Merge #39

39: Tools r=matklad a=matklad

closes #34 

bors r+

.cargo/config

@@ -1,3 +1,4 @@
 [alias]
 parse = "run --package tools --bin parse"
 gen = "run --package tools --bin gen"
+collect-tests = "run --package tools --bin collect-tests --"

appveyor.yml

@@ -10,6 +10,7 @@ install:
 build: false
 
 test_script:
+  - cargo collect-tests --verify
   - cargo test
 
 branches:

docs/TESTS.md

@@ -19,12 +19,26 @@ files to have the same name except for the leading number. In general,
 test suite should be append-only: old tests should not be modified,
 new tests should be created instead.
 
-
 Note that only `ok` tests are normative: `err` tests test error
 recovery and it is totally ok for a parser to not implement any error
 recovery at all. However, for libsyntax2.0 we do care about error
 recovery, and we do care about precise and useful error messages.
 
+There are also so-called "inline tests". They appear as the comments 
+with a `test` header in the source code, like this:
+
+```rust
+// test fn_basic
+// fn foo() {}
+fn fn_item(p: &mut Parser) {
+    // ...
+}
+``` 
+
+You can run `cargo collect-tests` command to collect all inline tests 
+into `tests/data/inline` directory. The main advantage of inline tests
+is that they help to illustrate what the relevant code is doing. 
+
 
 Contribution opportunity: design and implement testing infrastructure
 for validators.

