I'm trying to build a search function using FREETEXTTABLE to provide rankings of the results.
I've got it working on one of the tables using a query that looks something like this:
SELECT DISTINCT title, recordid, rank
FROM mn_records INNER JOIN FREETEXTTABLE(mn_records,*, 'sony') AS ft
ON mn_records.recordid = ft.[key]
ORDER BY rank DESC
The problem is that I want to search two tables mn_records and mn_items. A record has multiple items linked to it and some of the fields in the items table are different for the same record. I have the functionality working using FREETEXT:
SELECT DISTINCT title, recordid
FROM mn_records, mn_items
WHERE mn_records.recordid = mn_items.parentrecord
AND (FREETEXT (mn_records.*, 'sony')
OR FREETEXT (mn_items.*, 'sony'))
How can I properly join these two tables and the FREETEXTTABLE rank so that the search results will only contain distinct record_ids yet still search the information contained in mn_items?
I have used Full-Text Search in a stored procedure with English words whithout any problems..But when i use it with Arabic words it gives me THE Error:
Server: Msg 7619, Level 16, State 1, Procedure SearchProject, Line 19 Execution of a full-text operation failed. A clause of the query contained only ignored words.
The stored procedure :
CREATE PROCEDURE SearchProject ( @SearchString nvarchar(500), @CultureName nvarchar(50), @HowManyResults int OUTPUT ) AS CREATE TABLE #SearchTable ( FieldNO int, ProjectNO int, ProjectName nvarchar(200), ProjectDescription nvarchar(1000), ProjectImage nvarchar(1000), CultureID int )
INSERT INTO #SearchTable (FieldNO,ProjectNO,ProjectName,ProjectDescription,ProjectImage,CultureID) SELECT P.FieldNO, PL.ProjectNO,PL.ProjectName,PL.ProjectDescription,P.ProjectImage,PL.CultureID FROM FREETEXTTABLE(Project_Locale,*,@SearchString) AS FT JOIN Project_Locale AS PL ON FT.[KEY]=PL.ProjectCultureID JOIN Project AS P ON P.ProjectNO=PL.ProjectNO WHERE PL.CultureID=dbo.GetCultureID(@CultureName)
INSERT INTO #SearchTable (FieldNO,ProjectNO,ProjectName,ProjectDescription,ProjectImage,CultureID) SELECT P.FieldNO, PL.ProjectNO,PL.ProjectName,PL.ProjectDescription,P.ProjectImage,PL.CultureID FROM Project_Locale AS PL ,Project AS P, ProjectField_Locale AS FL, FREETEXTTABLE(ProjectField_Locale,*,@SearchString) AS FT2 WHERE FL.FieldNO=P.FieldNO AND FL.FieldCultureID=FT2.[KEY] AND PL.ProjectNO=P.ProjectNO AND PL.CultureID=dbo.GetCultureID(@CultureName) AND FL.CultureID=dbo.GetCultureID(@CultureName)
INSERT INTO #SearchTable (FieldNO,ProjectNO,ProjectName,ProjectDescription,ProjectImage,CultureID) SELECT P.FieldNO, PL.ProjectNO,PL.ProjectName,PL.ProjectDescription,P.ProjectImage,PL.CultureID FROM Project_Locale AS PL ,Project AS P, Feature_Locale AS PFEA, ProjectFeature AS FP, Feature AS F, FREETEXTTABLE(Feature_Locale,*,@SearchString) AS FT3 WHERE FT3.[KEY]=PFEA.FeatureCultureID AND FP.ProjectNO=P.ProjectNO AND FP.FeatureNO=PFEA.FeatureNO AND PL.ProjectNO=P.ProjectNO AND F.FeatureNO=PFEA.FeatureNO AND PFEA.CultureID=dbo.GetCultureID(@CultureName) AND PL.CultureID=dbo.GetCultureID(@CultureName)
SELECT @HowManyResults=COUNT(DISTINCT ProjectNO) FROM #SearchTable SELECT DISTINCT * FROM #SearchTable RETURN GO
I called the stored procedure using the following(in the Query Analayzer) :
USE nabeel1eagle DECLARE @HowManyResults int EXEC SearchProject 'ملÙ?ات','ar-SA',@HowManyResults OUTPUT
but when I use the previous code directly(without calling the stored procedure using EXEC) in the Query Analyzer it works fine.Like the following (part of the code)code:
SELECT P.FieldNO, PL.ProjectNO,PL.ProjectName,PL.ProjectDescription,P.ProjectImage,PL.CultureID FROM FREETEXTTABLE(Project_Locale,*,'ملÙ?ات') AS FT JOIN Project_Locale AS PL ON FT.[KEY]=PL.ProjectCultureID JOIN Project AS P ON P.ProjectNO=PL.ProjectNO WHERE PL.CultureID=dbo.GetCultureID('ar-SA')
Could any one help and tell me how to solve this problem?
