Improve qpdf --check

Fix exit status for case of errors without warnings, continue after
errors when possible, add test case for parsing a file with content
stream errors on some but not all pages.

6 files changed, 55 insertions, 19 deletions

diff --git a/ChangeLog b/ChangeLog
index d7a61d1f..3473a31e 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2013-01-25  Jay Berkenbilt  <ejb@ql.org>
+
+	* qpdf --check was exiting with status 0 in some rare cases even
+	when errors were found.  It now always exits with one of the
+	document error codes (0 for success, 2 for errors, 3 or warnings).
+
 2013-01-24  Jay Berkenbilt  <ejb@ql.org>
 
 	* qpdf --check now does syntactic checks all pages' content
@@ -5,10 +11,6 @@
 	errors are still not checked, and there are no plans to add
 	semantic checks.
 
-	* Bug fix: in versions 4.0.0 and 4.0.1, qpdf --check wasn't doing
-	as full of a check for linearized files as for non-linearized
-	files.
-
 2013-01-22  Jay Berkenbilt  <ejb@ql.org>
 
 	* Add QPDFObjectHandle::getTypeCode().  This method returns a
diff --git a/libqpdf/QPDFObjectHandle.cc b/libqpdf/QPDFObjectHandle.cc
index fca8191d..1a7dfc73 100644
--- a/libqpdf/QPDFObjectHandle.cc
+++ b/libqpdf/QPDFObjectHandle.cc
@@ -1012,10 +1012,10 @@ QPDFObjectHandle::parseInternal(PointerHolder<InputSource> input,
 	}
 	else if (! object.isInitialized())
 	{
-	    throw std::logic_error(
-		"INTERNAL ERROR: uninitialized object (token = " +
-		QUtil::int_to_string(token.getType()) +
-		", " + token.getValue() + ")");
+	    throw QPDFExc(qpdf_e_damaged_pdf, input->getName(),
+			  object_description,
+			  input->getLastOffset(),
+			  "parse error while reading object");
 	}
 	else
 	{
diff --git a/qpdf/qpdf.cc b/qpdf/qpdf.cc
index 102f4b7f..a416870a 100644
--- a/qpdf/qpdf.cc
+++ b/qpdf/qpdf.cc
@@ -1428,7 +1428,11 @@ int main(int argc, char* argv[])
 	    }
 	    if (check)
 	    {
-		bool okay = false;
+                // Code below may set okay to false but not to true.
+                // We assume okay until we prove otherwise but may
+                // continue to perform additional checks after finding
+                // errors.
+		bool okay = true;
 		std::cout << "checking " << infilename << std::endl;
 		try
 		{
@@ -1444,8 +1448,11 @@ int main(int argc, char* argv[])
 		    if (pdf.isLinearized())
 		    {
 			std::cout << "File is linearized\n";
-			okay = pdf.checkLinearization();
-			// any errors are reported by checkLinearization().
+			if (! pdf.checkLinearization())
+                        {
+                            // any errors are reported by checkLinearization()
+                            okay = false;
+                        }
 		    }
 		    else
 		    {
@@ -1453,8 +1460,8 @@ int main(int argc, char* argv[])
                     }
 
                     // Write the file no nowhere, uncompressing
-                    // streams.  This causes full file traversal
-                    // and decoding of all streams we can decode.
+                    // streams.  This causes full file traversal and
+                    // decoding of all streams we can decode.
                     QPDFWriter w(pdf);
                     Pl_Discard discard;
                     w.setOutputPipeline(&discard);
@@ -1464,19 +1471,30 @@ int main(int argc, char* argv[])
                     // Parse all content streams
                     std::vector<QPDFObjectHandle> pages = pdf.getAllPages();
                     DiscardContents discard_contents;
+                    int pageno = 0;
                     for (std::vector<QPDFObjectHandle>::iterator iter =
                              pages.begin();
                          iter != pages.end(); ++iter)
                     {
-                        QPDFObjectHandle::parseContentStream(
-                            (*iter).getKey("/Contents"), &discard_contents);
+                        ++pageno;
+                        try
+                        {
+                            QPDFObjectHandle::parseContentStream(
+                                (*iter).getKey("/Contents"),
+                                &discard_contents);
+                        }
+                        catch (QPDFExc& e)
+                        {
+                            okay = false;
+                            std::cout << "page " << pageno << ": "
+                                      << e.what() << std::endl;
+                        }
                     }
-
-                    okay = true;
 		}
 		catch (std::exception& e)
 		{
 		    std::cout << e.what() << std::endl;
+                    okay = false;
 		}
 		if (okay)
 		{
@@ -1493,6 +1511,10 @@ int main(int argc, char* argv[])
 				  << std::endl;
 		    }
 		}
+                else
+                {
+                    exit(EXIT_ERROR);
+                }
 	    }
 	}
 	else
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index 8d2b5cfc..f447bd83 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -199,7 +199,7 @@ $td->runtest("remove page we don't have",
 show_ntests();
 # ----------
 $td->notify("--- Miscellaneous Tests ---");
-$n_tests += 59;
+$n_tests += 60;
 
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -435,7 +435,7 @@ $td->runtest("EOF terminating literal tokens",
              $td->NORMALIZE_NEWLINES);
 $td->runtest("EOF reading token",
              {$td->COMMAND => "qpdf --check eof-reading-token.pdf"},
-             {$td->FILE => "eof-reading-token.out", $td->EXIT_STATUS => 0},
+             {$td->FILE => "eof-reading-token.out", $td->EXIT_STATUS => 2},
              $td->NORMALIZE_NEWLINES);
 $td->runtest("extra header text",
              {$td->COMMAND => "test_driver 32 minimal.pdf"},
@@ -478,6 +478,11 @@ $td->runtest("tokenize content streams",
              {$td->FILE => "tokenize-content-streams.out",
               $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("content stream errors",
+             {$td->COMMAND => "qpdf --check content-stream-errors.pdf"},
+             {$td->FILE => "content-stream-errors.out",
+              $td->EXIT_STATUS => 2},
+             $td->NORMALIZE_NEWLINES);
 
 show_ntests();
 # ----------
diff --git a/qpdf/qtest/qpdf/content-stream-errors.out b/qpdf/qtest/qpdf/content-stream-errors.out
new file mode 100644
index 00000000..a375f5e6
--- /dev/null
+++ b/qpdf/qtest/qpdf/content-stream-errors.out
@@ -0,0 +1,7 @@
+checking content-stream-errors.pdf
+PDF Version: 1.3
+File is not encrypted
+File is not linearized
+page 1: content stream object 7 0 (content, file position 52): parse error while reading object
+page 3: content stream object 15 0 (stream data, file position 117): EOF found while reading inline image
+page 4: content stream object 19 0 (content, file position 53): parse error while reading object
diff --git a/qpdf/qtest/qpdf/content-stream-errors.pdf b/qpdf/qtest/qpdf/content-stream-errors.pdf
new file mode 100644
index 00000000..abbb7d79
--- /dev/null
+++ b/qpdf/qtest/qpdf/content-stream-errors.pdf