rust/crates/base_db/src/input.rs

460 lines
14 KiB
Rust
Raw Normal View History

//! This module specifies the input to rust-analyzer. In some sense, this is
//! **the** most important module, because all other fancy stuff is strictly
//! derived from this input.
//!
//! Note that neither this module, nor any other part of the analyzer's core do
2020-02-18 05:33:16 -06:00
//! actual IO. See `vfs` and `project_model` in the `rust-analyzer` crate for how
//! actual IO is done and lowered to input.
2020-07-21 10:17:21 -05:00
use std::{fmt, iter::FromIterator, ops, str::FromStr, sync::Arc};
2020-08-13 03:19:09 -05:00
use cfg::CfgOptions;
2020-06-11 04:30:06 -05:00
use rustc_hash::{FxHashMap, FxHashSet};
2020-08-12 11:26:51 -05:00
use syntax::SmolStr;
2020-08-12 09:46:20 -05:00
use tt::TokenExpander;
use vfs::{file_set::FileSet, VfsPath};
2020-06-11 04:04:09 -05:00
pub use vfs::FileId;
2018-10-25 09:52:50 -05:00
2018-12-20 04:47:32 -06:00
/// Files are grouped into source roots. A source root is a directory on the
/// file systems which is watched for changes. Typically it corresponds to a
/// Rust crate. Source roots *might* be nested: in this case, a file belongs to
/// the nearest enclosing source root. Paths to files are always relative to a
/// source root, and the analyzer does not know the root path of the source root at
/// all. So, a file from one source root can't refer to a file in another source
2018-12-20 04:47:32 -06:00
/// root by path.
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash, PartialOrd, Ord)]
pub struct SourceRootId(pub u32);
#[derive(Clone, Debug, PartialEq, Eq)]
2018-12-20 04:47:32 -06:00
pub struct SourceRoot {
/// Sysroot or crates.io library.
///
/// Libraries are considered mostly immutable, this assumption is used to
/// optimize salsa's query structure
pub is_library: bool,
2020-06-11 04:04:09 -05:00
pub(crate) file_set: FileSet,
2018-12-20 04:47:32 -06:00
}
2018-10-25 09:52:50 -05:00
impl SourceRoot {
2020-06-11 04:04:09 -05:00
pub fn new_local(file_set: FileSet) -> SourceRoot {
SourceRoot { is_library: false, file_set }
2019-09-05 14:36:04 -05:00
}
2020-06-11 04:04:09 -05:00
pub fn new_library(file_set: FileSet) -> SourceRoot {
SourceRoot { is_library: true, file_set }
2019-09-05 14:36:04 -05:00
}
pub fn path_for_file(&self, file: &FileId) -> Option<&VfsPath> {
self.file_set.path_for_file(file)
}
pub fn file_for_path(&self, path: &VfsPath) -> Option<&FileId> {
self.file_set.file_for_path(path)
}
2020-06-11 04:04:09 -05:00
pub fn iter(&self) -> impl Iterator<Item = FileId> + '_ {
self.file_set.iter()
2019-10-11 03:37:54 -05:00
}
}
2018-12-20 04:47:32 -06:00
/// `CrateGraph` is a bit of information which turns a set of text files into a
/// number of Rust crates. Each crate is defined by the `FileId` of its root module,
/// the set of cfg flags (not yet implemented) and the set of dependencies. Note
2018-12-20 04:47:32 -06:00
/// that, due to cfg's, there might be several crates for a single `FileId`! As
/// in the rust-lang proper, a crate does not have a name. Instead, names are
/// specified on dependency edges. That is, a crate might be known under
/// different names in different dependent crates.
2018-12-20 04:47:32 -06:00
///
/// Note that `CrateGraph` is build-system agnostic: it's a concept of the Rust
/// language proper, not a concept of the build system. In practice, we get
2018-12-20 04:47:32 -06:00
/// `CrateGraph` by lowering `cargo metadata` output.
2018-10-25 09:52:50 -05:00
#[derive(Debug, Clone, Default, PartialEq, Eq)]
pub struct CrateGraph {
2018-12-05 07:01:18 -06:00
arena: FxHashMap<CrateId, CrateData>,
2018-10-25 09:52:50 -05:00
}
2018-12-20 04:47:32 -06:00
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub struct CrateId(pub u32);
2020-07-01 03:03:07 -05:00
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
2020-02-05 03:53:54 -06:00
pub struct CrateName(SmolStr);
2020-02-05 04:47:28 -06:00
impl CrateName {
2020-03-08 08:26:57 -05:00
/// Creates a crate name, checking for dashes in the string provided.
