Auto merge of #29072 - nagisa:place-arrow, r=pnkfelix

This commit generalises parsing of associative operators from left-associative
only (with some ugly hacks to support right-associative assignment) to properly
left/right-associative operators.

Parsing is still is not general enough to handle non-associative,
non-highest-precedence prefix or non-highest-precedence
postfix operators (e.g. `..` range syntax) and should be made to be.

Lastly, this commit adds support for parsing right-associative `<-` (left arrow)
operator with precedence higher than assignment as the operator for placement-in
feature.

---

This PR still needs various non-parser changes (e.g. src/grammar and tests) and I’m still working on these; the meat of the PR can already be reviewed, though, I think.

Please review carefully. I made sure that quirks I have discovered so far are preserved (see e.g. #29071) and am looking for more corner cases as I continue to work on tests et al, but there may be something I haven’t noticed or accounted for.

EDIT: I’m also not sure I managed to preserve all the semantics with the range operator inside non-trivial expressions since these are a mess at the moment. Crater runs would be nice.

Author: bors

src/grammar/RustLexer.g4

@@ -10,7 +10,7 @@ lexer grammar RustLexer;
 tokens {
     EQ, LT, LE, EQEQ, NE, GE, GT, ANDAND, OROR, NOT, TILDE, PLUS,
     MINUS, STAR, SLASH, PERCENT, CARET, AND, OR, SHL, SHR, BINOP,
-    BINOPEQ, AT, DOT, DOTDOT, DOTDOTDOT, COMMA, SEMI, COLON,
+    BINOPEQ, LARROW, AT, DOT, DOTDOT, DOTDOTDOT, COMMA, SEMI, COLON,
     MOD_SEP, RARROW, FAT_ARROW, LPAREN, RPAREN, LBRACKET, RBRACKET,
     LBRACE, RBRACE, POUND, DOLLAR, UNDERSCORE, LIT_CHAR, LIT_BYTE,
     LIT_INTEGER, LIT_FLOAT, LIT_STR, LIT_STR_RAW, LIT_BYTE_STR,
@@ -44,6 +44,7 @@ AND     : '&' ;
 OR      : '|' ;
 SHL     : '<<' ;
 SHR     : '>>' ;
+LARROW  : '<-' ;
 
 BINOP
     : PLUS
@@ -56,6 +57,7 @@ BINOP
     | OR
     | SHL
     | SHR
+    | LARROW
     ;
 
 BINOPEQ : BINOP EQ ;

src/grammar/lexer.l

@@ -317,6 +317,7 @@ r/#             {
 <str>\\[^n\nrt\\\x27\x220]  { return -1; }
 <str>(.|\n)                 { yymore(); }
 
+\<-  { return LARROW; }
 -\>  { return RARROW; }
 -    { return '-'; }
 -=   { return MINUSEQ; }

src/grammar/parser-lalr.y

@@ -45,6 +45,7 @@ extern char *yytext;
 %token DOTDOTDOT
 %token MOD_SEP
 %token RARROW
+%token LARROW
 %token FAT_ARROW
 %token LIT_BYTE
 %token LIT_CHAR
@@ -167,7 +168,8 @@ extern char *yytext;
 // prefix_exprs
 %precedence RETURN
 
