Using OLE¶

This chapter describes how you can use OLE 2.0 support in Adobe Acrobat for Microsoft Windows. Acrobat applications are OLE servers and also respond to a variety of OLE automation messages.

Since Acrobat provides the appropriate interfaces to be an OLE server, you can embed PDF documents into documents created by an application that is an OLE client, or link them to OLE containers. However, Acrobat does not perform in-place activation.

Acrobat supports the OLE automation methods described in this document. Acrobat Reader does not support OLE automation, except for the PDF browser controls provided in the AcroPDF object.

The best practical resources for Visual Basic or Visual C# programmers, besides the object browser, are the sample projects. The samples demonstrate use of the Acrobat OLE objects and contain comments describing the parameters for the more complicated methods. For more information, see the Acrobat SDK Samples Guide.

For more information on OLE 2.0 and OLE automation, see the OLE Automation Programmer’s Reference , ISBN 1-55615-851-3, Microsoft Press. You can also find numerous articles at http://msdn.microsoft.com .

OLE capabilities in Acrobat¶

For OLE automation, Acrobat provides three capabilities: rendering PDF documents, remotely controlling the application, and implementing PDF browser controls.

On-screen rendering¶

You can render PDF documents on the screen in two ways:

Use an interface similar to the Acrobat user interface.

In this approach, use the AVDoc object’s OpenInWindowEx method to open a PDF file in your application’s window. The window has vertical and horizontal scroll bars, and has buttons on the window’s perimeter for setting the zoom factor. Users interacting with this type of window find its operation similar to that of working in Acrobat. For example, links are active and the window can display any text annotation on a page.

The ActiveView sample in the Acrobat SDK Samples Guide demonstrates this approach.

Use the PDPage object’s DrawEx method.

In this approach, you provide a window and a device context, as well as a zoom factor. Acrobat renders the current page into your window. The application must manage the scroll bars and other items in the user interface.

The StaticView sample in the Acrobat SDK Samples Guide demonstrates this approach.

Remote control of Acrobat¶

You can control Acrobat remotely in two ways:

Given the exported interfaces, you can write an application that manipulates various aspects of PDF documents, such as pages, annotations, and bookmarks. Your application might use AVDoc, PDDoc, PDPage, and annotation methods, and might not provide any visual feedback that requires rendering into its application window.
You can launch Acrobat from your own application, which has set up the environment for the user. Your application can cause Acrobat to open a file, set the page location and zoom factor, and possibly even select some text. For example, this could be useful as part of a help system.

PDF browser controls¶

You can use the AcroPDF library to display a PDF document in applications using simplified browser controls. In this case, the PDF document is treated as an ActiveX document, and the interface is available in Acrobat Reader.

Load the document with the AcroPDF object’s LoadFile method. You can then implement browser controls for the following functionality:

To determine which page to display
To choose the display, view, and zoom modes
To display bookmarks, thumbs, scrollbars, and toolbars
To print pages using various options
To highlight a text selection

Development environment considerations¶

You have a choice of environments in which to integrate with Acrobat: Visual Basic, Visual C#, and Visual C++.

If possible, use Visual Basic or Visual C#. The run-time type checking offered by the CreateObject call in Visual Basic allows quick prototyping of an application, and in both of these languages the implementation details are simplified.

For comparison, consider the following examples, in which you can see strings with "AcroExch.App" and strings with "Acrobat.CAcroApp". The first is the form for the external string used by OLE clients to create an object of that type. The second is the form that is included in developer type libraries.

This example shows a Visual Basic subroutine to view a given page of an open document:

Viewing a page with Visual Basic

Private Sub myGoto(ByVal where As Integer)
    Dim app as Object, avdoc as Object, pageview as Object
    Set app = CreateObject("AcroExch.App")
    Set avdoc = app.GetActiveDoc
    Set pageview = avdoc.GetAVPageView
    pageview.Goto(where)
End Sub

The following example does the same, but in Visual C++:

