diff options
Diffstat (limited to 'src/libsyntax_expand/parse')
| -rw-r--r-- | src/libsyntax_expand/parse/lexer/tests.rs | 277 | ||||
| -rw-r--r-- | src/libsyntax_expand/parse/tests.rs | 338 |
2 files changed, 615 insertions, 0 deletions
diff --git a/src/libsyntax_expand/parse/lexer/tests.rs b/src/libsyntax_expand/parse/lexer/tests.rs new file mode 100644 index 00000000000..d8a1c359c95 --- /dev/null +++ b/src/libsyntax_expand/parse/lexer/tests.rs @@ -0,0 +1,277 @@ +use crate::config::process_configure_mod; + +use rustc_data_structures::sync::Lrc; +use rustc_parse::lexer::StringReader; +use syntax::token::{self, Token, TokenKind}; +use syntax::sess::ParseSess; +use syntax::source_map::{SourceMap, FilePathMapping}; +use syntax::util::comments::is_doc_comment; +use syntax::with_default_globals; +use syntax_pos::symbol::Symbol; +use syntax_pos::{BytePos, Span}; + +use errors::{Handler, emitter::EmitterWriter}; +use std::io; +use std::path::PathBuf; + +fn mk_sess(sm: Lrc<SourceMap>) -> ParseSess { + let emitter = EmitterWriter::new( + Box::new(io::sink()), + Some(sm.clone()), + false, + false, + false, + None, + false, + ); + ParseSess::with_span_handler( + Handler::with_emitter(true, None, Box::new(emitter)), + sm, + process_configure_mod, + ) +} + +// Creates a string reader for the given string. +fn setup<'a>(sm: &SourceMap, + sess: &'a ParseSess, + teststr: String) + -> StringReader<'a> { + let sf = sm.new_source_file(PathBuf::from(teststr.clone()).into(), teststr); + StringReader::new(sess, sf, None) +} + +#[test] +fn t1() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + let mut string_reader = setup( + &sm, + &sh, + "/* my source file */ fn main() { println!(\"zebra\"); }\n".to_string(), + ); + assert_eq!(string_reader.next_token(), token::Comment); + assert_eq!(string_reader.next_token(), token::Whitespace); + let tok1 = string_reader.next_token(); + let tok2 = Token::new( + mk_ident("fn"), + Span::with_root_ctxt(BytePos(21), BytePos(23)), + ); + assert_eq!(tok1.kind, tok2.kind); + assert_eq!(tok1.span, tok2.span); + assert_eq!(string_reader.next_token(), token::Whitespace); + // Read another token. + let tok3 = string_reader.next_token(); + assert_eq!(string_reader.pos.clone(), BytePos(28)); + let tok4 = Token::new( + mk_ident("main"), + Span::with_root_ctxt(BytePos(24), BytePos(28)), + ); + assert_eq!(tok3.kind, tok4.kind); + assert_eq!(tok3.span, tok4.span); + + assert_eq!(string_reader.next_token(), token::OpenDelim(token::Paren)); + assert_eq!(string_reader.pos.clone(), BytePos(29)) + }) +} + +// Checks that the given reader produces the desired stream +// of tokens (stop checking after exhausting `expected`). +fn check_tokenization(mut string_reader: StringReader<'_>, expected: Vec<TokenKind>) { + for expected_tok in &expected { + assert_eq!(&string_reader.next_token(), expected_tok); + } +} + +// Makes the identifier by looking up the string in the interner. +fn mk_ident(id: &str) -> TokenKind { + token::Ident(Symbol::intern(id), false) +} + +fn mk_lit(kind: token::LitKind, symbol: &str, suffix: Option<&str>) -> TokenKind { + TokenKind::lit(kind, Symbol::intern(symbol), suffix.map(Symbol::intern)) +} + +#[test] +fn doublecolon_parsing() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + check_tokenization( + setup(&sm, &sh, "a b".to_string()), + vec![mk_ident("a"), token::Whitespace, mk_ident("b")], + ); + }) +} + +#[test] +fn doublecolon_parsing_2() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + check_tokenization( + setup(&sm, &sh, "a::b".to_string()), + vec![mk_ident("a"), token::Colon, token::Colon, mk_ident("b")], + ); + }) +} + +#[test] +fn