2 releases (1 stable)
| 1.0.0 | Sep 27, 2020 |
|---|---|
| 0.1.0 | May 23, 2020 |
#1194 in Development tools
130KB
2K
SLoC
dpr
Evolution of... dynparser
Basic execution flow
Text -> Parsing -> Transform -> Text
More info about the peg syntax bellow.
Usage
Add to cargo.toml
[dependencies]
dpr = "0.1.0"
# dpr = {git = "https://github.com/jleahred/dpr" }
Wach examples below
Modifications
0.1.0 First version
TODO
- Adding external functions
- don't needed to be multiexpr pub(crate)struct Transf2Expr { pub(crate)mexpr: MultiExpr,
- remove and or multiexpr when only one option (and/or)
About
Giveng a peg grammar extended, it will verify the input and can generate an output based on transformation rules
But let's see by examples
Simple example
Starting with this peg
Peg:
main = char+
char = 'a' -> A
/ 'b' -> B
/ .
Given this input
Input:
aaacbbabdef
We got as result:
Output:
AAAcBBABdef
Addition calculator example
Peg:
main = expr
expr = num:num -> PUSH $(num)$(:endl)
(op:op expr:expr)? -> $(expr)EXEC $(op)$(:endl)
op = '+' -> ADD
/ '-' -> SUB
num = [0-9]+ ('.' [0-9])?
Input:
1+2-3
Output:
PUSH 1
PUSH 2
PUSH 3
EXEC SUB
EXEC ADD
Execution flow
Basic text trasnformation flow.
DSL flow
.--------.
| peg |
| user |
'--------'
|
v
.--------.
| GEN |
| rules |
'--------'
| .----------.
| | input |
| | user |
| '----------'
| |
| v
| .----------.
| | parse |
'--------------->| |
'----------'
|
v
.---------.
| replace |
| |
'---------'
|
v
.--------.
| OUTPUT |
| |
'--------'
The rust code for first example...
extern crate dpr;
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
"
main = char+
char = 'a' -> A
/ 'b' -> B
/ .
",
)
.gen_rules()?
.parse("aaacbbabdef")?
.replace()?
// ...
;
println!("{:#?}", result);
Ok(())
}
PEG rules grammar
You saw some examples, let see in detail
| token | Description |
|---|---|
= |
On left, symbol, on right expresion defining symbol |
symbol |
It's an string without quotes, no spaces, and ascii |
. |
Any char |
"..." |
Literal delimited by quotes |
<space> |
Separate tokens and Rule concatenation (and operation) |
/ |
Or operation |
(...) |
A expression composed of sub expresions |
? |
One optional |
* |
Repeat 0 or more |
+ |
Repeat 1 or more |
! |
negate expression, continue if not followed without consume |
& |
verify it follows..., but not consuming |
[...] |
Match chars. It's a list or ranges (or both) |
-> |
after the arrow, we have the transformation rule |
: |
To give a name, in order to use later in transformation |
| error(...) | This let's you to define an error message when this rule is satisfied |
Below there is the grammar witch define the valid peg inputs.
BTW, this grammar has been parsed to generate the code to parse itself ;-)
Let's see by example
Rules by example
A simple literal string.
main = "Hello world"
Concatenation (and)
main = "Hello " "world"
Referencing symbols
Symbol
main = hi
hi = "Hello world"
Or conditions /
main = "hello" / "hi"
Or multiline
main
= "hello"
/ "hi"
/ "hola"
Or multiline 2
main = "hello"
/ "hi"
/ "hola"
Or disorganized
main = "hello"
/ "hi" / "hola"
Parenthesis
main = ("hello" / "hi") " world"
Just multiline
Multiline1
main
= ("hello" / "hi") " world"
Multiline2
main
= ("hello" / "hi")
" world"
Multiline3
main = ("hello" / "hi")
" world"
It is recomended to use or operator / on each new line and = on first line, like
Multiline organized
main = ("hello" / "hi") " world"
/ "bye"
One optional
main = ("hello" / "hi") " world"?
