Community

Hi,

I am trying to find and replace following comma(,) and rest of the line on following three situation of select query and i got struck at the replacement speicification.

1) Line ended with comma(,)
Example: a.column3,
2) Line ended with spaces( ) after comma(,)
Example: a.column,\space\space\space
3) Line ended with comments(--some text spaces) after comma
Example: a.column2, -- comment,

Now i want above three scenarios to be replaced with text just before that comma(,) character.

I could able to find all of them with the following find specification.

Find What: ^\(.*\),[ ]*$\|^\(.*\),[[:blank:]]*--*.*$

Replace with: \1\2

This didn't work well when it finds the string by first expression itself.

I could do this in two search, but i would prefer to do it in single search itself.Please advise me on this....

Thanks
Thamark

I recommend Posix regular expression syntax, as it reduces the number of backslashes and increases readability:

Configure | Preferences | Editor

[X] Use POSIX regular expression syntax

Then:

Find what: ,[[:blank:]]*(--.*)?$
Replace with: [nothing]

[X] Regular expression

Replace All

Ben,

I tried following things and it didn't work.

Configure | Preferences | Editor

[X] Use POSIX regular expression syntax

Tried to do your find and replace for following sample data and it didn't work.

a.column1,
a.column2,
'',
1, -- comment,
'', -- comment,
'', -- comment,
'', -- comment

It didn't find those string at all. My textpad version is Textpad 4.7.3 : 32-bit version.

Result should be like

a.column1
a.column2
''
1
''
''
''

Thanks for your reply and i will be looking forward to see update in this regard...

I can assure you that it does work.

Make sure you remove any spurious trailing space that the web page may stick on the end of the search expression.

Ben,

I tried it again and it works perfectly. I think as you mentioned something added when I was copying from web to tool.

Thanks a lot and I never thought of this solution, since I was struck that I should be handling them as individual cases....

By the way do we have any answer for how to replace OR condition cases just curios

In the general case you can't do this.

Suppose you have an alternation (a regex with a | in it):
E11 (E12) E13 | E21 (E22) E23
so that, in a replacement, \1 is whatever E12 matches and \2 is whatever E22 matches.

Then if the first alternative ( E11 (E12) E13 ) matches, so that \1 is whatever E12 matches, then \2 simply doesn't exist, and if there is a \2 in the replacement it is included literally.

And if the second alternative ( E21 (E22) E23 ) matches, so that \2 is whatever E22 matches, then \1 simply doesn't exist, and if there is a \1 in the replacement it is included literally.

You could always capture everything before the first comma.

POSIX Example:

Find:::^([^,]+),.+$
Replace:::\1

Or if you needed everything before a second comma...

Find:::^([^,]+,[^,]+),.+$
Replace:::\1