Viewing a page with Visual C++

void goto(int where)
{
    CAcroApp app;
    CAcroAVDoc *avdoc = new CAcroAVDoc;
    CAcroAVPageView pageview;
    COleException e;
    app.CreateDispatch("AcroExch.App");
    avdoc->AttachDispatch(app.GetActiveDoc, TRUE);
    pageview->AttachDispatch(avdoc->GetAVPageView, TRUE);
    pageview->Goto(where);
}

The next example shows how to use PDF browser controls to view a page in Visual Basic:

Using AcroPDF browser controls with Visual Basic

Friend WithEvents AxAcroPDF1 As AxAcroPDFLib.AxAcroPDF
Me.AxAcroPDF1 = New AxAcroPDFLib.AxAcroPDF
'AxAcroPDF1
Me.AxAcroPDF1.Enabled = True
Me.AxAcroPDF1.Location = New System.Drawing.Point(24, 40)
Me.AxAcroPDF1.Name = "AxAcroPDF1"
Me.AxAcroPDF1.OcxState = CType(
       resources.GetObject("AxAcroPDF1.OcxState"),
       System.Windows.Forms.AxHost.State
)
Me.AxAcroPDF1.Size = New System.Drawing.Size(584, 600)
Me.AxAcroPDF1.TabIndex = 0
AxAcroPDF1.LoadFile("http://www.example.com/example.pdf")
AxAcroPDF1.setCurrentPage(TextBox2.Text)

The Visual Basic examples are simpler to read, write, and support, and the implementation details are similar to Visual C#.

In Visual C++, the CAcro classes hide much of the type checking that must be done. Using OLE automation objects in Visual C++ requires an understanding of the AttachDispatch and CreateDispatch methods of the COleDispatchDriver class. For more information, see Using the Acrobat OLE interfaces.

Note

The header files containing the values of constants that are required by C and C++ programmers to use OLE automation are located in the Acrobat SDK IAC directory. Visual Basic and Visual C# users do not need these header files, though it may be useful to refer to them in order to verify the constant definitions.

Environment configuration¶

The only requirement for using the OLE objects made available by Acrobat is to have the product installed on your system and the appropriate type library file included in the project references for your project. The Acrobat type library file is named Acrobat.tlb. This file is included in the InterAppCommunicationSupportHeaders folder in the SDK. Once you have the type library file included in your project, you can use the object browser to browse the OLE objects.

It is not sufficient to install just an ActiveX control or DLL to enable OLE automation. You must have the full Acrobat product installed.

If you are a Visual Basic programmer, it is helpful to include the iac.bas module in your project (included in the headers folder). This module defines the constant variables.

Necessary C knowledge¶

These documents, as well as the API, were designed with C programming in mind and programming with the API requires some familiarity with C concepts.

Although you do not need the header files provided in the SDK, you can use them to find the values of various constants, such as AV_DOC_VIEW, that are referenced in the documentation. The file iac.h contains most of these values.

Some of the methods, such as OpenInWindowEx, can be initially confusing when used in Visual Basic. OpenInWindowEx takes a long for the openflags parameter. The options for this parameter are:

AV_EXTERNAL_VIEW: Open the document with the toolbar visible.
AV_DOC_VIEW: Draw the page pane and scrollbars.
AV_PAGE_VIEW: Draw only the page pane.

If you were developing in C, these strings would be replaced by a numeric value prior to compilation; passing these strings to the method would not raise an error. When programming in Visual Basic, these strings correspond to constant variables defined in iac.bas.

In some situations, you need to apply a bitwise OR to multiple values and pass the resultant value to a method. For example, in iac.h the ntype parameter of the PDDocSave method is a bitwise OR of the following flags:

/* PDSaveFlags: used for PD-level Save
** All undefined flags should be set to zero.
** If either PDSaveCollectGarbage or PDSaveCopy are used, PDSaveFull must be used. */
typedef enum {
    PDSaveIncremental = 0x0000,  /* write changes only */
    PDSaveFull = 0x0001,         /* write entire file */
    PDSaveCopy = 0x0002,         /* write copy w/o affecting current state */
   PDSaveLinearized = 0x0004,   /* write the file linearized for
    **       page-served remote (net) access. */
   PDSaveBinaryOK = 0x0010, /* OK to store binary in file */
   PDSaveCollectGarbage = 0x0020  /* perform garbage collection on
   **       unreferenced objects */
} PDSaveFlags;

For example, if you would like to fully save the PDF file and optimize it for the Web (linearize it) within a Visual Basic application, pass PDSaveFull + PDSaveLinearized (both defined in iac.bas) into the ntype parameter; this is the equivalent of a binary OR of the PDSaveFull and PDSaveLinearized parameters.

