Regexp question

Kjetil Svendsen

Hi. I have a text input like this:

a: a-related text
b: b-related text
c: c-related text
d: d-related text
e: e-related text
f: f-related text

and a expression like this:

(a\:\s*)(?.*)(\r\n)(b\:\s*)(?.*)(\r\n)(c\:\s*)(?.*)(\r\n)(d\:\s*)(?.*)(\r\n)(e\:\s*)(?.*)(\r\n)(f\:\s*)(?.*)(\r\n)

So far so good. The problem is that I never know how many of the lines a-f which is present. The text might look like this:

a: a-related text
c: c-related text
f: f-related text

or:

a: a-related text
b: b-related text
some strange text
f: f-related text

Any suggestion on how to write my expression? Kjetil

leppie

Use multiline matching.

xacc.ide - now with TabsToSpaces support
IronScheme - 1.0 alpha 4a out now (29 May 2008)

Guffa

Don't write a pattern for matching each separate line, write a pattern that matches any of the lines that you want: (?:^|\r\n)([a-f]):\s*(.+?)(?:$|\r\n) This should give you a match for each line, containing the a-f character and the related text as captures.

Despite everything, the person most likely to be fooling you next is yourself.

Kjetil Svendsen

Beautiful expression. But how do I name my groups? Kjetil

Guffa

Without naming the groups, you can access them by index. You can name them like this: (?:^|\r\n)(?<section>[a-f]):\s*(?<description>.+?)(?:$|\r\n)

Despite everything, the person most likely to be fooling you next is yourself.

Kjetil Svendsen

Hi. I really appreciate your help, but I still need some help. RegExp is not my strongest side :-) I need to rephrase my question. My text input is more like: (I'm not allowed to use actual data from my project)

Breakfast: Sliced Bread
Lunch: At McDonald's
Dinner: Chicken
Supper: Mexican

and my task is to isolate what a person ate and save that info in my database. Different person might have different ways to report their meals. Each person has a consistent way though. There are ~150 persons and they all skip meals from time to time. Can you still help me ? Kjetil

Guffa

Instead of ([a-f]) in the pattern to match a single character, use (Breakfast|Lunch|Dinner|Supper) to match the strings.

Despite everything, the person most likely to be fooling you next is yourself.

Kjetil Svendsen

Hi. It still does not work. New problem: The user input is like this

Breakfast: Sliced Bread
Lunch: At McDonald's
(had some coffee and snack)
Dinner: Chicken
Supper: Mexican

and my match.Groups[0].Value is

Breakfast: Sliced Bread
Lunch: At McDonald's

instead of

Breakfast: Sliced Bread
Lunch: At McDonald's
Dinner: Chicken
Supper: Mexican

Why? This is starting to be embarrassing :) Kjetil

Guffa

No, that is not correct. If you check your result again, you will see that you are getting the breakfast and the dinner. That is because the pattern is consuming the line break before and after each match, so the next line won't match. Change the part of the pattern that matches the end of the line or end of the string from (?:$|\r\n) to (?=$|\r\n) to make a zero-width positive lookahead.

Despite everything, the person most likely to be fooling you next is yourself.