Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
765 changes: 632 additions & 133 deletions Cargo.lock

Large diffs are not rendered by default.

37 changes: 24 additions & 13 deletions Cargo.toml
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,9 @@ stable = ["json", "toml", "typescript", "tsx", "jsx", "go", "python", "yaml", "j
default = []
[dev-dependencies]
test-log = { workspace = true }
criterion = { version = "0.5", features = ["html_reports"] }
git2 = { version = "0.20.0" }
tempfile = {workspace = true}
[workspace]
members = [
"codegen-sdk-analyzer",
Expand Down Expand Up @@ -109,40 +112,48 @@ codegen-sdk-json = { path = "languages/codegen-sdk-json" }
tempfile = "3.16.0"
quote = "1.0.38"
proc-macro2 = "1.0.93"
derive_generic_visitor = "0.1.1"
insta = "1.42.1"
prettyplease = "0.2.29"
syn = { version = "2.0.98", features = ["proc-macro"] }
syn = { version = "2.0.98", features = ["proc-macro", "full"] }
derive_more = { version = "2.0.1", features = ["debug", "display"] }
salsa = "0.18.0"
salsa = {git = "https://github.com/salsa-rs/salsa", branch = "master"}
subenum = {git = "https://github.com/mrenow/subenum", branch = "main"}
indicatif-log-bridge = "0.2.3"
indicatif = { version = "0.17.11", features = ["rayon"] }
crossbeam-channel = "0.5.11"
rstest = "0.25.0"
indextree = "4.7.3"
thiserror = "2.0.11"

[profile.dev]
# codegen-backend = "cranelift"
split-debuginfo = "unpacked"
# split-debuginfo = "unpacked"

[profile.dev.package]
insta.opt-level = 3
similar.opt-level = 3
syn.opt-level = 3
convert_case.opt-level = 3
codegen-sdk-cst-generator.opt-level = 3
codegen-sdk-common.opt-level = 3
# codegen-sdk-cst-generator.opt-level = 3
# codegen-sdk-common.opt-level = 3

[profile.dev.build-override]
opt-level = 3
debug = true

[profile.test.package."codegen-sdk-cst"]
inherits = "dev"
opt-level = 0
debug = 0
strip = "none"
codegen-units = 256
incremental = true
# [profile.test.package."codegen-sdk-cst"]
# inherits = "dev"
# opt-level = 0
# debug = 0
# strip = "none"
# codegen-units = 256
# incremental = true
# codegen-backend = "cranelift"

[profile.test]
lto = false

[[bench]]
name = "parse"
harness = false
required-features = ["stable"]
36 changes: 36 additions & 0 deletions benches/parse.rs
Original file line number Diff line number Diff line change
@@ -0,0 +1,36 @@
use std::{hint::black_box, path::PathBuf};

use codegen_sdk_analyzer::Codebase;
use criterion::{Criterion, criterion_group, criterion_main};
fn clone_repo(url: String, name: String, tmp_dir: &tempfile::TempDir) -> PathBuf {
let repo_path = tmp_dir.path().join(name);
if !repo_path.exists() {
log::info!("Cloning repo: {} to {}", url, repo_path.display());
let mut fetch_opts = git2::FetchOptions::new();
fetch_opts.depth(1);
let _ = git2::build::RepoBuilder::new()
.fetch_options(fetch_opts)
.clone(&url, &repo_path)
.unwrap();
}
repo_path
}
fn parse_nest(path: &PathBuf) {
let _ = Codebase::new(path.clone());
}

fn criterion_benchmark(c: &mut Criterion) {
env_logger::init();
let temp_dir = tempfile::tempdir().unwrap();
let repo_path = clone_repo(
"https://github.com/nestjs/nest".to_string(),
"nest".to_string(),
&temp_dir,
);
c.bench_function("parse_nest", |b| {
b.iter(|| parse_nest(black_box(&repo_path)))
});
}

criterion_group!(benches, criterion_benchmark);
criterion_main!(benches);
11 changes: 6 additions & 5 deletions codegen-sdk-analyzer/src/codebase/discovery.rs
Original file line number Diff line number Diff line change
Expand Up @@ -22,14 +22,15 @@ pub fn log_languages() {

pub fn collect_files(db: &CodegenDatabase, dir: &PathBuf) -> FilesToParse {
let mut files = Vec::new();
let dir = dir.canonicalize().unwrap();
for language in LANGUAGES.iter() {
for extension in language.file_extensions.iter() {
files.extend(
glob(&format!(
"{dir}**/*.{extension}",
extension = extension,
dir = dir.display()
))
glob(
&dir.join(format!("**/*.{extension}", extension = extension))
.to_str()
.unwrap(),
)
.unwrap(),
);
}
Expand Down
10 changes: 5 additions & 5 deletions codegen-sdk-analyzer/src/parser.rs
Original file line number Diff line number Diff line change
@@ -1,18 +1,18 @@
use std::path::PathBuf;

use codegen_sdk_common::FileNodeId;
use codegen_sdk_cst::CSTLanguage;
use codegen_sdk_macros::{languages_ast, parse_language};
languages_ast!();