Repetitions
main = one_or_more_a / zero_or_many_b
one_or_more = "a"+
zero_or_many = "b"*
Negation will not move current possition
Next example will consume all chars till get an "a"
Negation
main = (!"a" .)* "a"
Consume till
comment = "//" (!"\n" .)*
/ "/*" (!"*/" .)* "*/"
Match a set of chars. Chars can be defined by range.
number = digit+ ("." digit+)?
digit = [0-9]
a_or_b = [ab]
id = [_a-zA-Z][_a-zA-Z0-9]*
a_or_b_or_digit = [ab0-9]
Simple recursion
one or more "a" recursive
as = "a" as
/ "a"
// simplified with `+`
ak = "a"+
Recursion to match parentheses
Recursion match par
match_par = "(" match_par ")"
/ "(" ")"
In order to produce custom errors, you have to use error(...) constructor
In next example, the system will complain with parenthesis error if they are unbalanced
parenth = '(' _ expr _ ( ')'
/ error("unbalanced parethesis: missing ')'")
)
As you can see, if you can run the rule to close properly the parenthesis, everything is OK, in other case, custom error message will be produced
Replacing
You can set the replace rules with ->
op = '+' -> ADD
/ '-' -> SUB
When + will be found and validated, it will be replaced by ADD
expr = num:num -> PUSH $(num)$(:endl)
(op:op expr:expr)? -> $(expr)EXEC $(op)$(:endl)
To refer to parsed chunk, you can name it using :
When refering to a symbol, you don't need to give a name
Next examples, are equivalent
expr = num:num -> PUSH $(num)$(:endl)
(op:op expr:expr)? -> $(expr)EXEC $(op)$(:endl)
expr = num -> PUSH $(num)$(:endl)
(op expr)? -> $(expr)EXEC $(op)$(:endl)
The arrow will work with current line. If you need to use trasnsformations
over some lines, you will have to use (...)
There is a grammar to parse the peg grammars that could be an example on file gcode/peg2code.rs
After the arrow, you will have the transformation rule.
Replacing tokens:
Things inside $(...) will be replaced.
Text outside it, will be written as it
Replacing tokens can refer to parsed text by name or by position.
-> $(num)
This will look for a name called num defined on left side to write it on output
Next line will also look for names, but on rep_symbol will not complain it it doesn't exists
rep_or_unary = atom_or_par rep_symbol? -> $(?rep_symbol)$(atom_or_par)
You can also refer an element by position
-> $(.1)
You can also refer to functions starting the replacing token with :
expr = num -> $(:endl)
Predefined functions are...
(Watch on replace.rs to see full replace functions)
"endl" => "\n",
"spc" => " ",
"_" => " ",
"tab" => "\t",
"(" => "\t",
// "now" => "pending",
_ => "?unknown_fn?",
Example
expr = num -> PUSH $(num)$(:endl)
(op expr)? -> $(.2)EXEC $(.1)$(:endl)
You can define your own functions (aka external functions)
In next example we created the replacement token el
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
"
main = char+
char = 'a' -> $(:el)A
/ 'b' -> $(:el)B
/ ch:. -> $(:el)$(ch)
",
)
.gen_rules()?
.parse("aaacbbabdef")?
.replace(Some(&dpr::FnCallBack(custom_funtions)))?
// ...
;
println!("{:#?}", result);
println!("{}", result.str());
Ok(())
}
fn custom_funtions(fn_txt: &str) -> Option<String> {
match fn_txt {
"el" => Some("\n".to_string()),
_ => None,
}
}
Full math expresion compiler example
What is a parser without an math expresion calculator?
Obiously, it's necessary to consider the operator priority, operator asociativity and parenthesis, and negative numbers and negative expresions
extern crate dpr;
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
r#"
main = expr
expr = term (
_ add_op _ term ->$(term)$(add_op)
)*
term = factor (
_ mult_op _ factor ->$(factor)$(mult_op)
)*
factor = pow (
_ pow_op _ subexpr ->$(subexpr)$(pow_op)
)*
pow = subexpr (
_ pow_op _ pow ->$(pow)$(pow_op)
)*
subexpr = '(' _ expr _ ->$(expr)
( ')' ->$(:none)
/ error("parenthesis error")
)
/ number ->PUSH $(number)$(:endl)
/ '-' _ subexpr ->PUSH 0$(:endl)$(subexpr)SUB$(:endl)
number = ([0-9]+ ('.' [0-9])?)
add_op = '+' ->EXEC ADD$(:endl)
/ '-' ->EXEC SUB$(:endl)
mult_op = '*' ->EXEC MUL$(:endl)
/ '/' ->EXEC DIV$(:endl)
pow_op = '^' ->EXEC POW$(:endl)
_ = ' '*
"#,
)
.gen_rules()?
