Add additional validation of document-level structures to `--check` option.

m-holger
1 parent 36e1c142
Showing 6 changed files with 37 additions and 3 deletions
libqpdf/QPDFJob.cc
manual/release-notes.rst
qpdf/qtest/outlines.test
qpdf/qtest/page-labels.test
qpdf/qtest/qpdf/outlines-with-loop-check.out
qpdf/qtest/qpdf/page-labels-num-tree-damaged-check.out
@@ -779,6 +779,14 @@ QPDFJob::doCheck(QPDF&amp; pdf)
             cout << "File is not linearized\n";
         }
+        // Create all document helper to trigger any validations they carry out.
+        auto& pages = pdf.pages();
+        (void)pdf.acroform();
+        (void)pdf.embedded_files();
+        (void)pdf.page_labels();
+        (void)pdf.outlines().resolveNamedDest(QPDFObjectHandle::newString("dummy"));
+        (void)pdf.outlines().getOutlinesForPage(pages.getAllPages().at(0));
+
         // Write the file to nowhere, uncompressing streams.  This causes full file traversal and
         // decoding of all streams we can decode.
         QPDFWriter w(pdf);
@@ -789,7 +797,7 @@ QPDFJob::doCheck(QPDF&amp; pdf)
         // Parse all content streams
         int pageno = 0;
-        for (auto& page: pdf.pages().getAllPages()) {
+        for (auto& page: pages.getAllPages()) {
             ++pageno;
             try {
                 page.parseContents(nullptr);
@@ -47,6 +47,9 @@ more detail.
      with the incompatible options :qpdf:ref:`--encrypt` or
      :qpdf:ref:`--copy-encryption`.
+   - Option :qpdf:ref:`--check` now includes additional basic checks of the
+     AcroForm, Dests, Outlines, and PageLabels structures.
+
   - Other enhancements
     - ``QPDFWriter`` will no longer add filters when writing empty streams.
@@ -21,7 +21,7 @@ my @outline_files = (
     'outlines-with-old-root-dests-dict',
     'outlines-with-loop',
     );
-my $n_tests = scalar(@outline_files);
+my $n_tests = scalar(@outline_files) + 1;
 foreach my $f (@outline_files)
 {
     $td->runtest("outlines: $f",
@@ -30,5 +30,10 @@ foreach my $f (@outline_files)
                  $td->NORMALIZE_NEWLINES);
 }
+$td->runtest("outlines: outlines-with-loop --check",
+             {$td->COMMAND => "qpdf --check outlines-with-loop.pdf"},
+             {$td->FILE => "outlines-with-loop-check.out", $td->EXIT_STATUS => 0},
+             $td->NORMALIZE_NEWLINES);
+
 cleanup();
 $td->report($n_tests);
@@ -14,7 +14,7 @@ cleanup();
 my $td = new TestDriver('page-labels');
-my $n_tests = 4;
+my $n_tests = 5;
 $td->runtest("complex page labels",
              {$td->COMMAND => "test_driver 47 page-labels-num-tree.pdf"},
@@ -38,6 +38,11 @@ $td-&gt;runtest(&quot;damaged page labels&quot;,
              {$td->FILE => "page-labels-num-tree-damaged.out", $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
+$td->runtest("damaged page labels --check",
+             {$td->COMMAND => "qpdf --check page-labels-num-tree-damaged.pdf"},
+             {$td->FILE => "page-labels-num-tree-damaged-check.out", $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+
 # --set-page-labels
 my @errors = (
     ["quack", ".*page label spec must be.*"],
+checking outlines-with-loop.pdf
+PDF Version: 1.3
+File is not encrypted
+File is not linearized
+No syntax or stream encoding errors found; the file may still contain
+errors that qpdf cannot detect
+checking page-labels-num-tree-damaged.pdf
+PDF Version: 1.3
+File is not encrypted
+File is not linearized
+WARNING: page-labels-num-tree-damaged.pdf (Name/Number tree node (object 2)): attempting to repair after error: page-labels-num-tree-damaged.pdf (Name/Number tree node (object 2)): keys are not sorted in validate
+WARNING: page-labels-num-tree-damaged.pdf (Name/Number tree node (object 37)): item 1 is invalid
+qpdf: operation succeeded with warnings