doublecolon_parsing_3() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + check_tokenization( + setup(&sm, &sh, "a ::b".to_string()), + vec![mk_ident("a"), token::Whitespace, token::Colon, token::Colon, mk_ident("b")], + ); + }) +} + +#[test] +fn doublecolon_parsing_4() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + check_tokenization( + setup(&sm, &sh, "a:: b".to_string()), + vec![mk_ident("a"), token::Colon, token::Colon, token::Whitespace, mk_ident("b")], + ); + }) +} + +#[test] +fn character_a() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + assert_eq!( + setup(&sm, &sh, "'a'".to_string()).next_token(), + mk_lit(token::Char, "a", None), + ); + }) +} + +#[test] +fn character_space() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + assert_eq!( + setup(&sm, &sh, "' '".to_string()).next_token(), + mk_lit(token::Char, " ", None), + ); + }) +} + +#[test] +fn character_escaped() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + assert_eq!( + setup(&sm, &sh, "'\\n'".to_string()).next_token(), + mk_lit(token::Char, "\\n", None), + ); + }) +} + +#[test] +fn lifetime_name() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + assert_eq!( + setup(&sm, &sh, "'abc".to_string()).next_token(), + token::Lifetime(Symbol::intern("'abc")), + ); + }) +} + +#[test] +fn raw_string() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + assert_eq!( + setup(&sm, &sh, "r###\"\"#a\\b\x00c\"\"###".to_string()).next_token(), + mk_lit(token::StrRaw(3), "\"#a\\b\x00c\"", None), + ); + }) +} + +#[test] +fn literal_suffixes() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + macro_rules! test { + ($input: expr, $tok_type: ident, $tok_contents: expr) => {{ + assert_eq!( + setup(&sm, &sh, format!("{}suffix", $input)).next_token(), + mk_lit(token::$tok_type, $tok_contents, Some("suffix")), + ); + // with a whitespace separator + assert_eq!( + setup(&sm, &sh, format!("{} suffix", $input)).next_token(), + mk_lit(token::$tok_type, $tok_contents, None), + ); + }} + } + + test!("'a'", Char, "a"); + test!("b'a'", Byte, "a"); + test!("\"a\"", Str, "a"); + test!("b\"a\"", ByteStr, "a"); + test!("1234", Integer, "1234"); + test!("0b101", Integer, "0b101"); + test!("0xABC", Integer, "0xABC"); + test!("1.0", Float, "1.0"); + test!("1.0e10", Float, "1.0e10"); + + assert_eq!( + setup(&sm, &sh, "2us".to_string()).next_token(), + mk_lit(token::Integer, "2", Some("us")), + ); + assert_eq!( + setup(&sm, &sh, "r###\"raw\"###suffix".to_string()).next_token(), + mk_lit(token::StrRaw(3), "raw", Some("suffix")), + ); + assert_eq!( + setup(&sm, &sh, "br###\"raw\"###suffix".to_string()).next_token(), + mk_lit(token::ByteStrRaw(3), "raw", Some("suffix")), + ); + }) +} + +#[test] +fn line_doc_comments() { + assert!(is_doc_comment("///")); + assert!(is_doc_comment("/// blah")); + assert!(!is_doc_comment("////")); +} + +#[test] +fn nested_block_comments() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + let mut lexer = setup(&sm, &sh, "/* /* */ */'a'".to_string()); + assert_eq!(lexer.next_token(), token::Comment); + assert_eq!(lexer.next_token(), mk_lit(token::Char, "a", None)); + }) +} + +#[test] +fn crlf_comments() { + with_default_globals(|| { + let sm = Lrc::new(SourceMap::new(FilePathMapping::empty())); + let sh = mk_sess(sm.clone()); + let mut lexer = setup(&sm, &sh, "// test\r\n/// test\r\n".to_string()); + let comment = lexer.next_token(); + assert_eq!(comment.kind, token::Comment); + assert_eq!((comment.span.lo(), comment.span.hi()), (BytePos(0), BytePos(7))); + assert_eq!(lexer.next_token(), token::Whitespace); + assert_eq!