X-Git-Url: https://code.octet-stream.net/hashgood/blobdiff_plain/0252a4d67a68e71c811a6f7ec8057b616e65bca7..93ceef0a11e57980c967222a6504da7005895a5b:/src/verify.rs diff --git a/src/verify.rs b/src/verify.rs index e519f8a..0f2d1d8 100644 --- a/src/verify.rs +++ b/src/verify.rs @@ -3,13 +3,12 @@ use super::{ VerificationSource, }; #[cfg(feature = "paste")] -use clipboard::{ClipboardContext, ClipboardProvider}; -use regex::Regex; +use copypasta::{ClipboardContext, ClipboardProvider}; use std::fs::File; use std::io; use std::io::prelude::*; use std::io::BufReader; -use std::path::PathBuf; +use std::path::Path; /// Calculate a list of candidate hashes based on the options specified. /// If no hash options have been specified returns None. @@ -45,7 +44,7 @@ fn get_by_parameter(param: &str) -> Result { fn get_from_clipboard() -> Result { #[cfg(feature = "paste")] { - let mut ctx: ClipboardContext = match ClipboardProvider::new() { + let mut ctx: ClipboardContext = match ClipboardContext::new() { Ok(ctx) => ctx, Err(e) => return Err(format!("Error getting system clipboard: {}", e)), }; @@ -62,20 +61,20 @@ fn get_from_clipboard() -> Result { filename: None, bytes, }; - return Ok(CandidateHashes { + Ok(CandidateHashes { alg, hashes: vec![candidate], source: VerificationSource::Clipboard, - }); + }) } #[cfg(not(feature = "paste"))] { - return Err("Paste not implemented".to_owned()); + Err("Paste not implemented".to_owned()) } } /// Generate a candidate hash from the digests file specified (could be "-" for STDIN), or throw an error. -fn get_from_file(path: &PathBuf) -> Result { +fn get_from_file(path: &Path) -> Result { // Get a reader for either standard input or the chosen path let reader: Box = if path.to_str() == Some("-") { Box::new(std::io::stdin()) @@ -97,7 +96,7 @@ fn get_from_file(path: &PathBuf) -> Result { let line = line.trim().to_owned(); // Does our first line look like a raw hash on its own? If so, use that - if let Some(candidate) = read_raw_candidate_from_file(&line, &path) { + if let Some(candidate) = read_raw_candidate_from_file(&line, path) { return Ok(candidate); } @@ -106,7 +105,7 @@ fn get_from_file(path: &PathBuf) -> Result { let full_lines = vec![Ok(line)].into_iter().chain(reader.lines()); // Does the entire file look like a coreutils-style digests file? (SHA1SUMS, etc.) - if let Some(candidate) = read_coreutils_digests_from_file(full_lines, &path) { + if let Some(candidate) = read_coreutils_digests_from_file(full_lines, path) { return Ok(candidate); } @@ -118,81 +117,80 @@ fn get_from_file(path: &PathBuf) -> Result { )) } -fn read_raw_candidate_from_file(line: &str, path: &PathBuf) -> Option { - // It is a little sad to use a dynamic regex in an otherwise nice Rust program - // These deserve to be replaced with a good old fashioned static parser - // But let's be honest: the impact is negligible - let re = Regex::new(r"^([[:xdigit:]]{32}|[[:xdigit:]]{40}|[[:xdigit:]]{64})$").unwrap(); - if re.is_match(line) { - // These should both always succeed due to the matching - let bytes = match hex::decode(line) { - Ok(bytes) => bytes, - _ => return None, - }; - let alg = match Algorithm::from_len(bytes.len()) { - Ok(alg) => alg, - _ => return None, - }; - return Some(CandidateHashes { - alg, - source: VerificationSource::RawFile(path.clone()), - hashes: vec![CandidateHash { - bytes, - filename: None, - }], - }); - } - None +fn try_parse_hash(s: &str) -> Option<(Algorithm, Vec)> { + let bytes = match hex::decode(s.trim()) { + Ok(bytes) => bytes, + _ => return None, + }; + let alg = match Algorithm::from_len(bytes.len()) { + Ok(alg) => alg, + _ => return None, + }; + Some((alg, bytes)) } -fn read_coreutils_digests_from_file(lines: I, path: &PathBuf) -> Option +fn read_raw_candidate_from_file(line: &str, path: &Path) -> Option { + let (alg, bytes) = try_parse_hash(line)?; + Some(CandidateHashes { + alg, + source: VerificationSource::RawFile(path.to_string_lossy().to_string()), + hashes: vec![CandidateHash { + bytes, + filename: None, + }], + }) +} + +fn read_coreutils_digests_from_file(lines: I, path: &Path) -> Option where - I: Iterator>, + I: Iterator>, + S: AsRef, { - let re = Regex::new( - r"^(?P([[:xdigit:]]{32}|[[:xdigit:]]{40}|[[:xdigit:]]{64})) .(?P.