In many instances, the numeric values are spelled out in comments in the Visual Basic sample code. However, knowledge of why the methods are structured in this way and how they are used in C can be useful to Visual Basic and Visual C# programmers.

Using the Acrobat OLE interfaces¶

This section describes using the CAcro classes and the COleDispatchDriver class. The CAcro classes are subclasses of COleDispatchDriver.

About the CAcro classes¶

OLE 2.0 support in Acrobat includes several classes whose names begin with “CAcro “, such as CAcroApp and CAcroPDDoc. Several files in the SDK encapsulate the definitions of these classes.

The CAcro classes are defined in the Acrobat type library acrobat.tlb. The OLEView tool in Visual Studio allows you to browse registered type libraries. Use acrobat.tlb when defining OLE automation for a project in Microsoft Visual C++. The files acrobat.h and acrobat.cpp are included in the Acrobat SDK, and implement a type-safe wrapper to the Acrobat automation server.

Note

Do not modify the acrobat.tlb, acrobat.h, and acrobat.cpp files in the SDK; these define Acrobat’s OLE automation interface.

The CAcro classes inherit from the MFC COleDispatchDriver class. Understanding this class makes it easier to write applications that use the CAcro classes and their methods.

See the Interapplication Communication API Reference for details on the CAcro classes and their methods.

About the COleDispatchDriver class¶

The COleDispatchDriver class implements the client side of OLE automation, providing most of the code needed to access automation objects. It provides the wrapper functions AttachDispatch, DetachDispatch, and ReleaseDispatch, as well as the convenience functions InvokeHelper, SetProperty, and GetProperty. You employ some of these methods when you use the Acrobat-provided automation objects. Other methods are used in the Acrobat implementation of these objects.

COleDispatchDriver is essentially a “class wrapper” for IDispatch, which is the OLE interface by which applications expose methods and properties so that other applications written in Visual Basic and Visual C# can use the application’s features. This provides OLE support for Acrobat applications.

Using COleDispatchDriver objects and methods¶

This section discusses how to use the classes exported by acrobat.cpp, and shows when to call the CreateDispatch and AttachDispatch methods.

The following is a section of code from acrobat.h that declares t he CAcroHiliteList class. CAcroHiliteList is a subclass of the COleDispatchDriver class, which means that it shares all the instance variables of COleDispatchDriver.

One of these variables is m_lpDispatch, which holds an LPDISPATCH for that object. An LPDISPATCH is a long pointer to an IDispatch, which can be considered an opaque data type representing a dispatch connection. m_lpDispatch can be used in functions that require an LPDISPATCH argument.

CAcroHiliteList class declaration

class CAcroHiliteList : public COleDispatchDriver
{
public:
   CAcroHiliteList() {}        // Calls COleDispatchDriver default constructor
   CAcroHiliteList(LPDISPATCH pDispatch) : COleDispatchDriver(pDispatch) {}
   CAcroHiliteList(const CAcroHiliteList& dispatchSrc) :

        COleDispatchDriver(dispatchSrc) {}

// Attributes
public:

// Operations
public:
   bool Add(short nOffset, short nLength);
};

The following is the related implementation section of the Add method from acrobat.cpp:

bool CAcroHiliteList::Add(short nOffset, short nLength)
{
   bool result;
   static BYTE parms[] =
       VTS_I2 VTS_I2;
   InvokeHelper(0x1, DISPATCH_METHOD, VT_I4, (void*)&result, parms,
       nOffset, nLength);
   return result;
}

When the Add method is called, such as with this code from the following example Using the COleDispatchDriver class,

hilite->Add(0, 10);

the InvokeHelper function is called. This COleDispatchDriver method takes a variable number of arguments. It eventually calls the Acrobat implementation for CAcroHiliteList object’s Add method. This happens across the virtual OLE “wires” and takes care of all the OLE details. The end result is that a page range is added to the CAcroHiliteList object.

