Improve wordlists, replacements, and test-cases (#16)

* Integrate CG wordlist.
finnbear · Oct 20, 2023 · 36d8cf5 · 36d8cf5
1 parent e593c7f
commit 36d8cf5
Show file tree

Hide file tree

Showing 12 changed files with 7,647 additions and 124 deletions.
diff --git a/Cargo.toml b/Cargo.toml
@@ -1,7 +1,7 @@
 [package]
 name = "rustrict"
 authors = ["Finn Bear"]
-version = "0.7.10"
+version = "0.7.11"
 edition = "2021"
 license = "MIT OR Apache-2.0"
 repository = "https://github.com/finnbear/rustrict/"
@@ -71,7 +71,7 @@ serde = {version = "1", features=["derive"], optional = true}
 rand = "0.8"
 csv = "1.1"
 censor_crate = {package = "censor", version = "0.3.0"}
-rustrict_old = {package = "rustrict", version = "0.7.5"}
+rustrict_old = {package = "rustrict", version = "0.7.10"}
 serial_test = "0.5"
 bincode = "1.3.3"
 serde_json = "1"
diff --git a/README.md b/README.md
@@ -172,7 +172,7 @@ is used as a dataset. Positive accuracy is the percentage of profanity detected
 
 | Crate | Accuracy | Positive Accuracy | Negative Accuracy | Time |
 |-------|----------|-------------------|-------------------|------|
-| [rustrict](https://crates.io/crates/rustrict) | 89.37%   | 92.79%            | 88.51%            | 8s   |
+| [rustrict](https://crates.io/crates/rustrict) | 88.32%   | 93.25%            | 87.09%            | 8s   |
 | [censor](https://crates.io/crates/censor) | 76.16%   | 72.76%            | 77.01%            | 23s  |
 
 ## Development

diff --git a/src/censor.rs b/src/censor.rs
@@ -1028,8 +1028,9 @@ mod tests {
     #[test]
     #[serial]
     fn censor() {
-        let censored = Censor::from_str("HELLO fučk Shit nigga WORLD!")
+        let censored = Censor::from_str("HELLO fučk Shit nudes WORLD!")
             .with_censor_replacement('#')
+            .with_censor_first_character_threshold(Type::SEXUAL & Type::SEVERE)
             .censor();
 
         assert_eq!(censored, "HELLO f### S### ##### WORLD!");
@@ -1139,7 +1140,7 @@ mod tests {
             "https://crates.io/crates/rustrict",
             rustrict,
             false, // true,
-            None,  // Some(rustrict_old),
+            None // Some(rustrict_old),
         );
         print_accuracy("https://crates.io/crates/censor", censor, false, None);
     }