Give up reading objects with too many consecutive errors

ChangeLog

@@ -1,3 +1,11 @@
+2019-06-15  Jay Berkenbilt  <ejb@ql.org>
+
+	* When parsing files, while reading an object, if there are too
+	many consecutive errors without enough intervening successes, give
+	up on the specific object. This reduces cases in which very badly
+	damaged files send qpdf into a tail spin reading one character at
+	a time and reporting warnings.
+
 2019-06-13  Jay Berkenbilt  <ejb@ql.org>
 
 	* Perform initial integration of Google's oss-fuzz project by

libqpdf/QPDFObjectHandle.cc

@@ -1628,8 +1628,11 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
     qpdf_offset_t offset = input->tell();
     offset_stack.push_back(offset);
     bool done = false;
+    int bad_count = 0;
+    int good_count = 0;
     while (! done)
     {
+        bool bad = false;
         std::vector<QPDFObjectHandle>& olist = olist_stack.back();
         parser_state_e state = state_stack.back();
         offset = offset_stack.back();
@@ -1651,6 +1654,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                              input->getLastOffset(),
                              "unexpected EOF"));
             }
+            bad = true;
             state = st_eof;
             break;
 
@@ -1661,6 +1665,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                          object_description,
                          input->getLastOffset(),
                          token.getErrorMessage()));
+            bad = true;
             object = newNull();
 	    break;
 
@@ -1672,6 +1677,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                          object_description,
                          input->getLastOffset(),
                          "treating unexpected brace token as null"));
+            bad = true;
             object = newNull();
 	    break;
 
@@ -1688,6 +1694,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                              object_description,
                              input->getLastOffset(),
                              "treating unexpected array close token as null"));
+                bad = true;
                 object = newNull();
 	    }
 	    break;
@@ -1705,6 +1712,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                              object_description,
                              input->getLastOffset(),
                              "unexpected dictionary close token"));
+                bad = true;
                 object = newNull();
 	    }
 	    break;
@@ -1719,6 +1727,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                              object_description,
                              input->getLastOffset(),
                              "ignoring excessively deeply nested data structure"));
+                bad = true;
                 object = newNull();
                 state = st_top;
             }
@@ -1800,6 +1809,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                                  input->getLastOffset(),
                                  "unknown token while reading object;"
                                  " treating as string"));
+                    bad = true;
                     object = newString(value);
 		}
 	    }
@@ -1824,6 +1834,7 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
                          input->getLastOffset(),
                          "treating unknown token type as null while "
                          "reading object"));
+            bad = true;
             object = newNull();
 	    break;
 	}
@@ -1839,6 +1850,32 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
             object = newNull();
         }
 
+        if (bad)
+        {
+            ++bad_count;
+            good_count = 0;
+        }
+        else
+        {
+            ++good_count;
+            if (good_count > 3)
+            {
+                bad_count = 0;
+            }
+        }
+        if (bad_count > 5)
+        {
+            // We had too many consecutive errors without enough
+            // intervening successful objects. Give up.
+            warn(context,
+                 QPDFExc(qpdf_e_damaged_pdf, input->getName(),
+                         object_description,
+                         input->getLastOffset(),
+                         "too many errors; giving up on reading object"));
+            state = st_top;
+            object = newNull();
+        }
+
         switch (state)
         {
           case st_eof:

qpdf/qtest/qpdf.test

@@ -624,7 +624,7 @@ my @bug_tests = (
     ["99", "object 0", 2],
     ["99b", "object 0", 2],
     ["100", "xref reconstruction loop", 2],
-    ["101", "resolve for exception text", 3],
+    ["101", "resolve for exception text", 2],
     ["117", "other infinite loop", 2],
     ["118", "other infinite loop", 2],
     ["119", "other infinite loop", 3],
@@ -639,7 +639,7 @@ my @bug_tests = (
     ["149", "xref prev pointer loop", 3],
     ["150", "integer overflow", 2],
     ["202", "even more deeply nested dictionary", 2],
-    ["263", "empty xref stream", 3],
+    ["263", "empty xref stream", 2],
     );
 $n_tests += scalar(@bug_tests);
 foreach my $d (@bug_tests)

