migrate grammar fetching/building code into helix-loader crate

This is a rather large refactor that moves most of the code for
loading, fetching, and building grammars into a new helix-loader
module. This works well with the [[grammars]] syntax for
languages.toml defined earlier: we only have to depend on the types
for GrammarConfiguration in helix-loader and can leave all the
[[language]] entries for helix-core.

4 files changed, 578 insertions, 0 deletions

diff --git a/helix-loader/Cargo.toml b/helix-loader/Cargo.toml
new file mode 100644
index 00000000..21b37333
--- /dev/null
+++ b/helix-loader/Cargo.toml
@@ -0,0 +1,23 @@
+[package]
+name = "helix-loader"
+version = "0.6.0"
+description = "A post-modern text editor."
+authors = ["Blaž Hrastnik <blaz@mxxn.io>"]
+edition = "2021"
+license = "MPL-2.0"
+categories = ["editor"]
+repository = "https://github.com/helix-editor/helix"
+homepage = "https://helix-editor.com"
+
+[dependencies]
+anyhow = "1"
+serde = { version = "1.0", features = ["derive"] }
+toml = "0.5"
+etcetera = "0.3"
+tree-sitter = "0.20"
+libloading = "0.7"
+once_cell = "1.9"
+
+# cloning/compiling tree-sitter grammars
+cc = { version = "1" }
+threadpool = { version = "1.0" }
diff --git a/helix-loader/build.rs b/helix-loader/build.rs
new file mode 100644
index 00000000..e0ebd1c4
--- /dev/null
+++ b/helix-loader/build.rs
@@ -0,0 +1,6 @@
+fn main() {
+    println!(
+        "cargo:rustc-env=BUILD_TARGET={}",
+        std::env::var("TARGET").unwrap()
+    );
+}
diff --git a/helix-loader/src/grammar.rs b/helix-loader/src/grammar.rs
new file mode 100644
index 00000000..61ef464f
--- /dev/null
+++ b/helix-loader/src/grammar.rs
@@ -0,0 +1,388 @@
+use anyhow::{anyhow, Context, Result};
+use libloading::{Library, Symbol};
+use serde::{Deserialize, Serialize};
+use std::fs;
+use std::time::SystemTime;
+use std::{
+    collections::HashSet,
+    path::{Path, PathBuf},
+    process::Command,
+    sync::mpsc::channel,
+};
+use tree_sitter::Language;
+
+#[cfg(unix)]
+const DYLIB_EXTENSION: &str = "so";
+
+#[cfg(windows)]
+const DYLIB_EXTENSION: &str = "dll";
+
+#[derive(Debug, Serialize, Deserialize)]
+struct Configuration {
+    #[serde(rename = "use-grammars")]
+    pub grammar_selection: Option<GrammarSelection>,
+    pub grammar: Vec<GrammarConfiguration>,
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "lowercase", untagged)]
+pub enum GrammarSelection {
+    Only(HashSet<String>),
+    Except(HashSet<String>),
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(deny_unknown_fields)]
+pub struct GrammarConfiguration {
+    #[serde(rename = "name")]
+    pub grammar_id: String,
+    pub source: GrammarSource,
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "lowercase", untagged)]
+pub enum GrammarSource {
+    Local {
+        path: String,
+    },
+    Git {
+        #[serde(rename = "git")]
+        remote: String,
+        #[serde(rename = "rev")]
+        revision: String,
+        subpath: Option<String>,
+    },
+}
+
+const BUILD_TARGET: &str = env!("BUILD_TARGET");
+const REMOTE_NAME: &str = "origin";
+
+pub fn get_language(name: &str) -> Result<Language> {
+    let name = name.to_ascii_lowercase();
+    let mut library_path = crate::runtime_dir().join("grammars").join(&name);
+    library_path.set_extension(DYLIB_EXTENSION);
+
+    let library = unsafe { Library::new(&library_path) }
+        .with_context(|| format!("Error opening dynamic library {library_path:?}"))?;
+    let language_fn_name = format!("tree_sitter_{}", name.replace('-', "_"));
+    let language = unsafe {
+        let language_fn: Symbol<unsafe extern "C" fn() -> Language> = library
+            .get(language_fn_name.as_bytes())
+            .with_context(|| format!("Failed to load symbol {language_fn_name}"))?;
+        language_fn()
+    };
+    std::mem::forget(library);
+    Ok(language)
+}
+
+pub fn fetch_grammars() -> Result<()> {
+    run_parallel(get_grammar_configs()?, fetch_grammar, "fetch")
+}
+
+pub fn build_grammars() -> Result<()> {
+    run_parallel(get_grammar_configs()?, build_grammar, "build")
+}
+
+// Returns the set of grammar configurations the user requests.
