[jira] [Commented] (TIKA-976) Inaccurate XLS detection trough POIFSContainerDetector

["Nick Burch (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TIKA-976?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14255450#comment-14255450 ] 

Nick Burch commented on TIKA-976:
---------------------------------

This is now being handled fully through TIKA-1490

> Inaccurate XLS detection trough POIFSContainerDetector
> ------------------------------------------------------
>
>                 Key: TIKA-976
>                 URL: https://issues.apache.org/jira/browse/TIKA-976
>             Project: Tika
>          Issue Type: Improvement
>          Components: mime
>    Affects Versions: 1.2
>            Reporter: Marco Quaranta
>              Labels: detection, mime, poi, xls
>             Fix For: 1.3
>
>         Attachments: test_book.xls
>
>
> I've found an inaccurate detection with the attached xls file. POIFSContainerDetector is unable to determine the exact mimetype (vnd.ms-excel) and returns the generic "x-tika-msoffice". This is due to the fact this file's root names are :[Book, DocumentSummaryInformation, SummaryInformation]. POIFSContainerDetector checks only that names contains "WorkBook".
> Could it be possible to add a further or-check like this:
> if (names.contains("Workbook") || names.contains("Book"))
> Thank you,
> Marco



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)