From 57c01ef81ff11b208188bbdedc98b6ce20c786b3 Mon Sep 17 00:00:00 2001
From: Jay Berkenbilt <ejb@ql.org>
Date: Sun, 26 Jan 2020 16:43:55 -0500
Subject: [PATCH] In qdf mode, don't write extra XRef streams (fixes #386)

fix-qdf assumes there is exactly one XRef stream and that it is at the
end of the file.
---
 ChangeLog                                    |   7 +++++++
 libqpdf/QPDFWriter.cc                        |  15 +++++++++++++++
 qpdf/qpdf.testcov                            |   1 +
 qpdf/qtest/qpdf.test                         |  12 +++++++++++-
 qpdf/qtest/qpdf/compress-objstm-xref-qdf.pdf | Bin 0 -> 1249 bytes
 5 files changed, 34 insertions(+), 1 deletion(-)
 create mode 100644 qpdf/qtest/qpdf/compress-objstm-xref-qdf.pdf

diff --git a/ChangeLog b/ChangeLog
index 38749752..b65581ba 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,12 @@
 2020-01-26  Jay Berkenbilt  <ejb@ql.org>
 
+	* Bug fix: in qdf mode, do not write out any XRef streams that may
+	have appeared in the original file. These are usually
+	unreferenced, but with --preserve-unreferenced, they could be
+	written out, which breaks fix-qdf's assumption that there is at
+	most one XRef stream and that it appears at the end of the file.
+	Fixes #386.
+
 	* Bug fix: when externalizing inline images, a colorspace value
 	that was a lookup key in the page's /Resource -> /ColorSpace
 	dictionary was not properly handled. Fixes #392.
diff --git a/libqpdf/QPDFWriter.cc b/libqpdf/QPDFWriter.cc
index 7b769ae4..5f97117d 100644
--- a/libqpdf/QPDFWriter.cc
+++ b/libqpdf/QPDFWriter.cc
@@ -1241,6 +1241,21 @@ QPDFWriter::enqueueObject(QPDFObjectHandle object)
                 " another file.");
         }
 
+        if (this->m->qdf_mode &&
+            object.isStream() && object.getDict().getKey("/Type").isName() &&
+            (object.getDict().getKey("/Type").getName() == "/XRef"))
+        {
+            // As a special case, do not output any extraneous XRef
+            // streams in QDF mode. Doing so will confuse fix-qdf,
+            // which expects to see only one XRef stream at the end of
+            // the file. This case can occur when creating a QDF from
+            // a file with object streams when preserving unreferenced
+            // objects since the old cross reference streams are not
+            // actually referenced by object number.
+            QTC::TC("qpdf", "QPDFWriter ignore XRef in qdf mode");
+            return;
+        }
+
 	QPDFObjGen og = object.getObjGen();
 
 	if (this->m->obj_renumber.count(og) == 0)
diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index db3de950..f16f0364 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -448,3 +448,4 @@ QPDFWriter stream in ostream 0
 QPDFObjectHandle duplicate dict key 0
 QPDFWriter no encryption sig contents 0
 QPDFPageObjectHelper colorspace lookup 0
+QPDFWriter ignore XRef in qdf mode 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index f9bab01f..e42e204a 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -2779,7 +2779,7 @@ 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
 
@@ -2800,6 +2800,16 @@ $td->runtest("compress objstm and xref",
 $td->runtest("check output",
 	     {$td->FILE => "a.pdf"},
 	     {$td->FILE => "compress-objstm-xref.pdf"});
+$td->runtest("qdf + preserved-unreferenced + xref streams",
+	     {$td->COMMAND => "qpdf --qdf --preserve-unreferenced" .
+                  " --static-id compress-objstm-xref.pdf a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "compress-objstm-xref-qdf.pdf"});
+$td->runtest("check fix-qdf idempotency",
+	     {$td->COMMAND => "fix-qdf a.pdf"},
+	     {$td->FILE => "a.pdf", $td->EXIT_STATUS => 0});
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/compress-objstm-xref-qdf.pdf b/qpdf/qtest/qpdf/compress-objstm-xref-qdf.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..d6b78a053f8f0a5b242fd053816b9373e6ff3da1
GIT binary patch
literal 1249
zcmah|&2G~`5KdZ7`v%<R5(y;0_U_t_OB9LbM^r!?5=%vjdT`@yaFlv0ds77-hZo^F
zcn9v>5Ho9=2+1u`#@^Xo&-cyDH@(?-`qF#lv)=b#-+r^+N9($bdEkP+-mt-d0XXrO
zI|a_g^=4si$$zi%wfPJo7v!CTPp@fKlm_|%8xC1%3YFY44lWZAg-inXIP*dJAq&74
z2-pLRr&b24yC(!kE`I@8o~k{#uR#~tI?I!7b9Od90bX;OC2LiJK%Vmh#Zh9CtzHv_
z2LVI_;I#vY8epV%xdD%Yzsu4Rmbe0tHQ?0&)G~P#QZ|sh9a0io3WIP}#YZ|fDmNvB
zO@bd(nkA=tU-NJ+o9Y>0ZHZeL(c3bcB?W$dD9>D#dRMHlVm3at3lSQJTn~a83INMT
z@tj$qR|{ppuTSiLtp=j|XwrU@v}jM-XH-!K6-cjG?G3wnxNAWZO}0l1>cX8SrLw6x
zZ`Jl*nQWC%_({IfX_l|Sxy<r6d6_jEba<ZJqP(ZJ)r;M=v1%vH+nq}M6qC|bd0LAb
zS2Q6+`ER#wbOCo&A4z{L^2L5K>M4C%wYnarzO6AL)ZtHKqK0kZ65<<NhaCSd#3`<y
z&9q63eh%@5P2v_(6pbLHq(jRy`Ukbj>eF0N=w_fL>B+f-B_J%CmC{^moJz|M4;R^2
zRH?mpHiqRuc+&GD9^g#(qc9R&^ku*UDMVlJF!G20zi~x%t+VdnuiJUrIl?De7wrh`
g80|^-_#Rh3(0+DWQZ7wWn0<i>d9F(nHMy9wKTU8&-T(jq

literal 0
HcmV?d00001