+// Grammars are configured in the default and user `languages.toml` and are
+// merged. The `grammar_selection` key of the config is then used to filter
+// down all grammars into a subset of the user's choosing.
+fn get_grammar_configs() -> Result<Vec<GrammarConfiguration>> {
+    let config: Configuration = crate::user_lang_config()
+        .context("Could not parse languages.toml")?
+        .try_into()?;
+
+    let grammars = match config.grammar_selection {
+        Some(GrammarSelection::Only(selections)) => config
+            .grammar
+            .into_iter()
+            .filter(|grammar| selections.contains(&grammar.grammar_id))
+            .collect(),
+        Some(GrammarSelection::Except(rejections)) => config
+            .grammar
+            .into_iter()
+            .filter(|grammar| !rejections.contains(&grammar.grammar_id))
+            .collect(),
+        None => config.grammar,
+    };
+
+    Ok(grammars)
+}
+
+fn run_parallel<F>(grammars: Vec<GrammarConfiguration>, job: F, action: &'static str) -> Result<()>
+where
+    F: Fn(GrammarConfiguration) -> Result<()> + std::marker::Send + 'static + Copy,
+{
+    let pool = threadpool::Builder::new().build();
+    let (tx, rx) = channel();
+
+    for grammar in grammars {
+        let tx = tx.clone();
+
+        pool.execute(move || {
+            tx.send(job(grammar)).unwrap();
+        });
+    }
+    pool.join();
+
+    // TODO: print all failures instead of the first one found.
+    if let Some(failure) = rx.try_iter().find_map(|result| result.err()) {
+        Err(anyhow!(
+            "Failed to {} some grammar(s).\n{}",
+            action,
+            failure
+        ))
+    } else {
+        Ok(())
+    }
+}
+
+fn fetch_grammar(grammar: GrammarConfiguration) -> Result<()> {
+    if let GrammarSource::Git {
+        remote, revision, ..
+    } = grammar.source
+    {
+        let grammar_dir = crate::runtime_dir()
+            .join("grammars/sources")
+            .join(&grammar.grammar_id);
+
+        fs::create_dir_all(&grammar_dir).context(format!(
+            "Could not create grammar directory {:?}",
+            grammar_dir
+        ))?;
+
+        // create the grammar dir contains a git directory
+        if !grammar_dir.join(".git").is_dir() {
+            git(&grammar_dir, ["init"])?;
+        }
+
+        // ensure the remote matches the configured remote
+        if get_remote_url(&grammar_dir).map_or(true, |s| s != remote) {
+            set_remote(&grammar_dir, &remote)?;
+        }
+
+        // ensure the revision matches the configured revision
+        if get_revision(&grammar_dir).map_or(true, |s| s != revision) {
+            // Fetch the exact revision from the remote.
+            // Supported by server-side git since v2.5.0 (July 2015),
+            // enabled by default on major git hosts.
+            git(&grammar_dir, ["fetch", REMOTE_NAME, &revision])?;
+            git(&grammar_dir, ["checkout", &revision])?;
+
+            println!(
+                "Grammar '{}' checked out at '{}'.",
+                grammar.grammar_id, revision
+            );
+            Ok(())
+        } else {
+            println!("Grammar '{}' is already up to date.", grammar.grammar_id);
+            Ok(())
+        }
+    } else {
+        println!("Skipping local grammar '{}'", grammar.grammar_id);
+        Ok(())
+    }
+}
+
+// Sets the remote for a repository to the given URL, creating the remote if
+// it does not yet exist.
