Add option to preserve unreferenced objects

Jay Berkenbilt
1 parent a94a729f
Showing 12 changed files with 212 additions and 1 deletions
ChangeLog
include/qpdf/QPDF.hh
include/qpdf/QPDFWriter.hh
libqpdf/QPDF.cc
libqpdf/QPDFWriter.cc
manual/qpdf-manual.xml
qpdf/qpdf.cc
qpdf/qpdf.testcov
qpdf/qtest/qpdf.test
qpdf/qtest/qpdf/unreferenced-dropped.pdf
qpdf/qtest/qpdf/unreferenced-objects.pdf
qpdf/qtest/qpdf/unreferenced-preserved.pdf
+2017-07-28  Jay Berkenbilt  <ejb@ql.org>
+
+	* Add --preserve-unreferenced command-line option and
+	setPreserveUnreferencedObjects method to QPDFWriter. This option
+	causes QPDFWriter to write all objects from the input file to the
+	output file regardless of whether the objects are referenced.
+	Objects are written to the output file in numerical order from the
+	input file. This option has no effect for linearized files.
+
 2017-07-27  Jay Berkenbilt  <ejb@ql.org>
  
 	* Add --precheck-streams command-line option and setStreamPrecheck
-	option to QPDFWriter to tell QPDFWriter to attempt decoding a
+	method to QPDFWriter to tell QPDFWriter to attempt decoding a
 	stream fully before deciding whether to filter it or not.
  
 	* Recover gracefully from streams that aren't filterable because
@@ -396,6 +396,12 @@ class QPDF
     QPDF_DLL
     void showXRefTable();
  
+    // Returns a list of indirect objects for every object in the xref
+    // table. Useful for discovering objects that are not otherwised
+    // referenced.
+    QPDF_DLL
+    std::vector<QPDFObjectHandle> getAllObjects();
+
     // Optimization support -- see doc/optimization.  Implemented in
     // QPDF_optimization.cc
  
@@ -155,6 +155,12 @@ class QPDFWriter
     QPDF_DLL
     void setPrecheckStreams(bool);
  
+    // Preserve unreferenced objects. The default behavior is to
+    // discard any object that is not visited during a traversal of
+    // the object structure from the trailer.
+    QPDF_DLL
+    void setPreserveUnreferencedObjects(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
@@ -427,6 +433,7 @@ class QPDFWriter
     qpdf_stream_data_e stream_data_mode;
     bool qdf_mode;
     bool precheck_streams;
+    bool preserve_unreferenced_objects;
     bool static_id;
     bool suppress_original_object_ids;
     bool direct_stream_lengths;
@@ -989,6 +989,22 @@ QPDF::showXRefTable()
     }
 }
  
+std::vector<QPDFObjectHandle>
+QPDF::getAllObjects()
+{
+    std::vector<QPDFObjectHandle> result;
+    for (std::map<QPDFObjGen, QPDFXRefEntry>::iterator iter =
+	     this->xref_table.begin();
+	 iter != this->xref_table.end(); ++iter)
+    {
+
+	QPDFObjGen const& og = (*iter).first;
+        result.push_back(QPDFObjectHandle::Factory::newIndirect(
+                             this, og.getObj(), og.getGen()));
+    }
+    return result;
+}
+
 void
 QPDF::setLastObjectDescription(std::string const& description,
 			       int objid, int generation)
@@ -58,6 +58,7 @@ QPDFWriter::init()
     stream_data_mode = qpdf_s_compress;
     qdf_mode = false;
     precheck_streams = false;
+    preserve_unreferenced_objects = false;
     static_id = false;
     suppress_original_object_ids = false;
     direct_stream_lengths = true;
@@ -184,6 +185,12 @@ QPDFWriter::setPrecheckStreams(bool val)
 }
  
 void
+QPDFWriter::setPreserveUnreferencedObjects(bool val)
+{
+    this->preserve_unreferenced_objects = val;
+}
+
+void
 QPDFWriter::setMinimumPDFVersion(std::string const& version)
 {
     setMinimumPDFVersion(version, 0);
@@ -3074,6 +3081,17 @@ QPDFWriter::writeStandard()
     writeHeader();
     writeString(this->extra_header_text);
  
