Handle direct page objects (fixes #164)

7 files changed, 144 insertions, 2 deletions

diff --git a/ChangeLog b/ChangeLog
index 180335f6..861ff64c 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2019-01-29  Jay Berkenbilt  <ejb@ql.org>
+
+	* Handle files with direct page objects, which is not allowed by
+	the PDF spec but has been seen in the wild. Fixes #164.
+
 2019-01-28  Jay Berkenbilt  <ejb@ql.org>
 
 	* Bug fix: when using --stream-data=compress, object streams and
diff --git a/libqpdf/QPDF_pages.cc b/libqpdf/QPDF_pages.cc
index 397c175a..01270652 100644
--- a/libqpdf/QPDF_pages.cc
+++ b/libqpdf/QPDF_pages.cc
@@ -105,7 +105,13 @@ QPDF::getAllPagesInternal3(QPDFObjectHandle cur_pages,
 	for (int i = 0; i < n; ++i)
 	{
             QPDFObjectHandle kid = kids.getArrayItem(i);
-            if (seen.count(kid.getObjGen()))
+            if (! kid.isIndirect())
+            {
+                QTC::TC("qpdf", "QPDF handle direct page object");
+                kid = makeIndirectObject(kid);
+                kids.setArrayItem(i, kid);
+            }
+            else if (seen.count(kid.getObjGen()))
             {
                 // Make a copy of the page. This does the same as
                 // shallowCopyPage in QPDFPageObjectHelper.
diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index 98857597..5150e567 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -429,3 +429,4 @@ QPDFPageObjectHelper non-trivial inheritance 0
 QPDFPageObjectHelper copy shared attribute 0
 qpdf from_nr from repeat_nr 0
 QPDF resolve duplicated page object 0
+QPDF handle direct page object 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index 85b9b8de..ca7ea12b 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -2461,7 +2461,7 @@ for (my $n = 16; $n <= 19; ++$n)
 show_ntests();
 # ----------
 $td->notify("--- Specific File Tests ---");
-$n_tests += 4;
+$n_tests += 7;
 
 # Special PDF files that caused problems at some point
 
@@ -2482,6 +2482,19 @@ $td->runtest("compress objstm and xref",
 $td->runtest("check output",
 	     {$td->FILE => "a.pdf"},
 	     {$td->FILE => "compress-objstm-xref.pdf"});
+$td->runtest("direct pages",
+             {$td->COMMAND =>
+                  "qpdf --static-id direct-pages.pdf --pages . -- a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "direct-pages-fixed.pdf"});
+$td->runtest("show direct pages",
+             {$td->COMMAND =>
+                  "qpdf --show-pages direct-pages.pdf"},
+             {$td->FILE => "direct-pages.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/direct-pages-fixed.pdf b/qpdf/qtest/qpdf/direct-pages-fixed.pdf
new file mode 100644
index 00000000..ab577e8c
--- /dev/null
+++ b/qpdf/qtest/qpdf/direct-pages-fixed.pdf
diff --git a/qpdf/qtest/qpdf/direct-pages.out b/qpdf/qtest/qpdf/direct-pages.out
new file mode 100644
index 00000000..4b4c3c47
--- /dev/null
+++ b/qpdf/qtest/qpdf/direct-pages.out
@@ -0,0 +1,6 @@
+page 1: 7 0 R
+  content:
+    3 0 R
+page 2: 8 0 R
+  content:
+    3 0 R
diff --git a/qpdf/qtest/qpdf/direct-pages.pdf b/qpdf/qtest/qpdf/direct-pages.pdf
new file mode 100644
index 00000000..c8f1c4ee
--- /dev/null
+++ b/qpdf/qtest/qpdf/direct-pages.pdf
@@ -0,0 +1,111 @@
+%PDF-1.3
+%¿÷¢þ
+%QDF-1.0
+
+%% Original object ID: 1 0
+1 0 obj
+<<
+  /Pages 2 0 R
+  /Type /Catalog
+>>
+endobj
+
+%% Original object ID: 2 0
+2 0 obj
+<<
+  /Count 2
+  /Kids [
+    <<
+      /Contents 3 0 R
+      /MediaBox [
+        0
+        0
+        612
+        792
+      ]
+      /Parent 2 0 R
+      /Resources <<
+        /Font <<
+          /F1 5 0 R
+        >>
+        /ProcSet 6 0 R
+      >>
+      /Type /Page
+    >>
+    <<
+      /Contents 3 0 R
+      /MediaBox [
+        0
+        0
+        612
+        792
+      ]
+      /Parent 2 0 R
+      /Resources <<
+        /Font <<
+          /F1 5 0 R
+        >>
+        /ProcSet 6 0 R
+      >>
+      /Type /Page
+    >>
+  ]
+  /Type /Pages
+>>
+endobj
+
+%% Contents for page 2
+%% Original object ID: 4 0
+3 0 obj
+<<
+  /Length 4 0 R
+>>
+stream
+BT
+  /F1 24 Tf
+  72 720 Td
+  (Potato) Tj
+ET
+endstream
+endobj
+
+4 0 obj
+44
+endobj
+
+%% Original object ID: 6 0
+5 0 obj
+<<
+  /BaseFont /Helvetica
+  /Encoding /WinAnsiEncoding
+  /Name /F1
+  /Subtype /Type1
+  /Type /Font
+>>
+endobj
+
+%% Original object ID: 7 0
+6 0 obj
+[
+  /PDF
+  /Text
+]
+endobj
+
+xref
+0 7
+0000000000 65535 f 
+0000000052 00000 n 
+0000000133 00000 n 
+0000000733 00000 n 
+0000000832 00000 n 
+0000000878 00000 n 
+0000001023 00000 n 
+trailer <<
+  /Root 1 0 R
+  /Size 7
+  /ID [<1323a5937c577a66735583a93698ce3c><372cbf44f6db88ab60d9263c0f0bd26a>]
+>>
+startxref
+1058
+%%EOF