qpdf/qtest/qpdf/issue-100.out

@@ -7,10 +7,8 @@ WARNING: issue-100.pdf (object 5 0, offset 289): unknown token while reading obj
 WARNING: issue-100.pdf (object 5 0, offset 294): unknown token while reading object; treating as string
 WARNING: issue-100.pdf (object 5 0, offset 297): unknown token while reading object; treating as string
 WARNING: issue-100.pdf (object 5 0, offset 304): unknown token while reading object; treating as string
-WARNING: issue-100.pdf (object 5 0, offset 308): unexpected )
+WARNING: issue-100.pdf (object 5 0, offset 304): too many errors; giving up on reading object
-WARNING: issue-100.pdf (object 5 0, offset 316): treating unexpected array close token as null
+WARNING: issue-100.pdf (object 5 0, offset 308): expected endobj
-WARNING: issue-100.pdf (object 5 0, offset 227): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-100.pdf (object 5 0, offset 321): expected endobj
 WARNING: issue-100.pdf (object 5 0, offset 418): /Length key in stream dictionary is not an integer
 WARNING: issue-100.pdf (object 5 0, offset 489): attempting to recover stream length
 WARNING: issue-100.pdf (object 5 0, offset 489): recovered stream length: 12

qpdf/qtest/qpdf/issue-101.out

@@ -38,116 +38,6 @@ WARNING: issue-101.pdf (object 11 0, offset 626): unknown token while reading ob
 WARNING: issue-101.pdf (object 11 0, offset 637): unknown token while reading object; treating as string
 WARNING: issue-101.pdf (object 11 0, offset 639): unknown token while reading object; treating as string
 WARNING: issue-101.pdf (object 11 0, offset 644): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 647): unknown token while reading object; treating as string
+WARNING: issue-101.pdf (object 11 0, offset 644): too many errors; giving up on reading object
-WARNING: issue-101.pdf (object 11 0, offset 687): unknown token while reading object; treating as string
+WARNING: issue-101.pdf (object 11 0, offset 647): expected endobj
-WARNING: issue-101.pdf (object 11 0, offset 691): unknown token while reading object; treating as string
+issue-101.pdf (offset 687): unable to find /Root dictionary
-WARNING: issue-101.pdf (object 11 0, offset 696): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 698): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 701): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 711): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 743): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 745): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 747): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 777): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 790): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 800): treating unexpected brace token as null
-WARNING: issue-101.pdf (object 11 0, offset 801): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 811): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 819): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 832): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 856): unexpected >
-WARNING: issue-101.pdf (object 11 0, offset 857): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 868): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 887): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 897): unexpected )
-WARNING: issue-101.pdf (object 11 0, offset 898): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 909): invalid character (¤) in hexstring
-WARNING: issue-101.pdf (object 11 0, offset 911): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 929): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 930): invalid character (²) in hexstring
-WARNING: issue-101.pdf (object 11 0, offset 932): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 944): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 947): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 970): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1046): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1067): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1075): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1080): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1084): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1102): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1112): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1124): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1133): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1145): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1148): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1150): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1151): unexpected )
-WARNING: issue-101.pdf (object 11 0, offset 1153): unexpected dictionary close token
-WARNING: issue-101.pdf (object 11 0, offset 1156): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1163): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1168): unexpected >
-WARNING: issue-101.pdf (object 11 0, offset 1170): invalid character (I) in hexstring
-WARNING: issue-101.pdf (object 11 0, offset 1167): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-101.pdf (object 11 0, offset 1167): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-101.pdf (object 11 0, offset 1167): expected dictionary key but found non-name object; inserting key /QPDFFake3
-WARNING: issue-101.pdf (object 11 0, offset 1176): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1180): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1184): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1190): unexpected >
-WARNING: issue-101.pdf (object 11 0, offset 1192): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1195): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1205): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1217): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1224): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1236): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1242): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-101.pdf (object 11 0, offset 1242): dictionary ended prematurely; using null as value for last key
-WARNING: issue-101.pdf (object 11 0, offset 1275): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1287): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1291): unexpected dictionary close token
-WARNING: issue-101.pdf (object 11 0, offset 1294): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1306): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1322): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1325): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1329): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1341): treating unexpected array close token as null
-WARNING: issue-101.pdf (object 11 0, offset 1312): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-101.pdf (object 11 0, offset 1312): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-101.pdf (object 11 0, offset 1312): expected dictionary key but found non-name object; inserting key /QPDFFake3
-WARNING: issue-101.pdf (object 11 0, offset 1312): expected dictionary key but found non-name object; inserting key /QPDFFake4
-WARNING: issue-101.pdf (object 11 0, offset 1312): dictionary ended prematurely; using null as value for last key
-WARNING: issue-101.pdf (object 11 0, offset 1349): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1353): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1357): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1359): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 11 0, offset 1368): unexpected )
-WARNING: issue-101.pdf (object 11 0, offset 1373): expected endobj
-WARNING: issue-101.pdf (object 2 0, offset 244): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3855): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3863): treating unexpected brace token as null
-WARNING: issue-101.pdf (object 7 0, offset 3864): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3866): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3873): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3879): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3888): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3901): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3905): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3913): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake3
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake4
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake5
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake6
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake7
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake8
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake9
-WARNING: issue-101.pdf (object 7 0, offset 3847): expected dictionary key but found non-name object; inserting key /QPDFFake10
-WARNING: issue-101.pdf (object 7 0, offset 3844): stream dictionary lacks /Length key
-WARNING: issue-101.pdf (object 7 0, offset 3962): attempting to recover stream length
-WARNING: issue-101.pdf (object 7 0, offset 3962): recovered stream length: 12
-WARNING: issue-101.pdf (object 8 0, offset 4067): invalid character ()) in hexstring
-WARNING: issue-101.pdf (object 8 0, offset 4069): expected endobj
-WARNING: issue-101.pdf (object 9 0, offset 2832): unknown token while reading object; treating as string
-WARNING: issue-101.pdf (object 9 0, offset 2834): expected endobj
-qpdf: operation succeeded with warnings; resulting file may have some problems

