Search a String

[S]

Any idea on how I would be able to do a search within C# that does
ranges or words
For example

I want to search for Chicken in the string
string s1 = "This is Great Chicken";
string s2 = "Chicken";
return s1.IndexOf(s2) >= 0

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

string s1 = "This is Great Chicken";
string s2 = "Chickin";

return ???

or

Worst yet
string s1 = "This is Great Chickin";
string s2 = "Chicken";

Person enter the search phrase misspelled the word Chicken;

string s1 = "This is Great Chickin";
string s2 = "Chicken";

[Author]

On Oct 2, 2:26*pm, S <scottremi...@y ahoo.comwrote:

Any idea on how I would be able to do a search within C# that does
ranges or words

For example

I want to search for Chicken in the string

string s1 = "This is Great Chicken";
string s2 = "Chicken";
return s1.IndexOf(s2) >= 0

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

string s1 = "This is Great Chicken";
string s2 = "Chickin";

return ???

or

Worst yet
string s1 = "This is Great Chickin";
string s2 = "Chicken";

Person enter the search phrase misspelled the word Chicken;

string s1 = "This is Great Chickin";
string s2 = "Chicken";

Check out the Regex namespace.

If you search something such as "Chickin" in a string which does not
contain this token, such as "This is Great Chicken", what would you
expect to return? I wouldn't expect it to return anything. Maybe you
wanna do something like google's key words suggestion if user entry
has some typos?

[Author]

On Oct 2, 2:51*pm, Author <gnewsgr...@gma il.comwrote:

On Oct 2, 2:26*pm, S <scottremi...@y ahoo.comwrote:

Any idea on how I would be able to do a search within C# that does
ranges or words

For example

I want to search for Chicken in the string

string s1 = "This is Great Chicken";
string s2 = "Chicken";
return s1.IndexOf(s2) >= 0

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

string s1 = "This is Great Chicken";
string s2 = "Chickin";

return ???

or

Worst yet
string s1 = "This is Great Chickin";
string s2 = "Chicken";

Person enter the search phrase misspelled the word Chicken;

string s1 = "This is Great Chickin";
string s2 = "Chicken";

Check out the Regex namespace.

Correction: I meant the System.Text.Reg ularExpressions Namespace
http://msdn.microsoft.com/en-us/libr...pressions.aspx

Regex is a class of this namespace.

If you search something such as "Chickin" in a string which does not
contain this token, such as "This is Great Chicken", what would you
expect to return? *I wouldn't expect it to return anything. *Maybe you
wanna do something like google's key words suggestion if user entry
has some typos?

[Peter Duniho]

On Thu, 02 Oct 2008 11:26:38 -0700, S <sc**********@y ahoo.comwrote:

Any idea on how I would be able to do a search within C# that does
ranges or words

There are a variety of techniques for dealing with partial matching as you
describe.

I've actually implemented a very simple approach myself in the past, in
which I do a sort of character-by-character merge between the target text
and the search word, finding the spot in the target text where the search
word fits the best with the least number of excluded characters. This
would work perfectly for the examples you gave, but whether that would
address your more general goals I can't say.

If that sounds promising, I might be able to dig up the code and post it
here. It's pretty inefficient, but it's small and simple so what the
heck? :)

You might also Google for "soundex". There's even a Wikipedia article on
the topic:
http://en.wikipedia.org/wiki/Soundex

That's a commonly used technique for matching words based on phonetics and
if you want a solution that is based more on language than on typography,
it might be a better approach.

A Google search on "dictionary spell check algorithm" also turned up a
number of promising links.

Pete

[Ignacio Machin ( .NET/ C# MVP )]

On Oct 2, 2:26 pm, S <scottremi...@y ahoo.comwrote:

Any idea on how I would be able to do a search within C# that does
ranges or words

For example

I want to search for Chicken in the string

string s1 = "This is Great Chicken";
string s2 = "Chicken";
return s1.IndexOf(s2) >= 0

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

string s1 = "This is Great Chicken";
string s2 = "Chickin";

return ???

or

Worst yet
string s1 = "This is Great Chickin";
string s2 = "Chicken";

Person enter the search phrase misspelled the word Chicken;

string s1 = "This is Great Chickin";
string s2 = "Chicken";

Hi,

I do not how to do it in code, I'm pretty sure that no version of the
framework provides such a feature.
In T-SQL you have this feature though take a look for SOUNDEX function

