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.

2 files changed, 28 insertions, 10 deletions

diff --git a/libqpdf/QPDF.cc b/libqpdf/QPDF.cc
index 00e13ca0..ad742daa 100644
--- a/libqpdf/QPDF.cc
+++ b/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
 	{
diff --git a/libqpdf/QPDF_encryption.cc b/libqpdf/QPDF_encryption.cc
index 51ed3072..88cd707f 100644
--- a/libqpdf/QPDF_encryption.cc
+++ b/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())
     {