+    if (this->preserve_unreferenced_objects)
+    {
+        QTC::TC("qpdf", "QPDFWriter preserve unreferenced standard");
+        std::vector<QPDFObjectHandle> all = this->pdf.getAllObjects();
+        for (std::vector<QPDFObjectHandle>::iterator iter = all.begin();
+             iter != all.end(); ++iter)
+        {
+            enqueueObject(*iter);
+        }
+    }
+
     // Put root first on queue.
     QPDFObjectHandle trailer = getTrimmedTrailer();
     enqueueObject(trailer.getKey("/Root"));
@@ -839,6 +839,27 @@ outfile.pdf&lt;/option&gt;
       </listitem>
      </varlistentry>
      <varlistentry>
+      <term><option>--preserve-unreferenced</option></term>
+      <listitem>
+       <para>
+        Tells qpdf to preserve objects that are not referenced when
+        writing the file. Ordinarily any object that is not referenced
+        in a traversal of the document from the trailer dictionary
+        will be discarded. This may be useful in working with some
+        damaged files or inspecting files with known unreferenced
+        objects.
+       </para>
+       <para>
+        This flag is ignored for linearized files and has the effect
+        of causing objects in the new file to be written in order by
+        object ID from the original file. This does not mean that
+        object numbers will be the same since qpdf may create stream
+        lengths as direct or indirect differently from the original
+        file, and the original file may have gaps in its numbering.
+       </para>
+      </listitem>
+     </varlistentry>
+     <varlistentry>
       <term><option>--qdf</option></term>
       <listitem>
        <para>
@@ -203,6 +203,7 @@ familiar with the PDF file format or who are PDF developers.\n\
 --object-streams=mode     controls handing of object streams\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\
 --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\
@@ -1030,6 +1031,7 @@ int main(int argc, char* argv[])
     bool ignore_xref_streams = false;
     bool qdf_mode = false;
     bool precheck_streams = false;
+    bool preserve_unreferenced_objects = false;
     std::string min_version;
     std::string force_version;
  
@@ -1219,6 +1221,10 @@ int main(int argc, char* argv[])
 	    {
 		precheck_streams = true;
 	    }
+	    else if (strcmp(arg, "preserve-unreferenced") == 0)
+	    {
+		preserve_unreferenced_objects = true;
+	    }
 	    else if (strcmp(arg, "min-version") == 0)
 	    {
 		if (parameter == 0)
@@ -1714,6 +1720,10 @@ int main(int argc, char* argv[])
             {
                 w.setPrecheckStreams(true);
             }
+            if (preserve_unreferenced_objects)
+            {
+                w.setPreserveUnreferencedObjects(true);
+            }
 	    if (normalize_set)
 	    {
 		w.setContentNormalization(normalize);
@@ -280,3 +280,4 @@ QPDFObjectHandle found fake 1
 QPDFObjectHandle no val for last key 0
 QPDF resolve failure to null 0
 QPDFWriter precheck stream 0
+QPDFWriter preserve unreferenced standard 0
@@ -743,6 +743,24 @@ $td-&gt;runtest(&quot;check output&quot;,
 	     {$td->FILE => "bad-data-precheck.pdf"});
 show_ntests();
 # ----------
+$td->notify("--- Preserve unreferenced objects ---");
+$n_tests += 4;
+
+$td->runtest("drop unused objects",
+	     {$td->COMMAND => "qpdf --static-id unreferenced-objects.pdf a.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "unreferenced-dropped.pdf"});
+$td->runtest("keep unused objects",
+	     {$td->COMMAND => "qpdf --static-id --preserve-unreferenced" .
+                  " unreferenced-objects.pdf a.pdf"},
+	     {$td->STRING => "", $td->EXIT_STATUS => 0});
+$td->runtest("check output",
+	     {$td->FILE => "a.pdf"},
+	     {$td->FILE => "unreferenced-preserved.pdf"});
+show_ntests();
+# ----------
 $td->notify("--- Copy Foreign Objects ---");
 $n_tests += 7;
  
+%PDF-1.3
+%¿÷¢þ
+%QDF-1.0
+
+2 0 obj
+<<
+  /Pages 1 0 R
+  /Type /Catalog
+>>
+endobj
+
+1 0 obj
+<<
+  /Count 1
+  /Kids [
+    3 0 R
+  ]
+  /Type /Pages
+>>
+endobj
+
+%% Page 1
+3 0 obj
+<<
+  /Contents 4 0 R
+  /MediaBox [
+    0
+    0
+    612
+    792
+  ]
+  /Parent 1 0 R
+  /Resources <<
+    /Font <<
+      /F1 6 0 R
+    >>
+    /ProcSet 9 0 R
+  >>
+  /Type /Page
+>>
+endobj
+
+%% Contents for page 1
+4 0 obj
+<<
+  /Length 5 0 R
+>>
+stream
+BT
+  /F1 24 Tf
+  72 720 Td
+  (Potato) Tj
+ET
+endstream
+endobj
+
+5 0 obj
+44
+endobj
+
+6 0 obj
+<<
+  /BaseFont /Helvetica
+  /Encoding /WinAnsiEncoding
+  /Name /F1
+  /Subtype /Type1
+  /Type /Font
+>>
+endobj
+
+7 0 obj
+[ 8 0 R ]
+endobj
+
+8 0 obj
+/Potato
+endobj
+
+9 0 obj
+[
+  /PDF
+  /Text
+]
+endobj
+
+xref
+0 10
+0000000000 65535 f 
+0000000079 00000 n 
+0000000025 00000 n 
+0000000161 00000 n 
+0000000376 00000 n 
+0000000475 00000 n 
+0000000494 00000 n 
+0000000612 00000 n 
+0000000638 00000 n 
+0000000662 00000 n 
+trailer <<
+  /Root 2 0 R
+  /Size 10
+  /ID [<5c2381b459937c988290150df782f1fd><5c2381b459937c988290150df782f1fd>]
+>>
+startxref
+697
+%%EOF