(lexer.next_token(), token::DocComment(Symbol::intern("/// test"))); + }) +} diff --git a/src/libsyntax_expand/parse/tests.rs b/src/libsyntax_expand/parse/tests.rs new file mode 100644 index 00000000000..60911617995 --- /dev/null +++ b/src/libsyntax_expand/parse/tests.rs @@ -0,0 +1,338 @@ +use crate::config::process_configure_mod; +use crate::tests::{matches_codepattern, string_to_stream, with_error_checking_parse}; + +use rustc_parse::new_parser_from_source_str; +use syntax::ast::{self, Name, PatKind}; +use syntax::attr::first_attr_value_str_by_name; +use syntax::sess::ParseSess; +use syntax::token::{self, Token}; +use syntax::print::pprust::item_to_string; +use syntax::ptr::P; +use syntax::source_map::FilePathMapping; +use syntax::symbol::{kw, sym}; +use syntax::tokenstream::{DelimSpan, TokenTree, TokenStream}; +use syntax::visit; +use syntax::with_default_globals; +use syntax_pos::{Span, BytePos, Pos, FileName}; +use errors::PResult; + +use std::path::PathBuf; + +fn sess() -> ParseSess { + ParseSess::new(FilePathMapping::empty(), process_configure_mod) +} + +/// Parses an item. +/// +/// Returns `Ok(Some(item))` when successful, `Ok(None)` when no item was found, and `Err` +/// when a syntax error occurred. +fn parse_item_from_source_str(name: FileName, source: String, sess: &ParseSess) + -> PResult<'_, Option<P<ast::Item>>> { + new_parser_from_source_str(sess, name, source).parse_item() +} + +// Produces a `syntax_pos::span`. +fn sp(a: u32, b: u32) -> Span { + Span::with_root_ctxt(BytePos(a), BytePos(b)) +} + +/// Parses a string, return an expression. +fn string_to_expr(source_str : String) -> P<ast::Expr> { + with_error_checking_parse(source_str, &sess(), |p| p.parse_expr()) +} + +/// Parses a string, returns an item. +fn string_to_item(source_str : String) -> Option<P<ast::Item>> { + with_error_checking_parse(source_str, &sess(), |p| p.parse_item()) +} + +#[should_panic] +#[test] fn bad_path_expr_1() { + with_default_globals(|| { + string_to_expr("::abc::def::return".to_string()); + }) +} + +// Checks the token-tree-ization of macros. +#[test] +fn string_to_tts_macro () { + with_default_globals(|| { + let tts: Vec<_> = + string_to_stream("macro_rules! zip (($a)=>($a))".to_string()).trees().collect(); + let tts: &[TokenTree] = &tts[..]; + + match tts { + [ + TokenTree::Token(Token { kind: token::Ident(name_macro_rules, false), .. }), + TokenTree::Token(Token { kind: token::Not, .. }), + TokenTree::Token(Token { kind: token::Ident(name_zip, false), .. }), + TokenTree::Delimited(_, macro_delim, macro_tts) + ] + if name_macro_rules == &sym::macro_rules && name_zip.as_str() == "zip" => { + let tts = ¯o_tts.trees().collect::<Vec<_>>(); + match &tts[..] { + [ + TokenTree::Delimited(_, first_delim, first_tts), + TokenTree::Token(Token { kind: token::FatArrow, .. }), + TokenTree::Delimited(_, second_delim, second_tts), + ] + if macro_delim == &token::Paren => { + let tts = &first_tts.trees().collect::<Vec<_>>(); + match &tts[..] { + [ + TokenTree::Token(Token { kind: token::Dollar, .. }), + TokenTree::Token(Token { kind: token::Ident(name, false), .. }), + ] + if first_delim == &token::Paren && name.as_str() == "a" => {}, + _ => panic!("value 3: {:?} {:?}", first_delim, first_tts), + } + let tts = &second_tts.trees().collect::<Vec<_>>(); + match &tts[..] { + [ + TokenTree::Token(Token { kind: token::Dollar, .. }), + TokenTree::Token(Token { kind: token::Ident(name, false), .. }), + ] + if second_delim == &token::Paren && name.as_str() == "a" => {}, + _ => panic!("value 4: {:?} {:?}", second_delim, second_tts), + } + }, + _ => panic!("value 2: {:?} {:?}", macro_delim, macro_tts), + } + }, + _ => panic!("value: {:?