+)$", - ) - .unwrap(); - let mut hashes = vec![]; let mut alg: Option = None; - for l in lines { - if let Ok(l) = l { - let l = l.trim(); - // Allow (ignore) blank lines - if l.is_empty() { - continue; - } - // If we can capture a valid line, use it - if let Some(captures) = re.captures(&l) { - let hash = &captures["hash"]; - let filename = &captures["filename"]; - // Decode the hex and algorithm for this line - let line_bytes = match hex::decode(hash) { - Ok(bytes) => bytes, - _ => return None, - }; - let line_alg = match Algorithm::from_len(line_bytes.len()) { - Ok(alg) => alg, - _ => return None, - }; - if alg.is_some() && alg != Some(line_alg) { - // Different algorithms in the same digest file are not supported - return None; + for l in lines.flatten() { + let l = l.as_ref().trim(); + // Allow (ignore) blank lines + if l.is_empty() { + continue; + } + // Expected format + // + let (line_alg, bytes, filename) = match l + .find(' ') + .and_then(|space_pos| { + // Char before filename should be space for text or * for binary + match l.chars().nth(space_pos + 1) { + Some(' ') | Some('*') => (l.get(..space_pos)).zip(l.get(space_pos + 2..)), + _ => None, + } + }) + .and_then(|(maybe_hash, filename)| { + // Filename should be in this position without extra whitespace + if filename.trim() == filename { + try_parse_hash(maybe_hash).map(|(alg, bytes)| (alg, bytes, filename)) } else { - // If we are the first line, we define the overall algorithm - alg = Some(line_alg); + None } - // So far so good - create an entry for this line - hashes.push(CandidateHash { - bytes: line_bytes, - filename: Some(filename.to_owned()), - }); - } else { - // But if we have a line with content we cannot parse, this is an error + }) { + Some(t) => t, + None => { + // if we have a line with content we cannot parse, this is an error return None; } + }; + if alg.is_some() && alg != Some(line_alg) { + // Different algorithms in the same digest file are not supported + return None; + } else { + // If we are the first line, we define the overall algorithm + alg = Some(line_alg); } + // So far so good - create an entry for this line + hashes.push(CandidateHash { + bytes, + filename: Some(filename.to_owned()), + }); } // It is a failure if we got zero hashes or we somehow don't know the algorithm @@ -207,7 +205,7 @@ where // Otherwise all is well and we can return our results Some(CandidateHashes { alg, - source: VerificationSource::DigestsFile(path.clone()), + source: VerificationSource::DigestsFile(path.to_string_lossy().to_string()), hashes, }) } @@ -281,3 +279,105 @@ pub fn verify_hash<'a>(calculated: &Hash, candidates: &'a CandidateHashes) -> Ve messages, } } + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn test_read_raw_inputs() { + let example_path = Path::new("some_file"); + let valid_md5 = "d229da563da18fe5d58cd95a6467d584"; + let valid_sha1 = "b314c7ebb7d599944981908b7f3ed33a30e78f3a"; + let valid_sha1_2 = valid_sha1.to_uppercase(); + let valid_sha256 = "1eb85fc97224598dad1852b5d6483bbcf0aa8608790dcc657a5a2a761ae9c8c6"; + + let invalid1 = "x"; + let invalid2 = "a"; + let invalid3 = "d229da563da18fe5d58cd95a6467d58"; + let invalid4 = "1eb85fc97224598dad1852b5d6483bbcf0aa8608790dcc657a5a2a761ae9c8c67"; + let invalid5 = "1eb85fc97224598dad1852b5d 483bbcf0aa8608790dcc657a5a2a761ae9c8c6"; + + assert!(matches!( + read_raw_candidate_from_file(valid_md5, example_path), + Some(CandidateHashes { + alg: Algorithm::Md5, + .. + }) + )); + assert!(matches!( + read_raw_candidate_from_file(valid_sha1, example_path), + Some(CandidateHashes { + alg: Algorithm::Sha1, + .. + }) + )); + assert!(matches!( + read_raw_candidate_from_file(&valid_sha1_2, example_path), + Some(CandidateHashes { + alg: Algorithm::Sha1, + .. + }) + )); + assert!(matches!( + read_raw_candidate_from_file(valid_sha256, example_path), + Some(CandidateHashes { + alg: Algorithm::Sha256, + .. + }) + )); + + for i in &[invalid1, invalid2, invalid3, invalid4, invalid5] { + assert!(read_raw_candidate_from_file(*i, example_path).is_none()); + } + } + + #[test] + fn test_read_shasums() { + let shasums = "4b91f7a387a6edd4a7c0afb2897f1ca968c9695b *cp + 75eb7420a9f5a260b04a3e8ad51e50f2838a17fc lel.txt + + fe6c26d485a3573a1cb0ad0682f5105325a1905f shasums"; + let lines = shasums.lines().map(std::io::Result::Ok); + let path = Path::new("SHASUMS"); + let candidates = read_coreutils_digests_from_file(lines, path); + + assert_eq!( + candidates, + Some(CandidateHashes { + alg: Algorithm::Sha1, + hashes: vec![ + CandidateHash { + bytes: hex::decode("4b91f7a387a6edd4a7c0afb2897f1ca968c9695b").unwrap(), + filename: Some("cp".to_owned()), + }, + CandidateHash { + bytes: hex::decode("75eb7420a9f5a260b04a3e8ad51e50f2838a17fc").unwrap(), + filename: Some("lel.txt".to_owned()), + }, + CandidateHash { + bytes: hex::decode("fe6c26d485a3573a1cb0ad0682f5105325a1905f").unwrap(), + filename: Some("shasums".to_owned()), + } + ], + source: VerificationSource::DigestsFile(path.to_string_lossy().to_string()), + }) + ); + } + + #[test] + fn test_invalid_shasums() { + let no_format = "4b91f7a387a6edd4a7c0afb2897f1ca968c9695b cp"; + let invalid_format = "4b91f7a387a6edd4a7c0afb2897f1ca968c9695b .cp"; + let extra_space = "4b91f7a387a6edd4a7c0afb2897f1ca968c9695b cp"; + + for digest in [no_format, invalid_format, extra_space] { + let lines = digest.lines().map(std::io::Result::Ok); + assert!( + read_coreutils_digests_from_file(lines, Path::new("SHASUMS")).is_none(), + "Should be invalid digest: {:?}", + digest + ); + } + } +}