Test json v2 with invalid stream data

5 files changed, 284 insertions, 3 deletions

diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index 23dce379..9f97540a 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -1110,7 +1110,7 @@ my @json_files = (
     ['V4-aes', ['--json-key=encrypt']],
     ['V4-aes', ['--json-key=encrypt', '--show-encryption-key']],
 );
-$n_tests += 2 * scalar(@json_files);
+$n_tests += 5 + (2 * scalar(@json_files));
 foreach my $d (@json_files)
 {
     my ($file, $xargs) = @$d;
@@ -1134,7 +1134,8 @@ foreach my $d (@json_files)
     my $in = "$file.pdf";
     $td->runtest("json v1 $out",
                  {$td->COMMAND =>
-                      ['qpdf', '--json=1', '--test-json-schema', @v1_xargs, $in]},
+                      ['qpdf', '--json=1', '--test-json-schema',
+                       @v1_xargs, $in]},
                  {$td->FILE => "$out-v1.out", $td->EXIT_STATUS => 0},
                  $td->NORMALIZE_NEWLINES);
     $td->runtest("json v2 $out",
@@ -1144,6 +1145,31 @@ foreach my $d (@json_files)
                  $td->NORMALIZE_NEWLINES);
 }
 
+$td->runtest("bad json stream data (inline)",
+             {$td->COMMAND =>
+                  "qpdf --json=2 --test-json-schema" .
+                  " --json-stream-data=inline bad-data.pdf > a.json"},
+             {$td->FILE => "bad-data-json.out", $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check (inline)",
+             {$td->FILE => "a.json"},
+             {$td->FILE => "json-bad-data-json-inline-v2.out"},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("bad json stream data (file)",
+             {$td->COMMAND =>
+                  "qpdf --json=2 --test-json-schema" .
+                  " --json-stream-data=file --json-stream-prefix=auto" .
+                  " bad-data.pdf > a.json"},
+             {$td->FILE => "bad-data-json.out", $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check (file)",
+             {$td->FILE => "a.json"},
+             {$td->FILE => "json-bad-data-json-file-v2.out"},
+             $td->NORMALIZE_NEWLINES);
+$td->runtest("check stream (file)",
+             {$td->FILE => "auto-4"},
+             {$td->FILE => "bad-data-4.out"});
+
 show_ntests();
 # ----------
 $td->notify("--- Page API Tests ---");
@@ -5778,6 +5804,6 @@ sub get_md5_checksum
 