}",tts), + } + }) +} + +#[test] +fn string_to_tts_1() { + with_default_globals(|| { + let tts = string_to_stream("fn a (b : i32) { b; }".to_string()); + + let expected = TokenStream::new(vec![ + TokenTree::token(token::Ident(kw::Fn, false), sp(0, 2)).into(), + TokenTree::token(token::Ident(Name::intern("a"), false), sp(3, 4)).into(), + TokenTree::Delimited( + DelimSpan::from_pair(sp(5, 6), sp(13, 14)), + token::DelimToken::Paren, + TokenStream::new(vec![ + TokenTree::token(token::Ident(Name::intern("b"), false), sp(6, 7)).into(), + TokenTree::token(token::Colon, sp(8, 9)).into(), + TokenTree::token(token::Ident(sym::i32, false), sp(10, 13)).into(), + ]).into(), + ).into(), + TokenTree::Delimited( + DelimSpan::from_pair(sp(15, 16), sp(20, 21)), + token::DelimToken::Brace, + TokenStream::new(vec![ + TokenTree::token(token::Ident(Name::intern("b"), false), sp(17, 18)).into(), + TokenTree::token(token::Semi, sp(18, 19)).into(), + ]).into(), + ).into() + ]); + + assert_eq!(tts, expected); + }) +} + +#[test] fn parse_use() { + with_default_globals(|| { + let use_s = "use foo::bar::baz;"; + let vitem = string_to_item(use_s.to_string()).unwrap(); + let vitem_s = item_to_string(&vitem); + assert_eq!(&vitem_s[..], use_s); + + let use_s = "use foo::bar as baz;"; + let vitem = string_to_item(use_s.to_string()).unwrap(); + let vitem_s = item_to_string(&vitem); + assert_eq!(&vitem_s[..], use_s); + }) +} + +#[test] fn parse_extern_crate() { + with_default_globals(|| { + let ex_s = "extern crate foo;"; + let vitem = string_to_item(ex_s.to_string()).unwrap(); + let vitem_s = item_to_string(&vitem); + assert_eq!(&vitem_s[..], ex_s); + + let ex_s = "extern crate foo as bar;"; + let vitem = string_to_item(ex_s.to_string()).unwrap(); + let vitem_s = item_to_string(&vitem); + assert_eq!(&vitem_s[..], ex_s); + }) +} + +fn get_spans_of_pat_idents(src: &str) -> Vec<Span> { + let item = string_to_item(src.to_string()).unwrap(); + + struct PatIdentVisitor { + spans: Vec<Span> + } + impl<'a> visit::Visitor<'a> for PatIdentVisitor { + fn visit_pat(&mut self, p: &'a ast::Pat) { + match p.kind { + PatKind::Ident(_ , ref ident, _) => { + self.spans.push(ident.span.clone()); + } + _ => { + visit::walk_pat(self, p); + } + } + } + } + let mut v = PatIdentVisitor { spans: Vec::new() }; + visit::walk_item(&mut v, &item); + return v.spans; +} + +#[test] fn span_of_self_arg_pat_idents_are_correct() { + with_default_globals(|| { + + let srcs = ["impl z { fn a (&self, &myarg: i32) {} }", + "impl z { fn a (&mut self, &myarg: i32) {} }", + "impl z { fn a (&'a self, &myarg: i32) {} }", + "impl z { fn a (self, &myarg: i32) {} }", + "impl z { fn a (self: Foo, &myarg: i32) {} }", + ]; + + for &src in &srcs { + let spans = get_spans_of_pat_idents(src); + let (lo, hi) = (spans[0].lo(), spans[0].hi()); + assert!("self" == &src[lo.to_usize()..hi.to_usize()], + "\"{}\" != \"self\". src=\"{}\"", + &src[lo.to_usize()..hi.to_usize()], src) + } + }) +} + +#[test] fn parse_exprs () { + with_default_globals(|| { + // just make sure that they parse.... + string_to_expr("3 + 4".to_string()); + string_to_expr("a::z.froob(b,&(987+3))".to_string()); + }) +} + +#[test] fn attrs_fix_bug () { + with_default_globals(|| { + string_to_item("pub fn mk_file_writer(path: &Path, flags: &[FileFlag]) + -> Result<Box<Writer>, String> { +#[cfg(windows)] +fn wb() -> c_int { + (O_WRONLY | libc::consts::os::extra::O_BINARY) as c_int +} + +#[cfg(unix)] +fn wb() -> c_int { O_WRONLY as c_int } + +let mut fflags: c_int = wb(); +}".to_string()); + }) +} + +#[test] fn crlf_doc_comments() { + with_default_globals(|| { + let sess = sess(); + + let name_1 = FileName::Custom("crlf_source_1".to_string()); + let source = "/// doc comment\r\nfn foo() {}".to_string(); + let item = parse_item_from_source_str(name_1, source, &sess) + .unwrap().unwrap(); + let doc = first_attr_value_str_by_name(&item.attrs, sym::doc).unwrap(); + assert_eq!