I have defined full-text indexing on a table a couple of months ago and the catalog was fully populated, it was working fine until now. But suddenly has stopped working, I have been scheduling incremental population on a regular basis and also have tried to remove and redefined full-text indexing but can't get it to work.
We are using freetexttable as our search function in our application and it seems to be partially working. I search for a word from a column of a table I included in the search catalog and it sometimes pickup the record and sometimes it doesn't. I got 2 records having 'business' as the keyword. And when I search that keyword, it only returned 1 record.
here is the codesnippet: select * from freetexttable(<tablename>, *, 'business')
i use two containstable (one of them with 'formsof') and one freetexttable attributes in a select command and i want to combine them with the logical 'or'.can i do this in the same command? what is the syntax?my code :select table1.field1,a.rank,b.rank,c.rankfrom table1containstable(table1,field,'"word"') as aor containstable(table1,field,'formsof(inflectional(word)') as b)or freetexttable(table1,field,'word') as cwhere table1.id=a.[key] and table1.id=b.[key] and table1.id=c.[key]the above syntax is wrong. i tried ',' instead of 'or' but the results were not right. thanks
I have the following code on a database that works perfectly (i.e. searching for 'print' will return results that have 'printer' in the field):
***************************************
INSERT INTO searchResults (articleId, articleTitle, articleSummary, articleType, reviewedBy, reviewedDate, submitDate, modifiedDate, appTitle, appVersion, rank, hasAccess) SELECT id, articleTitle, articleSummary, articleType, reviewedBy, reviewedDate, submitDate, modifiedDate, appTitle, appVersion, k.rank, hasAccess FROM articles As a INNER JOIN FREETEXTTABLE(articles, articlesummary, 'print') AS K ON a.id = K.[KEY]
***************************************
I am working on a new DB trying to accomplish the same thing, so I have the following code:
***************************************
SELECT id, firstName, lastName, k.rank FROM Pagers_Users As a INNER JOIN FREETEXTTABLE(Pagers_Users, lastName, 'smith') AS K ON a.id = K.[KEY] ORDER BY k.rank DESC
***************************************
However, this second block does not completely work (as I see it). When I enter the word 'smith' to serch for, it only brings back exact matches, not close calls, like 'smithe'.
I have also tried searching for other words (or partial words) and it only returns exact matches.
Is there something in setting up the Text Index that I forgot to turn on for "close" matches?
I am building a simple full text search engine for my site and I was wondering how would I retrieve rows 11-20 of the search result. This is required because I want to show my results only 10 at a time, like google does for instance. My query is as follows -
select top 10 ft_tbl.url, ft_tbl.title, ft_tbl.body, ft_tbl.date, (key_tbl.rank) from mytable as ft_tbl inner join freetexttable(mytable, (url, title, body), '".$searchstring."', 10) as key_tbl on ft_tbl.id = key_tbl.[key] order by (key_tbl.rank) DESC
In MySQL I would use LIMIT but I believe that doesnt exist in MS SQL
Can somebody help me in identifying the difference between FREETEXTABLE and CONTAINSTABLE statements?
I have recently started experimenting with the free text search and I feel that FREETEXTTABLE can look for the search string into multiple columns by breaking it into tokens.
