Add "n:/pdf-name" to qpdf JSON for binary names (fixes #1072)

3 files changed, 195 insertions, 0 deletions

diff --git a/qpdf/qtest/qpdf-json.test b/qpdf/qtest/qpdf-json.test
index 961b507a..9691d995 100644
--- a/qpdf/qtest/qpdf-json.test
+++ b/qpdf/qtest/qpdf-json.test
@@ -61,6 +61,7 @@ my @goodfiles = (
     'form-fields-and-annotations.pdf',
     'need-appearances.pdf',
     'fxo-blue.pdf',
+    'weird-tokens.pdf',
     );
 $n_tests += 6 * scalar(@goodfiles);
 
@@ -341,5 +342,21 @@ $td->runtest("check C API write to JSON stream",
              {$td->FILE => "auto-4"},
              {$td->FILE => "qpdf-ctest-47-4"});
 
+# Bugs #1072 and #1079 illustrate cases that qpdf-json got wrong. In
+# #1072, it was noticed that name tokens containing binary characters
+# (using #xx) would generate invalid JSON, even though qpdf's own JSON
+# parser would accept it. Also, the JSON spec allows real numbers in
+# scientific notation, but the PDF spec does not.
+$n_tests += 2;
+$td->runtest("handle binary names",
+             {$td->COMMAND =>
+                  "qpdf --json-output weird-tokens.pdf a.json"},
+             {$td->STRING => "", $td->EXIT_STATUS => 0});
+# Round-trip is tested above.
+$td->runtest("check json",
+             {$td->FILE => "a.json"},
+             {$td->FILE => "weird-tokens.json"},
+             $td->NORMALIZE_NEWLINES);
+
 cleanup();
 $td->report($n_tests);
diff --git a/qpdf/qtest/qpdf/weird-tokens.json b/qpdf/qtest/qpdf/weird-tokens.json
new file mode 100644
index 00000000..66f0ff06
--- /dev/null
+++ b/qpdf/qtest/qpdf/weird-tokens.json
@@ -0,0 +1,83 @@
+{
+  "qpdf": [
+    {
+      "jsonversion": 2,
+      "pdfversion": "2.0",
+      "pushedinheritedpageresources": false,
+      "calledgetallpages": false,
+      "maxobjectid": 6
+    },
+    {
+      "obj:1 0 R": {
+        "value": {
+          "/Extra": [
+            "u:Names with binary data",
+            "n:/ABCDEF+#ba#da#cc#e5",
+            "/ABCEDEF+Ï€",
+            "n:/one+#a0two",
+            "/text/plain",
+            "u:Very small/large reals",
+            0.00001,
+            1000000000000
+          ],
+          "/Pages": "2 0 R",
+          "/Type": "/Catalog"
+        }
+      },
+      "obj:2 0 R": {
+        "value": {
+          "/Count": 1,
+          "/Kids": [
+            "3 0 R"
+          ],
+          "/Type": "/Pages"
+        }
+      },
+      "obj:3 0 R": {
+        "value": {
+          "/Contents": "4 0 R",
+          "/MediaBox": [
+            0,
+            0,
+            612,
+            792
+          ],
+          "/Parent": "2 0 R",
+          "/Resources": {
+            "/Font": {
+              "/F1": "6 0 R"
+            }
+          },
+          "/Type": "/Page"
+        }
+      },
+      "obj:4 0 R": {
+        "stream": {
+          "data": "QlQKICAvRjEgMjQgVGYKICA3MiA3MjAgVGQKICAoUG90YXRvKSBUagpFVAo=",
+          "dict": {}
+        }
+      },
+      "obj:5 0 R": {
+        "value": 44
+      },
+      "obj:6 0 R": {
+        "value": {
+          "/BaseFont": "/Helvetica",
+          "/Encoding": "/WinAnsiEncoding",
+          "/Subtype": "/Type1",
+          "/Type": "/Font"
+        }
+      },
+      "trailer": {
+        "value": {
+          "/ID": [
+            "b:42841c13bbf709d79a200fa1691836f8",
+            "b:728c020f464c3cf7e02c12605fa7d88b"
+          ],
+          "/Root": "1 0 R",
+          "/Size": 7
+        }
+      }
+    }
+  ]
+}
diff --git a/qpdf/qtest/qpdf/weird-tokens.pdf b/qpdf/qtest/qpdf/weird-tokens.pdf
new file mode 100644
index 00000000..7c645df3
--- /dev/null
+++ b/qpdf/qtest/qpdf/weird-tokens.pdf
@@ -0,0 +1,95 @@
+%PDF-2.0
+%¿÷¢þ
+%QDF-1.0
+
+1 0 obj
+<<
+  /Extra [
+    (Names with binary data)
+    /ABCDEF+#ba#da#cc#e5
+    /ABCEDEF+#cf#80
+    /one+#a0two
+    /text#2fplain
+    (Very small/large reals)
+    0.00001
+    1000000000000
+  ]
+  /Pages 2 0 R
+  /Type /Catalog
+>>
+endobj
+
+2 0 obj
+<<
+  /Count 1
+  /Kids [
+    3 0 R
+  ]
+  /Type /Pages
+>>
+endobj
+
+%% Page 1
+3 0 obj
+<<
+  /Contents 4 0 R
+  /MediaBox [
+    0
+    0
+    612
+    792
+  ]
+  /Parent 2 0 R
+  /Resources <<
+    /Font <<
+      /F1 6 0 R
+    >>
+  >>
+  /Type /Page
+>>
+endobj
+
+%% Contents for page 1
+4 0 obj
+<<
+  /Length 5 0 R
+>>
+stream
+BT
+  /F1 24 Tf
+  72 720 Td
+  (Potato) Tj
+ET
+endstream
+endobj
+
+5 0 obj
+44
+endobj
+
+6 0 obj
+<<
+  /BaseFont /Helvetica
+  /Encoding /WinAnsiEncoding
+  /Subtype /Type1
+  /Type /Font
+>>
+endobj
+
+xref
+0 7
+0000000000 65535 f 
+0000000025 00000 n 
+0000000261 00000 n 
+0000000343 00000 n 
+0000000539 00000 n 
+0000000638 00000 n 
+0000000657 00000 n 
+trailer <<
+  /Root 1 0 R
+  /Size 7
+  /ID [<42841c13bbf709d79a200fa1691836f8><728c020f464c3cf7e02c12605fa7d88b>]
+>>
+startxref
+763
+%%EOF