On 12/08/2010 04:32 PM, Marius Dumitru Florea wrote:

Hi devs,

When importing an office document using the "Add> Page from Office" menu the page title is given by org.xwiki.officeimporter.document.XDOMOfficeDocument#getTitle() which looks for the first heading (org.xwiki.rendering.block.HeaderBlock) in the document content.

This is fine most of the time, but if the title of the office document is not set using heading style (e.g. a custom defined style is used instead) then:

(1) The page title is wrong. (2) The first heading is hidden because it equals the page title. This is very bad when there is text before the first heading.

WDYT about modifying XDOMOfficeDocument#getTitle() to take into account the first paragraph also when looking for the page title?

What happens if the document contains a header, like a date, time, address, company name, stuff like that, won't this break things even more, since all documents will be called "Finance Department, BigCompany, 42 Main Street"?

Note that I have many Word documents with this problem so fixing them to use heading style for their titles is not a good option for me.

-- Sergiu Dumitriu http://purl.org/net/sergiu/