Merge pull request #132 from epi052/reimplement-size-filters-using-filter-trait

Reimplement size-based filters using FeroxFilter trait
This commit is contained in:
epi
2020-11-19 14:44:47 -06:00
committed by GitHub
7 changed files with 402 additions and 47 deletions

View File

@@ -11,7 +11,7 @@ Long form explanations of most of the items below can be found in the [CONTRIBUT
## Static analysis checks
- [ ] All rust files are formatted using `cargo fmt`
- [ ] All `clippy` checks pass when running `cargo clippy --all-targets --all-features -- -D warnings -A clippy::unnecessary_unwrap`
- [ ] All `clippy` checks pass when running `cargo clippy --all-targets --all-features -- -D warnings -A clippy::deref_addrof`
- [ ] All existing tests pass
## Documentation

View File

@@ -1,6 +1,6 @@
[package]
name = "feroxbuster"
version = "1.6.2"
version = "1.6.3"
authors = ["Ben 'epi' Risher <epibar052@gmail.com>"]
license = "MIT"
edition = "2018"

View File

@@ -140,3 +140,152 @@ impl FeroxFilter for StatusCodeFilter {
self
}
}
/// Simple implementor of FeroxFilter; used to filter out responses based on the number of lines
/// in a Response body; specified using -N|--filter-lines
#[derive(Default, Debug, PartialEq)]
pub struct LinesFilter {
/// Number of lines in a Response's body that should be filtered
pub line_count: usize,
}
/// implementation of FeroxFilter for LinesFilter
impl FeroxFilter for LinesFilter {
/// Check `line_count` against what was passed in via -N|--filter-lines
fn should_filter_response(&self, response: &FeroxResponse) -> bool {
log::trace!("enter: should_filter_response({:?} {})", self, response);
let result = response.line_count() == self.line_count;
log::trace!("exit: should_filter_response -> {}", result);
result
}
/// Compare one LinesFilter to another
fn box_eq(&self, other: &dyn Any) -> bool {
other.downcast_ref::<Self>().map_or(false, |a| self == a)
}
/// Return self as Any for dynamic dispatch purposes
fn as_any(&self) -> &dyn Any {
self
}
}
/// Simple implementor of FeroxFilter; used to filter out responses based on the number of words
/// in a Response body; specified using -W|--filter-words
#[derive(Default, Debug, PartialEq)]
pub struct WordsFilter {
/// Number of words in a Response's body that should be filtered
pub word_count: usize,
}
/// implementation of FeroxFilter for WordsFilter
impl FeroxFilter for WordsFilter {
/// Check `word_count` against what was passed in via -W|--filter-words
fn should_filter_response(&self, response: &FeroxResponse) -> bool {
log::trace!("enter: should_filter_response({:?} {})", self, response);
let result = response.word_count() == self.word_count;
log::trace!("exit: should_filter_response -> {}", result);
result
}
/// Compare one WordsFilter to another
fn box_eq(&self, other: &dyn Any) -> bool {
other.downcast_ref::<Self>().map_or(false, |a| self == a)
}
/// Return self as Any for dynamic dispatch purposes
fn as_any(&self) -> &dyn Any {
self
}
}
/// Simple implementor of FeroxFilter; used to filter out responses based on the length of a
/// Response body; specified using -S|--filter-size
#[derive(Default, Debug, PartialEq)]
pub struct SizeFilter {
/// Overall length of a Response's body that should be filtered
pub content_length: u64,
}
/// implementation of FeroxFilter for SizeFilter
impl FeroxFilter for SizeFilter {
/// Check `content_length` against what was passed in via -S|--filter-size
fn should_filter_response(&self, response: &FeroxResponse) -> bool {
log::trace!("enter: should_filter_response({:?} {})", self, response);
let result = response.content_length() == self.content_length;
log::trace!("exit: should_filter_response -> {}", result);
result
}
/// Compare one SizeFilter to another
fn box_eq(&self, other: &dyn Any) -> bool {
other.downcast_ref::<Self>().map_or(false, |a| self == a)
}
/// Return self as Any for dynamic dispatch purposes
fn as_any(&self) -> &dyn Any {
self
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
/// just a simple test to increase code coverage by hitting as_any and the inner value
fn lines_filter_as_any() {
let filter = LinesFilter { line_count: 1 };
assert_eq!(filter.line_count, 1);
assert_eq!(
*filter.as_any().downcast_ref::<LinesFilter>().unwrap(),
filter
);
}
#[test]
/// just a simple test to increase code coverage by hitting as_any and the inner value
fn words_filter_as_any() {
let filter = WordsFilter { word_count: 1 };
assert_eq!(filter.word_count, 1);
assert_eq!(
*filter.as_any().downcast_ref::<WordsFilter>().unwrap(),
filter
);
}
#[test]
/// just a simple test to increase code coverage by hitting as_any and the inner value
fn size_filter_as_any() {
let filter = SizeFilter { content_length: 1 };
assert_eq!(filter.content_length, 1);
assert_eq!(
*filter.as_any().downcast_ref::<SizeFilter>().unwrap(),
filter
);
}
#[test]
/// just a simple test to increase code coverage by hitting as_any and the inner value
fn status_code_filter_as_any() {
let filter = StatusCodeFilter { filter_code: 200 };
assert_eq!(filter.filter_code, 200);
assert_eq!(
*filter.as_any().downcast_ref::<StatusCodeFilter>().unwrap(),
filter
);
}
}

