public override PdfObject GetDestinationPage(IDictionary <String, PdfObject> names)
{
PdfObject firstObj = ((PdfArray)GetPdfObject()).Get(0);
if (firstObj.IsDictionary())
{
PdfStructElem structElem = new PdfStructElem((PdfDictionary)firstObj);
while (true)
{
IList <IStructureNode> kids = structElem.GetKids();
IStructureNode firstKid = kids.Count > 0 ? kids[0] : null;
if (firstKid is PdfMcr)
{
return(((PdfMcr)firstKid).GetPageObject());
}
else
{
if (firstKid is PdfStructElem)
{
structElem = (PdfStructElem)firstKid;
}
else
{
break;
}
}
}
}
return(null);
}
/// <summary>Transforms root tags in a way that complies with the PDF References.</summary>
/// <remarks>
/// Transforms root tags in a way that complies with the PDF References.
/// <br/><br/>
/// PDF Reference
/// 10.7.3 Grouping Elements:
/// <br/><br/>
/// For most content extraction formats, the document must be a tree with a single top-level element;
/// the structure tree root (identified by the StructTreeRoot entry in the document catalog) must have
/// only one child in its K (kids) array. If the PDF file contains a complete document, the structure
/// type Document is recommended for this top-level element in the logical structure hierarchy. If the
/// file contains a well-formed document fragment, one of the structure types Part, Art, Sect, or Div
/// may be used instead.
/// </remarks>
public virtual void NormalizeDocumentRootTag()
{
// in this method we could deal with existing document, so we don't won't to throw exceptions here
bool forbid = forbidUnknownRoles;
forbidUnknownRoles = false;
IList <IPdfStructElem> rootKids = document.GetStructTreeRoot().GetKids();
if (rootKids.Count == 1 && allowedRootTagRoles.Contains(rootKids[0].GetRole()))
{
rootTagElement = (PdfStructElem)rootKids[0];
}
else
{
PdfStructElem prevRootTag = rootTagElement;
document.GetStructTreeRoot().GetPdfObject().Remove(PdfName.K);
if (prevRootTag == null)
{
rootTagElement = document.GetStructTreeRoot().AddKid(new PdfStructElem(document, PdfName.Document));
}
else
{
document.GetStructTreeRoot().AddKid(rootTagElement);
if (!PdfName.Document.Equals(rootTagElement.GetRole()))
{
WrapAllKidsInTag(rootTagElement, rootTagElement.GetRole());
rootTagElement.SetRole(PdfName.Document);
}
}
int originalRootKidsIndex = 0;
bool isBeforeOriginalRoot = true;
foreach (IPdfStructElem elem in rootKids)
{
// StructTreeRoot kids are always PdfStructElem, so we are save here to cast it
PdfStructElem kid = (PdfStructElem)elem;
if (kid.GetPdfObject() == rootTagElement.GetPdfObject())
{
isBeforeOriginalRoot = false;
continue;
}
bool kidIsDocument = PdfName.Document.Equals(kid.GetRole());
if (isBeforeOriginalRoot)
{
rootTagElement.AddKid(originalRootKidsIndex, kid);
originalRootKidsIndex += kidIsDocument ? kid.GetKids().Count : 1;
}
else
{
rootTagElement.AddKid(kid);
}
if (kidIsDocument)
{
RemoveOldRoot(kid);
}
}
}
forbidUnknownRoles = forbid;
}
private void WrapAllKidsInTag(PdfStructElem parent, PdfName wrapTagRole)
{
int kidsNum = parent.GetKids().Count;
TagTreePointer tagPointer = new TagTreePointer(document);
tagPointer.SetCurrentStructElem(parent).AddTag(0, wrapTagRole);
TagTreePointer newParentOfKids = new TagTreePointer(tagPointer);
tagPointer.MoveToParent();
for (int i = 0; i < kidsNum; ++i)
{
tagPointer.RelocateKid(1, newParentOfKids);
}
}
private void FlushStructElementAndItKids(PdfStructElem elem)
{
if (waitingTagToAssociatedObj.ContainsKey(elem.GetPdfObject()))
{
return;
}
foreach (IStructureNode kid in elem.GetKids())
{
if (kid is PdfStructElem)
{
FlushStructElementAndItKids((PdfStructElem)kid);
}
}
elem.Flush();
}
private void FlushStructElementAndItKids(PdfStructElem elem)
{
if (connectedStructToModel.ContainsKey(elem.GetPdfObject()))
{
return;
}
foreach (IPdfStructElem kid in elem.GetKids())
{
if (kid is PdfStructElem)
{
FlushStructElementAndItKids((PdfStructElem)kid);
}
}
elem.Flush();
}
/// <summary>Removes the current tag.</summary>
/// <remarks>
/// Removes the current tag. If it has kids, they will become kids of the current tag parent.