.parse("-(-1+2* 3^5 ^(- 2 ) -7)+8")?
.replace()?
// ...
;
println!("{:#?}", result);
println!("{}", result.str());
Ok(())
}
The output is a program for a stack machine, composed of a command with a parameter...
PUSH 0
PUSH 0
PUSH 1
EXEC SUB
PUSH 2
PUSH 3
PUSH 5
PUSH 0
PUSH 2
EXEC SUB
EXEC POW
EXEC POW
EXEC MUL
EXEC ADD
PUSH 7
EXEC SUB
EXEC SUB
PUSH 8
EXEC ADD
Full peg grammar doc spec
At the moment it's...
(for an updated reference, open peg2code.rs file :-)
fn text_peg2code() -> &'static str {
r#"
/* A peg grammar to parse peg grammars
*
*/
main = grammar -> $(grammar)EOP
grammar = rule+
symbol = [_a-zA-Z0-9] [_'"a-zA-Z0-9]*
rule = _ rule_name _ '=' _ expr _eol _ -> RULE$(:endl)$(rule_name)$(:endl)$(expr)
rule_name = symbol
expr = or -> OR$(:endl)$(or)CLOSE_MEXPR$(:endl)
or = _ and -> AND$(:endl)$(and)CLOSE_MEXPR$(:endl)
( _ '/' _ or )? -> $(or)
and = error
/ (andline transf2 and:(
_ ->$(:none)
!(rule_name _ ('=' / '{')) and )?) -> TRANSF2$(:endl)$(transf2)EOTRANSF2$(:endl)AND$(:endl)$(andline)CLOSE_MEXPR$(:endl)$(and)
/ andline (
( ' ' / comment )* eol+ _ -> $(:none)
!( rule_name _ ('=' / '{') ) and
)?
error = 'error' _ '(' _ literal _ ')' -> ERROR$(:endl)$(literal)$(:endl)
andline = andchunk (
' '+ ->$(:none)
( error / andchunk )
)*
andchunk = name e:rep_or_unary -> NAMED$(:endl)$(name)$(:endl)$(e)
/ rep_or_unary
// this is the and separator
_1 = ' ' / eol -> $(:none)
// repetitions or unary operator
rep_or_unary = atom_or_par rep_symbol? -> $(?rep_symbol)$(atom_or_par)
// atom_or_par -> $(atom_or_par)
/ '!' atom_or_par -> NEGATE$(:endl)$(atom_or_par)
/ '&' atom_or_par -> PEEK$(:endl)$(atom_or_par)
rep_symbol = '*' -> REPEAT$(:endl)0$(:endl)inf$(:endl)
/ '+' -> REPEAT$(:endl)1$(:endl)inf$(:endl)
/ '?' -> REPEAT$(:endl)0$(:endl)1$(:endl)
atom_or_par = atom / parenth
parenth = '(' _ expr _ -> $(expr)
( ')' -> $(:none)
/ error("unbalanced parethesis: missing ')'")
)
atom = a:literal -> ATOM$(:endl)LIT$(:endl)$(a)$(:endl)
/ a:match -> MATCH$(:endl)$(a)
/ a:rule_name -> ATOM$(:endl)RULREF$(:endl)$(a)$(:endl)
/ dot -> ATOM$(:endl)DOT$(:endl)
// as rule_name can start with a '.', dot has to be after rule_name
literal = lit_noesc / lit_esc
lit_noesc = _' l:( !_' . )* _' -> $(l)
_' = "'"
lit_esc = (_"
l:( esc_char
/ hex_char
/ !_" .
