Incremental improvement to word/character data.

Author: finnbear

Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "rustrict"
 authors = ["Finn Bear"]
-version = "0.7.35"
+version = "0.7.36"
 edition = "2021"
 license = "MIT OR Apache-2.0"
 repository = "https://github.com/finnbear/rustrict/"

README.md

@@ -177,7 +177,7 @@ 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) | 80.00%   | 94.01%            | 76.50%            | 9s   |
+| [rustrict](https://crates.io/crates/rustrict) | 79.99% | 94.02% | 76.49% | 10s |
 | [censor](https://crates.io/crates/censor) | 76.16%   | 72.76%            | 77.01%            | 23s  |
 | [stfu](https://crates.io/crates/stfu) | 91.74% | 77.69% | 95.25% | 45s |
 | [profane-rs](https://crates.io/crates/profane-rs) | 80.47% | 73.79% | 82.14% | 52s |

src/character_analyzer.rs

@@ -50,6 +50,8 @@ fn main() {
                 '𒐫' => 80,
                 '𒈙' => 35,
                 '༺' | '༻' => 25,
+                '𒀱' => 45,
+                '۞' => 19,
                 _ => {
                     let max_width = (max_width(c, &fonts) as f32 / 100f32).round() as u16;
                     if max_width > u8::MAX as u16 {

src/character_widths.bin

@@ -177,6 +177,7 @@ cancer
 castrates
 castrations
 ch
+child touche
 china flu
 china plague
 china virus

src/dictionary_blacklist.txt

@@ -51,6 +51,7 @@ betcha
 bigger s
 bishi
 blogger
+bob's
 bonjor
 of elchasai
 braig
@@ -125,6 +126,8 @@ hi tirp
 hiu
 hi u
 hi v
+hochst
+hoechst
 ho ho ho
 honkeytonk
 honkey tonk
@@ -228,6 +231,7 @@ shoehorn your
 shouldn't it
 since 1
 since l
+sombra
 speed*
 sperm whale
 spick and span
@@ -240,6 +244,7 @@ tally ho
 tally-ho
 tea the
 test test test
+than a lot
 then i guess
 then talk
 then, talk

src/dictionary_extra.txt

@@ -537,6 +537,7 @@ amphitrichate
 ample assured
 ample peer
 ampullar
+an a lot
 an ai
 an al
 an album
@@ -2735,6 +2736,7 @@ boards lut
 boards perm
 boards seeks
 boastful
+bob's
 bobbish
 bobsled
 bobsleigh
@@ -3494,6 +3496,17 @@ buzz hit
 buzz lut
 buzz off
 buzz perm
+buzz syd
+buzz sym
+buzz syn
+buzz syr
+buzz sys
+buzz ya
+buzz ye
+buzz yi
+buzz yo
+buzz yu
+buzzy
 bytes cumulative
 bytes ext
 bytes hilt
@@ -3816,6 +3829,7 @@ child dicke
 child licking
 child ongoing
 child stew
+child touched
 childrens cumulative
 childrens ext
 childrens hilt
@@ -7945,6 +7959,7 @@ hobbies seeks
 hoc
 hod
 hoecake
+hoechst
 hoed
 hoeful
 hoeing
@@ -11726,6 +11741,7 @@ muscles perm
 muscles seeks
 musklike
 muskroot
+muslim kill
 musselcracker
 mussuck
 mustard
@@ -16807,6 +16823,7 @@ solo ser
 solo vary
 solo zer
 somalia bomb
+sombra
 something hote
 something hotmail
 something hottest
@@ -19547,6 +19564,7 @@ weenier
 weeniest
 weet back
 weet dream
+weet girl
 weeweed
 weeweeing
 weightier

src/false_positives.txt

@@ -91,7 +91,7 @@ pub fn is_whitespace(c: char) -> bool {
         || c.is_format()
         || matches!(
             c,
-            '\u{115F}' | '\u{1160}' | '\u{2800}' | '\u{3164}' | '\u{FFA0}' | '\u{FFFC}'
+            '\u{115F}' | '\u{1160}' | '\u{20DD}' | '\u{2800}' | '\u{3164}' | '\u{FFA0}' | '\u{FFFC}'
         )
 }
 

src/lib.rs

@@ -8,7 +8,7 @@ lazy_static! {
     static ref EMAIL_ADDRESS : Regex = Regex::new(r#"(?i)[a-z0-9_\-]{3,}\s*(@|[\[\(\s]at[\s\)\]])\s*[a-z0-9_\-]{5,}\s*(\.|dot)\s*(com|net|org|gov|biz|co|us|ru|uk|de|se|to|tv|io|info|online|site)"#).unwrap();
     //static ref ADDRESS : Regex = Regex::new(r#"(?i)\d+[ ](?:[A-Za-z0-9\.-]+ )+(?:Avenue|Lane|Road|Boulevard|Drive|Street|Ave|Dr|Rd|Blvd|Ln|St)\.?(\s+#[0-9]{1,5})?"#).unwrap();
     static ref NAME : Regex = Regex::new(r#"(?i)(real\s)?name\s+is:?\s[a-zA-Z]+(\s[a-zA-z]+)?"#).unwrap();
-    static ref URL : Regex = Regex::new(r#"(?i)(https?:?/*)?[a-zA-Z0-9]{3,}\.(com|net|org|gov|biz|co|us|ru|uk|de|se|to|tv|io|info|online|site|link)"#).unwrap();
+    static ref URL : Regex = Regex::new(r#"(?i)(https?:?/*)?[a-zA-Z0-9]{3,}\.(com|net|org|gov|biz|co|us|ru|cc|uk|de|se|to|tv|io|gg|info|online|site|link)"#).unwrap();
 }
 
 /// Returns [`s`] with personally-identifiable information censored out, and a `true` if