/// This method call moves this
/// <c>TagTreePointer</c>
/// to the current tag parent.
/// <br /><br />
/// You cannot remove root tag, and also you cannot remove the tag if it's parent is already flushed;
/// in this two cases an exception will be thrown.
/// </remarks>
/// <returns>
/// this
/// <see cref="TagStructureContext"/>
/// instance.
/// </returns>
public virtual iText.Kernel.Pdf.Tagutils.TagTreePointer RemoveTag()
{
PdfStructElem currentStructElem = GetCurrentStructElem();
IStructureNode parentElem = currentStructElem.GetParent();
if (parentElem is PdfStructTreeRoot)
{
throw new PdfException(PdfException.CannotRemoveDocumentRootTag);
}
IList <IStructureNode> kids = currentStructElem.GetKids();
PdfStructElem parent = (PdfStructElem)parentElem;
if (parent.IsFlushed())
{
throw new PdfException(PdfException.CannotRemoveTagBecauseItsParentIsFlushed);
}
// remove waiting tag state if tag is removed
Object objForStructDict = tagStructureContext.GetWaitingTagsManager().GetObjForStructDict(currentStructElem
.GetPdfObject());
tagStructureContext.GetWaitingTagsManager().RemoveWaitingState(objForStructDict);
int removedKidIndex = parent.RemoveKid(currentStructElem);
PdfIndirectReference indRef = currentStructElem.GetPdfObject().GetIndirectReference();
if (indRef != null)
{
// TODO how about possible references to structure element from refs or structure destination for instance?
indRef.SetFree();
}
foreach (IStructureNode kid in kids)
{
if (kid is PdfStructElem)
{
parent.AddKid(removedKidIndex++, (PdfStructElem)kid);
}
else
{
PdfMcr mcr = PrepareMcrForMovingToNewParent((PdfMcr)kid, parent);
parent.AddKid(removedKidIndex++, mcr);
}
}
currentStructElem.GetPdfObject().Clear();
SetCurrentStructElem(parent);
return(this);
}
private void WrapAllKidsInTag(PdfStructElem parent, PdfName wrapTagRole, PdfNamespace wrapTagNs)
{
int kidsNum = parent.GetKids().Count;
TagTreePointer tagPointer = new TagTreePointer(parent, document);
tagPointer.AddTag(0, wrapTagRole.GetValue());
if (context.TargetTagStructureVersionIs2())
{
tagPointer.GetProperties().SetNamespace(wrapTagNs);
}
TagTreePointer newParentOfKids = new TagTreePointer(tagPointer);
tagPointer.MoveToParent();
for (int i = 0; i < kidsNum; ++i)
{
tagPointer.RelocateKid(1, newParentOfKids);
}
}
// it is StructTreeRoot
// should never happen as we always should have only one root tag and we don't remove it
private void FlushParentIfBelongsToPage(PdfStructElem parent, PdfPage currentPage)
{
if (parent.IsFlushed() || connectedStructToModel.ContainsKey(parent.GetPdfObject()) || parent.GetPdfObject
() == rootTagElement.GetPdfObject())
{
return;
}
IList <IPdfStructElem> kids = parent.GetKids();
bool allKidsBelongToPage = true;
foreach (IPdfStructElem kid in kids)
{
if (kid is PdfMcr)
{
PdfDictionary kidPage = ((PdfMcr)kid).GetPageObject();
if (!kidPage.IsFlushed() && !kidPage.Equals(currentPage.GetPdfObject()))
{
allKidsBelongToPage = false;
break;
}
}
else
{
if (kid is PdfStructElem)
{
// If kid is structElem and was already flushed then in kids list there will be null for it instead of
// PdfStructElem. And therefore if we get into this if clause it means that some StructElem wasn't flushed.
allKidsBelongToPage = false;
break;
}
}
}
if (allKidsBelongToPage)
{
IPdfStructElem parentsParent = parent.GetParent();
parent.Flush();
if (parentsParent is PdfStructElem)
{
FlushParentIfBelongsToPage((PdfStructElem)parentsParent, currentPage);
}
}
return;
}
internal virtual void FlushParentIfBelongsToPage(PdfStructElem parent, PdfPage currentPage)
{
if (parent.IsFlushed() || waitingTagsManager.GetObjForStructDict(parent.GetPdfObject()) != null || parent.