+fn set_remote(repository_dir: &Path, remote_url: &str) -> Result<String> {
+    git(
+        repository_dir,
+        ["remote", "set-url", REMOTE_NAME, remote_url],
+    )
+    .or_else(|_| git(repository_dir, ["remote", "add", REMOTE_NAME, remote_url]))
+}
+
+fn get_remote_url(repository_dir: &Path) -> Option<String> {
+    git(repository_dir, ["remote", "get-url", REMOTE_NAME]).ok()
+}
+
+fn get_revision(repository_dir: &Path) -> Option<String> {
+    git(repository_dir, ["rev-parse", "HEAD"]).ok()
+}
+
+// A wrapper around 'git' commands which returns stdout in success and a
+// helpful error message showing the command, stdout, and stderr in error.
+fn git<I, S>(repository_dir: &Path, args: I) -> Result<String>
+where
+    I: IntoIterator<Item = S>,
+    S: AsRef<std::ffi::OsStr>,
+{
+    let output = Command::new("git")
+        .args(args)
+        .current_dir(repository_dir)
+        .output()?;
+
+    if output.status.success() {
+        Ok(String::from_utf8_lossy(&output.stdout)
+            .trim_end()
+            .to_owned())
+    } else {
+        // TODO: figure out how to display the git command using `args`
+        Err(anyhow!(
+            "Git command failed.\nStdout: {}\nStderr: {}",
+            String::from_utf8_lossy(&output.stdout),
+            String::from_utf8_lossy(&output.stderr),
+        ))
+    }
+}
+
+fn build_grammar(grammar: GrammarConfiguration) -> Result<()> {
+    println!("{:#?}", grammar);
+    let grammar_dir = if let GrammarSource::Local { path } = &grammar.source {
+        PathBuf::from(&path)
+    } else {
+        crate::runtime_dir()
+            .join("grammars/sources")
+            .join(&grammar.grammar_id)
+    };
+
+    let grammar_dir_entries = grammar_dir.read_dir().with_context(|| {
+        format!("Failed to read directory {grammar_dir:?}. Did you use 'hx --fetch-grammars'?")
+    })?;
+
+    if grammar_dir_entries.count() == 0 {
+        return Err(anyhow!(
+            "Directory {grammar_dir:?} is empty. Did you use 'hx --fetch-grammars'?"
+        ));
+    };
+
+    let path = match &grammar.source {
+        GrammarSource::Git {
+            subpath: Some(subpath),
+            ..
+        } => grammar_dir.join(subpath),
+        _ => grammar_dir,
+    }
+    .join("src");
+
+    build_tree_sitter_library(&path, grammar)
+}
+
+fn build_tree_sitter_library(src_path: &Path, grammar: GrammarConfiguration) -> Result<()> {
+    let header_path = src_path;
+    let parser_path = src_path.join("parser.c");
+    let mut scanner_path = src_path.join("scanner.c");
+
+    let scanner_path = if scanner_path.exists() {
+        Some(scanner_path)
+    } else {
+        scanner_path.set_extension("cc");
+        if scanner_path.exists() {
+            Some(scanner_path)
+        } else {
+            None
+        }
+    };
+    let parser_lib_path = crate::runtime_dir().join("grammars");
+    let mut library_path = parser_lib_path.join(&grammar.grammar_id);
+    library_path.set_extension(DYLIB_EXTENSION);
+
+    let recompile = needs_recompile(&library_path, &parser_path, &scanner_path)
+        .context("Failed to compare source and binary timestamps")?;
+
+    if !recompile {
+        println!("Grammar '{}' is already built.", grammar.grammar_id);
+        return Ok(());
+    }
+
+    println!("Building grammar '{}'", grammar.grammar_id);
+
+    let mut config = cc::Build::new();
+    config
+        .cpp(true)
+        .opt_level(3)
+        .cargo_metadata(false)
+        .host(BUILD_TARGET)
+        .target(BUILD_TARGET);
+    let compiler = config.get_compiler();
+    let mut command = Command::new(compiler.path());
+    command.current_dir(src_path);
+    for (key, value) in compiler.env() {
+        command.env(key, value);
+    }
+
+    if cfg!(windows) {
+        command
+            .args(&["/nologo", "/LD", "/I"])
+            .arg(header_path)
+            .arg("/Od")
+            .arg("/utf-8");
+        if let Some(scanner_path) = scanner_path.as_ref() {
+            command.arg(scanner_path);
+        }
+
+        command