)*
_") -> $(l)
_" = '"'
esc_char = '\r'
/ '\n'
/ '\t'
/ '\\'
/ '\\"'
hex_char = '\0x' [0-9A-F] [0-9A-F]
eol = "\r\n" / "\n" / "\r"
_eol = (' ' / comment)* eol
match = '[' -> $(:none)
(
mchars b:(mbetween*) -> CHARS$(:endl)$(mchars)$(:endl)BETW$(:endl)$(b)EOBETW$(:endl)
/ b:(mbetween+) -> BETW$(:endl)$(b)EOBETW$(:endl)
)
']' -> $(:none)
mchars = (!']' !(. '-') .)+
mbetween = f:. '-' s:. -> $(f)$(:endl)$(s)$(:endl)
dot = '.'
_ = (
( ' '
/ eol
/ comment
)*
) -> $(:none)
comment = ( line_comment
/ mline_comment
) -> $(:none)
line_comment = '//' (!eol .)*
mline_comment = '/*' (!'*/' .)* '*/'
name = symbol ":" -> $(symbol)
transf2 = _1 _ '->' ' '* -> $(:none)
transf_rule -> $(transf_rule)
&eol
transf_rule = ( tmpl_text / tmpl_rule )+
tmpl_text = t:( (!("$(" / eol) .)+ ) -> TEXT$(:endl)$(t)$(:endl)
tmpl_rule = "$(" -> $(:none)
(
// by name optional
'?' symbol ->NAMED_OPT$(:endl)$(symbol)$(:endl)
// by name
/ symbol ->NAMED$(:endl)$(symbol)$(:endl)
// by pos
/ "." pos:([0-9]+) ->POS$(:endl)$(symbol)$(pos)$(:endl)
// by function
/ ":" ->$(:none)
fn:((!(")" / eol) .)+) ->FUNCT$(:endl)$(fn)$(:endl)
)
")" ->$(:none)
"#
}
Hacking the code
As you can see, the code to start parsing the peg input, is written in a text peg file
How is it possible?
At the moment, the rules_for_pegcode is...
r#"symbol"# => or!(and!(ematch!(chlist r#"_"# , from 'a', to 'z' , from 'A', to 'Z' , from '0', to '9' ), rep!(ematch!(chlist r#"_'""# , from 'a', to 'z' , from 'A', to 'Z' , from '0', to '9' ), 0)))
, r#"transf_rule"# => or!(and!(rep!(or!(and!(ref_rule!(r#"tmpl_text"#)), and!(ref_rule!(r#"tmpl_rule"#))), 1)))
, r#"mbetween"# => or!(and!(transf2!( and!( and!(named!("f", dot!()), lit!("-"), named!("s", dot!())) ) , t2rules!(t2_byname!("f"), t2_funct!("endl"), t2_byname!("s"), t2_funct!("endl"), ) )))
, r#"line_comment"# => or!(and!(lit!("//"), rep!(or!(and!(not!(ref_rule!(r#"eol"#)), dot!())), 0)))
, r#"lit_noesc"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"_'"#), named!("l", rep!(or!(and!(not!(ref_rule!(r#"_'"#)), dot!())), 0)), ref_rule!(r#"_'"#)) ) , t2rules!(t2_byname!("l"), ) )))
, r#"error"# => or!(and!(transf2!( and!( and!(lit!("error"), ref_rule!(r#"_"#), lit!("("), ref_rule!(r#"_"#), ref_rule!(r#"literal"#), ref_rule!(r#"_"#), lit!(")")) ) , t2rules!(t2_text!("ERROR"), t2_funct!("endl"), t2_byname!("literal"), t2_funct!("endl"), ) )))
, r#"atom_or_par"# => or!(and!(ref_rule!(r#"atom"#)), and!(ref_rule!(r#"parenth"#)))
, r#"tmpl_text"# => or!(and!(transf2!( and!( and!(named!("t", or!(and!(rep!(or!(and!(not!(or!(and!(lit!("$(")), and!(ref_rule!(r#"eol"#)))), dot!())), 1))))) ) , t2rules!(t2_text!("TEXT"), t2_funct!("endl"), t2_byname!("t"), t2_funct!("endl"), ) )))
, r#"atom"# => or!(and!(transf2!( and!( and!(named!("a", ref_rule!(r#"literal"#))) ) , t2rules!(t2_text!("ATOM"), t2_funct!("endl"), t2_text!("LIT"), t2_funct!("endl"), t2_byname!("a"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(named!("a", ref_rule!(r#"match"#))) ) , t2rules!(t2_text!("MATCH"), t2_funct!("endl"), t2_byname!("a"), ) )), and!(transf2!( and!( and!(named!("a", ref_rule!(r#"rule_name"#))) ) , t2rules!(t2_text!("ATOM"), t2_funct!("endl"), t2_text!("RULREF"), t2_funct!("endl"), t2_byname!("a"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(ref_rule!(r#"dot"#)) ) , t2rules!(t2_text!("ATOM"), t2_funct!("endl"), t2_text!("DOT"), t2_funct!("endl"), ) )))
, r#"grammar"# => or!(and!(rep!(ref_rule!(r#"rule"#), 1)))
, r#"dot"# => or!(and!(lit!(".")))
