tdf#63054: pdf_string_parser incorrectly handles escapes

This patch fixes the escapes handling by avoiding boost built-in
functions that skip whitespaces behind the scenes.

Change-Id: I9607ecd9e2e052aadcba69cf30ec25c981743c49
Reviewed-on: https://gerrit.libreoffice.org/15562
Tested-by: Jenkins <ci@libreoffice.org>
Tested-by: David Tardon <dtardon@redhat.com>
Reviewed-by: David Tardon <dtardon@redhat.com>

sdext/source/pdfimport/pdfparse/pdfparse.cxx

@@ -116,8 +116,8 @@ public:
                 else if( c == '\\' ) // ignore escaped braces
                 {
                     ++len;
-                    ++scan;
-                    if( scan.at_end() )
+                    ++scan.first;                 // tdf#63054: avoid skipping spaces
+                    if( scan.first == scan.last ) // tdf#63054: avoid skipping spaces
                         break;
                 }
                 ++len;