The following is an implementation of a method adapted from the ActiveView sample:

Using the COleDispatchDriver class

// This code demonstrates how to highlight words with
// either a word or page highlight list
void CActiveViewDoc::OnToolsHilitewords()
{
   CAcroAVPageView pageView;
   CAcroPDPage page;
   CAcroPDTextSelect* textSelect = new CAcroPDTextSelect;
   CAcroHiliteList* hilite = new CAcroHiliteList;
   char buf[255];
   long selectionSize;

   if ((BOOL) GetCurrentPageNum() > PDBeforeFirstPage) {

       // Obtain the AVPageView
       pageView.AttachDispatch(m_pAcroAVDoc->GetAVPageView(),TRUE);

       // Create the Hilite list object
       hilite->CreateDispatch("AcroExch.HiliteList");
       if (hilite) {

   // Add the first 10 words or characters of that page to the highlight list
           hilite->Add(0,10);
           page.AttachDispatch(pageView.GetPage(), TRUE);

           // Create text selection for either page or word highlight list
           textSelect->AttachDispatch(page.CreateWordHilite(hilite->m_lpDispatch));
           m_pAcroAVDoc->SetTextSelection(textSelect->m_lpDispatch);
           m_pAcroAVDoc->ShowTextSelect();

           // Extract the number of words and the first word of text selection
           selectionSize = textSelect->GetNumText();
           if (selectionSize)
               sprintf (buf, "# of words in text selection: %ldn1st word in text
                   selection = '%s'", selectionSize, textSelect->GetText(0));
           else
               sprintf (buf, "Failed to create text selection.");

           AfxMessageBox(buf);
       }
   }

   delete textSelect;
   delete hilite;
}

In the preceding example, the objects with the prefix CAcro are all CAcro class objects—and they are also COleDispatchDriver objects—because all the Acrobat CAcro classes are subclasses of COleDispatchDriver.

Instantiating a class is not sufficient to use it . Before you use an object, you must attach your object to the appropriate Acrobat object by using one of the Dispatch methods of the COleDispatchDriver class. These functions also initialize the m_lpDispatch instance variable for the object.

This code from the previous example shows how to attach an IDispatch that already exists:

CAcroAVPageView pageView;

// Obtain the AVPageView

pageView.AttachDispatch(m_pAcroAVDoc->GetAVPageView(), TRUE);

The GetAVPageView method of the CAcroAVDoc class returns an LPDISPATCH, which is what the AttachDispatch method is expecting for its first argument. The BOOL passed as the second argument indicates whether or not the IDispatch should be released when the object goes out of scope, and is typically TRUE. In general, when an LPDISPATCH is returned from a method such as GetAVPageView, you use AttachDispatch to attach it to an object.

The following code from the previous example uses the CreateDispatch method:

CAcroHiliteList *hilite = new CAcroHiliteList;
hilite->CreateDispatch("AcroExch.HiliteList");
hilite->Add(0, 10);

In this case, the CreateDispatch method both creates the IDispatch object and attaches it to the object. This code works fine; however, the following code would fail:

CAcroHiliteList *hilite = new CAcroHiliteList;
hilite->Add(0, 10);

This error is analogous to using an uninitialized variable. Until the IDispatch object is attached to the COleDispatchDriver object, it is not valid.

CreateDispatch takes a string parameter, such as "AcroExch.HiliteList", which represents a class. The following code is incorrect:

CAcroPDDoc doc = new CAcroPDDoc;
doc.CreateDispatch("AcroExch.Create");

This fails because Acrobat won’t respond to such a parameter. The parameter should be "AcroExch.PDDoc" instead.

