]> code.octet-stream.net Git - hashgood/blobdiff - src/verify.rs
Upgrade to maintained hashing libs
[hashgood] / src / verify.rs
index 6ad9ea9464a97870babac51567514250f8c15074..0f2d1d88325de4e19b90da50479bd75b6fd81b52 100644 (file)
@@ -2,14 +2,13 @@ use super::{
     Algorithm, CandidateHash, CandidateHashes, Hash, MatchLevel, MessageLevel, Opt, Verification,
     VerificationSource,
 };
-use clipboard::ClipboardContext;
-use clipboard::ClipboardProvider;
-use regex::Regex;
+#[cfg(feature = "paste")]
+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.
@@ -17,7 +16,7 @@ use std::path::PathBuf;
 pub fn get_candidate_hashes(opt: &Opt) -> Result<Option<CandidateHashes>, String> {
     if let Some(hash_string) = &opt.hash {
         return Ok(Some(get_by_parameter(hash_string)?));
-    } else if opt.paste {
+    } else if opt.get_paste() {
         return Ok(Some(get_from_clipboard()?));
     } else if let Some(hash_file) = &opt.hash_file {
         return Ok(Some(get_from_file(hash_file)?));
@@ -43,32 +42,39 @@ fn get_by_parameter(param: &str) -> Result<CandidateHashes, String> {
 
 /// Generate a candidate hash from the system clipboard, or throw an error.
 fn get_from_clipboard() -> Result<CandidateHashes, String> {
-    let mut ctx: ClipboardContext = match ClipboardProvider::new() {
-        Ok(ctx) => ctx,
-        Err(e) => return Err(format!("Error getting system clipboard: {}", e)),
-    };
+    #[cfg(feature = "paste")]
+    {
+        let mut ctx: ClipboardContext = match ClipboardContext::new() {
+            Ok(ctx) => ctx,
+            Err(e) => return Err(format!("Error getting system clipboard: {}", e)),
+        };
 
-    let possible_hash = match ctx.get_contents() {
-        Ok(value) => value,
-        Err(e) => format!("Error reading from clipboard: {}", e),
-    };
+        let possible_hash = match ctx.get_contents() {
+            Ok(value) => value,
+            Err(e) => format!("Error reading from clipboard: {}", e),
+        };
 
-    let bytes = hex::decode(&possible_hash)
-        .map_err(|_| "Clipboard contains invalid or truncated hex".to_owned())?;
-    let alg = Algorithm::from_len(bytes.len())?;
-    let candidate = CandidateHash {
-        filename: None,
-        bytes,
-    };
-    Ok(CandidateHashes {
-        alg,
-        hashes: vec![candidate],
-        source: VerificationSource::Clipboard,
-    })
+        let bytes = hex::decode(&possible_hash)
+            .map_err(|_| "Clipboard contains invalid or truncated hex".to_owned())?;
+        let alg = Algorithm::from_len(bytes.len())?;
+        let candidate = CandidateHash {
+            filename: None,
+            bytes,
+        };
+        Ok(CandidateHashes {
+            alg,
+            hashes: vec![candidate],
+            source: VerificationSource::Clipboard,
+        })
+    }
+    #[cfg(not(feature = "paste"))]
+    {
+        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<CandidateHashes, String> {
+fn get_from_file(path: &Path) -> Result<CandidateHashes, String> {
     // Get a reader for either standard input or the chosen path
     let reader: Box<dyn Read> = if path.to_str() == Some("-") {
         Box::new(std::io::stdin())
@@ -90,7 +96,7 @@ fn get_from_file(path: &PathBuf) -> Result<CandidateHashes, String> {
     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);
     }
 
@@ -99,7 +105,7 @@ fn get_from_file(path: &PathBuf) -> Result<CandidateHashes, String> {
     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);
     }
 
@@ -111,81 +117,80 @@ fn get_from_file(path: &PathBuf) -> Result<CandidateHashes, String> {
     ))
 }
 
-fn read_raw_candidate_from_file(line: &str, path: &PathBuf) -> Option<CandidateHashes> {
-    // 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<u8>)> {
+    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<I>(lines: I, path: &PathBuf) -> Option<CandidateHashes>
+fn read_raw_candidate_from_file(line: &str, path: &Path) -> Option<CandidateHashes> {
+    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<I, S>(lines: I, path: &Path) -> Option<CandidateHashes>
 where
-    I: Iterator<Item = io::Result<String>>,
+    I: Iterator<Item = io::Result<S>>,
+    S: AsRef<str>,
 {
-    let re = Regex::new(
-        r"^(?P<hash>([[:xdigit:]]{32}|[[:xdigit:]]{40}|[[:xdigit:]]{64})) .(?P<filename>.+)$",
-    )
-    .unwrap();
-
     let mut hashes = vec![];
     let mut alg: Option<Algorithm> = 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
+        // <valid-hash><space><space-or-*><filename>
+        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
@@ -200,7 +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,
     })
 }
@@ -274,3 +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
+            );
+        }
+    }
+}