#[salsa::tracked]
pub struct Parsed<'db> {
#[id]
path: PathBuf,
path: FileNodeId<'db>,
#[return_ref]
#[tracked]
pub file: Option<ParsedFile<'db>>,
}
#[salsa::tracked]
#[salsa::tracked(return_ref)]
pub fn parse_file(db: &dyn salsa::Database, file: codegen_sdk_ast::input::File) -> Parsed<'_> {
parse_language!();
Parsed::new(db, file.path(db), None)
Parsed::new(db, FileNodeId::new(db, file.path(db)), None)
}
52 changes: 31 additions & 21 deletions codegen-sdk-ast-generator/src/generator.rs
Original file line number Diff line number Diff line change
Expand Up @@ -8,8 +8,8 @@ fn get_definitions_impl(language: &Language) -> TokenStream {

impl<'db> codegen_sdk_ast::Definitions<'db> for #language_struct_name<'db> {
type Definitions = ();
fn definitions(self, _db: &'db dyn salsa::Database) -> Self::Definitions{
()
fn definitions(self, _db: &'db dyn salsa::Database) -> &'db Self::Definitions{
&()
}
}
};
Expand All @@ -18,13 +18,13 @@ fn get_definitions_impl(language: &Language) -> TokenStream {
#[salsa::tracked]
impl<'db> codegen_sdk_ast::Definitions<'db> for #language_struct_name<'db> {
type Definitions = Definitions<'db>;
#[salsa::tracked]
#[salsa::tracked(return_ref)]
fn definitions(self, db: &'db dyn salsa::Database) -> Self::Definitions {
let mut definitions = Definitions::default();
if let Some(program) = self.node(db) {
definitions = definitions.visit_by_val_infallible(&program);
return Definitions::visit(db, program);
} else {
return Definitions::default(db);
}
definitions
}
}
}
Expand All @@ -35,8 +35,8 @@ fn get_references_impl(language: &Language) -> TokenStream {
return quote! {
impl<'db> codegen_sdk_ast::References<'db> for #language_struct_name<'db> {
type References = ();
fn references(self, _db: &'db dyn salsa::Database) -> Self::References {
()
fn references(self, _db: &'db dyn salsa::Database) -> &'db Self::References {
&()
}
}
};
Expand All @@ -45,13 +45,13 @@ fn get_references_impl(language: &Language) -> TokenStream {
#[salsa::tracked]
impl<'db> codegen_sdk_ast::References<'db> for #language_struct_name<'db> {
type References = References<'db>;
#[salsa::tracked]
#[salsa::tracked(return_ref)]
fn references(self, db: &'db dyn salsa::Database) -> Self::References {
let mut references = References::default();
if let Some(program) = self.node(db) {
references = references.visit_by_val_infallible(&program);
}
references
if let Some(program) = self.node(db) {
return References::visit(db, program);
} else {
return References::default(db);
}
}
}
}
Expand All @@ -61,14 +61,15 @@ pub fn generate_ast(language: &Language) -> anyhow::Result<TokenStream> {
let language_name_str = language.name();
let definitions_impl = get_definitions_impl(language);
let references_impl = get_references_impl(language);
let program_id = format_ident!("{}", language.root_node());
let root_node_name = format_ident!("{}", language.root_node());
let content = quote! {
#[salsa::tracked]
pub struct #language_struct_name<'db> {
#[tracked]
#[return_ref]
node: Option<crate::cst::#program_id<'db>>,
pub node: Option<crate::cst::Parsed<'db>>,
#[id]
pub path: PathBuf,
pub id: codegen_sdk_common::FileNodeId<'db>,
}
// impl<'db> File for {language_struct_name}File<'db> {{
// fn path(&self) -> &PathBuf {{
Expand All @@ -77,15 +78,24 @@ pub fn generate_ast(language: &Language) -> anyhow::Result<TokenStream> {
// }}
pub fn parse(db: &dyn salsa::Database, input: codegen_sdk_ast::input::File) -> #language_struct_name<'_> {
log::debug!("Parsing {} file: {}", input.path(db).display(), #language_name_str);
let ast = crate::cst::parse_program_raw(db, input.contents(db));
#language_struct_name::new(db, ast, input.path(db).clone())
let ast = crate::cst::parse_program_raw(db, input.contents(db), input.path(db).clone());
let file_id = codegen_sdk_common::FileNodeId::new(db, input.path(db).clone());
#language_struct_name::new(db, ast, file_id)
}
#[salsa::tracked]
#[salsa::tracked(return_ref)]
pub fn parse_query(db: &dyn salsa::Database, input: codegen_sdk_ast::input::File) -> #language_struct_name<'_> {
parse(db, input)
}


impl<'db> #language_struct_name<'db> {
pub fn tree(&self, db: &'db dyn salsa::Database) -> &'db codegen_sdk_common::Tree<crate::cst::NodeTypes<'db>> {
self.node(db).unwrap().tree(db)
}
pub fn root(&self, db: &'db dyn salsa::Database) -> &'db crate::cst::#root_node_name<'db> {
let tree = self.tree(db);
tree.get(&self.node(db).unwrap().program(db)).unwrap().as_ref().try_into().unwrap()
}
}
#definitions_impl
#references_impl
// impl<'db> HasNode for {language_struct_name}File<'db> {
Expand Down
2 changes: 1 addition & 1 deletion codegen-sdk-ast-generator/src/lib.rs
Original file line number Diff line number Diff line change
Expand Up @@ -10,11 +10,11 @@ use syn::parse_quote;
pub fn generate_ast(language: &Language) -> anyhow::Result<()> {
let db = CSTDatabase::default();
let imports = quote! {
use derive_generic_visitor::{Visitor, Drive, Visit};
use codegen_sdk_common::*;
use std::path::PathBuf;
use codegen_sdk_cst::CSTLanguage;
use std::collections::BTreeMap;
use std::sync::mpsc::Sender;
};
let ast = generator::generate_ast(language)?;
let definition_visitor = visitor::generate_visitor(&db, language, "definition");
Expand Down
Loading