We have our on new XRef parsing/recovery mode

1 files changed, 0 insertions, 99 deletions

diff --git a/ALL_DIFF b/ALL_DIFF
index bb49249..0bff244 100644
--- a/ALL_DIFF
+++ b/ALL_DIFF
@@ -12,105 +12,6 @@ diff -ru xpdf-3.02/doc/pdftotext.1 xpdf-3.03/doc/pdftotext.1
  .B \-raw
  Keep the text in content stream order.  This is a hack which often
  "undoes" column formatting, etc.  Use of raw mode is no longer
-diff -ru xpdf-3.02/xpdf/PDFDoc.cc xpdf-3.03/xpdf/PDFDoc.cc
---- xpdf-3.02/xpdf/PDFDoc.cc	2007-02-27 23:05:52.000000000 +0100
-+++ xpdf-3.03/xpdf/PDFDoc.cc	2011-08-15 23:08:53.000000000 +0200
-@@ -183,38 +209,71 @@
-   // check header
-   checkHeader();
- 
-+  // read the xref and catalog
-+  if (!PDFDoc::setup2(ownerPassword, userPassword, gFalse)) {
-+    if (errCode == errDamaged || errCode == errBadCatalog) {
-+      // try repairing the xref table
-+      error(errSyntaxWarning, -1,
-+	    "PDF file is damaged - attempting to reconstruct xref table...");
-+      if (!PDFDoc::setup2(ownerPassword, userPassword, gTrue)) {
-+	return gFalse;
-+      }
-+    } else {
-+      return gFalse;
-+    }
-+  }
-+
-+#ifndef DISABLE_OUTLINE
-+  // read outline
-+  outline = new Outline(catalog->getOutline(), xref);
-+#endif
-+
-+  // done
-+  return gTrue;
-+}
-+
-+GBool PDFDoc::setup2(GString *ownerPassword, GString *userPassword,
-+		     GBool repairXRef) {
-   // read xref table
--  xref = new XRef(str);
-+  xref = new XRef(str, repairXRef);
-   if (!xref->isOk()) {
--    error(-1, "Couldn't read xref table");
-+    error(errSyntaxError, -1, "Couldn't read xref table");
-     errCode = xref->getErrorCode();
-+    delete xref;
-+    xref = NULL;
-     return gFalse;
-   }
- 
-   // check for encryption
-   if (!checkEncryption(ownerPassword, userPassword)) {
-     errCode = errEncrypted;
-+    delete xref;
-+    xref = NULL;
-     return gFalse;
-   }
- 
-   // read catalog
-   catalog = new Catalog(this);
-   if (!catalog->isOk()) {
--    error(-1, "Couldn't read page catalog");
-+    error(errSyntaxError, -1, "Couldn't read page catalog");
-     errCode = errBadCatalog;
-+    delete catalog;
-+    catalog = NULL;
-+    delete xref;
-+    xref = NULL;
-     return gFalse;
-   }
- 
--#ifndef DISABLE_OUTLINE
--  // read outline
--  outline = new Outline(catalog->getOutline(), xref);
--#endif
--
--  // done
-   return gTrue;
- }
- 
-diff -ru xpdf-3.02/xpdf/PDFDoc.h xpdf-3.03/xpdf/PDFDoc.h
---- xpdf-3.02/xpdf/PDFDoc.h	2007-02-27 23:05:52.000000000 +0100
-+++ xpdf-3.03/xpdf/PDFDoc.h	2011-08-15 23:08:53.000000000 +0200
-@@ -154,27 +162,44 @@
- private:
- 
-   GBool setup(GString *ownerPassword, GString *userPassword);
-+  GBool setup2(GString *ownerPassword, GString *userPassword,
-+	       GBool repairXRef);
-   void checkHeader();
-   GBool checkEncryption(GString *ownerPassword, GString *userPassword);
- 
-   GString *fileName;
- #ifdef WIN32
-   wchar_t *fileNameU;
- #endif
-   FILE *file;
-   BaseStream *str;
-   void *guiData;
-   double pdfVersion;
-   XRef *xref;
-   Catalog *catalog;
- #ifndef DISABLE_OUTLINE
-   Outline *outline;
- #endif
 diff -ru xpdf-3.02/xpdf/pdftotext.cc xpdf-3.03/xpdf/pdftotext.cc
 --- xpdf-3.02/xpdf/pdftotext.cc	2007-02-27 23:05:52.000000000 +0100
 +++ xpdf-3.03/xpdf/pdftotext.cc	2011-08-15 23:08:53.000000000 +0200