View File

@@ -3,7 +3,7 @@ use feroxbuster::{
banner,
config::{CONFIGURATION, PROGRESS_BAR, PROGRESS_PRINTER},
heuristics, logger, reporter,
scanner::{scan_url, PAUSE_SCAN},
scanner::{self, scan_url, PAUSE_SCAN},
utils::{ferox_print, get_current_depth, module_colorizer, status_colorizer},
FeroxError, FeroxResponse, FeroxResult, SLEEP_DURATION, VERSION,
};
@@ -112,6 +112,16 @@ async fn scan(
return Err(Box::new(err));
}
scanner::initialize(
words.len(),
CONFIGURATION.scan_limit,
&CONFIGURATION.extensions,
&CONFIGURATION.filter_status,
&CONFIGURATION.filter_line_count,
&CONFIGURATION.filter_word_count,
&CONFIGURATION.filter_size,
);
let mut tasks = vec![];
for target in targets {

View File

@@ -1,7 +1,9 @@
use crate::{
config::CONFIGURATION,
extractor::get_links,
filters::{FeroxFilter, StatusCodeFilter, WildcardFilter},
filters::{
FeroxFilter, LinesFilter, SizeFilter, StatusCodeFilter, WildcardFilter, WordsFilter,
},
heuristics, progress,
utils::{format_url, get_current_depth, make_request},
FeroxChannel, FeroxResponse, SLEEP_DURATION,
@@ -19,7 +21,7 @@ use std::{
convert::TryInto,
io::{stderr, Write},
ops::Deref,
sync::atomic::{AtomicBool, AtomicUsize, Ordering},
sync::atomic::{AtomicBool, AtomicU64, AtomicUsize, Ordering},
sync::{Arc, RwLock},
};
use tokio::{
@@ -34,6 +36,9 @@ use tokio::{
/// Single atomic number that gets incremented once, used to track first scan vs. all others
static CALL_COUNT: AtomicUsize = AtomicUsize::new(0);
/// Single atomic number that gets holds the number of requests to be sent per directory scanned
pub static NUMBER_OF_REQUESTS: AtomicU64 = AtomicU64::new(0);
/// Atomic boolean flag, used to determine whether or not a scan should pause or resume
pub static PAUSE_SCAN: AtomicBool = AtomicBool::new(false);
@@ -415,22 +420,6 @@ async fn try_recursion(
/// Simple helper to stay DRY; determines whether or not a given `FeroxResponse` should be reported
/// to the user or not.
pub fn should_filter_response(response: &FeroxResponse) -> bool {
if CONFIGURATION
.filter_size
.contains(&response.content_length())
|| CONFIGURATION
.filter_line_count
.contains(&response.line_count())
|| CONFIGURATION
.filter_word_count
.contains(&response.word_count())
{
// filtered value from --filter-size, size filters and wildcards are two separate filters
// and are applied independently
log::debug!("size filter: filtered out {}", response.url());
return true;
}
match FILTERS.read() {
Ok(filters) => {
for filter in filters.iter() {
@@ -594,14 +583,11 @@ pub async fn scan_url(
let (tx_dir, rx_dir): FeroxChannel<String> = mpsc::unbounded_channel();
let num_reqs_expected: u64 = if CONFIGURATION.extensions.is_empty() {
wordlist.len().try_into().unwrap()
} else {
let total = wordlist.len() * (CONFIGURATION.extensions.len() + 1);
total.try_into().unwrap()
};
let progress_bar = progress::add_bar(&target_url, num_reqs_expected, false);
let progress_bar = progress::add_bar(
&target_url,
NUMBER_OF_REQUESTS.load(Ordering::Relaxed),
false,
);
progress_bar.reset_elapsed();
if CALL_COUNT.load(Ordering::Relaxed) == 0 {
@@ -610,13 +596,6 @@ pub async fn scan_url(
// this protection allows us to add the first scanned url to SCANNED_URLS
// from within the scan_url function instead of the recursion handler
add_url_to_list_of_scanned_urls(&target_url, &SCANNED_URLS);
if CONFIGURATION.scan_limit == 0 {
// scan_limit == 0 means no limit should be imposed... however, scoping the Semaphore