For example
select ft_tbl.saon, ft_tbl.paon, ft_tbl.street, ft_tbl.postcode ,key_tbl.rank from temp as ft_tbl INNER JOIN freetextTABLE(temp, (saon, paon, street), '80 ridge avenue', 15) as key_tbl ON FT_tbl.ID = key_tbl.[key]
Now in my case €œ80€? is in PAON column and €œRidge Avenue€? is in STREET column. Among the resulting rows it also displays the required row. If I try to achieve this with CONTAINSTABLE using
select ft_tbl.saon, ft_tbl.paon, ft_tbl.street, ft_tbl.postcode ,key_tbl.rank from temp as ft_tbl INNER JOIN containsTABLE(temp, (saon, paon, street, postcode), '80 ridge avenue') as key_tbl ON FT_tbl.ID = key_tbl.[key]
It gives me error Syntax error near 'ridge' in the full-text search condition '80 ridge avenue'.
If I make search string as €˜€?80 ridge avenue€?€™ instead of €˜ridge avenue€™ then it doesn€™t give me error but also displays no results because 80 ridge avenue does not appear as whole in any single column. If I search by only specifying €œridge avenue€? in search string then I get a full list where street is ridge avenue.
My question is Does FREETEXTTABLE can search in all listed column by breaking the search string into tokens and CONTAINSTABLE only in one as whole?
Hi, I realised that I am not able to do a FREETEXT search on multiple table, example:SELECT * FROM [tStaffDir], [tStaffDir_ClientExp], [tStaffDir_CoreSpecs], [tStaffDir_GlobalExp], [tStaffDir_Lang], [tStaffDir_PrevEmp], [tStaffDir_TerEdu] WHERE FREETEXT(*, @Name) ORDER BY [Name] Can I use FREETEXTTABLE instead? How do I go about doing it?
I am using FREETEXTTABLE to enable users to search forum posts. It works extremely well (and fast!) and I am very happy with it.
I also have a top ten list of 'Most viewed' forum posts. Each time a post is accessed, a 'viewcount' value is incremented accordingly. The ten forum posts with the highest viewcount appear in the top ten list (by order of viewcount, of course).
Now... when a user searches on a search term, I would like the top ten list to also be affected by this search.
For example, the user searches on 'foo'. In response they get a list of results containing 'foo'. However, the top ten list also changes to reflect this. It now shows the ten forum posts with the highest viewcount which also contain the word 'foo'.
This is easy enough to do. The problem arises when we also add proximity searches.
For example, if I search on 'foo bar', FREETEXTTABLE will look for records which have both or either of these words. Records which contain both words will be ranked higher than records which contain either word. And of the records which contain both words, the closer together the two words are, the higher the ranking will be.
But how does this now work with the top ten list? For example, I may have two records:
- Record A has been viewed 10 times. It contains the words 'foo' and 'bar' and they are next to each other. - Record B has been viewed 10,000 times. But it only contains the word 'foo'.
Now, according to 'viewcount', record B should be ranked higher. But according to FREETEXTTABLE, record A should be ranked higher.
So, I need to combine both these rankings in some way. Regardless of how the 'top ten' records are chosen, they will be ordered by viewcount. But I'm not sure how to get the right top ten records in the first place.
I have an issue trying to pass a search text parameter to FREETEXTTABLE via Dataset. The following code works fine if you hardcode the search word/text as shown: SELECT KEY_TBL.RANK, FT_TBL.FaqQuestion, FT_TBL.FaqAnswer, FT_TBL.SearchFROM faq_table AS FT_TBL INNER JOIN FREETEXTTABLE(faq_table, Search, 'cool') AS KEY_TBL ON FT_TBL.FaqID = KEY_TBL.[KEY]ORDER BY KEY_TBL.RANK DESC Now, I want to do this: SELECT KEY_TBL.RANK, FT_TBL.FaqQuestion, FT_TBL.FaqAnswer, FT_TBL.SearchFROM faq_table AS FT_TBL INNER JOIN FREETEXTTABLE(faq_table, Search, @Search) AS KEY_TBL ON FT_TBL.FaqID = KEY_TBL.[KEY]ORDER BY KEY_TBL.RANK DESC The error I'm getting is @Search is not declared. How am I suppose to pass in a value? I have searched almost everywhere and nobody seemed to ask this precise question. I'm sure this is a huge problem. Can anyone help me please?
We have a query that uses the Full-text index on a view that's returning duplicate rows. We thought maybe it was the way we were joining, but we were able to simplify the query as much as possible and it still happens. Here's the query:
Code Snippet SELECT * FROM FREETEXTTABLE(vwSubtable, TitleSearch, 'Across five Aprils') AS KEY_TBL ORDER BY RANK DESC
vwSubtable is an indexed view that contains some of the columns in our original table, and is also filtering out some rows from the main table in a where clause. There are no joins in the view.
