Print documents
•Download documents
You can use Content Navigator to build a customized user experience. It supports many configuration options and includes a powerful API toolkit that you can use to extend the web client and build custom applications.
Figure 8-2 shows Content Navigator browsing a folder in Content Manager OnDemand.
Figure 8-2 Searching a Content Manager OnDemand folder with Content Navigator
Content Navigator is a full-feature client for Content Manager OnDemand. Its interface follows modern user interface styles, with a browser pane on the left that shows the available Content Manager OnDemand folders, and a search and result pane on the right. All components and data are dynamic, and they can be resized and changed.
|
Note: Content Navigator is a Web 2.0 Ajax-based client. These web applications rely on an up-to-date JavaScript engine, which is only available in newer browsers. Older browsers, such as Microsoft Internet Explorer Version 8, might not work correctly with Content Navigator.
|
Content Navigator, version 2.0.2 and later, provides many additional Content Manager OnDemand capabilities:
•AFP Viewer plug-in support
•External Data Services (EDS) support
•Favorites support for folders and documents
•Single and multiple AFP file download as PDF (with AFP2PDF enabled)
•Highlighted search result terms in full text searches
•Line2PDF conversion viewer
•XML viewer
Starting with Content Manager OnDemand V9.0 Content Navigator provides single sign-on (SSO) token pass-through to the client side. Date validation is no longer required. Support is provided for ‘t’ date expression and federated search across Content Manager OnDemand, FileNet P8, and IBM Content Manager repositories. Content Navigator is also the new CMIS packaging for Content Manager OnDemand.
Installing Content Navigator
Content Navigator must be installed natively with ODWEK and IBM WebSphere Application Server (or any other applicable web application server). Typically, Content Navigator is installed on a separate system in the web tier and not on the same system as the Content Manager OnDemand server.
The following prerequisites exist for a Content Navigator installation for Content Manager OnDemand:
•Native installation of the Content Navigator base software
•A database to store the Content Navigator configuration
•Web application server
•ODWEK
•Optional: AFP Transforms for AFP to PDF rendering
•Java Database Connectivity (JDBC) drivers (if not already present)
The Content Navigator database is relatively small, so a collocation with the Content Manager OnDemand database might be possible in small deployments. The installation manual provides SQL statements for creating the database and its table spaces.
After you install all of the components, run the Content Navigator Configuration and Deployment Tool to create a preconfigured web application and deploy it to the web application server.
The Configuration and Deployment Tool provides a wizard that leads you through the base setup process. You must provide details about your web application server and connection information to the configuration database. For the Content Manager OnDemand configuration, you must provide the location of your ODWEK installation. Run the deployment scripts at the end for deploying Content Navigator on your application server.
The installation is described in detail in the “Planning, installing, and configuring IBM Content Navigator” section of the IBM Content Manager OnDemand Knowledge Center:
Accessing the native libraries
The ODWEK Java API uses native libraries. To run Content Navigator (or any other web client that is based on the Java APIs), ensure that the web application server can access these libraries. To achieve this task, add the ODWEK directory into the PATH environment variable. On Windows platforms, you also must add the lib64 subdirectory of ODWEK into the PATH.
The following example shows the path to the directory in Windows:
PATH=%PATH%;C:Program FilesIBMOnDemandV9.5www;C:Program FilesIBMOnDemandV9.5in
On Linux and UNIX platforms, it is necessary to expand the LD_LIBRARY_PATH (LIBPATH on AIX) to include the ODWEK directory. This step must be performed in the environment on which the web application server is running by editing the start scripts.
For example, on Linux, you run this command:
export LD_LIBRARY_PATH="/opt/ibm/ondemand/V9.5/www:$LD_LIBRARY_PATH
The Content Navigator installer creates a shared native library in WebSphere Application Server. You can review this library in the Integrated Solution Console in the Environment, Shared libraries section. You need a library that has the class path set to the location of the ODApi.jar (for example, /opt/ibm/ondemand/V9.5/www/api/ODApi.jar) and the Native Library Path set to the ODWEK directory (for example, /opt/ibm/ondemand/V9.5/www). If you encounter any errors, ensure that these paths are valid.
|
Note: If multiple applications reference the same native library, the library gets loaded multiple times. But because the ODWEK library is a shared library, it can be loaded only one time for each JVM. So, if you are running multiple ODWEK web applications in one WebSphere Application Server, you must configure the shared library reference on the Class Loader level of the server itself instead of on the application level. You can use the Integrated Solution Console, which is in the class loader of the application server, for this task.
|
Administering Content Navigator
Content Navigator administration is performed in the admin desktop of the Content Navigator web application. For more information, see the “Administering IBM Content Navigator” section of the IBM Content Manager OnDemand Knowledge Center:
Adding a Content Manager OnDemand repository to Content Navigator
Multiple Content Manager OnDemand repositories can be added to a Content Navigator installation, exposing each repository to a defined set of users through the configuration of different desktops.