-%left '=' SHLEQ SHREQ MINUSEQ ANDEQ OREQ PLUSEQ STAREQ SLASHEQ CARETEQ PERCENTEQ
+%right '=' SHLEQ SHREQ MINUSEQ ANDEQ OREQ PLUSEQ STAREQ SLASHEQ CARETEQ PERCENTEQ
+%right LARROW
 %left OROR
 %left ANDAND
 %left EQEQ NE
@@ -1316,6 +1318,7 @@ nonblock_expr
 | RETURN expr                                                   { $$ = mk_node("ExprRet", 1, $2); }
 | BREAK                                                         { $$ = mk_node("ExprBreak", 0); }
 | BREAK lifetime                                                { $$ = mk_node("ExprBreak", 1, $2); }
+| nonblock_expr LARROW expr                                     { $$ = mk_node("ExprInPlace", 2, $1, $3); }
 | nonblock_expr '=' expr                                        { $$ = mk_node("ExprAssign", 2, $1, $3); }
 | nonblock_expr SHLEQ expr                                      { $$ = mk_node("ExprAssignShl", 2, $1, $3); }
 | nonblock_expr SHREQ expr                                      { $$ = mk_node("ExprAssignShr", 2, $1, $3); }
@@ -1375,6 +1378,7 @@ expr
 | RETURN expr                                         { $$ = mk_node("ExprRet", 1, $2); }
 | BREAK                                               { $$ = mk_node("ExprBreak", 0); }
 | BREAK ident                                         { $$ = mk_node("ExprBreak", 1, $2); }
+| expr LARROW expr                                    { $$ = mk_node("ExprInPlace", 2, $1, $3); }
 | expr '=' expr                                       { $$ = mk_node("ExprAssign", 2, $1, $3); }
 | expr SHLEQ expr                                     { $$ = mk_node("ExprAssignShl", 2, $1, $3); }
 | expr SHREQ expr                                     { $$ = mk_node("ExprAssignShr", 2, $1, $3); }
@@ -1435,6 +1439,7 @@ nonparen_expr
 | RETURN expr                                         { $$ = mk_node("ExprRet", 1, $2); }
 | BREAK                                               { $$ = mk_node("ExprBreak", 0); }
 | BREAK ident                                         { $$ = mk_node("ExprBreak", 1, $2); }
+| nonparen_expr LARROW nonparen_expr                  { $$ = mk_node("ExprInPlace", 2, $1, $3); }
 | nonparen_expr '=' nonparen_expr                     { $$ = mk_node("ExprAssign", 2, $1, $3); }
 | nonparen_expr SHLEQ nonparen_expr                   { $$ = mk_node("ExprAssignShl", 2, $1, $3); }
 | nonparen_expr SHREQ nonparen_expr                   { $$ = mk_node("ExprAssignShr", 2, $1, $3); }
@@ -1495,6 +1500,7 @@ expr_nostruct
 | RETURN expr                                         { $$ = mk_node("ExprRet", 1, $2); }
 | BREAK                                               { $$ = mk_node("ExprBreak", 0); }
 | BREAK ident                                         { $$ = mk_node("ExprBreak", 1, $2); }
+| expr_nostruct LARROW expr_nostruct                  { $$ = mk_node("ExprInPlace", 2, $1, $3); }
 | expr_nostruct '=' expr_nostruct                     { $$ = mk_node("ExprAssign", 2, $1, $3); }
 | expr_nostruct SHLEQ expr_nostruct                   { $$ = mk_node("ExprAssignShl", 2, $1, $3); }
 | expr_nostruct SHREQ expr_nostruct                   { $$ = mk_node("ExprAssignShr", 2, $1, $3); }
@@ -1794,6 +1800,7 @@ unpaired_token
 | GE                         { $$ = mk_atom(yytext); }
 | ANDAND                     { $$ = mk_atom(yytext); }
 | OROR                       { $$ = mk_atom(yytext); }
+| LARROW                     { $$ = mk_atom(yytext); }
 | SHLEQ                      { $$ = mk_atom(yytext); }
 | SHREQ                      { $$ = mk_atom(yytext); }
 | MINUSEQ                    { $$ = mk_atom(yytext); }

src/grammar/verify.rs

@@ -35,7 +35,7 @@ use syntax::parse::lexer::TokenAndSpan;
 
 fn parse_token_list(file: &str) -> HashMap<String, token::Token> {
     fn id() -> token::Token {
-        token::Ident(ast::Ident::with_empty_ctxt(Name(0))), token::Plain)
+        token::Ident(ast::Ident::with_empty_ctxt(Name(0)), token::Plain)
     }
 
     let mut res = HashMap::new();
@@ -208,7 +208,7 @@ fn parse_antlr_token(s: &str, tokens: &HashMap<String, token::Token>, surrogate_
         token::Literal(token::ByteStr(..), n)    => token::Literal(token::ByteStr(nm), n),
         token::Literal(token::ByteStrRaw(..), n) => token::Literal(token::ByteStrRaw(fix(content),
                                                                                 count(content)), n),
-        token::Ident(..)           => token::Ident(ast::Ident::with_empty_ctxt(nm)),
+        token::Ident(..)           => token::Ident(ast::Ident::with_empty_ctxt(nm),
                                                    token::ModName),
         token::Lifetime(..)        => token::Lifetime(ast::Ident::with_empty_ctxt(nm)),
         ref t => t.clone()

src/librustc_lint/unused.rs

@@ -369,6 +369,7 @@ impl EarlyLintPass for UnusedParens {
             ast::ExprRet(Some(ref value)) => (value, "`return` value", false),
             ast::ExprAssign(_, ref value) => (value, "assigned value", false),
             ast::ExprAssignOp(_, _, ref value) => (value, "assigned value", false),
+            ast::ExprInPlace(_, ref value) => (value, "emplacement value", false),
             _ => return
         };
         self.check_unused_parens_core(cx, &**value, msg, struct_lit_needs_parens);

src/libsyntax/ast_util.rs

@@ -242,26 +242,6 @@ pub fn struct_field_visibility(field: ast::StructField) -> Visibility {
     }
 }
 
-/// Maps a binary operator to its precedence
-pub fn operator_prec(op: ast::BinOp_) -> usize {
-  match op {
-      // 'as' sits here with 12
-      BiMul | BiDiv | BiRem     => 11,
-      BiAdd | BiSub             => 10,
-      BiShl | BiShr             =>  9,
-      BiBitAnd                  =>  8,
-      BiBitXor                  =>  7,
-      BiBitOr                   =>  6,
-      BiLt | BiLe | BiGe | BiGt | BiEq | BiNe => 3,
-      BiAnd                     =>  2,
-      BiOr                      =>  1
-  }
-}
-
-/// Precedence of the `as` operator, which is a binary operator
-/// not appearing in the prior table.
-pub const AS_PREC: usize = 12;
-
 pub fn empty_generics() -> Generics {
     Generics {
         lifetimes: Vec::new(),

src/libsyntax/lib.rs

@@ -66,6 +66,7 @@ pub mod util {
     #[cfg(test)]
     pub mod parser_testing;
     pub mod small_vector;
+    pub mod parser;
 }
 
 pub mod diagnostics {