[TIKA-516] Excel 5 files are inconsistently detected as either "application/msword" or "application/vnd.ms-excel"

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Affects Version/s: 0.7
• Fix Version/s: 0.8
• Component/s: parser
• Assignee: Unassigned
• Reporter: Victor Kazakov
• Created: 15/Sep/10 22:42
• Updated: 06/Oct/10 12:39
• Resolved: 06/Oct/10 12:39

Description

Using the AutoDetectParser on an Excel 5 file inconsistently detects it as either "application/msword" or "application/vnd.ms-excel"

See the following code:

public static void main(String[] args) throws Exception {
FileInputStream stream = null;
try {
for (int i = 0; i < 10; i++)

{ File file = new File("excel5.xls"); stream = new FileInputStream(file); AutoDetectParser parser = new AutoDetectParser(); Metadata metadata = new Metadata(); metadata.set(Metadata.RESOURCE_NAME_KEY, file.getName()); parser.parse(stream, new DefaultHandler(), metadata); System.out.println(metadata.get(Metadata.CONTENT_TYPE)); }

} finally {
if (stream != null)

{ stream.close(); }

}
}

an example output is:
application/vnd.ms-excel
application/msword
application/msword
application/vnd.ms-excel
application/vnd.ms-excel
application/vnd.ms-excel
application/vnd.ms-excel
application/msword
application/vnd.ms-excel
application/msword

The excel 5 file I used is attached to this bug.