initial prototype of the tool to generate copyright notices

Cargo.lock

@@ -1498,6 +1498,15 @@ dependencies = [
  "termcolor",
 ]
 
+[[package]]
+name = "generate-copyright"
+version = "0.1.0"
+dependencies = [
+ "anyhow",
+ "serde",
+ "serde_json",
+]
+
 [[package]]
 name = "generic-array"
 version = "0.14.4"

Cargo.toml

@@ -40,6 +40,7 @@ members = [
   "src/tools/replace-version-placeholder",
   "src/tools/lld-wrapper",
   "src/tools/collect-license-metadata",
+  "src/tools/generate-copyright",
 ]
 
 exclude = [

src/bootstrap/builder.rs

@@ -754,6 +754,7 @@ macro_rules! describe {
                 run::ReplaceVersionPlaceholder,
                 run::Miri,
                 run::CollectLicenseMetadata,
+                run::GenerateCopyright,
             ),
             // These commands either don't use paths, or they're special-cased in Build::build()
             Kind::Clean | Kind::Format | Kind::Setup => vec![],

src/bootstrap/run.rs

@@ -222,3 +222,33 @@ fn run(self, builder: &Builder<'_>) -> Self::Output {
         dest
     }
 }
+
+#[derive(Debug, PartialOrd, Ord, Copy, Clone, Hash, PartialEq, Eq)]
+pub struct GenerateCopyright;
+
+impl Step for GenerateCopyright {
+    type Output = PathBuf;
+    const ONLY_HOSTS: bool = true;
+
+    fn should_run(run: ShouldRun<'_>) -> ShouldRun<'_> {
+        run.path("src/tools/generate-copyright")
+    }
+
+    fn make_run(run: RunConfig<'_>) {
+        run.builder.ensure(GenerateCopyright);
+    }
+
+    fn run(self, builder: &Builder<'_>) -> Self::Output {
+        let license_metadata = builder.ensure(CollectLicenseMetadata);
+
+        // Temporary location, it will be moved to the proper one once it's accurate.
+        let dest = builder.out.join("COPYRIGHT.md");
+
+        let mut cmd = builder.tool_cmd(Tool::GenerateCopyright);
+        cmd.env("LICENSE_METADATA", &license_metadata);
+        cmd.env("DEST", &dest);
+        builder.run(&mut cmd);
+
+        dest
+    }
+}

src/bootstrap/tool.rs

@@ -381,6 +381,7 @@ fn run(self, builder: &Builder<'_>) -> PathBuf {
     BumpStage0, "src/tools/bump-stage0", "bump-stage0";
     ReplaceVersionPlaceholder, "src/tools/replace-version-placeholder", "replace-version-placeholder";
     CollectLicenseMetadata, "src/tools/collect-license-metadata", "collect-license-metadata";
+    GenerateCopyright, "src/tools/generate-copyright", "generate-copyright";
 );
 
 #[derive(Debug, Copy, Clone, Hash, PartialEq, Eq, Ord, PartialOrd)]

src/tools/generate-copyright/Cargo.toml

@@ -0,0 +1,11 @@
+[package]
+name = "generate-copyright"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+anyhow = "1.0.65"
+serde = { version = "1.0.147", features = ["derive"] }
+serde_json = "1.0.85"

src/tools/generate-copyright/src/main.rs

@@ -0,0 +1,94 @@
+use anyhow::Error;
+use std::io::Write;
+use std::path::PathBuf;
+
+fn main() -> Result<(), Error> {
+    let dest = env_path("DEST")?;
+    let license_metadata = env_path("LICENSE_METADATA")?;
+
+    let metadata: Metadata = serde_json::from_slice(&std::fs::read(&license_metadata)?)?;
+
+    let mut buffer = Vec::new();
+    render_recursive(&metadata.files, &mut buffer, 0)?;
+
+    std::fs::write(&dest, &buffer)?;
+
+    Ok(())
+}
+
+fn render_recursive(node: &Node, buffer: &mut Vec<u8>, depth: usize) -> Result<(), Error> {
+    let prefix = std::iter::repeat("> ").take(depth + 1).collect::<String>();
+
+    match node {
+        Node::Root { childs } => {
+            for child in childs {
+                render_recursive(child, buffer, depth)?;
+            }
+        }
+        Node::Directory { name, childs, license } => {
+            render_license(&prefix, std::iter::once(name), license, buffer)?;
+            if !childs.is_empty() {
+                writeln!(buffer, "{prefix}")?;
+                writeln!(buffer, "{prefix}*Exceptions:*")?;
+                for child in childs {
+                    writeln!(buffer, "{prefix}")?;
+                    render_recursive(child, buffer, depth + 1)?;
+                }
+            }
+        }
+        Node::FileGroup { names, license } => {
+            render_license(&prefix, names.iter(), license, buffer)?;
+        }
+        Node::File { name, license } => {
+            render_license(&prefix, std::iter::once(name), license, buffer)?;
+        }
+    }
+
+    Ok(())
+}
+
+fn render_license<'a>(
+    prefix: &str,
+    names: impl Iterator<Item = &'a String>,
+    license: &License,
+    buffer: &mut Vec<u8>,
+) -> Result<(), Error> {
+    for name in names {
+        writeln!(buffer, "{prefix}**`{name}`**  ")?;
+    }
+    writeln!(buffer, "{prefix}License: `{}`  ", license.spdx)?;
+    for (i, copyright) in license.copyright.iter().enumerate() {
+        let suffix = if i == license.copyright.len() - 1 { "" } else { "  " };
+        writeln!(buffer, "{prefix}Copyright: {copyright}{suffix}")?;
+    }
+
+    Ok(())
+}
+
+#[derive(serde::Deserialize)]
+struct Metadata {
+    files: Node,
+}
+
+#[derive(serde::Deserialize)]
+#[serde(rename_all = "kebab-case", tag = "type")]
+pub(crate) enum Node {
+    Root { childs: Vec<Node> },
+    Directory { name: String, childs: Vec<Node>, license: License },
+    File { name: String, license: License },
+    FileGroup { names: Vec<String>, license: License },
+}
+
+#[derive(serde::Deserialize)]
+struct License {
+    spdx: String,
+    copyright: Vec<String>,
+}
+
+fn env_path(var: &str) -> Result<PathBuf, Error> {
+    if let Some(var) = std::env::var_os(var) {
+        Ok(var.into())
+    } else {
+        anyhow::bail!("missing environment variable {var}")
+    }
+}