Strange behavior with regex lookarounds

[David Wimmer]

I'm having a real strange behavior using regex lookarounds for Lexems inside the EditControl.

I defined my Lexem as follows:

new Lexem() { StartText = "hell(?=o)", IsRegex = true, LexemType = EditTokenType.Keyword, FormatName = "KeywordFormat" }

For testing purpose this is the only Lexem in my CustomLanguage class which inherits from ProceduralLanguageBase.
In the EditControl the Lexem matches "hellO", but not "hello". CaseSensitive is set to false in my CustomLanguage class.
When i change my regex to "hell(?=O)" it matches "hello" but not "hellO".

Apart from that, the correct match would be "hell" in both cases.

How can i fix this?

My CustomLanguage class:

public class CustomLanguage : ProceduralLanguageBase
    {
        private List<EditFormats> formats = new List<EditFormats>()
        {
            new EditFormats() { Foreground = Brushes.Red, FormatName = "KeywordFormat" }
        };

        private List<Lexem> lexems = new List<Lexem>()
        {
            new Lexem() { StartText = "hell(?=o)", IsRegex = true, LexemType = EditTokenType.Keyword, FormatName = "KeywordFormat" }
        };

        public CustomLanguage(EditControl control)
            : base(control)
        {
            Name = "Custom";
            FileExtension = "cus";
            ApplyColoring = true;
            SupportsIntellisense = false;
            SupportsOutlining = true;
            CaseSensitive = false;
            TextForeground = Brushes.Black;
            Lexem = new LexemCollection(lexems);
            Formats = new FormatsCollection(formats);
        }
    }

MainWindow code-behind (setting the DocumentLanguage):

public MainWindow()
        {
            InitializeComponent();
            var customLanguage = new CustomLanguage(editControl1);
            editControl1.DocumentLanguage = Syncfusion.Windows.Edit.Languages.Custom;
            editControl1.CustomLanguage = customLanguage;
        }

I'm using .Net 4.6 and Syncfusion 14.4.0.15

Attachment: EditControlSample_c94ee2e5.7z

[Kannan R]

Hi David,   

  

Thank you for contacting Syncfusion support.    

  

Query 1: How to provide Case insensitive Regex Match?   

   

We must use Inline Case Insensitive option to achieve this reported behavior. Please make use of below MSDN link and code snippet to achieve this requirement.    

  

Link: https://msdn.microsoft.com/en-us/library/yd1hzczs(v=vs.110).aspx#Case    

  

Code Snippet: [C#]   

   

   

new Lexem() { StartText = "hell(?i)(o)", IsRegex = true, LexemType =EditTokenType.Keyword, FormatName = "KeywordFormat" },   

   

   

Screenshot   

   

    

   

Query 2: Usage of default Case Insensitive option in Custom Language.   

   

In this Case, Sensitive option in Custom Language is to enable / disable Case sensitivity for all other items like Keyword and remaining options but not for Regex Input. 

  

For Example   

   

Here, we have added public as Keyword and set as case insensitive. So, it will enable user to highlight any public, even if user enter it as different mixed cases but not for the Regex option. Please refer with the below example.    

  

Screenshot   

   

    

   

We have also modified your sample; it helps to achieve this requirement and it can be downloaded from below location.    

  

Sample: https://www.syncfusion.com/downloads/support/forum/128517/ze/EditControlSample-1696361432   

   

Query 3: Word hell must match all the three cases.    

   

No, as we have provided an option after word “hell”, Regex would not match to it and expect the next character “o” input in it.  

 

Kindly check with the above solution and let us know if it is helpful.   

  

Regards   

Kannan   

 

[David Wimmer]

Hi Kannan,

Query 1: I don't want to match case insensitive.

Your regex hell(?i)(o) has a different meaning than hell(?=o).

Your regex means you want to match hello or hellO, but I want to match hell when it is followed by an o. So it only matches hell in hello, but not hellO.
But when I'm using the regex hell(?=o) as StartText in Lexem it matches whole word hellO and not hello, though it should match hell when it is followed by an o.

See http://regexr.com/3f5nl to text the regex or test it with System.Text.RegularExpressions.Regex from the .Net Framework

Query 2: Is clear to me.

Query 3: Explained in query 1.

Kind Regards
David

[Vijayalakshmi Roopkumar]

 Hi David,

From your update, we come to know that your reported issue with EditControl as follows:

Using the regex you have provided "hell(?=o)" , the word hello matches with the Lower case "o" , whereas the word matches with Upper case "O" at our end.

We have also taken the video for the same:

Video: https://www.syncfusion.com/downloads/support/forum/128517/ze/Editcontrol_video1708501745 

Please go through the video and confirm us whether you have reporting this issue. Also let us know that you want to match the word hell with "o" irrespective of Case Sensitive?

If not please brief about the issue , so that it would be helpful for us to proceed further on this.

Regards,
Vijayalakshmi V.R. 

[David Wimmer]

Hi Vijayalakshm,

yes, I'm reporting the issue as shown in the video.
And I want to match the word hell with "o" considering Case Sensitive.

Examples of the EditControl-behavior I expected using the regex "hell(?=o)":

Input	IsMatch	Formatted text
hello	Yes	hello
hellO	No	hellO
hell	No	hell
Hello	No	Hello
Chello	Yes	Chello

Hope this helps!

Regards,
David

[Vijayalakshmi Roopkumar]

Hi David,

We have confirmed the reported issue with EditControl is a defect and have logged a defect report regarding this. A support incident to track the status of this defect has been created under your account. Please log on to our support website to check for further updates.

https://www.syncfusion.com/account/login?ReturnUrl=%2fsupport%2fdirecttrac%2fincidents 

Regards,
Vijayalakshmi V.R.