+            .arg(parser_path)
+            .arg("/link")
+            .arg(format!("/out:{}", library_path.to_str().unwrap()));
+    } else {
+        command
+            .arg("-shared")
+            .arg("-fPIC")
+            .arg("-fno-exceptions")
+            .arg("-g")
+            .arg("-I")
+            .arg(header_path)
+            .arg("-o")
+            .arg(&library_path)
+            .arg("-O3");
+        if let Some(scanner_path) = scanner_path.as_ref() {
+            if scanner_path.extension() == Some("c".as_ref()) {
+                command.arg("-xc").arg("-std=c99").arg(scanner_path);
+            } else {
+                command.arg(scanner_path);
+            }
+        }
+        command.arg("-xc").arg(parser_path);
+        if cfg!(all(unix, not(target_os = "macos"))) {
+            command.arg("-Wl,-z,relro,-z,now");
+        }
+    }
+
+    let output = command.output().context("Failed to execute C compiler")?;
+    if !output.status.success() {
+        return Err(anyhow!(
+            "Parser compilation failed.\nStdout: {}\nStderr: {}",
+            String::from_utf8_lossy(&output.stdout),
+            String::from_utf8_lossy(&output.stderr)
+        ));
+    }
+
+    Ok(())
+}
+
+fn needs_recompile(
+    lib_path: &Path,
+    parser_c_path: &Path,
+    scanner_path: &Option<PathBuf>,
+) -> Result<bool> {
+    if !lib_path.exists() {
+        return Ok(true);
+    }
+    let lib_mtime = mtime(lib_path)?;
+    if mtime(parser_c_path)? > lib_mtime {
+        return Ok(true);
+    }
+    if let Some(scanner_path) = scanner_path {
+        if mtime(scanner_path)? > lib_mtime {
+            return Ok(true);
+        }
+    }
+    Ok(false)
+}
+
+fn mtime(path: &Path) -> Result<SystemTime> {
+    Ok(fs::metadata(path)?.modified()?)
+}
+
+/// Gives the contents of a file from a language's `runtime/queries/<lang>`
+/// directory
+pub fn load_runtime_file(language: &str, filename: &str) -> Result<String, std::io::Error> {
+    let path = crate::RUNTIME_DIR
+        .join("queries")
+        .join(language)
+        .join(filename);
+    std::fs::read_to_string(&path)
+}
diff --git a/helix-loader/src/lib.rs b/helix-loader/src/lib.rs
new file mode 100644
index 00000000..a2c4d96f
--- /dev/null
+++ b/helix-loader/src/lib.rs
@@ -0,0 +1,161 @@
+pub mod grammar;
+
+use etcetera::base_strategy::{choose_base_strategy, BaseStrategy};
+
+pub static RUNTIME_DIR: once_cell::sync::Lazy<std::path::PathBuf> =
+    once_cell::sync::Lazy::new(runtime_dir);
+
+pub fn runtime_dir() -> std::path::PathBuf {
+    if let Ok(dir) = std::env::var("HELIX_RUNTIME") {
+        return dir.into();
+    }
+
+    const RT_DIR: &str = "runtime";
+    let conf_dir = config_dir().join(RT_DIR);
+    if conf_dir.exists() {
+        return conf_dir;
+    }
+
+    if let Ok(dir) = std::env::var("CARGO_MANIFEST_DIR") {
+        // this is the directory of the crate being run by cargo, we need the workspace path so we take the parent
+        return std::path::PathBuf::from(dir).parent().unwrap().join(RT_DIR);
+    }
+
+    // fallback to location of the executable being run
+    std::env::current_exe()
+        .ok()
+        .and_then(|path| path.parent().map(|path| path.to_path_buf().join(RT_DIR)))
+        .unwrap()
+}
+
+pub fn config_dir() -> std::path::PathBuf {
+    // TODO: allow env var override
+    let strategy = choose_base_strategy().expect("Unable to find the config directory!");
+    let mut path = strategy.config_dir();
+    path.push("helix");
+    path
+}
+
+pub fn cache_dir() -> std::path::PathBuf {
+    // TODO: allow env var override
+    let strategy = choose_base_strategy().expect("Unable to find the config directory!");
+    let mut path = strategy.cache_dir();
+    path.push("helix");
+    path
+}
+
+pub fn config_file() -> std::path::PathBuf {
+    config_dir().join("config.toml")
+}
+
+pub fn lang_config_file() -> std::path::PathBuf {
+    config_dir().join("languages.toml")
+}
+
+pub fn log_file() -> std::path::PathBuf {
+    cache_dir().join("helix.log")
+}
+
+/// Default bultin-in languages.toml.
