cl_abap_matcher and [[:unicode:]]

[former_member418756]

Hello experts,

I'm having difficulty using ABAP class cl_abap_matcher with special characters.  Specifically, I'm trying to find any unicode characters in a string using cl_abap_matcher with the following code:

p_instring = 'THIS IS A TEST.  你好 THIS IS ONLY A TEST.'

matcher = cl_abap_matcher=>create( pattern = '[[:unicode:]]'

                                                        text     = p_instring ).

if matcher->match( ).

"match found

...

However, if I use the "find regex" version:

p_instring = 'THIS IS A TEST.  你好 THIS IS ONLY A TEST.'

find regex '[[:unicode:]]' in p_instring

this works fine.

Thanks in advance for any help!

Bruce

[Sandra_Rossi]

It's because matcher->match( ) does not return a boolean but a structure of type MATCH_RESULT !

Instead, you may use if cl_abap_matcher=>matches( pattern = '[[:Unicode:]]' text = 'THIS IS A TEST ... ONLY A TEST.' ).

[former_member418756]

Hi Sandra,

Thank you for the response but matcher->match( ) returns a type ABAP_BOOL value from local variable "SUCCESS":

For a test, I also tried to use cl_abap_matcher=>matches... but got the same results - false for [[:unicode:]] test:

lv_match = cl_abap_matcher=>matches( pattern = <fs_str_rpl>-srch_regex

                                                             text     = 'THIS IS A TEST.  你好 THIS IS ONLY A TEST.' ).

Thanks again for the reply!

Bruce

[Sandra_Rossi]

Sorry, I made a confusion with matcher->get_match( ).

Horst is right! It works with: if cl_abap_matcher=>matches( pattern = '.*[[:unicode:]].*' text = p_instring ).

[horst_keller]

Where .* is the most simple but sometimes not the best way, because it matches everything and is greedy etc. I normally prefer [^ ... ]* expressions.

[horst_keller]

Hey, isn't that simply because a matcher matches the whole string while FIND looks for a portion in a string??

Without putting something appropriate around the current regex, of course it won't work. Use DEMO_REGEX for trying.

B.T.W. there are also built-in functions MATCH and MATCHES.

ABAP Keyword Documentation

[former_member418756]

Thank you both for the feedback!  My requirement is to handle both a find and replace scenario using cl_abap_matcher given the same regex input.  Since the replace feature works with [[:unicode:]] just fine, I didn't want to use a different search string when using match or matches.

I didn't realize that those to methods used the entire string.  Now, rather than using match or matches, I'm simply using find_next and the same regex expression can work for both find and replace search types.

...

lc_matcher = cl_abap_matcher=>create( pattern = '[[:unicode:]]'

                                                            text     = 'THIS IS A TEST.  你好 THIS IS ONLY A TEST.' ).

if lc_matcher->find_next( ) = abap_true.

...

This works perfectly.  Thanks again for the feedback!  Wouldn't have figured it out without the help!

Bruce