fixed weights of probabilities for various xeger options #58
Conversation
updated xeger (definition of that charset) updated regexp - use that set of chars instead of "printable ascii" used before + create function for print usedcharset for any regexp
define charset for .*, \w, \W, \d, \D, \s, \S
|
|
||
| namespace Fare | ||
| { | ||
| public static class RandomGenerator |
There was a problem hiding this comment.
What's the advantage of this PRNG compared to others? E.g. SplitMix, PCG, etc.
There was a problem hiding this comment.
Oh no idea,didn't give it much thought and this was just the natural implementation :-)
|
|
||
| // Assert | ||
| Assert.All(result, regex => Assert.Matches(pattern, regex)); | ||
| Assert.All(result, regex => Assert.Matches(pattern, regex)); |
There was a problem hiding this comment.
| Assert.All(result, regex => Assert.Matches(pattern, regex)); | |
| Assert.All(result, regex => Assert.Matches(pattern, regex)); |
| { | ||
| public static class RandomGenerator | ||
| { | ||
| private static Random RandomInstance = new Random(); |
|
|
||
| namespace Fare | ||
| { | ||
| public static class RandomGenerator |
| var index = RandomInstance.Next(sum); | ||
| return items[borders.IndexOf(x => x > index)]; | ||
| } | ||
|
|
| @@ -87,7 +127,7 @@ public void GeneratedTextIsCorrectWithRexEngine(string pattern) | |||
| "x[0-9A-Z]", | |||
| "[^A-M]in", | |||
| ".gr", | |||
| @"\(.*l", | |||
| //@"\(.*l", - on "classic" went sometimes overflow (even before my changes) | |||
There was a problem hiding this comment.
Can you put it back? (Uncomment) We can always tackle the erratic tests in a separate PR.
| @@ -121,9 +161,72 @@ public void GeneratedTextIsCorrectWithRexEngine(string pattern) | |||
| @"\Sabc\S{3}111", | |||
| @"^\S\S (\S)+$", | |||
| @"\\abc\\d", | |||
| @"\w+1\w{4}", | |||
| //@"\w+1\w{4}", | |||
There was a problem hiding this comment.
Can you put it back? (Uncomment) We can always tackle the erratic tests in a separate PR.
| {"x[0-9A-Z]","x0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ"}, | ||
| {"[^A-M]in",null}, | ||
| {".gr",null}, | ||
| //@"\(.*l", - on "classic" went sometimes overflow (even before my changes) |
There was a problem hiding this comment.
Based on this commented-out case, it looks like this data theory list is a duplicate of the existing one(?) If so, why?
|
@Licho1, I can see that you're updating the PR, which is great 👍 What is the current state of the PR? |
Current solution gives every continuous set of characters same weight. For example if you have
"[abcdefghijklmnopqrstuvxyzč]" 50% chars were from a-z and 50% were the č