The valid strings for CreateDispatch are as follows:

Class	String
CAcroPoint	“AcroExch.Point”
CAcroRect	“AcroExch.Rect”
CAcroTime	“AcroExch.Time”
CAcroApp	“AcroExch.App”
CAcroPDDoc	“AcroExch.PDDoc”
CAcroAVDoc	“AcroExch.AVDoc”
CAcroHiliteList	“AcroExch.HiliteList”
CAcroPDBookmark	“AcroExch.PDBookmark”
CAcroMatrix	“AcroExch.Matrix”
AcroPDF	“AxAcroPDFLib.AxAcroPDF”

Refer again to this code from the previous example:

CAcroPDPage page;
page.AttachDispatch(pageView.GetPage(), TRUE);

A PDPage object is required because the purpose of this code is to highlight words on the current page. Since it is a CAcro variable, it is necessary to attach to the OLE object before using its methods. CreateDispatch cannot be used to create a PDPage object because "AcroExch.PDPage" is not a valid string for CreateDispatch. However, the AVPageView method GetPage returns an LPDISPATCH pointer for a PDPage object. This is passed as the first argument to the AttachDispatch method of the page object. The TRUE argument indicates that the object is to be released automatically when it goes out of scope.

CAcroPDTextSelect* textSelect = new CAcroPDTextSelect;
textSelect->AttachDispatch
   (page.CreateWordHilite(hilite->m_lpDispatch));
m_pAcroAVDoc->SetTextSelection (textSelect->m_lpDispatch);
m_pAcroAVDoc->ShowTextSelect();

This code performs the following steps:

Declares a text selection object textSelect.
Calls the CAcroPDPage method CreateWordHilite, which returns an LPDISPATCH for a PDTextSelect. CreateWordHilite takes an LPDISPATCH argument representing a CAcroHilite list. The hilite variable already contains a CAcroHiliteList object, and its instance variable m_lpDispatch contains the LPDISPATCH pointer for the object.
Calls the CAcroAVDoc object’s SetTextSelection method to select the first ten words on the current page.
Calls the AcroAVDoc's ShowTextSelect method to cause the visual update on the screen.

Using the JSObject interface¶

Acrobat provides a rich set of JavaScript programming interfaces that can be used from within the Acrobat environment. It also provides the JSObject interface, which allows external clients to access the same functionality from environments such as Visual Basic.

In precise terms, JSObject is an interpretation layer between an OLE automation client, such as a Visual Basic application, and the JavaScript functionality provided by Acrobat. From a developer’s point of view, programming JSObject in a Visual Basic environment is similar to programming in JavaScript using the Acrobat console.

This section explains how to extend Acrobat using JavaScript in a Visual Basic programming environment. It provides a set of examples to illustrate the key concepts.

Whenever possible, you should take advantage of these capabilities by using the JSObject interface available within the AcroExch.PDDoc object. To obtain the interface, invoke the object’s GetJSObject method.

Adding a reference to the Acrobat type library¶

This procedure adds a reference to the Acrobat type library so that you can access the Acrobat automation APIs, including JSObject, in Visual Basic. Do this before using the JSObject interface, as in the examples that follow.

To add a reference to the Acrobat type library:
Install Acrobat and Visual Basic.
Create a new Visual Basic project from the Windows Application template. This provides a blank form and project workspace.
Select Project > Add Reference and click the COM tab.
From the list of available references, select Adobe Acrobat <version> Type Library and click OK.

Creating a simple application¶

This example provides the minimum code to display “Hello, Acrobat!” in the Acrobat JavaScript console.

To set up and run the “Hello, Acrobat!” example:
Open the source code window for the default form by clicking View > Code.
Select (Form1 Events) from the selection box in the upper left corner of that window.

The selection box in the upper right corner now shows all the functions available to the Form1 object.

Select Load from the functions selection box. This creates an empty function stub. The Form1 Load function is called when Form1 is first displayed, so this is a good place to add the initialization code.
Add the following code to define some global variables before the subroutine.

Dim gApp As Acrobat.CAcroApp
Dim gPDDoc As Acrobat.CAcroPDDoc
Dim jso As Object

Add the following code to the private Form1_Load subroutine.

gApp = CreateObject("AcroExch.App")
gPDDoc = CreateObject("AcroExch.PDDoc")
If gPDDoc.Open("c:example.pdf") Then
    jso = gPDDoc.GetJSObject
    jso.console.Show
    jso.console.Clear
    jso.console.println ("Hello, Acrobat!")
    gApp.Show
End If

Create a file called example.pdf at the root level of the C: drive.
Save and run the project.

