Handle root /Pages pointing to other than page tree root (fixes #398)

Jay Berkenbilt
1 parent 6d2b4d8f
Showing 6 changed files with 126 additions and 3 deletions
ChangeLog
libqpdf/QPDF_pages.cc
qpdf/qtest/qpdf.test
qpdf/qtest/qpdf/pages-is-page-out.pdf
qpdf/qtest/qpdf/pages-is-page.out
qpdf/qtest/qpdf/pages-is-page.pdf
+2020-02-22  Jay Berkenbilt  <ejb@ql.org>
+
+	* Detect, warn, and correct the case of /Pages in the document
+	catalog incorrectly pointing to a page or intermediate node
+	instead of the root of the pages tree. Fixes #398.
+
 2020-01-26  Jay Berkenbilt  <ejb@ql.org>
 	* 9.1.1: release
@@ -49,8 +49,36 @@ QPDF::getAllPages()
     {
         std::set<QPDFObjGen> visited;
         std::set<QPDFObjGen> seen;
-	getAllPagesInternal(getRoot().getKey("/Pages"), this->m->all_pages,
-                            visited, seen);
+        QPDFObjectHandle pages = getRoot().getKey("/Pages");
+        bool warned = false;
+        bool changed_pages = false;
+        while (pages.isDictionary() && pages.hasKey("/Parent"))
+        {
+            if (seen.count(pages.getObjGen()))
+            {
+                // loop -- will be detected again and reported later
+                break;
+            }
+            // Files have been found in the wild where /Pages in the
+            // catalog points to the first page. Try to work around
+            // this and similar cases with this heuristic.
+            if (! warned)
+            {
+                getRoot().warnIfPossible(
+                    "document page tree root (root -> /Pages) doesn't point"
+                    " to the root of the page tree; attempting to correct");
+                warned = true;
+            }
+            seen.insert(pages.getObjGen());
+            changed_pages = true;
+            pages = pages.getKey("/Parent");
+        }
+        if (changed_pages)
+        {
+            getRoot().replaceKey("/Pages", pages);
+        }
+        seen.clear();
+	getAllPagesInternal(pages, this->m->all_pages, visited, seen);
     }
     return this->m->all_pages;
 }
@@ -2779,7 +2779,7 @@ for (my $n = 16; $n &lt;= 19; ++$n)
 show_ntests();
 # ----------
 $td->notify("--- Specific File Tests ---");
-$n_tests += 7;
+$n_tests += 9;
 # Special PDF files that caused problems at some point
@@ -2810,6 +2810,14 @@ $td-&gt;runtest(&quot;check output&quot;,
 $td->runtest("check fix-qdf idempotency",
 	     {$td->COMMAND => "fix-qdf a.pdf"},
 	     {$td->FILE => "a.pdf", $td->EXIT_STATUS => 0});
+$td->runtest("pages points to page",
+             {$td->COMMAND =>
+                  "qpdf --static-id --linearize pages-is-page.pdf a.pdf"},
+             {$td->FILE => "pages-is-page.out", $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "pages-is-page-out.pdf"});
 show_ntests();
 # ----------
+WARNING: pages-is-page.pdf, object 1 0 at offset 19: document page tree root (root -> /Pages) doesn't point to the root of the page tree; attempting to correct
+qpdf: operation succeeded with warnings; resulting file may have some problems
+%PDF-1.3
+1 0 obj
+<<
+  /Type /Catalog
+  /Pages 3 0 R
+>>
+endobj
+
+2 0 obj
+<<
+  /Type /Pages
+  /Kids [
+    3 0 R
+  ]
+  /Count 1
+>>
+endobj
+
+3 0 obj
+<<
+  /Type /Page
+  /Parent 2 0 R
+  /MediaBox [0 0 612 792]
+  /Contents 4 0 R
+  /Resources <<
+    /ProcSet 5 0 R
+    /Font <<
+      /F1 6 0 R
+    >>
+  >>
+>>
+endobj
+
+4 0 obj
+<<
+  /Length 44
+>>
+stream
+BT
+  /F1 24 Tf
+  72 720 Td
+  (Potato) Tj
+ET
+endstream
+endobj
+
+5 0 obj
+[
+  /PDF
+  /Text
+]
+endobj
+
+6 0 obj
+<<
+  /Type /Font
+  /Subtype /Type1
+  /Name /F1
+  /BaseFont /Helvetica
+  /Encoding /WinAnsiEncoding
+>>
+endobj
+
+xref
+0 7
+0000000000 65535 f 
+0000000009 00000 n 
+0000000063 00000 n 
+0000000135 00000 n 
+0000000307 00000 n 
+0000000403 00000 n 
+0000000438 00000 n 
+trailer <<
+  /Size 7
+  /Root 1 0 R
+>>
+startxref
+556
+%%EOF