qpdf/qtest/qpdf/issue-263.out

@@ -8,31 +8,5 @@ WARNING: issue-263.pdf (trailer, offset 79): unknown token while reading object;
 WARNING: issue-263.pdf (trailer, offset 82): unexpected )
 WARNING: issue-263.pdf (trailer, offset 83): unknown token while reading object; treating as string
 WARNING: issue-263.pdf (trailer, offset 87): unexpected >
-WARNING: issue-263.pdf (trailer, offset 89): unexpected dictionary close token
+WARNING: issue-263.pdf (trailer, offset 87): too many errors; giving up on reading object
-WARNING: issue-263.pdf (trailer, offset 92): unexpected >
+issue-263.pdf: unable to find trailer dictionary while recovering damaged file
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake3
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake4
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake5
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake6
-WARNING: issue-263.pdf (trailer, offset 40): expected dictionary key but found non-name object; inserting key /QPDFFake7
-WARNING: issue-263.pdf (trailer, offset 98): unknown token while reading object; treating as string
-WARNING: issue-263.pdf (trailer, offset 103): unexpected )
-WARNING: issue-263.pdf (trailer, offset 107): unknown token while reading object; treating as string
-WARNING: issue-263.pdf (trailer, offset 119): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-263.pdf (trailer, offset 163): unknown token while reading object; treating as string
-WARNING: issue-263.pdf (trailer, offset 173): unknown token while reading object; treating as string
-WARNING: issue-263.pdf (trailer, offset 113): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-263.pdf (trailer, offset 113): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-263.pdf (trailer, offset 113): dictionary ended prematurely; using null as value for last key
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake1
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake2
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake3
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake4
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake5
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake6
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake7
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake8
-WARNING: issue-263.pdf (trailer, offset 18): expected dictionary key but found non-name object; inserting key /QPDFFake9
-qpdf: operation succeeded with warnings; resulting file may have some problems