// permit is tricky, so as a workaround, we'll add a ridiculous number of permits to
// the semaphore (1,152,921,504,606,846,975 to be exact) and call that 'unlimited'
SCAN_LIMITER.add_permits(usize::MAX >> 4);
}
}
// When acquire is called and the semaphore has remaining permits, the function immediately
@@ -653,15 +632,6 @@ pub async fn scan_url(
add_filter_to_list_of_ferox_filters(filter, FILTERS.clone());
// add any status code filters to `FILTERS`
for code_filter in &CONFIGURATION.filter_status {
let filter = StatusCodeFilter {
filter_code: *code_filter,
};
let boxed_filter = Box::new(filter);
add_filter_to_list_of_ferox_filters(boxed_filter, FILTERS.clone());
}
// producer tasks (mp of mpsc); responsible for making requests
let producers = stream::iter(looping_words.deref().to_owned())
.map(|word| {
@@ -715,6 +685,84 @@ pub async fn scan_url(
log::trace!("exit: scan_url");
}
/// Perform steps necessary to run scans that only need to be performed once (warming up the
/// engine, as it were)
pub fn initialize(
num_words: usize,
scan_limit: usize,
extensions: &[String],
status_code_filters: &[u16],
lines_filters: &[usize],
words_filters: &[usize],
size_filters: &[u64],
) {
log::trace!(
"enter: initialize({}, {}, {:?}, {:?}, {:?}, {:?}, {:?})",
num_words,
scan_limit,
extensions,
status_code_filters,
lines_filters,
words_filters,
size_filters,
);
// number of requests only needs to be calculated once, and then can be reused
let num_reqs_expected: u64 = if extensions.is_empty() {
num_words.try_into().unwrap()
} else {
let total = num_words * (extensions.len() + 1);
total.try_into().unwrap()
};
NUMBER_OF_REQUESTS.store(num_reqs_expected, Ordering::Relaxed);
// add any status code filters to `FILTERS` (-C|--filter-status)
for code_filter in status_code_filters {
let filter = StatusCodeFilter {
filter_code: *code_filter,
};
let boxed_filter = Box::new(filter);
add_filter_to_list_of_ferox_filters(boxed_filter, FILTERS.clone());
}
// add any line count filters to `FILTERS` (-N|--filter-lines)
for lines_filter in lines_filters {
let filter = LinesFilter {
line_count: *lines_filter,
};
let boxed_filter = Box::new(filter);
add_filter_to_list_of_ferox_filters(boxed_filter, FILTERS.clone());
}
// add any line count filters to `FILTERS` (-W|--filter-words)
for words_filter in words_filters {
let filter = WordsFilter {
word_count: *words_filter,
};
let boxed_filter = Box::new(filter);
add_filter_to_list_of_ferox_filters(boxed_filter, FILTERS.clone());
}
// add any line count filters to `FILTERS` (-S|--filter-size)
for size_filter in size_filters {
let filter = SizeFilter {
content_length: *size_filter,
};
let boxed_filter = Box::new(filter);
add_filter_to_list_of_ferox_filters(boxed_filter, FILTERS.clone());
}
if scan_limit == 0 {
// scan_limit == 0 means no limit should be imposed... however, scoping the Semaphore
// permit is tricky, so as a workaround, we'll add a ridiculous number of permits to
// the semaphore (1,152,921,504,606,846,975 to be exact) and call that 'unlimited'
SCAN_LIMITER.add_permits(usize::MAX >> 4);
}
log::trace!("exit: initialize");
}
#[cfg(test)]
mod tests {
use super::*;

View File

@@ -92,8 +92,8 @@ pub fn get_url_path_length(url: &Url) -> u64 {
let path = url.path();
let segments = if path.starts_with('/') {
path[1..].split_terminator('/')
let segments = if let Some(split) = path.strip_prefix('/') {
split.split_terminator('/')
} else {
log::trace!("exit: get_url_path_length -> 0");
return 0;