docs/TOOLS.md

@@ -17,14 +17,20 @@ cargo tool
 ```
 
 
-# Tool: `gen`
+## Tool: `gen`
 
 This tool reads a "grammar" from [grammar.ron](../grammar.ron) and
 generates the `syntax_kinds.rs` file. You should run this tool if you 
 add new keywords or syntax elements.
 
 
-# Tool: 'parse'
+## Tool: `parse`
 
 This tool reads rust source code from the standard input, parses it,
 and prints the result to stdout.
+
+
+## Tool: `collect-tests`
+
+This tools collect inline tests from comments in libsyntax2 source code
+and places them into `tests/data/inline` directory.

src/parser/event_parser/grammar/items/mod.rs

@@ -52,11 +52,15 @@ fn item(p: &mut Parser) {
             STATIC_ITEM
         }
         CONST_KW => match p.nth(1) {
+            // test const_fn
+            // const fn foo() {}
             FN_KW => {
                 p.bump();
                 fn_item(p);
                 FN_ITEM
             }
+            // test const_unsafe_fn
+            // const unsafe fn foo() {}
             UNSAFE_KW if p.nth(2) == FN_KW => {
                 p.bump();
                 p.bump();

tests/data/parser/inline/0001_const_unsafe_fn.rs

@@ -0,0 +1 @@
+const unsafe fn foo() {}

tests/data/parser/inline/0001_const_unsafe_fn.txt

@@ -0,0 +1,15 @@
+FILE@[0; 25)
+  FN_ITEM@[0; 25)
+    CONST_KW@[0; 5)
+    WHITESPACE@[5; 6)
+    UNSAFE_KW@[6; 12)
+    WHITESPACE@[12; 13)
+    FN_KW@[13; 15)
+    WHITESPACE@[15; 16)
+    IDENT@[16; 19) "foo"
+    L_PAREN@[19; 20)
+    R_PAREN@[20; 21)
+    WHITESPACE@[21; 22)
+    L_CURLY@[22; 23)
+    R_CURLY@[23; 24)
+    WHITESPACE@[24; 25)

tests/data/parser/inline/0002_const_fn.rs

@@ -0,0 +1 @@
+const fn foo() {}

tests/data/parser/inline/0002_const_fn.txt

@@ -0,0 +1,13 @@
+FILE@[0; 18)
+  FN_ITEM@[0; 18)
+    CONST_KW@[0; 5)
+    WHITESPACE@[5; 6)
+    FN_KW@[6; 8)
+    WHITESPACE@[8; 9)
+    IDENT@[9; 12) "foo"
+    L_PAREN@[12; 13)
+    R_PAREN@[13; 14)
+    WHITESPACE@[14; 15)
+    L_CURLY@[15; 16)
+    R_CURLY@[16; 17)
+    WHITESPACE@[17; 18)

tests/data/parser/ok/0024_const_fn.rs

@@ -1,5 +0,0 @@
-const fn foo() {
-}
-
-const unsafe fn foo() {
-}

tests/data/parser/ok/0024_const_fn.txt

@@ -1,29 +0,0 @@
-FILE@[0; 46)
-  FN_ITEM@[0; 20)
-    CONST_KW@[0; 5)
-    WHITESPACE@[5; 6)
-    FN_KW@[6; 8)
-    WHITESPACE@[8; 9)
-    IDENT@[9; 12) "foo"
-    L_PAREN@[12; 13)
-    R_PAREN@[13; 14)
-    WHITESPACE@[14; 15)
-    L_CURLY@[15; 16)
-    WHITESPACE@[16; 17)
-    R_CURLY@[17; 18)
-    WHITESPACE@[18; 20)
-  FN_ITEM@[20; 46)
-    CONST_KW@[20; 25)
-    WHITESPACE@[25; 26)
-    UNSAFE_KW@[26; 32)
-    WHITESPACE@[32; 33)
-    FN_KW@[33; 35)
-    WHITESPACE@[35; 36)
-    IDENT@[36; 39) "foo"
-    L_PAREN@[39; 40)
-    R_PAREN@[40; 41)
-    WHITESPACE@[41; 42)
-    L_CURLY@[42; 43)
-    WHITESPACE@[43; 44)
-    R_CURLY@[44; 45)
-    WHITESPACE@[45; 46)

tests/parser.rs

@@ -7,7 +7,7 @@ use testutils::dir_tests;
 
 #[test]
 fn parser_tests() {
-    dir_tests(&["parser/ok", "parser/err"], |text| {
+    dir_tests(&["parser/inline", "parser/ok", "parser/err"], |text| {
         let tokens = tokenize(text);
         let file = parse(text.to_string(), &tokens);
         dump_tree(&file)

tools/Cargo.toml

@@ -9,4 +9,6 @@ serde = "1.0.26"
 serde_derive = "1.0.26"
 file = "1.1.1"
 ron = "0.1.5"
+walkdir = "2"
+itertools = "0.7"
 libsyntax2 = { path = "../" }

tools/src/bin/collect-tests.rs

@@ -0,0 +1,130 @@
+extern crate file;
+extern crate itertools;
+extern crate walkdir;
+
+use walkdir::WalkDir;
+use itertools::Itertools;
+
+use std::path::{Path, PathBuf};
+use std::collections::HashSet;
+use std::fs;
+
+fn main() {
+    let verify = ::std::env::args().any(|arg| arg == "--verify");
+
+    let d = grammar_dir();
+    let tests = tests_from_dir(&d);
+    let existing = existing_tests();
+
+    for t in existing.difference(&tests) {
+        panic!("Test is deleted: {}\n{}", t.name, t.text);
+    }
+
+    let new_tests = tests.difference(&existing);
+    for (i, t) in new_tests.enumerate() {
+        if verify {
+            panic!("Inline test is not recorded: {}", t.name);
+        }
+
+        let name = format!("{:04}_{}.rs", existing.len() + i + 1, t.name);
+        println!("Creating {}", name);
+        let path = inline_tests_dir().join(name);
+        file::put_text(&path, &t.text).unwrap();
+    }
+}
+
+#[derive(Debug, Eq)]
+struct Test {
+    name: String,
+    text: String,
+}
+
+impl PartialEq for Test {
+    fn eq(&self, other: &Test) -> bool {
+        self.name.eq(&other.name)
+    }
+}
+
+impl ::std::hash::Hash for Test {
+    fn hash<H: ::std::hash::Hasher>(&self, state: &mut H) {
+        self.name.hash(state)
+    }
+}
+
+fn tests_from_dir(dir: &Path) -> HashSet<Test> {
+    let mut res = HashSet::new();
+    for entry in WalkDir::new(dir) {
+        let entry = entry.unwrap();
+        if !entry.file_type().is_file() {
+            continue;
+        }
+        if entry.path().extension().unwrap_or_default() != "rs" {
+            continue;
+        }
+        let text = file::get_text(entry.path()).unwrap();
+
+        for test in collect_tests(&text) {
+            if let Some(old_test) = res.replace(test) {
+                panic!("Duplicate test: {}", old_test.name)
+            }
+        }
+    }
+    res
+}
+
+fn collect_tests(s: &str) -> Vec<Test> {
+    let mut res = vec![];
+    let prefix = "// ";
+    let comment_blocks = s.lines()
+        .map(str::trim_left)
+        .group_by(|line| line.starts_with(prefix));
+
+    for (is_comment, block) in comment_blocks.into_iter() {
+        if !is_comment {
+            continue;
+        }
+        let mut block = block.map(|line| &line[prefix.len()..]);
+        let first = block.next().unwrap();
+        if !first.starts_with("test ") {
+            continue;
+        }
+        let name = first["test ".len()..].to_string();
+        let text: String = itertools::join(block.chain(::std::iter::once("")), "\n");
+        assert!(!text.trim().is_empty() && text.ends_with("\n"));
+        res.push(Test { name, text })
+    }
+    res
+}
+
+fn existing_tests() -> HashSet<Test> {
+    let mut res = HashSet::new();
+    for file in fs::read_dir(&inline_tests_dir()).unwrap() {
+        let file = file.unwrap();
+        let path = file.path();
+        if path.extension().unwrap_or_default() != "rs" {
+            continue;
+        }
+        let name = path.file_name().unwrap().to_str().unwrap();
+        let name = name["0000_".len()..name.len() - 3].to_string();
+        let text = file::get_text(&path).unwrap();
+        res.insert(Test { name, text });
+    }
+    res
+}
+
+fn inline_tests_dir() -> PathBuf {
+    let res = base_dir().join("tests/data/parser/inline");
+    if !res.is_dir() {
+        fs::create_dir_all(&res).unwrap();
+    }
+    res
+}
+
+fn grammar_dir() -> PathBuf {
+    base_dir().join("src/parser/event_parser/grammar")
+}
+
+fn base_dir() -> PathBuf {
+    let dir = env!("CARGO_MANIFEST_DIR");
+    PathBuf::from(dir).parent().unwrap().to_owned()
+}