For the configuration of Content Manager OnDemand repositories, you need the following parameters:
•Display name: The depository name that is displayed to the users.
•Server name: IP or host name of the Content Manager OnDemand server.
•Port number: Instance port (the default is 1445).
•If you want an encrypted connection between ODWEK and the Content Manager OnDemand server, enable Secure Sockets Layer (SSL) and provide an SSL key ring database and stash file. Enabling SSL consumes additional resources on both systems (Content Manager OnDemand and the web tier).
|
Note: This option does not affect the SSL security of the web application, for example, between the web server and the browser. It only encrypts the API communication between the web tier and the Content Manager OnDemand server.
|
•If you want to use AFP Transforms or another transform filter through generic transforms, you must specify the path to the correct configuration files.
You can specify additional configuration parameters, for example, in the ODConfig class in the Java API. For more information, see the Javadoc of ODApi or IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646.
Content Navigator viewer options
For each Content Navigator Desktop, a different viewer map can be active. Within a viewer map, for each content type, a different viewer can be configured. Several viewers are available to Content Manager OnDemand repositories in Content Navigator:
•Content Navigator uses the viewers that ship with ODWEK, for example, the line data applet. Repository-specific features can be handled only by ODWEK viewers.
•ODWEK performs conversions, for example, an AFP to PDF conversion.
•Built-in viewers for Content Navigator:
– Ajax viewer and a simple PDF and HTML conversion
– Web browser pass-through
– PDF-inline viewer for addressing the Adobe Acrobat viewer browser plug-in
– Generic Applet viewer
– IBM Daeja™ ViewONE Virtual viewer
For a full listing of the viewers, see the IBM Knowledge Center:
•Content Manager OnDemand plug-in viewers for Content Navigator:
– AFP Viewer plug-in
– FileNet Content Federation Services Viewer plug-in
– XML Viewer plug-in
•Third-party viewers can be integrated into Content Navigator. For IBM Production Imaging Edition, for example, a third-party viewer is integrated with Content Navigator. You can integrate your own viewer by using the Content Navigator plug-in architecture.
The generic applet viewer (“applet viewer”) is a Java applet, which can handle various types of documents, such as PDF and Microsoft Office documents (which it renders), images, line data, and AFP documents. The generic applet viewer might be an option if you work with images that are stored in Content Manager OnDemand.
If you want to avoid the use of Java applets and your content is viewable by browsers (for example, certain image types or textual data), try the browser pass-through viewer, which lets the browser handle the data natively. If you work with AFP and must use the AFP browser plug-in, register the Content Navigator plug-in, AFPViewerPlugin.jar, and configure the viewer map that is assigned to your Content Navigator desktop to use the AFP viewer for the application/afp MIME type. The AFPViewerPlugin.jar file ships with Content Navigator. You must choose the web browser pass-through viewer.
The Ajax viewer is a Web 2.0 JavaScript application that provides basic document functions, such as page-wise browsing, rotation, or zoom. It is not a Java applet.
The generic applet viewer, the built-in PDF and HTML conversion, and the Ajax viewer can all work with various data types:
•Images (such as TIFF, JPEG, and DICOM)
•Office documents
•PDF
•Most line data documents
•Certain AFP data
However, they all use a rendering engine to display Office, PDF, and AFP data into an image. This rendering might work well with certain Office and PDF files, but it fails on most non-basic AFP data streams.
For more information, see 8.1.1, “Viewer options” on page 186.
|
Note: Content Navigator is a Web 2.0 client and relies on HTML 5 and JavaScript for its core client functionality and especially for the Ajax viewers. Not all browsers are suitable for running Content Navigator fast and efficiently, especially for Microsoft Internet Explorer browsers before version 9. Test Content Navigator with your user browser thoroughly before you consider a deployment.
|
Extending Content Navigator
Content Navigator is not designed as a client that is dedicated solely to Content Manager OnDemand, so a more complex configuration is necessary than with simpler client options. Content Navigator provides many configuration and customization options through its API and plug-in methodology. For more information about the customization options of Content Navigator, see Customizing and Extending IBM Content Navigator, SG24-8055.