When you run the application, Acrobat is launched, Form1 is displayed, and the JavaScript Debugger window is opened, displaying “Hello, Acrobat!”.

Displaying “Hello, Acrobat!” in the JavaScript console

Dim gApp As Acrobat.CAcroApp
Dim gPDDoc As Acrobat.CAcroPDDoc
Dim jso As Object

Private Sub Form1_Load(ByVal sender As Object, ByVal e As System.EventArgs)

         Handles Me.Load
    gApp = CreateObject("AcroExch.App")
    gPDDoc = CreateObject("AcroExch.PDDoc")
    If gPDDoc.Open("c:example.pdf") Then
        jso = gPDDoc.GetJSObject
        jso.console.Show
        jso.console.Clear
        jso.console.println ("Hello, Acrobat!")
        gApp.Show
    End If
End Sub

The Visual Basic program attaches to the Acrobat automation interface using the CreateObject call, and then shows the main window using the App object’s Show command.

You may have a few questions after studying the code. For example, why is jso declared as an Object, while gApp and gPDDoc are declared as types found in the Acrobat type library? Is there a real type for JSObject ?

The answer is no, JSObject does not appear in the type library, except in the context of the CAcroPDDoc.GetJSObject call. The COM interface used to export JavaScript functionality through JSObject is known as an IDispatch interface, which in Visual Basic is more commonly known simply as an “Object” type. This means that the methods available to the programmer are not particularly well-defined. For example, if you replace the call to

jso.console.clear

with

jso.ThisCantPossiblyCompileCanIt("Yes it can!")

the compiler compiles the code, but fails at run time. Visual Basic has no type information for JSObject, so Visual Basic does not know if a particular call is syntactically valid until run-time, and will compile any function call to a JSObject. For that reason, you must rely on the documentation to know what functionality is available through the JSObject interface. For details, see the JavaScript for Acrobat API Reference.

You may also wonder why it is necessary to open a PDDoc before creating a JSObject. Running the program shows that no document appears onscreen, and suggests that using the JavaScript console should be possible without a PDDoc. However, JSObject is designed to work closely with a particular document, as most of the available features operate at the document level. There are some application-level features in JavaScript (and therefore in JSObject), but they are of secondary interest. In practice, a JSObject is always associated with a particular document.

When working with a large number of documents, you must structure your code so that a new JSObject is acquired for each document, rather than creating a single JSObject to work on every document.

Working with annotations¶

This example uses the JSObject interface to open a PDF file, add a predefined annotation to it, and save the file back to disk.

To set up and run the annotations example:
Create a new Visual Basic project and add the Adobe Acrobat type library to the project.
From the Toolbox, drag the OpenFileDialog control to the form.
Drag a Button to your form.

Select View > Code and set up the following source code:
Adding an annotation

Dim gApp As Acrobat.CAcroApp

Private Sub Form1_Load(ByVal sender As System.Object, ByVal e As

    System.EventArgs) Handles MyBase.Load
    gApp = CreateObject("AcroExch.App")
End Sub

Private Sub Form1_Closed(Cancel As Integer)
    If Not gApp Is Nothing Then
        gApp.Exit
    End If
    gApp = Nothing
End Sub

Private Sub Button1_Click(ByVal sender As System.Object, ByVal e As

    System.EventArgs) Handles Button1.Click
    Dim pdDoc As Acrobat.CAcroPDDoc
    Dim page As Acrobat.CAcroPDPage
    Dim jso As Object
    Dim path As String
    Dim point(1) As Integer
    Dim popupRect(3) As Integer
    Dim pageRect As Object
    Dim annot As Object
    Dim props As Object

    OpenFileDialog1.ShowDialog()
    path = OpenFileDialog1.FileName

    pdDoc = CreateObject("AcroExch.PDDoc")
    If pdDoc.Open(path) Then
        jso = pdDoc.GetJSObject
        If Not jso Is Nothing Then

            ' Get size for page 0 and set up arrays
            page = pdDoc.AcquirePage(0)
            pageRect = page.GetSize
            point(0) = 0
            point(1) = pageRect.y
            popupRect(0) = 0
            popupRect(1) = pageRect.y - 100
            popupRect(2) = 200
            popupRect(3) = pageRect.y