GetParent() is PdfStructTreeRoot)
{
return;
}
IList <IStructureNode> kids = parent.GetKids();
bool readyToBeFlushed = true;
foreach (IStructureNode kid in kids)
{
if (kid is PdfMcr)
{
PdfDictionary kidPage = ((PdfMcr)kid).GetPageObject();
if (!kidPage.IsFlushed() && (currentPage == null || !kidPage.Equals(currentPage.GetPdfObject())))
{
readyToBeFlushed = false;
break;
}
}
else
{
if (kid is PdfStructElem)
{
// If kid is structElem and was already flushed then in kids list there will be null for it instead of
// PdfStructElement. And therefore if we get into this if-clause it means that some StructElem wasn't flushed.
readyToBeFlushed = false;
break;
}
}
}
if (readyToBeFlushed)
{
IStructureNode parentsParent = parent.GetParent();
parent.Flush();
if (parentsParent is PdfStructElem)
{
FlushParentIfBelongsToPage((PdfStructElem)parentsParent, currentPage);
}
}
}
protected internal virtual void InspectKid(IStructureNode kid)
{
try {
if (kid is PdfStructElem)
{
PdfStructElem structElemKid = (PdfStructElem)kid;
PdfName s = structElemKid.GetRole();
String tagN = s.GetValue();
String tag = FixTagName(tagN);
@out.Write("<");
@out.Write(tag);
InspectAttributes(structElemKid);
@out.Write(">" + Environment.NewLine);
PdfString alt = (structElemKid).GetAlt();
if (alt != null)
{
@out.Write("<alt><![CDATA[");
@out.Write(iText.IO.Util.StringUtil.ReplaceAll(alt.GetValue(), "[\\000]*", ""));
@out.Write("]]></alt>" + Environment.NewLine);
}
InspectKids(structElemKid.GetKids());
@out.Write("</");
@out.Write(tag);
@out.Write(">" + Environment.NewLine);
}
else
{
if (kid is PdfMcr)
{
ParseTag((PdfMcr)kid);
}
else
{
@out.Write(" <flushedKid/> ");
}
}
}
catch (System.IO.IOException e) {
throw new iText.IO.IOException(iText.IO.IOException.UnknownIOException, e);
}
}
private void AddStructTreeRootKidsToTheRootTag(IList <IStructureNode> rootKids)
{
int originalRootKidsIndex = 0;
bool isBeforeOriginalRoot = true;
foreach (IStructureNode elem in rootKids)
{
// StructTreeRoot kids are always PdfStructElement, so we are save here to cast it
PdfStructElem kid = (PdfStructElem)elem;
if (kid.GetPdfObject() == rootTagElement.GetPdfObject())
{
isBeforeOriginalRoot = false;
continue;
}
// This boolean is used to "flatten" possible deep "stacking" of the tag structure in case of the multiple pages copying operations.
// This could happen due to the wrapping of all the kids in the createNewRootTag or ensureExistingRootTagIsDocument methods.
// And therefore, we don't need here to resolve mappings, because we exactly know which role we set.
bool kidIsDocument = PdfName.Document.Equals(kid.GetRole());
if (kidIsDocument && kid.GetNamespace() != null && context.TargetTagStructureVersionIs2())
{
// we flatten only tags of document role in standard structure namespace
String kidNamespaceName = kid.GetNamespace().GetNamespaceName();
kidIsDocument = StandardNamespaces.PDF_1_7.Equals(kidNamespaceName) || StandardNamespaces.PDF_2_0.Equals(kidNamespaceName
);
}
if (isBeforeOriginalRoot)
{
rootTagElement.AddKid(originalRootKidsIndex, kid);
originalRootKidsIndex += kidIsDocument ? kid.GetKids().Count : 1;
}
else
{
rootTagElement.AddKid(kid);
}
if (kidIsDocument)
{
RemoveOldRoot(kid);
}
}
}