8.2.2 Content Manager OnDemand Windows client
The Content Manager OnDemand Windows client is a full function, feature-rich client that meets the needs of line-of-business application areas and customer service representatives. The Windows client displays content in its native format and is considered a corporate internal access client. Many technical aspects of the Windows client are described in 8.1.1, “Viewer options” on page 186 and 8.1.2, “Client infrastructure options” on page 190.
Figure 8-3 shows a user that is logged in to a folder that performed a search and received the results list. Figure 8-3 shows the indication of a note or hold and also the location of the document. On the right side of the hit list, the load date and document size are displayed.
Figure 8-3 Content Manager OnDemand results list in the Windows client
As the full function client for Content Manager OnDemand, the Windows client provides various business functions and features that can be selected at the document level, as shown in Figure 8-4 on page 199.
Figure 8-4 Windows client capabilities
You also can show the pages within a document or report as thumbnails, which provide you with a visual representation of the report.
8.2.3 CICS Client
The CICS 3270-based interface was the original user interface for Content Manager OnDemand z/OS. It was the predecessor to the Windows and web technology clients that are used today by most Content Manager OnDemand customers. Customers still request the CICS Client to use it to meet their production needs, typically as they migrate their user base (and applications) from a host environment to a client/server or web architecture. The CICS client was developed to meet this need. The CICS Client provides a functional subset of the windows and web clients. The CICS Client is English only. It is included in the Content Manager OnDemand maintenance. It does not ship in the Content Manager OnDemand package, so it must be downloaded and installed separately.
The CICS Client can be downloaded from the following website:
Figure 8-5 on page 200 shows the Content Manager OnDemand CICS Client login panel, which requires the standard login credentials.
.gif)
Figure 8-5 Content Manager OnDemand CICS Client login panel
The CICS Client provides viewing capabilities for line data reports and a “best fit” model for fully composed AFP documents. Viewing a standard line data report is shown in Figure 8-6.
.gif)
Figure 8-6 Viewing a standard line data report
8.2.4 Integration with other Enterprise Content Manager products
Content Manager OnDemand provides integration points with other IBM Enterprise Content Manager software on many different levels. Integration can occur on the client level (for example, by using another product’s user interface (UI) as the client for Content Manager OnDemand). You can use an infrastructural integration in which another product accesses Content Manager OnDemand and information is exchanged between the products at a lower level.
For more information about the most common integrations, see Federated Content Management: Accessing Content from Disparate Repositories with IBM Content Federation Services and IBM Content Integrator, SG24-7742.
8.2.5 Federated search with IBM Information Integrator
Information Integrator is an IBM Enterprise Content Manager product that is available for all Enterprise Content Manager customers. Although it has many functions, it is primarily a federation system.
It can connect to various systems, such as Content Manager OnDemand, Content Manager, FileNet P8, and content management systems by other vendors. You can create a virtual archive, spanning across all connected systems and document models. Users can search in one system and the search is propagated to multiple back-end repositories. Information Integrator maps virtual fields to folder fields in Content Manager OnDemand (or respective models in other systems) and delivers a consistent hit list of documents to the user.
Content Integrator might be an option for you if you use separate Content Manager OnDemand systems (instances or physical systems) and must provide a cross-system search (for example, for eDiscovery or legal inquiries). Another use case is to provide repository-neutral services with access to multiple content management systems.
|
Note: Information Integrator is an abstraction layer. You lose Content Manager OnDemand specific functionality, because the virtual archive provides only the common functionality that can be implemented by all archives. Always check your use case to verify that a virtual archive meets your needs for functional compatibility and performance.
|
8.2.6 Integration with IBM FileNet P8
Integration exists between IBM FileNet P8 and Content Manager OnDemand through FileNet Content Federation Services. Content Manager OnDemand documents can be federated into FileNet P8, making them accessible like any other FileNet P8 documents for FileNet P8 users.
This federation differs compared to Information Integrator. In Content Federation Services, for each Content Manager OnDemand document, a virtual document is created in FileNet P8 (resulting in database records in FileNet P8). So, these documents act as FileNet P8 documents from a FileNet P8 user’s perspective. Information Integrator does not have its own database and does not create virtual documents, but it instead calls Content Manager OnDemand for searches and passes on the result list. A search in FileNet P8 never starts a search in Content Manager OnDemand, but it can find only federated Content Manager OnDemand documents, which are cataloged in the FileNet P8 database.