            ' Create a new text annot
            annot = jso.AddAnnot
            props = annot.getProps
            props.Type = "Text"
            annot.setProps props

            ' Fill in a few fields
            props = annot.getProps
            props.page = 0
            props.point = point
            props.popupRect = popupRect
            props.author = "John Doe"
            props.noteIcon = "Comment"
            props.strokeColor = jso.Color.red
            props.Contents = "I added this comment from Visual Basic!"
            annot.setProps props
        End If
        pdDoc.Close
        MsgBox "Annotation added to " & path
    Else
        MsgBox "Failed to open " & path
    End If

    pdDoc = Nothing
End Sub

Save and run the application.

The code in the Form_Load and Form_Closed routines initializes and shuts down the Acrobat automation interface. More interesting work happens in the Command button’s click routine. The first lines declare local variables and show the Windows Open dialog box, which allows the user to select a file to be annotated. The code then opens the PDF file’s PDDoc object and obtains a JSObject interface to that document.

Some standard Acrobat automation methods are used to determine the size of the first page in the document. These numbers are critical to achieving the correct layout, because the PDF coordinate system is based in the lower-left corner of the page, but the annotation will be anchored at the upper left corner of the page.

The lines following the "Create a new text annot" comment do exactly that, but this block of code bears additional explanation.

First, addAnnot looks as if it is a method of JSObject, but the JavaScript reference shows that the method is associated with the doc object. You might expect the syntax to be jso.doc.addAnnot. However, jso is the Doc object, so jso.addAnnot is correct. All of the properties and methods in the Doc object are used in this manner.

Second, observe the use of annot.getProps and annot.setProps. The Annot object is implemented with a separate properties object, meaning that you cannot set the properties directly. For example, you cannot do the following:

annot = jso.AddAnnot
annot.Type = "Text"
annot.page = 0
...

Instead, you must obtain the properties object of Annot using annot.getProps, and use that object for read or write access. To save changes back to the original Annot, call annot.setProps with the modified properties object.

Third, note the use of JSObject ‘s color property. This object defines several simple colors such as red, green, and blue. In working with colors, you may need a greater range of colors than is available through this object. Also, there is a performance hit associated with every call to JSObject. To set colors more efficiently, you can use code such as the following, which sets the annot’s strokeColor to red directly, bypassing the color object.

dim color(0 to 3) as Variant
color(0) = "RGB"
color(1) = 1#
color(2) = 0#
color(3) = 0#
annot.strokeColor = color

You can use this technique anywhere a color array is needed as a parameter to a JSObject routine. The example sets the colorspace to RGB and specifies floating point values ranging from 0 to 1 for red, green, and blue. Note the use of the # character following the color values. These are required, since they tell Visual Basic that the array element should be set to a floating point value, rather than an integer. It is also important to declare the array as containing Variants, because it contains both strings and floating point values. The other color spaces (“T”, “G”, “CMYK”) have varying requirements for array length. For more information, refer to the Color object in the JavaScript for Acrobat API Reference.

Note

If you want users to be able to edit annotations, set the JavaScript property Collab.showAnnotsToolsWhenNoCollab to true.

Spell-checking a document¶

Acrobat includes a plug-in that can scan a document for spelling errors. The plug-in also provides JavaScript methods that can be accessed using JSObject. In this example, you start with the source code from the example Adding an annotation and make the following changes:

Add a List View control to the main form. Keep the default name ListView1 for the control.
Replace the code in the existing Command1_Click routine with the following:
Spell-checking a document

Private Sub Button1_Click(ByVal sender As System.Object, ByVal e As

         System.EventArgs) Handles Button1.Click
    Dim pdDoc As Acrobat.CAcroPDDoc
    Dim jso As Object
    Dim path As String
    Dim count As Integer
    Dim i As Integer, j As Integer
    Dim word As Variant
    Dim result As Variant
    Dim foundErr As Boolean

    OpenFileDialog1.ShowDialog()
    path = OpenFileDialog1.FileName
    foundErr = False
    pdDoc = CreateObject("AcroExch.PDDoc")

