Properly handle pages with no contents (fixes #194)

Remove calls to assertPageObject(). All cases in the library that
called assertPageObject() work fine if you don't call
assertPageObject() because nothing assumes anything that was being
checked by that call. Removing the calls enables more files to be
successfully processed.

2 files changed, 23 insertions, 11 deletions

diff --git a/libqpdf/QPDFObjectHandle.cc b/libqpdf/QPDFObjectHandle.cc
index 2e9cc996..c178a492 100644
--- a/libqpdf/QPDFObjectHandle.cc
+++ b/libqpdf/QPDFObjectHandle.cc
@@ -932,8 +932,6 @@ QPDFObjectHandle::getGeneration() const
 std::map<std::string, QPDFObjectHandle>
 QPDFObjectHandle::getPageImages()
 {
-    assertPageObject();
-
     // Note: this code doesn't handle inherited resources.  If this
     // page dictionary doesn't have a /Resources key or has one whose
     // value is null or an empty dictionary, you are supposed to walk
@@ -1081,7 +1079,6 @@ QPDFObjectHandle::addPageContents(QPDFObjectHandle new_contents, bool first)
 void
 QPDFObjectHandle::rotatePage(int angle, bool relative)
 {
-    assertPageObject();
     if ((angle % 90) != 0)
     {
         throw std::runtime_error(
@@ -1137,7 +1134,6 @@ QPDFObjectHandle::rotatePage(int angle, bool relative)
 void
 QPDFObjectHandle::coalesceContentStreams()
 {
-    assertPageObject();
     QPDFObjectHandle contents = this->getKey("/Contents");
     if (contents.isStream())
     {
@@ -1218,7 +1214,6 @@ QPDFObjectHandle::parse(std::string const& object_str,
 void
 QPDFObjectHandle::pipePageContents(Pipeline* p)
 {
-    assertPageObject();
     std::string description = "page object " +
         QUtil::int_to_string(this->m->objid) + " " +
         QUtil::int_to_string(this->m->generation);
@@ -1256,7 +1251,6 @@ QPDFObjectHandle::pipeContentStreams(
 void
 QPDFObjectHandle::parsePageContents(ParserCallbacks* callbacks)
 {
-    assertPageObject();
     std::string description = "page object " +
         QUtil::int_to_string(this->m->objid) + " " +
         QUtil::int_to_string(this->m->generation);
@@ -1267,7 +1261,6 @@ QPDFObjectHandle::parsePageContents(ParserCallbacks* callbacks)
 void
 QPDFObjectHandle::filterPageContents(TokenFilter* filter, Pipeline* next)
 {
-    assertPageObject();
     std::string description = "token filter for page object " +
         QUtil::int_to_string(this->m->objid) + " " +
         QUtil::int_to_string(this->m->generation);
@@ -2222,8 +2215,29 @@ QPDFObjectHandle::assertNumber()
 bool
 QPDFObjectHandle::isPageObject()
 {
-    // Some PDF files have /Type broken on pages.
-    return (this->isDictionary() && this->hasKey("/Contents"));
+    // See comments in QPDFObjectHandle.hh.
+    if (! this->isDictionary())
+    {
+        return false;
+    }
+    if (this->hasKey("/Type"))
+    {
+        QPDFObjectHandle type = this->getKey("/Type");
+        if (type.isName() && (type.getName() == "/Page"))
+        {
+            return true;
+        }
+        // Files have been seen in the wild that have /Type (Page)
+        if (type.isString() && (type.getStringValue() == "Page"))
+        {
+            return true;
+        }
+    }
+    if (this->hasKey("/Contents"))
+    {
+        return true;
+    }
+    return false;
 }
 
 bool
diff --git a/libqpdf/QPDF_pages.cc b/libqpdf/QPDF_pages.cc
index 31307e4e..ea5afdb5 100644
--- a/libqpdf/QPDF_pages.cc
+++ b/libqpdf/QPDF_pages.cc
@@ -191,7 +191,6 @@ QPDF::insertPage(QPDFObjectHandle newpage, int pos)
     // pos = npages adds to the end.
 
     flattenPagesTree();
-    newpage.assertPageObject();
 
     if (! newpage.isIndirect())
     {
@@ -288,7 +287,6 @@ QPDF::addPage(QPDFObjectHandle newpage, bool first)
 int
 QPDF::findPage(QPDFObjectHandle& page)
 {
-    page.assertPageObject();
     return findPage(page.getObjGen());
 }