StartsWith/EndsWith vs Like

[mobi12]

Is there any difference between following commands?

If mwd.EndsWith("ous") Then mwd = mwd.Replace("ous","ás")

If mwd Like "*ous" Then mwd = mwd.Replace("ous","ás")

[jmcilhinney]

Specifically, yes. Practically, no. Personally, I never use the VB Like operator. You'll likely find that most who use C# as well don't. I'd suggest you go with the former but, whatever you do choose, be consistent. The fact that the former is the same in C# and is therefore consistent is one reason I do use it.

[mobi12]

Intesting the situation with C#.

I find LIKE more versatile because I can compare like words ending with cise or cize in one command LIKE "*ci[sz]e".

The corresponding code with EndsWith will be more complex to write!

[jmcilhinney]

Intesting the situation with C#.

I find LIKE more versatile because I can compare like words ending with cise or cize in one command LIKE "*ci[sz]e".

The corresponding code with EndsWith will be more complex to write!

You wouldn't use EndsWith for that in C# but rather a Regex. Regular expressions are more versatile still, but do need a bit of work to learn. Like a I said, just be consistent. If you consistently want to use the simplest option then that would be EndsWith where it is possible, otherwise Like where it is possible, otherwise Regex. Writing the most readable code should always be one of your highest priorities and EndsWith is more self-documenting than Like so that's a point in its favour where it can be used.

[dday9]

Something to consider is the following statement from the Like operator's documentation: https://docs.microsoft.com/en-us/dot.../like-operator

The Like operator is currently not supported in .NET Core and .NET Standard projects.

Something else to consider, back to JMcIlhinney's comments on consistency, is that the Like operator uses a limited pattern syntax that isn't a one-to-one match to its closest (more widespread) alternative RegEx. For example, the pattern for a character not in range for the Like operator is [!abc] whereas the pattern in RegEx is [^abc].

Speaking to your specific "ends with cise or cize" example, here is an alternative using RegEx:

ci[sz]e$

Usage in Visual Basic .NET:

Dim pattern = New Regex("ci[sz]e$")
pattern.IsMatch("string to test")

Fiddle: https://dotnetfiddle.net/RLkE0A

[i00]

I would just use a regex replace and not do the check...
Also with your examples if they end with ous every occurrence of ous in the string will be replaced!

With the $ at the end in your pattern only the one at the end of the string will get replaced.