This seems like it's about as simple a query as we could get. It will return some rows twice (ie. the same primary key row is returned back as two separate rows in the resultset). This is a problem since we're filling a datagrid, which is throwing a ConcurrencyException because the primary key is already in there.
I made sure we have SP2 installed on my SQL Server. Any ideas on what might be happening?
I am trying to do a freetext filter with mutiple columns using a column list, but I can't get the syntax down for multiple column list. First, am I am going about this the right way...Do I need to be doing both? Second why doesn't mutiple columns work. I can't find any good samples online. What I am trying to accomplish is a refined search stored procedure that uses the freetext to do the search refinement. Any help would be appreciated.
select
b.rank,
a.ProductID,
a.ProductName,
a.Sequence,
a.ProductImage,
a.ItemID,
a.ItemName,
a.ManufacturerItemCode,
a.ItemImage,
a.ItemSourceID,
a.PackageID,
a.BrandID,
a.BrandName,
a.ManufacturerID,
a.ManufacturerName,
a.ProductCategoryID,
a.CategoryID,
a.CategoryName,
d.CustomerGroupName,
isnull(h.PackageDescription,a.ItemPKG) as PKG,
case g.StockStatus
when 1 then 'Yes'
when 0 then 'No'
else ''
end as StockStatus,
isnull(g.StandardUnitPrice,a.ListPrice) as Price,
isnull(j.SupplierAbbreviation,a.ManufacturerAbbreviation) as ItemSource
from
dbo.vw_mcProductItem a
inner join freetexttable(dbo.vw_mcProductItem, (ProductName,ItemName,ManufacturerItemCode,ItemPKG,BrandName,ManufacturerName,ManufacturerAbbreviation,CategoryName), @SearchWord) as b ON a.ItemID = b.[KEY]
inner join [dbo].[mcCustomerGroupItem] c on c.ItemID = a.ItemID
inner join [dbo].[mcCustomerGroup] d on d.CustomerGroupID = c.CustomerGroupID
inner join [dbo].[mcCustomerGroupCustomer] e on e.CustomerGroupID = d.CustomerGroupID
inner join [dbo].[mcCustomerUser] f on f.CustomerID = e.CustomerID
left outer join [dbo].[mcSupplierItem] g on g.ItemID = a.ItemID
left outer join [dbo].[mcPackage] h on h.PackageID = g.SellingPackageID
left outer join [dbo].[mcItemSource] i on i.ItemSourceId = a.ItemSourceId
left outer join [dbo].[mcSupplier] j on j.SupplierID = g.SupplierID
where
d.CustomerGroupID = @CustomerGroupID
and f.UserID = @UserID
and FREETEXT(BrandName,ManufacturerName,CategoryName, @SearchWord)
Please have a look at the following two queries, the purpose of which is to find which ten users (represented by 'Username') have created the most records which contain the term 'foo':
SELECT TOP 10 Username, COUNT(*) AS [Count] FROM Options
WHERE FREETEXT(*, 'foo')
GROUP BY Username
ORDER BY [Count] DESC
SELECT TOP 10 Username, COUNT(*) AS [Count] FROM Options
JOIN FREETEXTTABLE (Options, *, 'foo', 500) ct
ON OptionID = ct.[KEY]
GROUP BY Username
ORDER BY [Count] DESC
They both produce the same result set. However, I am wondering which is more performant. At first glance, it would seem the first one would be. It doesn't involve a JOIN and should, therefore, be more efficient.
But this depends on how the FREETEXT expression is evaluated. My concern is that internally, SQL Server would generate an entire recordset based on 'WHERE FREETEXT(*, 'foo')', which could be thousands of records, and only then restrict this to the TOP 10 by COUNT.
If this does happen, then it would be better to join to a FREETEXTTABLE, where I can at least restrict the result set using the 'top_n_by_rank' parameter (which is set as '500' in this case, as this seems a good balance of performance against the likely number of duplicates I will get in my FREETEXTTABLE results).
So... I am worrying about this unnecessarily? Should I just use the simpler first version?