2020-02-05 04:47:28 -06:00
/// Dashes are not allowed in the crate names,
/// hence the input string is returned as `Err` for those cases.
pub fn new(name: &str) -> Result<CrateName, &str> {
if name.contains('-') {
Err(name)
} else {
Ok(Self(SmolStr::new(name)))
}
}
2020-03-08 08:26:57 -05:00
/// Creates a crate name, unconditionally replacing the dashes with underscores.
2020-02-05 04:47:28 -06:00
pub fn normalize_dashes(name: &str) -> CrateName {
Self(SmolStr::new(name.replace('-', "_")))
2020-02-05 03:53:54 -06:00
}
2020-03-16 05:03:43 -05:00
}
2020-03-16 04:47:52 -05:00
2020-06-11 04:30:06 -05:00
impl fmt::Display for CrateName {
2020-03-16 05:03:43 -05:00
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{}", self.0)
2020-03-16 04:47:52 -05:00
}
2020-02-05 03:53:54 -06:00
}
2020-07-01 02:53:53 -05:00
impl ops::Deref for CrateName {
type Target = str;
fn deref(&self) -> &Self::Target {
&*self.0
}
}
2020-03-18 07:56:46 -05:00
#[derive(Debug, Copy, Clone, PartialEq, Eq, Hash)]
2020-03-26 11:41:44 -05:00
pub struct ProcMacroId(pub u32);
#[derive(Debug, Clone)]
pub struct ProcMacro {
pub name: SmolStr,
pub expander: Arc<dyn TokenExpander>,
}
impl Eq for ProcMacro {}
impl PartialEq for ProcMacro {
fn eq(&self, other: &ProcMacro) -> bool {
self.name == other.name && Arc::ptr_eq(&self.expander, &other.expander)
}
}
2020-03-18 07:56:46 -05:00
2019-11-22 05:12:45 -06:00
#[derive(Debug, Clone, PartialEq, Eq)]
2020-03-09 04:26:46 -05:00
pub struct CrateData {
pub root_file_id: FileId,
pub edition: Edition,
/// The name to display to the end user.
/// This actual crate name can be different in a particular dependent crate
/// or may even be missing for some cases, such as a dummy crate for the code snippet.
2020-10-02 13:16:22 -05:00
pub display_name: Option<CrateName>,
2020-03-09 05:14:51 -05:00
pub cfg_options: CfgOptions,
pub env: Env,
2020-03-09 04:26:46 -05:00
pub dependencies: Vec<Dependency>,
2020-03-18 07:56:46 -05:00
pub proc_macro: Vec<ProcMacro>,
}
2019-02-10 15:34:29 -06:00
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
pub enum Edition {
Edition2018,
Edition2015,
}
#[derive(Default, Debug, Clone, PartialEq, Eq)]
pub struct Env {
entries: FxHashMap<String, String>,
2020-03-10 22:04:02 -05:00
}
2020-03-10 08:59:12 -05:00
2018-12-05 07:01:18 -06:00
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct Dependency {
2018-12-08 16:05:49 -06:00
pub crate_id: CrateId,
2020-07-01 02:53:53 -05:00
pub name: CrateName,
2018-12-08 15:51:06 -06:00
}
2018-12-05 07:01:18 -06:00
impl CrateGraph {
pub fn add_crate_root(
&mut self,
file_id: FileId,
edition: Edition,
2020-10-02 13:16:22 -05:00
display_name: Option<CrateName>,
cfg_options: CfgOptions,
env: Env,
2020-08-12 09:46:20 -05:00
proc_macro: Vec<(SmolStr, Arc<dyn tt::TokenExpander>)>,
) -> CrateId {
2020-03-26 11:41:44 -05:00
let proc_macro =
proc_macro.into_iter().map(|(name, it)| ProcMacro { name, expander: it }).collect();
2020-03-09 05:17:39 -05:00
let data = CrateData {
root_file_id: file_id,
edition,
display_name,
cfg_options,
env,
2020-03-18 07:56:46 -05:00
proc_macro,
2020-03-09 05:17:39 -05:00
dependencies: Vec::new(),
};
2018-12-05 07:01:18 -06:00
let crate_id = CrateId(self.arena.len() as u32);
let prev = self.arena.insert(crate_id, data);
2018-10-25 09:52:50 -05:00
assert!(prev.is_none());
crate_id
}
pub fn add_dep(
&mut self,
from: CrateId,
2020-02-05 03:53:54 -06:00
name: CrateName,
