Merge pull request #471 from ratmice/clone_parse_param

Relax `%parse-param` bounds from `Copy` to `Clone`

Cargo.toml

@@ -11,6 +11,7 @@ members=[
     "lrpar/examples/calc_ast",
     "lrpar/examples/calc_parsetree",
     "lrpar/examples/start_states",
+    "lrpar/examples/mut_param",
     "lrtable",
     "nimbleparse",
 ]

doc/src/actioncode.md

@@ -51,9 +51,9 @@ make use of the following:
 
 A single extra parameter can be passed to action functions if the `%parse-param
 <var>: <type>` declaration is used. The variable `<var>` is then visible in all
-action code. `<type>` must implement the [`Copy`
-trait](https://doc.rust-lang.org/std/marker/trait.Copy.html) (note that `&`
-references implement `Copy`).
+action code. `<type>` must implement the [`Clone`
+trait](https://doc.rust-lang.org/stable/std/clone/trait.Clone.html) (note that `Copy`
+bounds imply `Clone`, and `&` references implement `Copy`).
 
 For example if a grammar has a declaration:
 

lrpar/cttests/src/lib.rs

@@ -34,6 +34,9 @@ lrpar_mod!("multitypes.y");
 lrlex_mod!("parseparam.l");
 lrpar_mod!("parseparam.y");
 
+lrlex_mod!("parseparam_copy.l");
+lrpar_mod!("parseparam_copy.y");
+
 lrlex_mod!("passthrough.l");
 lrpar_mod!("passthrough.y");
 
@@ -247,6 +250,16 @@ fn test_parseparam() {
     }
 }
 
+#[test]
+fn test_parseparam_copy() {
+    let lexerdef = parseparam_copy_l::lexerdef();
+    let lexer = lexerdef.lexer("101");
+    match parseparam_copy_y::parse(&lexer, 3) {
+        (Some(104), _) => (),
+        _ => unreachable!(),
+    }
+}
+
 #[test]
 fn test_passthrough() {
     let lexerdef = passthrough_l::lexerdef();

lrpar/cttests/src/parseparam_copy.test

@@ -0,0 +1,18 @@
+name: Test %parse-param copy
+yacckind: Grmtools
+grammar: |
+    %start S
+    %parse-param p: u64
+    %%
+    S -> u64:
+        // Previously %parse-param required a `Copy` bounds.
+        // Since then we relaxed the bounds to require `Clone`.
+        // This tests backwards compatibility of actions that
+        // rely on the older copy bounds.
+        'INT' { (move |_| {})(p); check_copy(p); p + $lexer.span_str($1.unwrap().span()).parse::<u64>().unwrap() }
+    ;
+    %%
+    fn check_copy<T: Copy>(_: T){}
+lexer: |
+    %%
+    [0-9]+ 'INT'

lrpar/examples/clone_param/Cargo.toml

@@ -0,0 +1,20 @@
+[package]
+name = "mut_param"
+version = "0.1.0"
+authors = ["Matt Rice <[email protected]>"]
+edition = "2021"
+license = "Apache-2.0/MIT"
+
+[[bin]]
+doc = false
+name = "mut_param"
+
+[build-dependencies]
+cfgrammar = { path="../../../cfgrammar" }
+lrlex = { path="../../../lrlex" }
+lrpar = { path="../.." }
+
+[dependencies]
+cfgrammar = { path="../../../cfgrammar" }
+lrlex = { path="../../../lrlex" }
+lrpar = { path="../.." }

lrpar/examples/clone_param/README.md

@@ -0,0 +1,20 @@
+# `clone_param`
+
+## Description
+Example which shows how to use interior mutability with the `%parse-param` directive.
+As a parameter the parse function accepts a `Rc<RefCell<u64>>`.
+
+## Input 
+For input the parser accepts a positive or negative integer e.g. `-1`, `42`, etc followed
+by any sequence of `+`, or `-` characters. Except for the initial `-` on a negative integer,
+`+` or `-` are treated as `Increment` and `Decrement` operators.
+
+## Evaluation
+Rather than building an AST, the param is directly mutated by the actions.
+As such an input sequence like `-3++-` will evalute to `-2`.
+
+## Example 
+```
+>>> -3++-
+Evaluated: RefCell { value: -2 }
+```

