Merge branch 'main' of github.com:jamesturk/jellyfish

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "jellyfish"
-version = "1.0.2"
+version = "1.0.3"
 edition = "2021"
 description = "Approximate and phonetic matching of strings."
 authors = ["James Turk <dev@jamesturk.net>"]
@@ -15,7 +15,7 @@ name = "jellyfish"
 crate-type = ["cdylib"]
 
 [dependencies]
-pyo3 = "0.18.0"
+pyo3 = "0.20.0"
 unicode-segmentation = "^1.6.0"
 unicode-normalization = "^0.1"
 smallvec = "1.10.0"

python/jellyfish/_jellyfish.py

@@ -340,6 +340,10 @@ def match_rating_codex(s):
 
     # we ignore spaces
     s = s.upper().replace(" ", "")
+    # any remaining non-alphabetic characters are invalid
+    if not s.isalpha():
+        raise ValueError("string must be alphabetic")
+
     codex = []
 
     prev = None

src/match_rating.rs

@@ -2,12 +2,17 @@ use crate::common::FastVec;
 use std::cmp;
 use unicode_segmentation::UnicodeSegmentation;
 
-pub fn match_rating_codex(s: &str) -> String {
+pub fn match_rating_codex(s: &str) -> Result<String, String> {
     // match rating only really makes sense on strings
     let s = &s.to_uppercase()[..];
     let v = UnicodeSegmentation::graphemes(s, true).collect::<FastVec<&str>>();
     let mut codex = String::new();
     let mut prev = "~tmp~";
+    let is_alpha = s.chars().all(|c| c.is_alphabetic() || c == ' ');
+
+    if !is_alpha {
+        return Err(String::from("Strings must only contain alphabetical characters"));
+    }
 
     for (i, c) in v.iter().enumerate() {
         let vowel = *c == "A" || *c == "E" || *c == "I" || *c == "O" || *c == "U";
@@ -22,15 +27,15 @@ pub fn match_rating_codex(s: &str) -> String {
         let mut newcodex = String::new();
         newcodex.push_str(codex.get(..3).unwrap());
         newcodex.push_str(codex.get(codex.len() - 3..).unwrap());
-        return newcodex;
+        return Ok(newcodex);
     }
 
-    codex
+    Ok(codex)
 }
 
 pub fn match_rating_comparison(s1: &str, s2: &str) -> Result<bool, String> {
-    let codex1 = match_rating_codex(s1);
-    let codex2 = match_rating_codex(s2);
+    let codex1 = match_rating_codex(s1)?;
+    let codex2 = match_rating_codex(s2)?;
 
     // need to know which is longer for comparisons later
     let (longer, shorter) = if codex1.len() > codex2.len() {
@@ -97,9 +102,13 @@ pub fn match_rating_comparison(s1: &str, s2: &str) -> Result<bool, String> {
 mod test {
     use super::*;
     use crate::testutils::testutils;
+    pub fn mrc_unwrapped(s: &str) -> String {
+        return match_rating_codex(s).unwrap();
+    }
+
     #[test]
     fn test_match_rating() {
-        testutils::test_str_func("testdata/match_rating_codex.csv", match_rating_codex);
+        testutils::test_str_func("testdata/match_rating_codex.csv", mrc_unwrapped);
     }
 
     #[test]
@@ -117,4 +126,10 @@ mod test {
         let result = match_rating_comparison("Tim", "Timothy");
         assert_eq!(result.is_err(), true);
     }
+
+    #[test]
+    fn test_match_rating_codex_bad_str() {
+        let result = match_rating_codex("i’m going home");
+        assert!(result.is_err());
+    }
 }

src/rustyfish.rs

@@ -9,6 +9,7 @@ use crate::match_rating_comparison as _mr_comparison;
 use crate::metaphone as _metaphone;
 use crate::nysiis as _nysiis;
 use crate::soundex as _soundex;
+use pyo3::exceptions::PyValueError;
 use pyo3::prelude::*;
 
 /// Calculates the Damerau-Levenshtein distance between two strings.
@@ -47,7 +48,8 @@ fn levenshtein_distance(a: &str, b: &str) -> PyResult<usize> {
 // Calculates the Match Rating Approach code for a string.
 #[pyfunction]
 fn match_rating_codex(a: &str) -> PyResult<String> {
-    Ok(_mr_codex(a))
+    // convert to ValueError
+    _mr_codex(a).map_err(|e| PyErr::new::<PyValueError, _>(format!("{}", e)))
 }
 
 // Calculates the Match Rating Approach comparison for two strings.

tests/test_jellyfish.py

@@ -1,5 +1,4 @@
 import csv
-import platform
 import pytest
 
 open_kwargs = {"encoding": "utf8"}
@@ -158,3 +157,9 @@ def test_mr_codex_type(jf):
     assert jf.match_rating_codex("abc") == "ABC"
     with pytest.raises(TypeError) as exc:
         jf.match_rating_codex(b"abc")
+
+
+def test_mr_codex_bad_string(jf):
+    with pytest.raises(ValueError) as exc:
+        res = jf.match_rating_codex("i’m")
+        print(res)