Apache
/
poi
mirror of https://github.com/apache/poi.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

#59739 When testing for a valid NFIB in a FileInformationBlock, accept common "nearby" values as found from CommonCrawl 

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@1750866 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Nick Burch 2016-06-30 21:20:23 +00:00
parent e616b8cf12
commit 3680dc8992
1 changed files with 27 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

45
src/scratchpad/src/org/apache/poi/hwpf/model/FileInformationBlock.java
View File

@ -143,36 +143,45 @@ public final class FileInformationBlock implements Cloneable
private void assertCbRgFcLcb() private void assertCbRgFcLcb()
{ {
switch ( getNFib() ) int nfib = getNFib();
String nfibHex = String.format("%04X", nfib);
// Note - CommonCrawl shows there's more variation in these than
// the documentation suggests, so accept common clusters around
// the "correct" value as well
switch ( nfib )
{ {
case 0x00C1: case 0x0071:
assertCbRgFcLcb( "0x00C1", 0x005D, "0x005D", _cbRgFcLcb ); // Found in CommonCrawl corpus but not in the docs...
break; break;
case 0x00D9: case 0x00BE:
assertCbRgFcLcb( "0x00D9", 0x006C, "0x006C", _cbRgFcLcb ); case 0x00BF:
case 0x00C0:
case 0x00C1: // Docs "official"
case 0x00C2:
case 0x00C3:
assertCbRgFcLcb(nfibHex, 0x005D, "0x005D", _cbRgFcLcb );
break;
case 0x00D8:
case 0x00D9: // Docs "official"
assertCbRgFcLcb(nfibHex, 0x006C, "0x006C", _cbRgFcLcb );
break; break;
case 0x0101: case 0x0101:
assertCbRgFcLcb( "0x0101", 0x0088, "0x0088", _cbRgFcLcb ); assertCbRgFcLcb( "0x0101", 0x0088, "0x0088", _cbRgFcLcb );
break; break;
case 0x010C: // TODO Is CommonCrawl 265 = 0x109 the one above or below?
assertCbRgFcLcb( "0x010C", 0x00A4, "0x00A4", _cbRgFcLcb ); case 0x010B:
case 0x010C: // Docs "official"
assertCbRgFcLcb(nfibHex, 0x00A4, "0x00A4", _cbRgFcLcb );
break; break;
case 0x0112: case 0x0112:
assertCbRgFcLcb( "0x0112", 0x00B7, "0x00B7", _cbRgFcLcb ); assertCbRgFcLcb( "0x0112", 0x00B7, "0x00B7", _cbRgFcLcb );
break; break;
default: default:
/* /* The Word spec has a much smaller list of "valid" values
When running with the large CommonCrawl corpus we found the following ids in documents that are processed fine: * to what the large CommonCrawl corpus contains!
java.lang.IllegalStateException: Invalid file format version number: 113
java.lang.IllegalStateException: Invalid file format version number: 191
java.lang.IllegalStateException: Invalid file format version number: 192
java.lang.IllegalStateException: Invalid file format version number: 194
java.lang.IllegalStateException: Invalid file format version number: 195
java.lang.IllegalStateException: Invalid file format version number: 216
java.lang.IllegalStateException: Invalid file format version number: 265
java.lang.IllegalStateException: Invalid file format version number: 267
*/ */
logger.log(POILogger.WARN, "Invalid file format version number: " + getNFib()); logger.log(POILogger.WARN, "Invalid file format version number: " + nfib + "("+nfibHex+")");
} }
} }