Shell Programming and Scripting

I'm just learning Regex and while testing my understanding I received some unexpected results.

I created example.txt with the text "abcddd". Running the command

Code:

grep --color '[^abc]d' example.txt

I received the results:

Code:

"abcddd"

with the first and second letter d highlighted in red.


So based on this return my statement meant- search the file example.txt for a string where the first character is not a,b, or c that is followed by the character d and highlight the found characters.

What is confusing is that in my LPIC guide the character ^ represents the start of line. In practice it seems to represent "exclude the following characters". Is there a difference in using it inside the brackets vs. outside brackets?

---------- Post updated at 01:44 PM ---------- Previous update was at 01:25 PM ----------

Apologize I actually found the answer in after reading more carefully.

The use of carat "^a" specifies that the line start with the character "a".

The use of ^a is context sensitive. If ^a appears at the start of an RE, it matches a input whose first character is a, but if ^ is the first character in a bracket expression (such as [^abc], it makes the bracket expression a non-matching bracket expression (in this case matching any single character that is not a, not b, and not c). So, in your original command, the first place in the string abcddd that matched the BRE [^abc]d was the dd after the c. In this case the first d is a character that was not a, b, or c and the second d matches the d in the BRE after the non-matching bracket expression. To get what you originally requested, I assume that you came up with something like:

Code:

grep --color '^[^abc]d' example.txt

where the first circumflex in the BRE is an anchor and the 2nd circumflex identifies the bracket expression as a non-matching bracket expression.