Allow QPDFTokenizer::readToken to return bad tokens

Sometimes we want to ignore bad tokens rather than having them throw
an exception. A coverage case is commented out here and added in a
later commit.

2 files changed, 16 insertions, 7 deletions

diff --git a/libqpdf/QPDF.cc b/libqpdf/QPDF.cc
index cfadaa3a..3a8dc875 100644
--- a/libqpdf/QPDF.cc
+++ b/libqpdf/QPDF.cc
@@ -1329,9 +1329,10 @@ QPDF::recoverStreamLength(PointerHolder<InputSource> input,
 }
 
 QPDFTokenizer::Token
-QPDF::readToken(PointerHolder<InputSource> input)
+QPDF::readToken(PointerHolder<InputSource> input, bool allow_bad)
 {
-    return this->tokenizer.readToken(input, this->last_object_description);
+    return this->tokenizer.readToken(
+        input, this->last_object_description, allow_bad);
 }
 
 QPDFObjectHandle
diff --git a/libqpdf/QPDFTokenizer.cc b/libqpdf/QPDFTokenizer.cc
index f7124e49..c0e09279 100644
--- a/libqpdf/QPDFTokenizer.cc
+++ b/libqpdf/QPDFTokenizer.cc
@@ -475,7 +475,8 @@ QPDFTokenizer::betweenTokens()
 
 QPDFTokenizer::Token
 QPDFTokenizer::readToken(PointerHolder<InputSource> input,
-                         std::string const& context)
+                         std::string const& context,
+                         bool allow_bad)
 {
     qpdf_offset_t offset = input->tell();
     Token token;
@@ -514,13 +515,20 @@ QPDFTokenizer::readToken(PointerHolder<InputSource> input,
 	input->unreadCh(char_to_unread);
     }
 
+    input->setLastOffset(offset);
+
     if (token.getType() == tt_bad)
     {
-	throw QPDFExc(qpdf_e_damaged_pdf, input->getName(),
-		      context, offset, token.getErrorMessage());
+        if (allow_bad)
+        {
+//            QTC::TC("qpdf", "QPDFTokenizer allowing bad token");
+        }
+        else
+        {
+            throw QPDFExc(qpdf_e_damaged_pdf, input->getName(),
+                          context, offset, token.getErrorMessage());
+        }
     }
 
-    input->setLastOffset(offset);
-
     return token;
 }