to: CrateId,
2019-11-22 05:12:45 -06:00
) -> Result<(), CyclicDependenciesError> {
if self.dfs_find(from, to, &mut FxHashSet::default()) {
2019-11-22 05:12:45 -06:00
return Err(CyclicDependenciesError);
}
2020-07-01 02:53:53 -05:00
self.arena.get_mut(&from).unwrap().add_dep(name, to);
2019-07-04 21:59:28 -05:00
Ok(())
2018-12-05 07:01:18 -06:00
}
2018-12-21 10:13:26 -06:00
pub fn is_empty(&self) -> bool {
self.arena.is_empty()
}
pub fn iter(&self) -> impl Iterator<Item = CrateId> + '_ {
2019-07-04 12:26:44 -05:00
self.arena.keys().copied()
}
/// Returns an iterator over all transitive dependencies of the given crate.
pub fn transitive_deps(&self, of: CrateId) -> impl Iterator<Item = CrateId> + '_ {
let mut worklist = vec![of];
let mut deps = FxHashSet::default();
while let Some(krate) = worklist.pop() {
if !deps.insert(krate) {
continue;
}
worklist.extend(self[krate].dependencies.iter().map(|dep| dep.crate_id));
}
deps.remove(&of);
deps.into_iter()
}
2019-03-23 02:53:48 -05:00
// FIXME: this only finds one crate with the given root; we could have multiple
pub fn crate_id_for_crate_root(&self, file_id: FileId) -> Option<CrateId> {
2020-03-09 04:26:46 -05:00
let (&crate_id, _) =
self.arena.iter().find(|(_crate_id, data)| data.root_file_id == file_id)?;
Some(crate_id)
2018-11-26 15:12:43 -06:00
}
/// Extends this crate graph by adding a complete disjoint second crate
/// graph.
///
/// The ids of the crates in the `other` graph are shifted by the return
/// amount.
pub fn extend(&mut self, other: CrateGraph) -> u32 {
let start = self.arena.len() as u32;
self.arena.extend(other.arena.into_iter().map(|(id, mut data)| {
let new_id = id.shift(start);
for dep in &mut data.dependencies {
dep.crate_id = dep.crate_id.shift(start);
}
(new_id, data)
}));
start
}
2018-12-22 01:30:58 -06:00
fn dfs_find(&self, target: CrateId, from: CrateId, visited: &mut FxHashSet<CrateId>) -> bool {
if !visited.insert(from) {
2018-12-22 01:30:58 -06:00
return false;
}
if target == from {
return true;
}
2020-03-09 05:11:59 -05:00
for dep in &self[from].dependencies {
2020-03-09 05:18:41 -05:00
let crate_id = dep.crate_id;
2018-12-22 01:30:58 -06:00
if self.dfs_find(target, crate_id, visited) {
return true;
}
}
2019-07-04 12:26:44 -05:00
false
}
2018-10-25 09:52:50 -05:00
}
2020-03-09 05:11:59 -05:00
impl ops::Index<CrateId> for CrateGraph {
type Output = CrateData;
fn index(&self, crate_id: CrateId) -> &CrateData {
&self.arena[&crate_id]
}
}
2019-11-22 05:12:45 -06:00
impl CrateId {
pub fn shift(self, amount: u32) -> CrateId {
CrateId(self.0 + amount)
}
}
impl CrateData {
2020-07-01 02:53:53 -05:00
fn add_dep(&mut self, name: CrateName, crate_id: CrateId) {
2019-11-22 05:12:45 -06:00
self.dependencies.push(Dependency { name, crate_id })
}
2019-11-22 05:08:18 -06:00
}
impl FromStr for Edition {
type Err = ParseEditionError;
fn from_str(s: &str) -> Result<Self, Self::Err> {
let res = match s {
"2015" => Edition::Edition2015,
"2018" => Edition::Edition2018,
_ => return Err(ParseEditionError { invalid_input: s.to_string() }),
2019-11-22 05:08:18 -06:00
};
Ok(res)
}
}
impl fmt::Display for Edition {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
f.write_str(match self {
Edition::Edition2015 => "2015",
Edition::Edition2018 => "2018",
})
}
}
2020-07-21 10:17:21 -05:00
impl FromIterator<(String, String)> for Env {
fn from_iter<T: IntoIterator<Item = (String, String)>>(iter: T) -> Self {