If a FileNet P8 system is installed in your environment that serves as your primary content management system and reports need to be available to users without their knowing that those reports are in a different system, this integration might suit your needs. The same situation applies to the use of FileNet P8 Records Management, which can be applied to Content Manager OnDemand documents as well, therefore bringing a level of federated records management capability to your documents.
When you plan your integration with FileNet P8, remember this federation is active: Content Manager OnDemand actively publishes document links into a FileNet P8 system. You must consider both volumes (FileNet P8 systems usually are smaller than Content Manager OnDemand systems) and the active federation process.
For more information about Content Manager OnDemand and FileNet P8 integration, see IBM FileNet Content Federation Services for Content Manager OnDemand, SC19-2711.
8.3 Client API overview
With various client options, multiple API options are available to navigate through the system and access Content Manager OnDemand documents. Although the Java API that is provided by Content Manager ODWEK is the API that is used most by clients and the basis for most development projects, other APIs are available and used for a limited range of scenarios.
The following list shows the APIs that are available for Content Manager OnDemand:
•Content Manager ODWEK: The Java API for Content Manager OnDemand
•SOAP and Representational State Transfer (REST) web services that follow the CMIS standard
•Windows OLE (ActiveX control) that is provided by the Windows client
•XML administrative API through the ARSXML server command
•Structured APIs on z/OS environments
•The standard Content Manager OnDemand server commands that serve as a console-based API to work with Content Manager OnDemand documents
8.3.1 Content Manager OnDemand Web Enablement Kit
ODWEK provides a Java API to access Content Manager OnDemand servers and their documents. It is the strategic client API that provides the largest feature set of any Content Manager OnDemand API. It is used by web clients, such as Content Navigator or WEBi, by abstraction layers, such as Information Integrator, or by API components, such as CMIS.
The ODWEK Java API and its use to develop Content Manager OnDemand clients are described in detail in IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646. This section covers only a basic overview and focuses on client considerations about ODWEK. Developers are encouraged to read the referenced book before they plan a client development that is based on ODWEK.
Scope
ODWEK is a Content Manager OnDemand component that can be used by all Content Manager OnDemand customers. It is focused on typical client use cases, such as searching for and accessing data that is stored in a Content Manager OnDemand archive. It also has web viewers, such as the line data applet and Content Manager OnDemand AFP viewer.
For more information about ODWEK viewers and conversion support, see “Windows client viewers” on page 187.
Before Content Manager OnDemand Web Enablement Kit (ODWEK) Java API V9.5, the only API that allowed documents to be added to the Content Manager OnDemand archive was the ODFolder.storeDocument API, which resulted in an archive request to the Content Manager OnDemand server for each document. This API is suitable for low-volume ad hoc storage.
In ODWEK V9.5, new APIs were introduced to allow documents to be loaded in bulk, which provides high-volume storage similar to the arsload command. To accomplish bulk loading by using the ODWEK Java API, you perform these steps:
1. Call the ODServer.loadInit API to initiate the load process.
2. For each document to load, call the ODServer.loadAddDoc API, which passes the number of pages, a hash table of index values to store, and the document data.
3. Call the ODServer.loadCommit API, which specifies the application group and application to send the load data and load request to the Content Manager OnDemand server.
For special client needs, the Java API provides access to the object model (application group and application) of Content Manager OnDemand and facilitates an ARSXML pass-through, which can be used to perform administrative tasks.
Native library dependency
Because of the nature of the Content Manager OnDemand architecture, ODWEK requires the use of native libraries.
In addition to the physical presence on the system, Java applications must be aware of the native libraries. The ODWEK native libraries are loaded as shared memory objects and cannot be reloaded multiple times. If you run multiple ODWEK applications in one web application server, consider this restriction.
For a description of how the native library reference is managed for the ODWEK client in IBM Content Navigator in IBM WebSphere Application Server, see “Accessing the native libraries” on page 195.
ODWEK web client design considerations
When you design a web client for Content Manager OnDemand that is based on ODWEK, consider the following items:
•Dependency on a native shared library affects deployment and general options, such as the message language, which can be set only for the whole environment.
•Be careful with multithreading document access. Access to a single session with the Content Manager OnDemand server must be in a single-threaded fashion. Only one thread can access objects of a specified Content Manager OnDemand session at a time.
•Every session that is established with a Content Manager OnDemand server consumes memory on the ODWEK system. For high-usage applications that support many concurrent users, for example, web clients that work with non-named users, we suggest the use of connection pooling.
