Add newline-before-endstream option (fixes #103)

Jay Berkenbilt
1 parent 6a7d53ad
Showing 6 changed files with 41 additions and 2 deletions
ChangeLog
include/qpdf/QPDFWriter.hh
libqpdf/QPDFWriter.cc
qpdf/qpdf.cc
qpdf/qtest/qpdf.test
qpdf/qtest/qpdf/newline-before-endstream.pdf
 2017-07-29  Jay Berkenbilt  <ejb@ql.org>
+	* Add --newline-before-endstream command-line option and
+	setNewlineBeforeEndstream method to QPDFWriter. This forces qpdf
+	to always add a newline before the endstream keyword. It is a
+	necessary but not sufficient condition for PDF/A compliance. Fixes
+	#103.
+
 	* Handle zlib data errors when decoding streams. Fixes #106.
 	* Improve handling of files where the "stream" keyword is not
@@ -161,6 +161,11 @@ class QPDFWriter
     QPDF_DLL
     void setPreserveUnreferencedObjects(bool);
+    // Always write a newline before the endstream keyword. This helps
+    // with PDF/A compliance, though it is not sufficient for it.
+    QPDF_DLL
+    void setNewlineBeforeEndstream(bool);
+
     // Set the minimum PDF version.  If the PDF version of the input
     // file (or previously set minimum version) is less than the
     // version passed to this method, the PDF version of the output
@@ -434,6 +439,7 @@ class QPDFWriter
     bool qdf_mode;
     bool precheck_streams;
     bool preserve_unreferenced_objects;
+    bool newline_before_endstream;
     bool static_id;
     bool suppress_original_object_ids;
     bool direct_stream_lengths;
@@ -59,6 +59,7 @@ QPDFWriter::init()
     qdf_mode = false;
     precheck_streams = false;
     preserve_unreferenced_objects = false;
+    newline_before_endstream = false;
     static_id = false;
     suppress_original_object_ids = false;
     direct_stream_lengths = true;
@@ -191,6 +192,12 @@ QPDFWriter::setPreserveUnreferencedObjects(bool val)
 }
 void
+QPDFWriter::setNewlineBeforeEndstream(bool val)
+{
+    this->newline_before_endstream = val;
+}
+
+void
 QPDFWriter::setMinimumPDFVersion(std::string const& version)
 {
     setMinimumPDFVersion(version, 0);
@@ -1580,7 +1587,7 @@ QPDFWriter::unparseObject(QPDFObjectHandle object, int level,
 	char last_char = this->pipeline->getLastChar();
 	popPipelineStack();
-	if (this->qdf_mode)
+	if (this->qdf_mode || this->newline_before_endstream)
 	{
 	    if (last_char != '\n')
 	    {
@@ -204,6 +204,7 @@ familiar with the PDF file format or who are PDF developers.\n\
 --ignore-xref-streams     tells qpdf to ignore any cross-reference streams\n\
 --precheck-streams        precheck ability to decode streams\n\
 --preserve-unreferenced   preserve unreferenced objects\n\
+--newline-before-endstream  always put a newline before endstream\n\
 --qdf                     turns on \"QDF mode\" (below)\n\
 --min-version=version     sets the minimum PDF version of the output file\n\
 --force-version=version   forces this to be the PDF version of the output file\n\
@@ -1032,6 +1033,7 @@ int main(int argc, char* argv[])
     bool qdf_mode = false;
     bool precheck_streams = false;
     bool preserve_unreferenced_objects = false;
+    bool newline_before_endstream = false;
     std::string min_version;
     std::string force_version;
@@ -1225,6 +1227,10 @@ int main(int argc, char* argv[])
 	    {
 		preserve_unreferenced_objects = true;
 	    }
+	    else if (strcmp(arg, "newline-before-endstream") == 0)
+	    {
+		newline_before_endstream = true;
+	    }
 	    else if (strcmp(arg, "min-version") == 0)
 	    {
 		if (parameter == 0)
@@ -1724,6 +1730,10 @@ int main(int argc, char* argv[])
             {
                 w.setPreserveUnreferencedObjects(true);
             }
+            if (newline_before_endstream)
+            {
+                w.setNewlineBeforeEndstream(true);
+            }
 	    if (normalize_set)
 	    {
 		w.setContentNormalization(normalize);
@@ -206,7 +206,7 @@ $td-&gt;runtest(&quot;remove page we don&#39;t have&quot;,
 show_ntests();
 # ----------
 $td->notify("--- Miscellaneous Tests ---");
-$n_tests += 89;
+$n_tests += 91;
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -618,6 +618,16 @@ $td-&gt;runtest(&quot;split content stream errors&quot;,
               $td->EXIT_STATUS => 3},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("newline before endstream",
+             {$td->COMMAND =>
+                  "qpdf --static-id --newline-before-endstream" .
+                  " minimal.pdf a.pdf"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check output",
+             {$td->FILE => "a.pdf"},
+             {$td->FILE => "newline-before-endstream.pdf"});
+
 show_ntests();
 # ----------