Allow modifying macro tokens at parse time

bindgen-integration/build.rs

@@ -1,7 +1,7 @@
 extern crate bindgen;
 
 use bindgen::callbacks::{
-    DeriveInfo, IntKind, MacroParsingBehavior, ParseCallbacks,
+    DeriveInfo, IntKind, MacroParsingBehavior, ParseCallbacks, Token, TokenKind,
 };
 use bindgen::{Builder, EnumVariation, Formatter};
 use std::collections::HashSet;
@@ -145,6 +145,48 @@ impl ParseCallbacks for MacroCallback {
             vec![]
         }
     }
+
+    fn modify_macro(&self, _name: &str, tokens: &mut Vec<Token>) {
+        // Handle macros dealing with bit positions of the format HI:LO
+        if tokens.len() == 4 && tokens[2].kind == TokenKind::Punctuation {
+            if let Ok(colon) = std::str::from_utf8(&tokens[2].raw) {
+                if colon != ":" {
+                    return;
+                }
+                let high = match std::str::from_utf8(&tokens[1].raw) {
+                    Ok(s) => {
+                        if let Ok(val) = s.parse::<u16>() {
+                            val
+                        } else {
+                            return;
+                        }
+                    }
+                    Err(_) => {
+                        return;
+                    }
+                };
+
+                let low = match std::str::from_utf8(&tokens[3].raw) {
+                    Ok(s) => {
+                        if let Ok(val) = s.parse::<u16>() {
+                            val
+                        } else {
+                            return;
+                        }
+                    }
+                    Err(_) => {
+                        return;
+                    }
+                };
+                let value: u32 = ((high as u32) << 16) | low as u32;
+                tokens[1] = Token::from((
+                    TokenKind::Literal,
+                    value.to_string().as_bytes(),
+                ));
+                tokens.truncate(2);
+            }
+        }
+    }
 }
 
 impl Drop for MacroCallback {

bindgen-integration/cpp/Test.h

@@ -22,6 +22,8 @@
 #define TESTMACRO_STRING_EXPR ("string")
 #define TESTMACRO_STRING_FUNC_NON_UTF8(x) (x "ÿÿ") /* invalid UTF-8 on purpose */
 
+#define TESTMACRO_COLON_VALUE 1:2
+
 enum {
   MY_ANNOYING_MACRO =
 #define MY_ANNOYING_MACRO 1

bindgen-integration/src/lib.rs

@@ -345,3 +345,9 @@ fn test_wrap_static_fns() {
         extern_bindings::wrap_as_variadic_fn2_wrapped(1, 2);
     }
 }
+
+#[test]
+fn test_colon_define() {
+    let gold: u32 = (1u32 << 16) | 2;
+    assert_eq!(gold, bindings::TESTMACRO_COLON_VALUE);
+}

bindgen/callbacks.rs

@@ -4,6 +4,8 @@ pub use crate::ir::analysis::DeriveTrait;
 pub use crate::ir::derive::CanDerive as ImplementsTrait;
 pub use crate::ir::enum_ty::{EnumVariantCustomBehavior, EnumVariantValue};
 pub use crate::ir::int::IntKind;
+pub use cexpr::token::Kind as TokenKind;
+pub use cexpr::token::Token;
 use std::fmt;
 
 /// An enum to allow ignoring parsing of macros.
@@ -49,6 +51,9 @@ pub trait ParseCallbacks: fmt::Debug {
         None
     }
 
+    /// Modify the contents of a macro
+    fn modify_macro(&self, _name: &str, _tokens: &mut Vec<Token>) {}
+
     /// The integer kind an integer macro should have, given a name and the
     /// value of that macro, or `None` if you want the default to be chosen.
     fn int_macro(&self, _name: &str, _value: i64) -> Option<IntKind> {

bindgen/ir/var.rs

@@ -427,7 +427,11 @@ fn parse_macro(
 ) -> Option<(Vec<u8>, cexpr::expr::EvalResult)> {
     use cexpr::expr;
 
-    let cexpr_tokens = cursor.cexpr_tokens();
+    let mut cexpr_tokens = cursor.cexpr_tokens();
+
+    for callbacks in &ctx.options().parse_callbacks {
+        callbacks.modify_macro(&cursor.spelling(), &mut cexpr_tokens);
+    }
 
     let parser = expr::IdentifierParser::new(ctx.parsed_macros());