Hi,
I quite agree with vtam,
>From p.376 of "Internet Routing Architectures" Sec. Edition,
ip as-path access-list 4 permit ^1 ?[0-9]*$
This has some side effects, which allow as-paths such as 109, 110, 120
like vtam has pointed out, in addition to "all the AS paths that start
with 1 and that are of length 2-that is, AS1 and its direct customers"
I think the right answer would be as follows:
ip as-path access-list 4 permit ^1 [0-9]*$
ip as-path access-list 4 permit ^1$
The first line is for its direct cumstomers, and the second line for
its local networks. I'm afraid there's no one line version of this.
Am I right?
On 1 Dec 2000 11:41:56 -0500, [EMAIL PROTECTED] ("michael champion")
wrote:
>Although I am by no means a regular expression expert, I do not believe that
>^1 [0-9]*$ will match an AS-path of 1; the space forces a requirement of a
>maximum of two path elements, with the first element a 1. A single AS-path
>of 1 would not match. Multiple AS-path elements are separated by a space .
>Thus, the expression would match 1 100, 1 200, 1 1, 1 {any AS number}, but
>would not match just 1. What is confusing about regular expressions is that
>every element in the expression is an operator, and sometimes you have
>operators acting on other operators. Thus .* matches all paths because . is
>an operator representing any single character and * is an operator
>representing . repeated as many times as necessary (not just the particular
>character that . matched). This is thus equivalent to .., ..., ...., etc. as
>many times as necessary.
>
>Powerful, but cryptic, and not well-documented at all by Cisco (who knows
>which type of regular expression engine they are using?) This all comes from
>the Unix world, which is why you see it in Perl, Tcl, egrep, etc.
>
>Regards,
>MLC
>"vtam" <[EMAIL PROTECTED]> wrote in message 907i83$9l6$[EMAIL PROTECTED]">news:907i83$9l6$[EMAIL PROTECTED]...
>> The space is the books. Because it said it match all the AS_paths that
>start
>> with 1 and of length 2, such as AS_path=1, =1 10, =1 200, etc. As you
>said,
>> ? # 0 or 1 of the preceding characters (in this case a space). So if ?=0,
>it
>> is ^1[0-9]*$; if ?=1, it is ^1 [0-9]*$.
>> Do i misread your meaning?
>>
>>
>> "Drew Simonis" <[EMAIL PROTECTED]> wrote in message
>> [EMAIL PROTECTED]">news:[EMAIL PROTECTED]...
>> >
>> > ^1 ?[0-9]*$
>> > ^
>> > ^
>> >
>> > Is that space yours or the books? Broken apart, that regex matches
>> > (assuming standard egrep'ish metachars)
>> >
>> > ^ # beginning of line
>> > 1 # followed by the digit 1
>> > (space) # followed by a space
>> > ? # 0 or 1 of the preceding characters (in this case a space)
>> > [0-9] # a single digit within the range of 0-9
>> > * # 0 or more of the preceding characters, up to the end of
>> > # the pattern
>> > $ # end of line char
>> >
>> > So, is this equivalent to ^1[0-9]*$? I don't think so. Assuming
>> > that the pattern with a space was a typo, we are allowed an
>> > optional 1. Assuming it wasn't a typo, we are allowed the space
>> > character. Neither of these options would be matched by your more
>> > restrictive pattern. As for the specific pattern to match, you
>> > can't really say without knowing what you are matching with.
>> >
>> > Different regex engines support different metachars.
>>
>>
>>
>> _________________________________
>> FAQ, list archives, and subscription info:
>http://www.groupstudy.com/list/cisco.html
>> Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
>>
>
>
>_________________________________
>FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
>Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
_________________________________
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
