>>How do you save the data in the database - is it always the same single quote character or can be anything from these three types?
>
>It could be any combination at the save an any combination at the search.
>
>>Assuming you always convert bad quotes to a single quote when saving data, then you just need to adjust your search expression to replace different kind of quote with a single quote.
>
>There are many sources of incoming saves. So, we have to consider the fact that there could be several types of quotes.
>
>>If you have data with different types of quotes, then you may use replace(replace(...) like replace(replace...
>>to convert your ColumnData and Search data to uniform quote. The good solution for this case will be to have computed indexed column in a database with that character converted to a single quote. So, instead of using a column, you will be using 'CleanedData' computed column.
>
>Yes, this is what I needed. I didn't think about that. But, this gives me an impression that I could lift any field in my data dictionary to identify a table field with its relational table field for the search usability. Then, if I set the property to True, when doing the search, that could be used.
For what it's worth, this is what I'm doing in foxpro
I have an extra field that is equal to the original field with
(1) All non-alpha and non-digits stripped ( including your quotes)
(2) All accents dropped (grave, aigue, circonflex, cedille, ...)
(3) then converted to uppercase
eg
Jean d'Éléphant
becomes
JEANDELEPHANT
I do the same with the expression to search
The query is simpler then
Gregory