Add support for banned word exceptions

Cargo.lock

@@ -1431,7 +1431,7 @@ checksum = "df94ce210e5bc13cb6651479fa48d14f601d9858cfe0467f43ae157023b938d3"
 
 [[package]]
 name = "payphone"
-version = "4.0.9"
+version = "4.0.11"
 dependencies = [
  "async-trait",
  "axum",

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "payphone"
-version = "4.0.9"
+version = "4.0.11"
 edition = "2024"
 default-run = "payphone"
 

playbooks/08-payphone.yml

@@ -4,8 +4,7 @@
   tasks:
     - name: Bump Version
       ansible.builtin.shell: >
-        cargo release commit --execute --no-confirm && \
+        cargo set-version --bump patch
-        cargo release patch --no-verify --no-publish --execute --no-confirm
       changed_when: false
 
     - name: Get Cargo Project Metadata

release.sh

@@ -0,0 +1 @@
+uv run ansible-playbook -i inventory.ini playbooks/08-payphone.yml --vault-password-file playbooks/00-vault-pwd.txt

src/groups/calls/lists/explicit-sexual-language.txt

@@ -59,7 +59,7 @@ huz
 slut
 dldo
 dildo
-goon
+goon / go on
 gooning
 i will fuh you
 i will fuck you

src/groups/calls/moderation.rs

@@ -247,7 +247,7 @@ fn regex_from_phrase(phrase: &str) -> String {
 ///
 /// Returns:
 /// [`Regex`] - The merged regex for the phrases.
-fn regex_from_phrases_list(phrases: &[&str]) -> Regex {
+fn regex_from_phrases_list(phrases: &[String]) -> Regex {
     let mut pattern = String::new();
 
     for phrase in phrases {
@@ -405,31 +405,43 @@ pub trait Rule: Send + Sync {
 /// * `file_content` - The content of the file to parse.
 ///
 /// Returns:
-/// [`Vec<&str>`] - The list of phrases found in the file.
+/// [`(Vec<String>, Vec<String>)`] - The list of phrases found in the file.
-fn parse_phrases_file(file_content: &str) -> Vec<&str> {
+fn parse_phrases_file(file_content: &str) -> (Vec<String>, Vec<String>) {
-    file_content
+    let mut banned = vec![];
-        .lines()
+    let mut exceptions = vec![];
-        .filter_map(|line| {
+
-            let trimmed = line.trim();
+    for line in file_content.lines() {
+        let trimmed = line.trim().to_ascii_lowercase();
+
         if !trimmed.is_empty() && !trimmed.starts_with("//") {
-                Some(trimmed)
+            if let Some((phrase, phrase_exceptions)) = trimmed.split_once("/") {
+                banned.push(phrase.trim().to_string());
+                exceptions.extend(phrase_exceptions.split(',').map(|s| s.trim().to_string()));
             } else {
-                None
+                banned.push(trimmed.to_string());
+            }
         }
-        })
-        .collect()
     }
 
-struct HateSpeechAndSlursRule(Regex);
+    (banned, exceptions)
+}
+
+struct HateSpeechAndSlursRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl HateSpeechAndSlursRule {
     fn new() -> Self {
-        let phrases: Vec<&str> =
+        let (phrases, exceptions) =
             parse_phrases_file(include_str!("lists/hate-speech-and-slurs.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -445,24 +457,43 @@ impl Rule for HateSpeechAndSlursRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
 }
 
-struct ThreatsAndSelfHarmRule(Regex);
+struct ThreatsAndSelfHarmRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl ThreatsAndSelfHarmRule {
     fn new() -> Self {
-        let phrases: Vec<&str> =
+        let (phrases, exceptions) =
             parse_phrases_file(include_str!("lists/threats-and-self-harm.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -478,24 +509,43 @@ impl Rule for ThreatsAndSelfHarmRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
 }
 
-struct ExplicitSexualLanguageRule(Regex);
+struct ExplicitSexualLanguageRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl ExplicitSexualLanguageRule {
     fn new() -> Self {
-        let phrases: Vec<&str> =
+        let (phrases, exceptions) =
             parse_phrases_file(include_str!("lists/explicit-sexual-language.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -511,23 +561,43 @@ impl Rule for ExplicitSexualLanguageRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
 }
 
-struct SuggestiveLanguageRule(Regex);
+struct SuggestiveLanguageRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl SuggestiveLanguageRule {
     fn new() -> Self {
-        let phrases: Vec<&str> = parse_phrases_file(include_str!("lists/suggestive-language.txt"));
+        let (phrases, exceptions) =
+            parse_phrases_file(include_str!("lists/suggestive-language.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -543,9 +613,22 @@ impl Rule for SuggestiveLanguageRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
@@ -580,17 +663,23 @@ impl Rule for AdvertisingRule {
     }
 }
 
-struct SensitiveTopicsAndTriggeringContentRule(Regex);
+struct SensitiveTopicsAndTriggeringContentRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl SensitiveTopicsAndTriggeringContentRule {
     fn new() -> Self {
-        let phrases: Vec<&str> = parse_phrases_file(include_str!(
+        let (phrases, exceptions) = parse_phrases_file(include_str!(
             "lists/sensitive-topics-and-triggering-content.txt"
         ));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -606,23 +695,42 @@ impl Rule for SensitiveTopicsAndTriggeringContentRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
 }
 
-pub struct MildSwearWordsRule(Regex);
+pub struct MildSwearWordsRule {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl MildSwearWordsRule {
     fn new() -> Self {
-        let phrases: Vec<&str> = parse_phrases_file(include_str!("lists/mild-swear-words.txt"));
+        let (phrases, exceptions) = parse_phrases_file(include_str!("lists/mild-swear-words.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -638,24 +746,43 @@ impl Rule for MildSwearWordsRule {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }
 }
 
-pub struct CasualInsultsAndTeasing(Regex);
+pub struct CasualInsultsAndTeasing {
+    pattern: Regex,
+    exceptions: Vec<String>,
+}
 
 impl CasualInsultsAndTeasing {
     fn new() -> Self {
-        let phrases: Vec<&str> =
+        let (phrases, exceptions) =
             parse_phrases_file(include_str!("lists/casual-insults-and-teasing.txt"));
 
         let regex = regex_from_phrases_list(&phrases);
 
-        Self(regex)
+        Self {
+            pattern: regex,
+            exceptions,
+        }
     }
 }
 
@@ -671,9 +798,22 @@ impl Rule for CasualInsultsAndTeasing {
     fn detect(&self, _original: &str, normalized: &str) -> (Vec<String>, Vec<String>) {
         (
             vec![],
-            self.0
+            self.pattern
                 .find_iter(normalized)
-                .map(|m| normalized[m.start()..m.end()].to_string())
+                .filter_map(|m| {
+                    let string = normalized[m.start()..m.end()].to_string();
+
+                    if self.exceptions.contains(
+                        &string
+                            .split_ascii_whitespace()
+                            .collect::<String>()
+                            .to_ascii_lowercase(),
+                    ) {
+                        None
+                    } else {
+                        Some(string)
+                    }
+                })
                 .collect(),
         )
     }