lrpar/examples/clone_param/build.rs

@@ -0,0 +1,20 @@
+#![deny(rust_2018_idioms)]
+use cfgrammar::yacc::YaccKind;
+use lrlex::CTLexerBuilder;
+
+fn main() {
+    // Since we're using both lrlex and lrpar, we use lrlex's `lrpar_config` convenience function
+    // that makes it easy to a) create a lexer and parser and b) link them together.
+    CTLexerBuilder::new()
+        .rust_edition(lrlex::RustEdition::Rust2021)
+        .lrpar_config(|ctp| {
+            ctp.yacckind(YaccKind::Grmtools)
+                .rust_edition(lrpar::RustEdition::Rust2021)
+                .grammar_in_src_dir("param.y")
+                .unwrap()
+        })
+        .lexer_in_src_dir("param.l")
+        .unwrap()
+        .build()
+        .unwrap();
+}

lrpar/examples/clone_param/src/main.rs

@@ -0,0 +1,40 @@
+#![allow(clippy::unnecessary_wraps)]
+
+use std::io::{self, BufRead, Write};
+use std::{ rc::Rc, cell::RefCell };
+use lrlex::lrlex_mod;
+use lrpar::lrpar_mod;
+
+// Using `lrlex_mod!` brings the lexer for `param.l` into scope. By default the module name will be
+// `param_l` (i.e. the file name, minus any extensions, with a suffix of `_l`).
+lrlex_mod!("param.l");
+// Using `lrpar_mod!` brings the parser for `param.y` into scope. By default the module name will be
+// `param_y` (i.e. the file name, minus any extensions, with a suffix of `_y`).
+lrpar_mod!("param.y");
+
+fn main() {
+    // Get the `LexerDef` for the `param` language.
+    let lexerdef = param_l::lexerdef();
+    let stdin = io::stdin();
+    loop {
+        print!(">>> ");
+        io::stdout().flush().ok();
+        match stdin.lock().lines().next() {
+            Some(Ok(ref l)) => {
+                if l.trim().is_empty() {
+                    continue;
+                }
+                // Now we create a lexer with the `lexer` method with which we can lex an input.
+                let lexer = lexerdef.lexer(l);
+                let param = Rc::new(RefCell::new(0));
+                // Pass the lexer to the parser and lex and parse the input.
+                let (_opt , errs) = param_y::parse(&lexer, param.clone());
+                for e in errs {
+                    println!("{}", e.pp(&lexer, &param_y::token_epp));
+                }
+                println!("Evaluated: {:?}", &param); 
+            }
+            _ => break,
+        }
+    }
+}

lrpar/examples/clone_param/src/param.l

@@ -0,0 +1,5 @@
+%%
+(\-?)[0-9]+ "INT"
+\- "Decr"
+\+ "Incr"
+[\n\t\ ] ;

lrpar/examples/clone_param/src/param.y

@@ -0,0 +1,21 @@
+%token Incr Decr
+%parse-param val: Rc<RefCell<i64>>
+%%
+Expr -> () : "INT" Ops {
+	*val.borrow_mut() += parse_int($lexer.span_str($1.map_err(|_| "<evaluation aborted>").unwrap().span())).unwrap()
+	};
+Ops -> (): 
+	%empty {}
+	| Ops Incr { *val.borrow_mut() += 1; }
+	| Ops Decr { *val.borrow_mut() -= 1; };
+%%
+use std::{ rc::Rc, cell::RefCell, error::Error };
+
+fn parse_int(s: &str) -> Result<i64, Box<dyn Error>> {
+    match s.parse::<i64>() {
+        Ok(val) => Ok(val),
+        Err(_) => {
+            Err(Box::from(format!("{} cannot be represented as a i64", s)))
+        }
+    }
+}

