canonicalize UE4 name into legal rust token

src/lib.rs

@@ -4,7 +4,8 @@ pub mod types;
 
 pub mod sdk {
     use std::{
-        collections::{hash_map::Entry, HashMap},
+        borrow::Cow,
+        collections::{hash_map::Entry, BTreeSet, HashMap, HashSet},
         sync::Mutex,
     };
 
@@ -155,7 +156,9 @@ pub mod sdk {
                     UAnyType::UField(_) => {}
                     UAnyType::UScriptStruct(_) => {}
                     UAnyType::UProperty(_) => {}
-                    UAnyType::UEnum(obj) => Self::process_enum(obj),
+                    UAnyType::UEnum(obj) => {
+                        let enm = Self::process_enum(obj)?;
+                    }
                     UAnyType::UStruct(_) => {}
                     UAnyType::UFunction(_) => {}
                 }
@@ -164,15 +167,136 @@ pub mod sdk {
             Ok(())
         }
 
-        fn process_enum(enm: UEnum) {
-            let names = enm
+        fn process_enum(enm: UEnum) -> anyhow::Result<Enum> {
+            let values = enm
                 .get_names()
                 .iter()
-                .map(|name| name.get_name().unwrap_or("<invalid-name>".to_string()))
+                .map(|name| name.get_name().unwrap_or("AnonymousVariant".to_string()))
+                .map(|name| canonicalize_name(&name).to_string())
                 .collect::<Vec<_>>();
 
-            log::info!("enum: {}", enm.as_uobject().get_full_name_or_default());
-            log::info!("{names:#?}");
+            let name = enm
+                .as_uobject()
+                .get_full_name()
+                .context("enum name could not be found")?;
+            let name = canonicalize_name(&name).to_string();
+
+            Ok(Enum { name, values })
         }
     }
+
+    fn keywords() -> HashSet<&'static str> {
+        let mut keywords = HashSet::new();
+
+        // rust keywords
+        keywords.extend([
+            "as", "break", "const", "continue", "crate", "else", "enum", "extern", "false", "fn",
+            "for", "if", "impl", "in", "let", "loop", "match", "mod", "move", "mut", "pub", "ref",
+            "return", "self", "Self", "static", "struct", "super", "trait", "true", "type",
+            "unsafe", "use", "where", "while", "async", "await", "dyn", "abstract", "become",
+            "box", "do", "final", "macro", "override", "priv", "typeof", "unsized", "virtual",
+            "yield", "try",
+        ]);
+
+        // rust types
+        keywords.extend([
+            "bool", "f64", "f32", "str", "char", "u8", "u16", "u32", "u64", "u128", "i8", "i16",
+            "i32", "i64", "i128", "usize", "isize",
+        ]);
+
+        keywords
+    }
+
+    fn token_chars() -> HashSet<char> {
+        let mut chars = HashSet::new();
+
+        chars.extend([
+            ' ', '?', '+', '-', ':', '/', '^', '(', ')', '[', ']', '<', '>', '&', '.', '#', '\'',
+            '"', '%',
+        ]);
+
+        chars
+    }
+
+    struct SplitResult<'a> {
+        start: Option<&'a str>,
+        middle: usize,
+        end: Option<&'a str>,
+    }
+
+    impl<'a> SplitResult<'a> {
+        pub fn is_valid(&self) -> bool {
+            self.start.is_some() && self.middle == 0 && self.end.is_none()
+        }
+
+        pub fn into_valid(self) -> Cow<'a, str> {
+            if self.is_valid() {
+                Cow::Borrowed(self.start.unwrap())
+            } else {
+                let mut valid = self.start.map(|s| s.to_string()).unwrap_or_default();
+                valid.extend(core::iter::repeat('_').take(self.middle));
+
+                match self.end {
+                    Some(end) => {
+                        valid.push_str(&split_at_illegal_char(end).into_valid());
+                    }
+                    None => {}
+                }
+
+                Cow::Owned(valid)
+            }
+        }
+    }
+
+    fn empty_or_some(s: &str) -> Option<&str> {
+        if s.is_empty() {
+            None
+        } else {
+            Some(s)
+        }
+    }
+
+    fn split_at_illegal_char(input: &str) -> SplitResult {
+        let illegal_chars = token_chars();
+        if let Some(pos) = input.find(|c| illegal_chars.contains(&c)) {
+            let start = empty_or_some(&input[..pos]);
+            // skip the illegal char
+            let rest = &input[pos + 1..];
+
+            if let Some(pos2) = rest.find(|c| !illegal_chars.contains(&c)) {
+                SplitResult {
+                    start,
+                    middle: pos2 + 1,
+                    end: empty_or_some(&rest[pos2..]),
+                }
+            } else {
+                SplitResult {
+                    start,
+                    middle: 1,
+                    end: empty_or_some(rest),
+                }
+            }
+        } else {
+            SplitResult {
+                start: empty_or_some(input),
+                middle: 0,
+                end: None,
+            }
+        }
+    }
+
+    fn canonicalize_name(name: &str) -> Cow<str> {
+        let valid = split_at_illegal_char(name).into_valid();
+        if keywords().contains(valid.as_ref()) {
+            Cow::Owned(format!("_{}", &valid))
+        } else {
+            valid
+        }
+    }
+
+    #[derive(Debug)]
+    struct Enum {
+        name: String,
+        values: Vec<String>,
+    }
 }