Question

I’m running a FullTextSqlQuery where TrimDuplicates is set to true and have been searching for days on why certain items are not in the results. I just found out today that the result appears when I set TrimDuplicates=false.

Is this a known SharePoint search bug?

My code is simple:

using (var fullTextSqlQuery = new FullTextSqlQuery(_searchServiceApplicationProxy))
{
    fullTextSqlQuery.QueryText = querytext;
    fullTextSqlQuery.ResultsProvider = SearchProvider.Default;
    fullTextSqlQuery.TrimDuplicates = true;
    fullTextSqlQuery.EnableStemming = true;
    fullTextSqlQuery.EnableNicknames = true;
    fullTextSqlQuery.IgnoreAllNoiseQuery = true;
    fullTextSqlQuery.ResultTypes |= ResultType.RelevantResults;

    if (pageSize.HasValue && pageSize.Value > 0)
    {
        fullTextSqlQuery.RowLimit = pageSize.Value;
        fullTextSqlQuery.TotalRowsExactMinimum = pageSize.Value;

        if (selectedPage.HasValue && selectedPage.Value > 0)
            fullTextSqlQuery.StartRow = (selectedPage.Value - 1) * pageSize.Value;
    }

    searchResults = fullTextSqlQuery.Execute();
}

Thanks in advance for your answers.

Answer 1

Found a partial answer in: https://sharepoint.stackexchange.com/questions/14460/sharepoint-2010-search-when-is-a-document-a-duplicate

It seems like TrimDuplicates removes search results if the data is similar to a certain degree (does not have to be 100%). Though the actual % is not known.

In short, there’s no bug in Trim Duplicates, just not a lot of understanding on how this feature behaves.

Now my problem is figuring out which data it is comparing. Since it seems like even when I make the data in my custom columns unique, do an index reset and another full crawl, the page I’m looking for is still trimmed off.