lrpar/src/lib/cpctplus.rs

@@ -106,7 +106,7 @@ struct CPCTPlus<
     StorageT: 'static + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT: 'a,
-    ParamT: Copy,
+    ParamT: Clone,
 > where
     usize: AsPrimitive<StorageT>,
 {
@@ -118,7 +118,7 @@ pub(super) fn recoverer<
     StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT: 'a,
-    ParamT: Copy,
+    ParamT: Clone,
 >(
     parser: &'a Parser<StorageT, LexerTypesT, ActionT, ParamT>,
 ) -> Box<dyn Recoverer<StorageT, LexerTypesT, ActionT, ParamT> + 'a>
@@ -135,7 +135,7 @@ impl<
         StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
         LexerTypesT: LexerTypes<StorageT = StorageT>,
         ActionT: 'a,
-        ParamT: Copy,
+        ParamT: Clone,
     > Recoverer<StorageT, LexerTypesT, ActionT, ParamT>
     for CPCTPlus<'a, 'b, 'input, StorageT, LexerTypesT, ActionT, ParamT>
 where
@@ -270,7 +270,7 @@ impl<
         StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
         LexerTypesT: LexerTypes<StorageT = StorageT>,
         ActionT: 'a,
-        ParamT: Copy,
+        ParamT: Clone,
     > CPCTPlus<'a, 'b, 'input, StorageT, LexerTypesT, ActionT, ParamT>
 where
     usize: AsPrimitive<StorageT>,
@@ -457,7 +457,7 @@ fn apply_repairs<
     StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT: 'a,
-    ParamT: Copy,
+    ParamT: Clone,
 >(
     parser: &Parser<StorageT, LexerTypesT, ActionT, ParamT>,
     mut laidx: usize,
@@ -496,7 +496,7 @@ fn simplify_repairs<
     StorageT: 'static + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT,
-    ParamT: Copy,
+    ParamT: Clone,
 >(
     parser: &Parser<StorageT, LexerTypesT, ActionT, ParamT>,
     all_rprs: &mut Vec<Vec<ParseRepair<LexerTypesT::LexemeT, StorageT>>>,
@@ -557,7 +557,7 @@ fn rank_cnds<
     StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT: 'a,
-    ParamT: Copy,
+    ParamT: Clone,
 >(
     parser: &Parser<StorageT, LexerTypesT, ActionT, ParamT>,
     finish_by: Instant,

lrpar/src/lib/parser.rs

@@ -88,7 +88,7 @@ pub(super) struct Parser<
     StorageT: 'static + Eq + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT: 'a,
-    ParamT: Copy,
+    ParamT: Clone,
 > where
     usize: AsPrimitive<StorageT>,
 {
@@ -241,7 +241,7 @@ impl<
         StorageT: 'static + Debug + Eq + Hash + PrimInt + Unsigned,
         LexerTypesT: LexerTypes<StorageT = StorageT>,
         ActionT: 'a,
-        ParamT: Copy,
+        ParamT: Clone,
     > Parser<'a, 'b, 'input, StorageT, LexerTypesT, ActionT, ParamT>
 where
     usize: AsPrimitive<StorageT>,
@@ -329,7 +329,7 @@ where
                         self.lexer,
                         span,
                         astack.drain(pop_idx - 1..),
-                        self.param,
+                        self.param.clone(),
                     ));
                     astack.push(v);
                 }
@@ -447,7 +447,7 @@ where
                                 self.lexer,
                                 span,
                                 astack_uw.drain(pop_idx - 1..),
-                                self.param,
+                                self.param.clone(),
                             ));
                             astack_uw.push(v);
                         } else {
@@ -595,7 +595,7 @@ pub(super) trait Recoverer<
     StorageT: 'static + Debug + Hash + PrimInt + Unsigned,
     LexerTypesT: LexerTypes<StorageT = StorageT>,
     ActionT,
-    ParamT: Copy,
+    ParamT: Clone,
 > where
     usize: AsPrimitive<StorageT>,
 {
@@ -877,7 +877,7 @@ where
     /// (`None, [...]`), errors and a value (`Some(...), [...]`), as well as a value and no errors
     /// (`Some(...), []`). Errors are sorted by the position they were found in the input and can
     /// be a mix of lexing and parsing errors.
-    pub fn parse_actions<'b: 'a, 'input: 'b, ActionT: 'a, ParamT: Copy>(
+    pub fn parse_actions<'b: 'a, 'input: 'b, ActionT: 'a, ParamT: Clone>(
         &self,
         lexer: &'b dyn NonStreamingLexer<'input, LexerTypesT>,
         actions: &'a [ActionFn<'a, 'b, 'input, StorageT, LexerTypesT, ActionT, ParamT>],