Env { entries: FromIterator::from_iter(iter) }
}
}
impl Env {
pub fn set(&mut self, env: &str, value: String) {
self.entries.insert(env.to_owned(), value);
}
pub fn get(&self, env: &str) -> Option<String> {
self.entries.get(env).cloned()
}
2020-03-10 22:04:02 -05:00
}
2020-03-10 08:59:12 -05:00
2019-11-22 05:12:45 -06:00
#[derive(Debug)]
pub struct ParseEditionError {
invalid_input: String,
}
2019-11-22 05:08:18 -06:00
impl fmt::Display for ParseEditionError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(f, "invalid edition: {:?}", self.invalid_input)
}
}
impl std::error::Error for ParseEditionError {}
2019-11-22 05:12:45 -06:00
#[derive(Debug)]
pub struct CyclicDependenciesError;
#[cfg(test)]
mod tests {
2020-02-05 04:47:28 -06:00
use super::{CfgOptions, CrateGraph, CrateName, Dependency, Edition::Edition2018, Env, FileId};
#[test]
fn detect_cyclic_dependency_indirect() {
let mut graph = CrateGraph::default();
2020-03-08 08:26:57 -05:00
let crate1 = graph.add_crate_root(
FileId(1u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
let crate2 = graph.add_crate_root(
FileId(2u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
let crate3 = graph.add_crate_root(
FileId(3u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
2020-02-05 04:47:28 -06:00
assert!(graph.add_dep(crate1, CrateName::new("crate2").unwrap(), crate2).is_ok());
assert!(graph.add_dep(crate2, CrateName::new("crate3").unwrap(), crate3).is_ok());
assert!(graph.add_dep(crate3, CrateName::new("crate1").unwrap(), crate1).is_err());
}
#[test]
fn detect_cyclic_dependency_direct() {
let mut graph = CrateGraph::default();
let crate1 = graph.add_crate_root(
FileId(1u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
Default::default(),
);
let crate2 = graph.add_crate_root(
FileId(2u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
Default::default(),
);
assert!(graph.add_dep(crate1, CrateName::new("crate2").unwrap(), crate2).is_ok());
assert!(graph.add_dep(crate2, CrateName::new("crate2").unwrap(), crate2).is_err());
}
#[test]
fn it_works() {
let mut graph = CrateGraph::default();
2020-03-08 08:26:57 -05:00
let crate1 = graph.add_crate_root(
FileId(1u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
let crate2 = graph.add_crate_root(
FileId(2u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
let crate3 = graph.add_crate_root(
FileId(3u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
2020-02-05 04:47:28 -06:00
assert!(graph.add_dep(crate1, CrateName::new("crate2").unwrap(), crate2).is_ok());
assert!(graph.add_dep(crate2, CrateName::new("crate3").unwrap(), crate3).is_ok());
2020-02-05 03:53:54 -06:00
}
#[test]
fn dashes_are_normalized() {
let mut graph = CrateGraph::default();
2020-03-08 08:26:57 -05:00
let crate1 = graph.add_crate_root(
FileId(1u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
let crate2 = graph.add_crate_root(
FileId(2u32),
Edition2018,
None,
CfgOptions::default(),
Env::default(),
2020-03-10 22:04:02 -05:00
Default::default(),
2020-03-08 08:26:57 -05:00
);
2020-02-05 04:47:28 -06:00
assert!(graph
.add_dep(crate1, CrateName::normalize_dashes("crate-name-with-dashes"), crate2)
.is_ok());
2020-02-05 03:53:54 -06:00
assert_eq!(
2020-03-09 05:11:59 -05:00
graph[crate1].dependencies,
2020-07-01 02:53:53 -05:00
vec![Dependency {
crate_id: crate2,
name: CrateName::new("crate_name_with_dashes").unwrap()
}]
2020-02-05 03:53:54 -06:00
);
}
}