View File

@@ -55,3 +55,151 @@ fn filters_status_code_should_filter_response() {
assert_eq!(mock_two.times_called(), 1);
teardown_tmp_directory(tmp_dir);
}
#[test]
/// create a FeroxResponse that should elicit a true from
/// LinesFilter::should_filter_response
fn filters_lines_should_filter_response() {
let srv = MockServer::start();
let (tmp_dir, file) =
setup_tmp_directory(&["LICENSE".to_string(), "file.js".to_string()], "wordlist").unwrap();
let mock = Mock::new()
.expect_method(GET)
.expect_path("/LICENSE")
.return_status(302)
.return_body("this is a test")
.create_on(&srv);
let mock_two = Mock::new()
.expect_method(GET)
.expect_path("/file.js")
.return_status(200)
.return_body("this is also a test of some import\nwith 2 lines, no less")
.create_on(&srv);
let cmd = Command::cargo_bin("feroxbuster")
.unwrap()
.arg("--url")
.arg(srv.url("/"))
.arg("--wordlist")
.arg(file.as_os_str())
.arg("--filter-lines")
.arg("2")
.unwrap();
cmd.assert().success().stdout(
predicate::str::contains("/LICENSE")
.and(predicate::str::contains("302"))
.and(predicate::str::contains("14"))
.and(predicate::str::contains("/file.js"))
.not()
.and(predicate::str::contains("200"))
.not()
.and(predicate::str::contains("2l"))
.not(),
);
assert_eq!(mock.times_called(), 1);
assert_eq!(mock_two.times_called(), 1);
teardown_tmp_directory(tmp_dir);
}
#[test]
/// create a FeroxResponse that should elicit a true from
/// WordsFilter::should_filter_response
fn filters_words_should_filter_response() {
let srv = MockServer::start();
let (tmp_dir, file) =
setup_tmp_directory(&["LICENSE".to_string(), "file.js".to_string()], "wordlist").unwrap();
let mock = Mock::new()
.expect_method(GET)
.expect_path("/LICENSE")
.return_status(302)
.return_body("this is a test")
.create_on(&srv);
let mock_two = Mock::new()
.expect_method(GET)
.expect_path("/file.js")
.return_status(200)
.return_body("this is also a test of some import\nwith 2 lines, no less")
.create_on(&srv);
let cmd = Command::cargo_bin("feroxbuster")
.unwrap()
.arg("--url")
.arg(srv.url("/"))
.arg("--wordlist")
.arg(file.as_os_str())
.arg("--filter-words")
.arg("13")
.unwrap();
cmd.assert().success().stdout(
predicate::str::contains("/LICENSE")
.and(predicate::str::contains("302"))
.and(predicate::str::contains("14"))
.and(predicate::str::contains("/file.js"))
.not()
.and(predicate::str::contains("200"))
.not()
.and(predicate::str::contains("13w"))
.not(),
);
assert_eq!(mock.times_called(), 1);
assert_eq!(mock_two.times_called(), 1);
teardown_tmp_directory(tmp_dir);
}
#[test]
/// create a FeroxResponse that should elicit a true from
/// SizeFilter::should_filter_response
fn filters_size_should_filter_response() {
let srv = MockServer::start();
let (tmp_dir, file) =
setup_tmp_directory(&["LICENSE".to_string(), "file.js".to_string()], "wordlist").unwrap();
let mock = Mock::new()
.expect_method(GET)
.expect_path("/LICENSE")
.return_status(302)
.return_body("this is a test")
.create_on(&srv);
let mock_two = Mock::new()
.expect_method(GET)
.expect_path("/file.js")
.return_status(200)
.return_body("this is also a test of some import\nwith 2 lines, no less")
.create_on(&srv);
let cmd = Command::cargo_bin("feroxbuster")
.unwrap()
.arg("--url")
.arg(srv.url("/"))
.arg("--wordlist")
.arg(file.as_os_str())
.arg("--filter-size")
.arg("56")
.unwrap();
cmd.assert().success().stdout(
predicate::str::contains("/LICENSE")
.and(predicate::str::contains("302"))
.and(predicate::str::contains("14"))
.and(predicate::str::contains("/file.js"))
.not()
.and(predicate::str::contains("200"))
.not()
.and(predicate::str::contains("56c"))
.not(),
);
assert_eq!(mock.times_called(), 1);
assert_eq!(mock_two.times_called(), 1);
teardown_tmp_directory(tmp_dir);
}