Using a dot in replace()

[PJCarr]

It seems that using a dot character, following another character e.g. ',.' in a replace function replaces all instances of the first character plus the one following it.

so if field2 = 'line1, line2, line3,.line4'

Code: Select all

set field1 = replace(field2, ',.', '.') ;

sets field1 to 'line1.line2.line3.line4' and not the intended 'line1, line2, line3.line4'

Also

Code: Select all

set field1 = replace(field2, ',\.', '.') ;

(verify actually complains at this but the code still runs).

doesn't give 'line1, line2, line3.line4'
but
'line1.line2.line3.line4' as before.

So, how do you replace ',.' with '.' using the replace function?

Is this a feature, bug or unintended consequence?

Any and all help appreciated.

[AJDulk]

You need to use Regular Expression syntax. Thus

Code: Select all

set field1 = replace(field2, ',.', '.') ;

becomes

Code: Select all

set field1 = replace(field2, ',[.]', '.') ;

Regards,
Antony

[PJCarr]

AJDulk - thanks for the reply.

Do you know if this is documented anywhere?

What else can one do with it?

[AJDulk]

For a general description see the Regular Expression article on Wikipedia. There are various links available in the article, including one to Regular Expression Examples.

For even more reading, do a search in your favourite search engine.

Regards,
Antony

[PJCarr]

Antony,

I should have been more specfic. What I meant was is this documented anywhere within ebase and which bits of RE can be used within replace().

It's nicer to have this stuff documented than waste time experimenting ...

Perhaps one of the Ebase team could enlighten me?

[AJDulk]

Ah, okay.

The only documentation I can find in Ebase about Regular Expressions is the System Validators section of the documentation (.../doc/validators.htm). It is, frankly, a very basic summary.

I am interested to see what light our friends at Ebase can throw on the subject.

Regards,
Antony