Handle more broken files

Space rather than newline after xref, missing /ID in trailer for
encrypted file.  This enables qpdf to handle some files that xpdf can
handle.  Adobe reader can't necessarily handle them.

ChangeLog

@@ -1,3 +1,8 @@
+2013-06-15  Jay Berkenbilt  <ejb@ql.org>
+
+	* Handle some additional broken files with missing /ID in trailer
+	for encrypted files and with space rather than newline after xref.
+
 2013-06-14  Jay Berkenbilt  <ejb@ql.org>
 
 	* Detect and correct /Outlines dictionary being a direct object

libqpdf/QPDF.cc

@@ -430,11 +430,22 @@ QPDF::read_xref(qpdf_offset_t xref_offset)
     std::map<int, int> free_table;
     while (xref_offset)
     {
+        char buf[7];
+        memset(buf, 0, sizeof(buf));
 	this->file->seek(xref_offset, SEEK_SET);
-	std::string line = this->file->readLine(50);
-	if (line == "xref")
+	this->file->read(buf, sizeof(buf) - 1);
+        // The PDF spec says xref must be followed by a line
+        // terminator, but files exist in the wild where it is
+        // terminated by arbitrary whitespace.
+        PCRE xref_re("^xref\\s+");
+	PCRE::Match m = xref_re.match(buf);
+	if (m)
 	{
-	    xref_offset = read_xrefTable(this->file->tell());
+            QTC::TC("qpdf", "QPDF xref space",
+                    ((buf[4] == '\n') ? 0 :
+                     (buf[4] == '\r') ? 1 :
+                     (buf[4] == ' ') ? 2 : 9999));
+	    xref_offset = read_xrefTable(xref_offset + m.getMatch(0).length());
 	}
 	else
 	{

libqpdf/QPDF_encryption.cc

@@ -791,17 +791,24 @@ QPDF::initializeEncryption()
     // encryption dictionary.
     this->encrypted = true;
 
+    std::string id1;
     QPDFObjectHandle id_obj = this->trailer.getKey("/ID");
-    if (! (id_obj.isArray() &&
-	   (id_obj.getArrayNItems() == 2) &&
-	   id_obj.getArrayItem(0).isString()))
+    if ((id_obj.isArray() &&
+         (id_obj.getArrayNItems() == 2) &&
+         id_obj.getArrayItem(0).isString()))
     {
-	throw QPDFExc(qpdf_e_damaged_pdf, this->file->getName(),
-		      "trailer", this->file->getLastOffset(),
-		      "invalid /ID in trailer dictionary");
+        id1 = id_obj.getArrayItem(0).getStringValue();
+    }
+    else
+    {
+        // Treating a missing ID as the empty string enables qpdf to
+        // decrypt some invalid encrypted files with no /ID that
+        // poppler can read but Adobe Reader can't.
+	warn(QPDFExc(qpdf_e_damaged_pdf, this->file->getName(),
+                     "trailer", this->file->getLastOffset(),
+                     "invalid /ID in trailer dictionary"));
     }
 
-    std::string id1 = id_obj.getArrayItem(0).getStringValue();
     QPDFObjectHandle encryption_dict = this->trailer.getKey("/Encrypt");
     if (! encryption_dict.isDictionary())
     {

qpdf/qpdf.testcov

@@ -264,3 +264,4 @@ QPDFObjectHandle inline image token 0
 QPDF not caching overridden objstm object 0
 QPDFWriter original obj non-zero gen 0
 QPDF_optimization indirect outlines 0
+QPDF xref space 2

qpdf/qtest/qpdf.test

@@ -199,7 +199,7 @@ $td->runtest("remove page we don't have",
 show_ntests();
 # ----------
 $td->notify("--- Miscellaneous Tests ---");
-$n_tests += 64;
+$n_tests += 65;
 
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -509,6 +509,14 @@ $td->runtest("check file",
              {$td->FILE => "a.pdf"},
              {$td->FILE => "gen1.qdf"});
 
+# This file, from a user, is missing /ID in its trailer even though it
+# is encrypted and also has a space instead of a newline after its
+# xref keyword.  xpdf can open it, but Adobe reader can't.
+$td->runtest("check broken file",
+             {$td->COMMAND => "qpdf --check invalid-id-xref.pdf"},
+             {$td->FILE => "invalid-id-xref.out", $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+
 show_ntests();
 # ----------
 $td->notify("--- Numeric range parsing tests ---");

qpdf/qtest/qpdf/invalid-id-xref.out

@@ -0,0 +1,16 @@
+WARNING: invalid-id-xref.pdf (trailer, file position 2493795): invalid /ID in trailer dictionary
+checking invalid-id-xref.pdf
+PDF Version: 1.1
+R = 3
+P = -1804
+User password = 
+extract for accessibility: not allowed
+extract for any purpose: allowed
+print low resolution: allowed
+print high resolution: allowed
+modify document assembly: not allowed
+modify forms: not allowed
+modify annotations: allowed
+modify other: not allowed
+modify anything: not allowed
+File is not linearized