[Ignacio Machin ( .NET/ C# MVP )]

On Oct 2, 2:51 pm, Author <gnewsgr...@gma il.comwrote:

On Oct 2, 2:26 pm, S <scottremi...@y ahoo.comwrote:

Any idea on how I would be able to do a search within C# that does
ranges or words

For example

I want to search for Chicken in the string

string s1 = "This is Great Chicken";
string s2 = "Chicken";
return s1.IndexOf(s2) >= 0

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

string s1 = "This is Great Chicken";
string s2 = "Chickin";

return ???

or

Worst yet
string s1 = "This is Great Chickin";
string s2 = "Chicken";

Person enter the search phrase misspelled the word Chicken;

string s1 = "This is Great Chickin";
string s2 = "Chicken";

Check out the Regex namespace.

If you search something such as "Chickin" in a string which does not
contain this token, such as "This is Great Chicken", what would you
expect to return? I wouldn't expect it to return anything. Maybe you
wanna do something like google's key words suggestion if user entry
has some typos?

I do not know how to use a Regex in this case.can you elaborate?

[Jeff Johnson]

"S" <sc**********@y ahoo.comwrote in message
news:de******** *************** ***********@64g 2000hsm.googleg roups.com...

The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

The answer is that what you're looking for is not at all simple. I'd
recommend you do a Google search to see if you can find sample code for a
spell checker.

[Arto Viitanen]

Jeff Johnson kirjoitti:

"S" <sc**********@y ahoo.comwrote in message
news:de******** *************** ***********@64g 2000hsm.googleg roups.com...

>The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

The answer is that what you're looking for is not at all simple. I'd
recommend you do a Google search to see if you can find sample code for a
spell checker.

Spelling checker does not help. For example the text contains "Cut" and
the user searches for "Cat". Both are spelled right, so the search gives
no result.

But, with addition to already mentioned Soundex, you might like to check
Levenshtein distance
(http://en.wikipedia.org/wiki/Levenshtein_distance). It tells how
similar two texts are; i.e. how many simple edit commands are needed to
make the second text from the first. For example distance between
Chicken and Chikin is one.

About Soundex: I have not studied it much, but I have a feeling Knuth
built it for english words. How well does it apply to other languages
or names?
-
Arto Viitanen

[Jeff Johnson]

"Arto Viitanen" <ar***********@ microteam.fiwro te in message
news:48******** *************** @news.tdc.fi...

>>The trick is how would I search for: Person doing the search
misspelled the word Chickin (Happens all the time);

The answer is that what you're looking for is not at all simple. I'd
recommend you do a Google search to see if you can find sample code for a
spell checker.

Spelling checker does not help. For example the text contains "Cut" and
the user searches for "Cat". Both are spelled right, so the search gives
no result.

I agree, IF you say, "Hey, Mr. Spell Checker algorithm, tell me what's
misspelled in this sentence." However, if you are in control of the code,
I'd think you'd be able to say, "Hey, Mr. Spell Checker algorithm, make a
list of all the probable misspellings of 'chicken' and tell me if this
string contains any of them." See where I'm going?

But, with addition to already mentioned Soundex, you might like to check
Levenshtein distance
(http://en.wikipedia.org/wiki/Levenshtein_distance). It tells how
similar two texts are; i.e. how many simple edit commands are needed to
make the second text from the first. For example distance between
Chicken and Chikin is one.

Sounds good.

About Soundex: I have not studied it much, but I have a feeling Knuth
built it for english words. How well does it apply to other languages
or names?

I've heard people say that Soundex is an ancient crutch and shouldn't be
relied upon too much.

[=?UTF-8?B?IkZlcm5hbmRvIEEuIEfDs21leiBGLiI=?=]

Peter Duniho wrote:

On Thu, 02 Oct 2008 11:26:38 -0700, S <sc**********@y ahoo.comwrote:

>Any idea on how I would be able to do a search within C# that does
ranges or words

There are a variety of techniques for dealing with partial matching as
you describe.

I've actually implemented a very simple approach myself in the past, in
which I do a sort of character-by-character merge between the target
text and the search word, finding the spot in the target text where the
search word fits the best with the least number of excluded characters.
This would work perfectly for the examples you gave, but whether that
would address your more general goals I can't say.

[SNIP]

Pete

That sounds interesting. Is there any chance of seeing it as a
CodeProject article? :D