Mirror/zeek
1
0
Fork 0
mirror of https://github.com/zeek/zeek.git synced 2025-10-09 18:18:19 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

Updates for the urls.bro script. Fixes BIT-1404.

Browse source
This commit is contained in:
Seth Hall 2015-06-01 11:38:26 -04:00
parent 2b1cd66f17
commit 097354a43f
3 changed files with 48 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

34
scripts/base/utils/urls.bro
View file

@ -6,23 +6,23 @@ const url_regex = /^([a-zA-Z\-]{3,5})(:\/\/[^\/?#"'\r\n><]*)([^?#"'\r\n><]*)([^[
## A URI, as parsed by :bro:id:`decompose_uri`.
type URI: record {
## The URL's scheme..
scheme: string &optional;
scheme: string &optional;
## The location, which could be a domain name or an IP address. Left empty if not
## specified.
netlocation: string;
netlocation: string;
## Port number, if included in URI.
portnum: count &optional;
portnum: count &optional;
## Full including the file name. Will be '/' if there's not path given.
path: string;
path: string;
## Full file name, including extension, if there is a file name.
file_name: string &optional;
file_name: string &optional;
## The base filename, without extension, if there is a file name.
file_base: string &optional;
file_base: string &optional;
## The filename's extension, if there is a file name.
file_ext: string &optional;
file_ext: string &optional;
## A table of all query parameters, mapping their keys to values, if there's a
## query.
params: table[string] of string &optional;
params: table[string] of string &optional;
};
## Extracts URLs discovered in arbitrary text.
@ -46,19 +46,19 @@ function find_all_urls_without_scheme(s: string): string_set
return return_urls;
}
function decompose_uri(s: string): URI
function decompose_uri(uri: string): URI
{
local parts: string_vec;
local u: URI = [$netlocation="", $path="/"];
local u = URI($netlocation="", $path="/");
local s = uri;
if ( /\?/ in s)
if ( /\?/ in s )
{
# Parse query.
u$params = table();
parts = split_string1(s, /\?/);
s = parts[0];
local query: string = parts[1];
local query = parts[1];
if ( /&/ in query )
{
@ -73,7 +73,7 @@ function decompose_uri(s: string): URI
}
}
}
else
else if ( /=/ in query )
{
parts = split_string1(query, /=/);
u$params[parts[0]] = parts[1];
@ -97,14 +97,14 @@ function decompose_uri(s: string): URI
if ( |u$path| > 1 && u$path[|u$path| - 1] != "/" )
{
local last_token: string = find_last(u$path, /\/.+/);
local last_token = find_last(u$path, /\/.+/);
local full_filename = split_string1(last_token, /\//)[1];
if ( /\./ in full_filename )
{
u$file_name = full_filename;
u$file_base = split_string1(full_filename, /\./)[0];
u$file_ext = split_string1(full_filename, /\./)[1];
u$file_ext = split_string1(full_filename, /\./)[1];
}
else
{
@ -122,7 +122,9 @@ function decompose_uri(s: string): URI
u$portnum = to_count(parts[1]);
}
else
{
u$netlocation = s;
}
return u;
}