+pub fn default_lang_config() -> toml::Value {
+    toml::from_slice(include_bytes!("../../languages.toml"))
+        .expect("Could not parse bultin-in languages.toml to valid toml")
+}
+
+/// User configured languages.toml file, merged with the default config.
+pub fn user_lang_config() -> Result<toml::Value, toml::de::Error> {
+    let def_lang_conf = default_lang_config();
+    let data = std::fs::read(crate::config_dir().join("languages.toml"));
+    let user_lang_conf = match data {
+        Ok(raw) => {
+            let value = toml::from_slice(&raw)?;
+            merge_toml_values(def_lang_conf, value)
+        }
+        Err(_) => def_lang_conf,
+    };
+
+    Ok(user_lang_conf)
+}
+
+// right overrides left
+pub fn merge_toml_values(left: toml::Value, right: toml::Value) -> toml::Value {
+    use toml::Value;
+
+    fn get_name(v: &Value) -> Option<&str> {
+        v.get("name").and_then(Value::as_str)
+    }
+
+    match (left, right) {
+        (Value::Array(mut left_items), Value::Array(right_items)) => {
+            left_items.reserve(right_items.len());
+            for rvalue in right_items {
+                let lvalue = get_name(&rvalue)
+                    .and_then(|rname| left_items.iter().position(|v| get_name(v) == Some(rname)))
+                    .map(|lpos| left_items.remove(lpos));
+                let mvalue = match lvalue {
+                    Some(lvalue) => merge_toml_values(lvalue, rvalue),
+                    None => rvalue,
+                };
+                left_items.push(mvalue);
+            }
+            Value::Array(left_items)
+        }
+        (Value::Table(mut left_map), Value::Table(right_map)) => {
+            for (rname, rvalue) in right_map {
+                match left_map.remove(&rname) {
+                    Some(lvalue) => {
+                        let merged_value = merge_toml_values(lvalue, rvalue);
+                        left_map.insert(rname, merged_value);
+                    }
+                    None => {
+                        left_map.insert(rname, rvalue);
+                    }
+                }
+            }
+            Value::Table(left_map)
+        }
+        // Catch everything else we didn't handle, and use the right value
+        (_, value) => value,
+    }
+}
+
+#[cfg(test)]
+mod merge_toml_tests {
+    use super::merge_toml_values;
+
+    #[test]
+    fn language_tomls() {
+        use toml::Value;
+
+        const USER: &str = "
+        [[language]]
+        name = \"nix\"
+        test = \"bbb\"
+        indent = { tab-width = 4, unit = \"    \", test = \"aaa\" }
+        ";
+
+        let base: Value = toml::from_slice(include_bytes!("../../languages.toml"))
+            .expect("Couldn't parse built-in languages config");
+        let user: Value = toml::from_str(USER).unwrap();
+
+        let merged = merge_toml_values(base, user);
+        let languages = merged.get("language").unwrap().as_array().unwrap();
+        let nix = languages
+            .iter()
+            .find(|v| v.get("name").unwrap().as_str().unwrap() == "nix")
+            .unwrap();
+        let nix_indent = nix.get("indent").unwrap();
+
+        // We changed tab-width and unit in indent so check them if they are the new values
+        assert_eq!(
+            nix_indent.get("tab-width").unwrap().as_integer().unwrap(),
+            4
+        );
+        assert_eq!(nix_indent.get("unit").unwrap().as_str().unwrap(), "    ");
+        // We added a new keys, so check them
+        assert_eq!(nix.get("test").unwrap().as_str().unwrap(), "bbb");
+        assert_eq!(nix_indent.get("test").unwrap().as_str().unwrap(), "aaa");
+        // We didn't change comment-token so it should be same
+        assert_eq!(nix.get("comment-token").unwrap().as_str().unwrap(), "#");
+    }
+}