•Ensure that a timeout concept is implemented in your application that meets the Content Manager OnDemand user activity timeout. Sessions that do not time out might lead to memory leaks or high memory consumption on the Content Manager OnDemand and ODWEK machines.
|
Note: Starting with version 9 of ODWEK, additional functions were added to reset the inactivity timeout counter of a user session. This enhancement simplifies the design of connection pooling and timeout scenarios.
|
For a connection pooling sample that covers the topics of thread safety, resource consumption, and timeouts in detail, see Chapter 6, “Connection pooling and connection handling”, in IBM Content Manager OnDemand Web Enablement Kit Java APIs: The Basics and Beyond, SG24-7646.
8.3.2 Content Management Interoperability Services
CMIS is an open standard for accessing content management repositories. It is an OASIS specification and it is supported by various applications from different vendors, including IBM (with FileNet P8, Content Manager, and Content Manager OnDemand).
CMIS provides a common access interface for searching, retrieving, and in the case of document management systems, modifying and deleting documents. It is a web services interface that is implemented in either SOAP web services and REST (Atom) services.
For more information about CMIS, see the CMIS page on the OASIS website, the CMIS overview page at the IBM Enterprise Content Manager website, and the technical documentation that is available:
•Implementing Web Applications with CM Information Integrator for Content and OnDemand Web Enablement Kit, SG24-6338
•Content Management Interoperability Services for Content Manager OnDemand is installed as part of the IBM Content Navigator installation. For more information, see “Installing Content Navigator” on page 194.
When you consider implementing your own software on CMIS, remember CMIS is used for accessing document management systems, but not necessarily high-volume report archives, such as Content Manager OnDemand.
The methodology of accessing documents is based on folders and subfolders with documents in it (such as in a file system) and partially emulated by Content Manager OnDemand with its different object model. The use of CMIS must be considered as an abstraction layer that might have an impact on throughput and feature exposure. Also, much of the CMIS API is not supported by Content Manager OnDemand (such as the storage and deletion functions).
8.3.3 Other client-based API options
Other client-based API options include Windows ActiveX API, structured API on z/OS, server commands, and XML Administration interface (ARSXML).
Windows ActiveX API
The Windows client ships an ActiveX control, which can be used in its own application for accessing Content Manager OnDemand servers and documents through the functions that are provided by the Windows client. It is a development API that enables the development of custom applications that use an installed Windows client as the API provider. The ActiveX API covers only a basic operation subset.
For more information about the Windows client-based API, see Windows Client Customization Guide, SC19-3357.
Structured API on z/OS
In z/OS environments, Content Manager OnDemand includes Structured APIs that provide custom applications in CICS, IBM IMS™, TSO, or batch environments with the ability to connect to Content Manage OnDemand servers. The Structured APIs support only the basic read operations (log on, open folder, search, and retrieve documents and annotations).
Structured APIs are handled by a dedicated component of Content Manager OnDemand that is called MidServer. MidServer relies on ODWEK and its API to access the Content Manager OnDemand server.
Structured APIs are available only on z/OS, and they are called from COBOL or C applications in the same manner as MVS calls. Because ODWEK is used as the access path to the Content Manager OnDemand server, the Structured APIs can be used to access non-z/OS Content Manager OnDemand servers, as well.
Server commands
In addition to the API options, which are exposed through Java, OLE, or Web Services, Content Manager OnDemand provides console (command-line) applications that provide specific functions, such as searching, retrieving, or deleting documents, and sophisticated functions, such as placing holds and working with the full text engine. Most of this functionality is exposed through the ARSDOC application.
Simpler custom applications, for example, shell scripts, can use these server console applications to interact with Content Manager OnDemand systems. The applications are available only as part of a Content Manager OnDemand server installation. Because most of them (namely ARSDOC) communicate with the server through TCP/IP, you can connect and interact with Content Manager OnDemand servers remotely on other platforms. When you call remote servers, ensure that the local installation that provides the ARS applications and the actual Content Manager OnDemand server are on the same version level.
For more information about the administrative commands, see the specific command descriptions in the IBM Content Manager OnDemand Knowledge Center:
XML administration interface: ARSXML
In addition to the user client APIs, the ARSXML server command provides an interface for administrative users and applications to access the Content Manager OnDemand data model. By using ARSXML, folders, application groups, applications, and users can be exported, created, deleted, and modified. It works on XML documents by describing the change, action, or selection criteria and the resulting output XML document.
ARSXML is a console application that is available on the Content Manager OnDemand server. It can work with remote servers if they are at the same release level.
XMLs can be passed to and from ARSXML through the ODWEK Java API, which enables Java applications to programmatically call ARSXML and obtain access to administrative data model functions.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.