Add header_type and custom_type to extract config

Extract configuration is now emitted in the output config, so
tooling can load and perform their own tasks on extracted assets
without having to parse YAML.

`header_type`:
- `symbol` (default): Emit a full symbol declaration.
- `raw`: Emit raw array data (for wrapping in your own declaration)
- `none`: Don't emit a header at all. (For custom processing)

`custom_type`/`custom_data`: Passed through to the output config
as-is for custom tasks.

Author: encounter

Cargo.lock

@@ -3,7 +3,7 @@ name = "decomp-toolkit"
 description = "Yet another GameCube/Wii decompilation toolkit."
 authors = ["Luke Street <luke@street.dev>"]
 license = "MIT OR Apache-2.0"
-version = "1.1.4"
+version = "1.2.0"
 edition = "2021"
 publish = false
 repository = "https://github.com/encounter/decomp-toolkit"

Cargo.toml

@@ -5,6 +5,7 @@ use std::{
     fs::DirBuilder,
     io::{Cursor, Seek, Write},
     mem::take,
+    str::FromStr,
     time::Instant,
 };
 
@@ -36,7 +37,7 @@ use crate::{
     },
     util::{
         asm::write_asm,
-        bin2c::bin2c,
+        bin2c::{bin2c, HeaderKind},
         comment::MWComment,
         config::{
             apply_splits_file, apply_symbols_file, is_auto_symbol, signed_hex_serde,
@@ -303,6 +304,20 @@ pub struct ExtractConfig {
     /// Path is relative to `out_dir/include`.
     #[serde(with = "unix_path_serde_option", default, skip_serializing_if = "Option::is_none")]
     pub header: Option<Utf8UnixPathBuf>,
+    /// The type for the extracted symbol in the header file. By default, the header will emit
+    /// a full symbol declaration (a.k.a. `symbol`), but this can be set to `raw` to emit the raw
+    /// data as a byte array. `none` avoids emitting a header entirely, in which case the `header`
+    /// field can be used by external asset processing.
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub header_type: Option<String>,
+    /// A user-defined type for use with external asset processing. This value is simply passed
+    /// through to the `custom_type` field in the output config.
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub custom_type: Option<String>,
+    /// User-defined data for use with external asset processing. This value is simply passed
+    /// through to the `custom_data` field in the output config.
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub custom_data: Option<serde_json::Value>,
 }
 
 /// A relocation that should be blocked.
@@ -364,6 +379,19 @@ pub struct OutputModule {
     pub ldscript: Utf8UnixPathBuf,
     pub entry: Option<String>,
     pub units: Vec<OutputUnit>,
+    pub extract: Vec<OutputExtract>,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, Default)]
+pub struct OutputExtract {
+    pub symbol: String,
+    #[serde(with = "unix_path_serde_option")]
+    pub binary: Option<Utf8UnixPathBuf>,
+    #[serde(with = "unix_path_serde_option")]
+    pub header: Option<Utf8UnixPathBuf>,
+    pub header_type: String,
+    pub custom_type: Option<String>,
+    pub custom_data: Option<serde_json::Value>,
 }
 
 #[derive(Serialize, Deserialize, Debug, Clone, Default, PartialEq, Eq, Hash)]
@@ -938,6 +966,7 @@ fn split_write_obj(
         ldscript: out_dir.join("ldscript.lcf").with_unix_encoding(),
         units: Vec::with_capacity(split_objs.len()),
         entry,
+        extract: Vec::with_capacity(module.config.extract.len()),
     };
     for (unit, split_obj) in module.obj.link_order.iter().zip(&split_objs) {
         let out_obj = write_elf(split_obj, config.export_all)?;
@@ -975,14 +1004,34 @@ fn split_write_obj(
             write_if_changed(&out_path, data)?;
         }
 
-        if let Some(header) = &extract.header {
-            let header_string = bin2c(symbol, section, data);
-            let out_path = base_dir.join("include").join(header.with_encoding());
-            if let Some(parent) = out_path.parent() {
-                DirBuilder::new().recursive(true).create(parent)?;
+        let header_kind = match extract.header_type.as_deref() {
+            Some(value) => match HeaderKind::from_str(value) {
+                Ok(kind) => kind,
+                Err(()) => bail!("Invalid header type '{}'", value),
+            },
+            _ => HeaderKind::Symbol,
+        };
+
+        if header_kind != HeaderKind::None {
+            if let Some(header) = &extract.header {
+                let header_string = bin2c(symbol, section, data, header_kind);
+                let out_path = base_dir.join("include").join(header.with_encoding());
+                if let Some(parent) = out_path.parent() {
+                    DirBuilder::new().recursive(true).create(parent)?;
+                }
+                write_if_changed(&out_path, header_string.as_bytes())?;
             }
-            write_if_changed(&out_path, header_string.as_bytes())?;
         }
+
+        // Copy to output config
+        out_config.extract.push(OutputExtract {
+            symbol: symbol.name.clone(),
+            binary: extract.binary.clone(),
+            header: extract.header.clone(),
+            header_type: header_kind.to_string(),
+            custom_type: extract.custom_type.clone(),
+            custom_data: extract.custom_data.clone(),
+        });
     }
 
     // Generate ldscript.lcf

src/cmd/dol.rs

@@ -1,3 +1,5 @@
+use std::{fmt, str::FromStr};
+
 use crate::obj::{ObjSection, ObjSectionKind, ObjSymbol};
 
 const PROLOGUE: &str = r#"
@@ -13,8 +15,50 @@ const PROLOGUE: &str = r#"
 
 "#;
 
+/// The output header type.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub enum HeaderKind {
+    /// Do not generate a header. (Used for custom processing)
+    None,
+    /// A full symbol definition.
+    Symbol,
+    /// Raw array data.
+    Raw,
+}
+
+impl FromStr for HeaderKind {
+    type Err = ();
+
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        match s {
+            "none" => Ok(Self::None),
+            "symbol" => Ok(Self::Symbol),
+            "raw" => Ok(Self::Raw),
+            _ => Err(()),
+        }
+    }
+}
+
+impl fmt::Display for HeaderKind {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::None => write!(f, "none"),
+            Self::Symbol => write!(f, "symbol"),
+            Self::Raw => write!(f, "raw"),
+        }
+    }
+}
+
 /// Converts a binary blob into a C array.
-pub fn bin2c(symbol: &ObjSymbol, section: &ObjSection, data: &[u8]) -> String {
+pub fn bin2c(symbol: &ObjSymbol, section: &ObjSection, data: &[u8], kind: HeaderKind) -> String {
+    match kind {
+        HeaderKind::None => String::new(),
+        HeaderKind::Symbol => bin2c_symbol(symbol, section, data),
+        HeaderKind::Raw => bin2c_raw(data),
+    }
+}
+
+fn bin2c_symbol(symbol: &ObjSymbol, section: &ObjSection, data: &[u8]) -> String {
     let mut output = String::new();
     output.push_str(PROLOGUE);
     output.push_str(&format!(
@@ -41,3 +85,19 @@ pub fn bin2c(symbol: &ObjSymbol, section: &ObjSection, data: &[u8]) -> String {
     output.push_str("\n};\n");
     output
 }
+
+fn bin2c_raw(data: &[u8]) -> String {
+    let mut output = String::new();
+    for (i, byte) in data.iter().enumerate() {
+        if i > 0 {
+            if i % 16 == 0 {
+                output.push('\n');
+            } else {
+                output.push(' ');
+            }
+        }
+        output.push_str(&format!("0x{:02X},", byte));
+    }
+    output.push('\n');
+    output
+}