(doc.as_str(), "/// doc comment"); + + let name_2 = FileName::Custom("crlf_source_2".to_string()); + let source = "/// doc comment\r\n/// line 2\r\nfn foo() {}".to_string(); + let item = parse_item_from_source_str(name_2, source, &sess) + .unwrap().unwrap(); + let docs = item.attrs.iter().filter(|a| a.has_name(sym::doc)) + .map(|a| a.value_str().unwrap().to_string()).collect::<Vec<_>>(); + let b: &[_] = &["/// doc comment".to_string(), "/// line 2".to_string()]; + assert_eq!(&docs[..], b); + + let name_3 = FileName::Custom("clrf_source_3".to_string()); + let source = "/** doc comment\r\n * with CRLF */\r\nfn foo() {}".to_string(); + let item = parse_item_from_source_str(name_3, source, &sess).unwrap().unwrap(); + let doc = first_attr_value_str_by_name(&item.attrs, sym::doc).unwrap(); + assert_eq!(doc.as_str(), "/** doc comment\n * with CRLF */"); + }); +} + +#[test] +fn ttdelim_span() { + fn parse_expr_from_source_str( + name: FileName, source: String, sess: &ParseSess + ) -> PResult<'_, P<ast::Expr>> { + new_parser_from_source_str(sess, name, source).parse_expr() + } + + with_default_globals(|| { + let sess = sess(); + let expr = parse_expr_from_source_str(PathBuf::from("foo").into(), + "foo!( fn main() { body } )".to_string(), &sess).unwrap(); + + let tts: Vec<_> = match expr.kind { + ast::ExprKind::Mac(ref mac) => mac.stream().trees().collect(), + _ => panic!("not a macro"), + }; + + let span = tts.iter().rev().next().unwrap().span(); + + match sess.source_map().span_to_snippet(span) { + Ok(s) => assert_eq!(&s[..], "{ body }"), + Err(_) => panic!("could not get snippet"), + } + }); +} + +// This tests that when parsing a string (rather than a file) we don't try +// and read in a file for a module declaration and just parse a stub. +// See `recurse_into_file_modules` in the parser. +#[test] +fn out_of_line_mod() { + with_default_globals(|| { + let item = parse_item_from_source_str( + PathBuf::from("foo").into(), + "mod foo { struct S; mod this_does_not_exist; }".to_owned(), + &sess(), + ).unwrap().unwrap(); + + if let ast::ItemKind::Mod(ref m) = item.kind { + assert!(m.items.len() == 2); + } else { + panic!(); + } + }); +} + +#[test] +fn eqmodws() { + assert_eq!(matches_codepattern("",""),true); + assert_eq!(matches_codepattern("","a"),false); + assert_eq!(matches_codepattern("a",""),false); + assert_eq!(matches_codepattern("a","a"),true); + assert_eq!(matches_codepattern("a b","a \n\t\r b"),true); + assert_eq!(matches_codepattern("a b ","a \n\t\r b"),true); + assert_eq!(matches_codepattern("a b","a \n\t\r b "),false); + assert_eq!(matches_codepattern("a b","a b"),true); + assert_eq!(matches_codepattern("ab","a b"),false); + assert_eq!(matches_codepattern("a b","ab"),true); + assert_eq!(matches_codepattern(" a b","ab"),true); +} + +#[test] +fn pattern_whitespace() { + assert_eq!(matches_codepattern("","\x0C"), false); + assert_eq!(matches_codepattern("a b ","a \u{0085}\n\t\r b"),true); + assert_eq!(matches_codepattern("a b","a \u{0085}\n\t\r b "),false); +} + +#[test] +fn non_pattern_whitespace() { + // These have the property 'White_Space' but not 'Pattern_White_Space' + assert_eq!(matches_codepattern("a b","a\u{2002}b"), false); + assert_eq!(matches_codepattern("a b","a\u{2002}b"), false); + assert_eq!(matches_codepattern("\u{205F}a b","ab"), false); + assert_eq!(matches_codepattern("a \u{3000}b","ab"), false); +} |