Read xref table without PCRE

Also accept more errors than before.

4 files changed, 106 insertions, 1 deletions

diff --git a/qpdf/qpdf.testcov b/qpdf/qpdf.testcov
index 2860f55e..c08ed721 100644
--- a/qpdf/qpdf.testcov
+++ b/qpdf/qpdf.testcov
@@ -289,3 +289,7 @@ qpdf single-pages %d 0
 qpdf single-pages .pdf 0
 qpdf single-pages other 0
 QPDFTokenizer allowing bad token 0
+QPDF ignore first space in xref entry 0
+QPDF ignore first extra space in xref entry 0
+QPDF ignore second extra space in xref entry 0
+QPDF ignore length error xref entry 0
diff --git a/qpdf/qtest/qpdf.test b/qpdf/qtest/qpdf.test
index b3d13bde..d0cba589 100644
--- a/qpdf/qtest/qpdf.test
+++ b/qpdf/qtest/qpdf.test
@@ -232,7 +232,7 @@ foreach my $d (@bug_tests)
 show_ntests();
 # ----------
 $td->notify("--- Miscellaneous Tests ---");
-$n_tests += 86;
+$n_tests += 87;
 
 $td->runtest("qpdf version",
 	     {$td->COMMAND => "qpdf --version"},
@@ -669,6 +669,13 @@ $td->runtest("ignore bad token",
               $td->EXIT_STATUS => 0},
              $td->NORMALIZE_NEWLINES);
 
+$td->runtest("recoverable xref errors",
+             {$td->COMMAND =>
+                  "qpdf --check --show-xref xref-errors.pdf"},
+             {$td->FILE => "xref-errors.out",
+              $td->EXIT_STATUS => 3},
+             $td->NORMALIZE_NEWLINES);
+
 show_ntests();
 # ----------
 $td->notify("--- Single Page ---");
diff --git a/qpdf/qtest/qpdf/xref-errors.out b/qpdf/qtest/qpdf/xref-errors.out
new file mode 100644
index 00000000..7a2cf384
--- /dev/null
+++ b/qpdf/qtest/qpdf/xref-errors.out
@@ -0,0 +1,15 @@
+WARNING: xref-errors.pdf (xref table, file position 585): accepting invalid xref table entry
+WARNING: xref-errors.pdf (xref table, file position 606): accepting invalid xref table entry
+WARNING: xref-errors.pdf (xref table, file position 627): accepting invalid xref table entry
+WARNING: xref-errors.pdf (xref table, file position 648): accepting invalid xref table entry
+WARNING: xref-errors.pdf (xref table, file position 667): accepting invalid xref table entry
+checking xref-errors.pdf
+PDF Version: 1.3
+File is not encrypted
+File is not linearized
+1/0: uncompressed; offset = 9
+2/0: uncompressed; offset = 63
+3/0: uncompressed; offset = 135
+4/0: uncompressed; offset = 307
+5/0: uncompressed; offset = 403
+6/0: uncompressed; offset = 438
diff --git a/qpdf/qtest/qpdf/xref-errors.pdf b/qpdf/qtest/qpdf/xref-errors.pdf
new file mode 100644
index 00000000..1778ab92
--- /dev/null
+++ b/qpdf/qtest/qpdf/xref-errors.pdf
@@ -0,0 +1,79 @@
+%PDF-1.3
+1 0 obj
+<<
+  /Type /Catalog
+  /Pages 2 0 R
+>>
+endobj
+
+2 0 obj
+<<
+  /Type /Pages
+  /Kids [
+    3 0 R
+  ]
+  /Count 1
+>>
+endobj
+
+3 0 obj
+<<
+  /Type /Page
+  /Parent 2 0 R
+  /MediaBox [0 0 612 792]
+  /Contents 4 0 R
+  /Resources <<
+    /ProcSet 5 0 R
+    /Font <<
+      /F1 6 0 R
+    >>
+  >>
+>>
+endobj
+
+4 0 obj
+<<
+  /Length 44
+>>
+stream
+BT
+  /F1 24 Tf
+  72 720 Td
+  (Potato) Tj
+ET
+endstream
+endobj
+
+5 0 obj
+[
+  /PDF
+  /Text
+]
+endobj
+
+6 0 obj
+<<
+  /Type /Font
+  /Subtype /Type1
+  /Name /F1
+  /BaseFont /Helvetica
+  /Encoding /WinAnsiEncoding
+>>
+endobj
+
+xref
+0 7
+0000000000 65535 f 
+ 0000000009 00000 n 
+0000000063  00000 n 
+0000000135 00000  n 
+000000307 00000 n 
+0000000403 0000 n 
+0000000438 00000 n 
+trailer <<
+  /Size 7
+  /Root 1 0 R
+>>
+startxref
+556
+%%EOF