Bugfix for hash generation of non utf8 files

This commit is contained in:
Daniel Heule
2026-01-09 09:54:52 +01:00
parent 95b3076fe2
commit 9bd7685dc3
4 changed files with 47 additions and 20 deletions

View File

@@ -8,9 +8,9 @@ use std::path::Path;
pub(crate) fn filesum_filtered( pub(crate) fn filesum_filtered(
path: &Path, path: &Path,
oldfile: &mut String, oldfile: &mut Vec<u8>,
diff_string: &mut String, diff_string: &mut String,
re: Option<&regex::Regex>, re: Option<&regex::bytes::Regex>,
) -> io::Result<(bool, String)> { ) -> io::Result<(bool, String)> {
// Open file for hashing // Open file for hashing
match re { match re {
@@ -18,30 +18,37 @@ pub(crate) fn filesum_filtered(
None => trace!("try to open file {:?} for hashing", path), None => trace!("try to open file {:?} for hashing", path),
} }
let mut hasher = Sha3_256::new(); let mut hasher = Sha3_256::new();
let mut filedata = String::with_capacity(2048); let mut filedata = Vec::with_capacity(2048);
let mut changed = false; let mut changed = false;
// we read only smal files, so its fast to read the whole file to memory, so we can also du a diff // we read only smal files, so its fast to read the whole file to memory, so we can also du a diff
if let Ok(file_contents) = fs::read_to_string(path) { if let Ok(file_contents) = fs::read(path) {
if let Some(re) = re { if let Some(re) = re {
trace!("Filter lines with regex {:?}", re); trace!("Filter lines with regex {:?}", re);
for line in file_contents.lines() { for line in file_contents.split_inclusive(|&b| b == b'\n') {
if re.is_match(line) { if re.is_match(line) {
trace!("line {} skipped by filter regex", line); trace!("line {} skipped by filter regex", String::from_utf8_lossy(line).trim());
continue; continue;
} }
// Update the hasher with the bytes // Update the hasher with the bytes
filedata.push_str(line); filedata.extend_from_slice(line);
} }
} else { } else {
// we do not have a filter regex, so we could simply paste the file to the hasher // we do not have a filter regex, so we could simply paste the file to the hasher
trace!("Hash file without filter regex"); trace!("Hash file without filter regex");
filedata = file_contents; filedata = file_contents;
} }
hasher.update(format!("{}\n", filedata).as_bytes());
// Debug to find errors
//trace!("Hasher input data {:?}",String::from_utf8_lossy(&filedata));
// Update the hasher with the bytes
hasher.update(&filedata);
if !oldfile.is_empty() && *oldfile != filedata { if !oldfile.is_empty() && *oldfile != filedata {
diff_string.clear(); diff_string.clear();
for diff in diff::lines(oldfile, &filedata) { let ofs = String::from_utf8_lossy(oldfile);
let nfs = String::from_utf8_lossy(&filedata);
for diff in diff::lines(&ofs, &nfs) {
match diff { match diff {
Result::Left(l) => { Result::Left(l) => {
trace!("Diff - {}", l); // Removed line trace!("Diff - {}", l); // Removed line

View File

@@ -1,5 +1,6 @@
use log::{debug, error, info}; use log::{debug, error, info};
use regex::Regex; use regex::Regex;
use regex::bytes::Regex as BRegex;
pub(crate) fn compile_re(regex: Option<String>, name: &str) -> Option<Regex> { pub(crate) fn compile_re(regex: Option<String>, name: &str) -> Option<Regex> {
if let Some(r) = regex { if let Some(r) = regex {
@@ -19,3 +20,22 @@ pub(crate) fn compile_re(regex: Option<String>, name: &str) -> Option<Regex> {
None None
} }
} }
pub(crate) fn compile_re_bin(regex: Option<String>, name: &str) -> Option<BRegex> {
if let Some(r) = regex {
let re = BRegex::new(&r);
match re {
Ok(r) => {
debug!("Sucessfull compiled {} filter regex: {:?}", name, r);
Some(r)
}
Err(e) => {
error!("Error compiling {} filter regex: {:?}", name, e);
None
}
}
} else {
info!("No filter regex for {} supplied", name);
None
}
}

View File

@@ -19,7 +19,7 @@ pub mod snmp;
use bonding::bonding_status; use bonding::bonding_status;
use config::DataFunctionsFilesum; use config::DataFunctionsFilesum;
use filesum::filesum_filtered; use filesum::filesum_filtered;
use helper::compile_re; use helper::{compile_re, compile_re_bin};
use multipath::multipath_status; use multipath::multipath_status;
use processes::Ptypes; use processes::Ptypes;
use snmp::{Oid, OidData, SnmpData}; use snmp::{Oid, OidData, SnmpData};
@@ -189,10 +189,10 @@ fn t_filesum(
#[cfg(target_os = "linux")] #[cfg(target_os = "linux")]
{ {
// allocate some strings for holding file contents between check runs // allocate some strings for holding file contents between check runs
let mut oldpasswd = String::with_capacity(2048); let mut oldpasswd = Vec::with_capacity(2048);
let mut oldshadow = String::with_capacity(2048); let mut oldshadow = Vec::with_capacity(2048);
let mut oldgroup = String::with_capacity(2048); let mut oldgroup = Vec::with_capacity(2048);
let mut oldauthkey = String::with_capacity(2048); let mut oldauthkey = Vec::with_capacity(2048);
let mut hash_passwd = String::with_capacity(128); let mut hash_passwd = String::with_capacity(128);
let mut diff_passwd = String::with_capacity(2048); let mut diff_passwd = String::with_capacity(2048);
let mut hash_shadow = String::with_capacity(128); let mut hash_shadow = String::with_capacity(128);
@@ -203,10 +203,10 @@ fn t_filesum(
let mut diff_authkey = String::with_capacity(2048); let mut diff_authkey = String::with_capacity(2048);
// allocate Option<Regex> for our regex ... // allocate Option<Regex> for our regex ...
let re_passwd = compile_re(options.passwd, "passwd"); let re_passwd = compile_re_bin(options.passwd, "passwd");
let re_shadow = compile_re(options.shadow, "shadow"); let re_shadow = compile_re_bin(options.shadow, "shadow");
let re_group = compile_re(options.group, "group"); let re_group = compile_re_bin(options.group, "group");
let re_authkey = compile_re(options.authorized_keys, "authorized_keys"); let re_authkey = compile_re_bin(options.authorized_keys, "authorized_keys");
// prepare variables which we use in the whole function // prepare variables which we use in the whole function
let oid_filesum_time = Oid::from_str("6.1.0").unwrap(); let oid_filesum_time = Oid::from_str("6.1.0").unwrap();

View File

@@ -49,7 +49,7 @@ impl Oid {
x x
} }
/* pub fn add_suffix(&self, o: &Oid) -> Oid { /* pub fn add_suffix(&self, o: &Oid) -> Oid {
let mut x = self.clone(); let mut x = self.clone();
x.0.extend(o.clone().0); x.0.extend(o.clone().0);
x x