 sub cleanup
 {
-    system("rm -rf *.ps *.pnm ?.pdf ?.qdf *.enc* tif1 tif2 tiff-cache");
+    system("rm -rf a.json *.ps *.pnm ?.pdf ?.qdf *.enc* tif1 tif2 tiff-cache");
     system("rm -rf *split-out* ???-kfo.pdf *.tmpout \@file.pdf auto-*");
 }
diff --git a/qpdf/qtest/qpdf/bad-data-4.out b/qpdf/qtest/qpdf/bad-data-4.out
new file mode 100644
index 00000000..ba572187
--- /dev/null
+++ b/qpdf/qtest/qpdf/bad-data-4.out
diff --git a/qpdf/qtest/qpdf/bad-data-json.out b/qpdf/qtest/qpdf/bad-data-json.out
new file mode 100644
index 00000000..00019fec
--- /dev/null
+++ b/qpdf/qtest/qpdf/bad-data-json.out
@@ -0,0 +1,3 @@
+WARNING: bad-data.pdf (offset 319): error decoding stream data for object 4 0: LZWDecoder: bad code received
+WARNING: bad-data.pdf (offset 319): stream will be re-processed without filtering to avoid data loss
+qpdf: operation succeeded with warnings
diff --git a/qpdf/qtest/qpdf/json-bad-data-json-file-v2.out b/qpdf/qtest/qpdf/json-bad-data-json-file-v2.out
new file mode 100644
index 00000000..147dc372
--- /dev/null
+++ b/qpdf/qtest/qpdf/json-bad-data-json-file-v2.out
@@ -0,0 +1,126 @@
+{
+  "version": 2,
+  "parameters": {
+    "decodelevel": "generalized"
+  },
+  "pages": [
+    {
+      "contents": [
+        "4 0 R"
+      ],
+      "images": [],
+      "label": null,
+      "object": "3 0 R",
+      "outlines": [],
+      "pageposfrom1": 1
+    }
+  ],
+  "pagelabels": [],
+  "acroform": {
+    "fields": [],
+    "hasacroform": false,
+    "needappearances": false
+  },
+  "attachments": {},
+  "encrypt": {
+    "capabilities": {
+      "accessibility": true,
+      "extract": true,
+      "modify": true,
+      "modifyannotations": true,
+      "modifyassembly": true,
+      "modifyforms": true,
+      "modifyother": true,
+      "printhigh": true,
+      "printlow": true
+    },
+    "encrypted": false,
+    "ownerpasswordmatched": false,
+    "parameters": {
+      "P": 0,
+      "R": 0,
+      "V": 0,
+      "bits": 0,
+      "filemethod": "none",
+      "key": null,
+      "method": "none",
+      "streammethod": "none",
+      "stringmethod": "none"
+    },
+    "userpasswordmatched": false
+  },
+  "outlines": [],
+  "qpdf": {
+    "jsonversion": 2,
+    "pdfversion": "1.3",
+    "objects": {
+      "obj:1 0 R": {
+        "value": {
+          "/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": "5 0 R"
+            },
+            "/ProcSet": "6 0 R"
+          },
+          "/Type": "/Page"
+        }
+      },
+      "obj:4 0 R": {
+        "stream": {
+          "datafile": "auto-4",
+          "dict": {
+            "/Filter": "/LZWDecode"
+          }
+        }
+      },
+      "obj:5 0 R": {
+        "value": {
+          "/BaseFont": "/Helvetica",
+          "/Encoding": "/WinAnsiEncoding",
+          "/Name": "/F1",
+          "/Subtype": "/Type1",
+          "/Type": "/Font"
+        }
+      },
+      "obj:6 0 R": {
+        "value": [
+          "/PDF",
+          "/Text"
+        ]
+      },
+      "trailer": {
+        "value": {
+          "/ID": [
+            "b:4a5e383aee3970ed44eb87234d7f93f1",
+            "b:4a5e383aee3970ed44eb87234d7f93f1"
+          ],
+          "/Root": "1 0 R",
+          "/Size": 7
+        }
+      }
+    }
+  }
+}
diff --git a/qpdf/qtest/qpdf/json-bad-data-json-inline-v2.out b/qpdf/qtest/qpdf/json-bad-data-json-inline-v2.out
new file mode 100644
index 00000000..bfd7ad36
--- /dev/null
+++ b/qpdf/qtest/qpdf/json-bad-data-json-inline-v2.out
@@ -0,0 +1,126 @@
+{
+  "version": 2,
+  "parameters": {
+    "decodelevel": "generalized"
+  },
+  "pages": [
+    {
+      "contents": [
+        "4 0 R"
+      ],
+      "images": [],
+      "label": null,
+      "object": "3 0 R",
+      "outlines": [],
+      "pageposfrom1": 1
+    }
+  ],
+  "pagelabels": [],
+  "acroform": {
+    "fields": [],
+    "hasacroform": false,
+    "needappearances": false
+  },
+  "attachments": {},
+  "encrypt": {
+    "capabilities": {
+      "accessibility": true,
+      "extract": true,
+      "modify": true,
+      "modifyannotations": true,
+      "modifyassembly": true,
+      "modifyforms": true,
+      "modifyother": true,
+      "printhigh": true,
+      "printlow": true
+    },
+    "encrypted": false,
+    "ownerpasswordmatched": false,
+    "parameters": {
+      "P": 0,
+      "R": 0,
+      "V": 0,
+      "bits": 0,
+      "filemethod": "none",
+      "key": null,
+      "method": "none",
+      "streammethod": "none",
+      "stringmethod": "none"
+    },
+    "userpasswordmatched": false
+  },
+  "outlines": [],
+  "qpdf": {
+    "jsonversion": 2,
+    "pdfversion": "1.3",
+    "objects": {
+      "obj:1 0 R": {
+        "value": {
+          "/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": "5 0 R"
+            },
+            "/ProcSet": "6 0 R"
+          },
+          "/Type": "/Page"
+        }
+      },
+      "obj:4 0 R": {
+        "stream": {
+          "data": "eJxzCuFSUNB3M1QwMlEISQOyzY2AyEAhJAXI1gjIL0ksyddUCMnicg3hAgDLAQnI",
+          "dict": {
+            "/Filter": "/LZWDecode"
+          }
+        }
+      },
+      "obj:5 0 R": {
+        "value": {
+          "/BaseFont": "/Helvetica",
+          "/Encoding": "/WinAnsiEncoding",
+          "/Name": "/F1",
+          "/Subtype": "/Type1",
+          "/Type": "/Font"
+        }
+      },
+      "obj:6 0 R": {
+        "value": [
+          "/PDF",
+          "/Text"
+        ]
+      },
+      "trailer": {
+        "value": {
+          "/ID": [
+            "b:4a5e383aee3970ed44eb87234d7f93f1",
+            "b:4a5e383aee3970ed44eb87234d7f93f1"
+          ],
+          "/Root": "1 0 R",
+          "/Size": 7
+        }
+      }
+    }
+  }
+}