    If pdDoc.Open(path) Then
        jso = pdDoc.GetJSObject
        If Not jso Is Nothing Then
            count = jso.getPageNumWords(0)
            For i = 0 To count - 1
                word = jso.getPageNthWord(0, i)
                If VarType(word) = vbString Then
                    result = jso.spell.checkWord(word)
                    If IsArray(result) Then
                        foundErr = True
                        ListView1.Items.Add (word & " is misspelled.")
                        ListView1.Items.Add ("Suggestions:")
                        For j = LBound(result) To UBound(result)
                            ListView1.Items.Add (result(j))
                        Next j
                        ListView1.Items.Add ("")
                    End If
                End If
            Next i
            jso = Nothing
            pdDoc.Close

            If Not foundErr Then
                ListView1.Items.Add ("No spelling errors found in " & path)
            End If
        End If
    Else
        MsgBox "Failed to open " & path
    End If

    pdDoc = Nothing
End Sub

In this example, note the use of the Spell object’s check method. As described in the JavaScript for Acrobat API Reference, this method takes a word as input, and returns a null object if the word is found in the dictionary, or an array of suggested words if the word is not found.

The safest approach when storing the return value of a JSObject method call is to use a Variant. You can use the IsArray function to determine if the Variant is an array, and write code to handle that situation accordingly. In this simple example, if the program finds an array of suggested words, it dumps them out to the List View control.

Tips for translating JavaScript to JSObject¶

Covering every method available to JSObject is beyond the scope of this document. However, the JavaScript for Acrobat API Reference covers the subject in detail, and much can be inferred from the reference by keeping a few basic facts in mind:

Most of the objects and methods in the reference are available in Visual Basic, but not all. In particular, any JavaScript object that requires the new operator for construction cannot be created in Visual Basic. This includes the Report object.
The Annots object is unusual in that it requires JSObject to set and get its properties as a separate object using the getProps and setProps methods.
If you are unsure what type to use to declare a variable, declare it as a Variant. This gives Visual Basic more flexibility for type conversion, and helps prevent runtime errors.
JSObject cannot add new properties, methods, or objects to JavaScript. Due to this limitation, the global.setPersistent property is not meaningful.
JSObject is case-insensitive. Visual Basic often capitalizes leading characters of an identifier and prevents you from changing its case. Don’t be concerned about this, since JSObject ignores case when matching the identifier to its JavaScript equivalent.
JSObject always returns values as Variants. This includes property gets as well as return values from method calls. An empty Variant is used when a null return value is expected. When JSObject returns an array, each element in the array is a Variant. To determine the actual data type of a Variant, use the utility functions IsArray, IsNumeric, IsEmpty, IsObject, and VarType from the Information module of the Visual Basic for Applications (VBA) library.
JSObject can process most elemental Visual Basic types for setting properties and for and input parameters for method calls, including Variant, Array, Boolean, String, Date, Double, Long, Integer, and Byte. JSObject can accept Object parameters, but only when the Object is the result of a property get or method call to a JSObject. JSObject fails to accept values of type Error and Currency.

Summary of OLE objects and methods¶

OLE automation support is provided by a set of classes in the Acrobat API.

The following diagram shows the objects and methods that are used in OLE. The arrows indicate bridge methods, which are methods that can get an object from a related object of a different layer. For example, if you want to get the PDDoc associated with a particular AVDoc object, you can use the GetPDDoc method in the AcroExch.AVDoc object.

OLE objects and methods

Using OLE¶

OLE capabilities in Acrobat¶

On-screen rendering¶

Remote control of Acrobat¶

PDF browser controls¶

Development environment considerations¶

Environment configuration¶

Necessary C knowledge¶

Using the Acrobat OLE interfaces¶

About the CAcro classes¶

About the COleDispatchDriver class¶

Using COleDispatchDriver objects and methods¶

Using the JSObject interface¶

Adding a reference to the Acrobat type library¶

Creating a simple application¶

Working with annotations¶

Spell-checking a document¶

Tips for translating JavaScript to JSObject¶

Other development topics¶

Synchronous messaging¶

MDI applications¶

Event handling in child windows¶

Determining if an Acrobat application is running¶

Exiting from an application¶

Summary of OLE objects and methods¶