, r#"eol"# => or!(and!(lit!("\r\n")), and!(lit!("\n")), and!(lit!("\r")))
, r#"expr"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"or"#)) ) , t2rules!(t2_text!("OR"), t2_funct!("endl"), t2_byname!("or"), t2_text!("CLOSE_MEXPR"), t2_funct!("endl"), ) )))
, r#"name"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"symbol"#), lit!(":")) ) , t2rules!(t2_byname!("symbol"), ) )))
, r#"literal"# => or!(and!(ref_rule!(r#"lit_noesc"#)), and!(ref_rule!(r#"lit_esc"#)))
, r#"rule"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"_"#), ref_rule!(r#"rule_name"#), ref_rule!(r#"_"#), lit!("="), ref_rule!(r#"_"#), ref_rule!(r#"expr"#), ref_rule!(r#"_eol"#), ref_rule!(r#"_"#)) ) , t2rules!(t2_text!("RULE"), t2_funct!("endl"), t2_byname!("rule_name"), t2_funct!("endl"), t2_byname!("expr"), ) )))
, r#"andchunk"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"name"#), named!("e", ref_rule!(r#"rep_or_unary"#))) ) , t2rules!(t2_text!("NAMED"), t2_funct!("endl"), t2_byname!("name"), t2_funct!("endl"), t2_byname!("e"), ) )), and!(ref_rule!(r#"rep_or_unary"#)))
, r#"_""# => or!(and!(lit!("\"")))
, r#"mline_comment"# => or!(and!(lit!("/*"), rep!(or!(and!(not!(lit!("*/")), dot!())), 0), lit!("*/")))
, r#"andline"# => or!(and!(ref_rule!(r#"andchunk"#), rep!(or!(and!(transf2!( and!( and!(rep!(lit!(" "), 1)) ) , t2rules!(t2_funct!("none"), ) ), or!(and!(ref_rule!(r#"error"#)), and!(ref_rule!(r#"andchunk"#))))), 0)))
, r#"hex_char"# => or!(and!(lit!("\0x"), ematch!(chlist r#""# , from '0', to '9' , from 'A', to 'F' ), ematch!(chlist r#""# , from '0', to '9' , from 'A', to 'F' )))
, r#"mchars"# => or!(and!(rep!(or!(and!(not!(lit!("]")), not!(or!(and!(dot!(), lit!("-")))), dot!())), 1)))
, r#"transf2"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"_1"#), ref_rule!(r#"_"#), lit!("->"), rep!(lit!(" "), 0)) ) , t2rules!(t2_funct!("none"), ) ), transf2!( and!( and!(ref_rule!(r#"transf_rule"#)) ) , t2rules!(t2_byname!("transf_rule"), ) ), peek!(ref_rule!(r#"eol"#))))
, r#"_'"# => or!(and!(lit!("'")))
, r#"and"# => or!(and!(ref_rule!(r#"error"#)), and!(transf2!( and!( and!(or!(and!(ref_rule!(r#"andline"#), ref_rule!(r#"transf2"#), named!("and", rep!(or!(and!(transf2!( and!( and!(ref_rule!(r#"_"#)) ) , t2rules!(t2_funct!("none"), ) ), not!(or!(and!(ref_rule!(r#"rule_name"#), ref_rule!(r#"_"#), or!(and!(lit!("=")), and!(lit!("{")))))), ref_rule!(r#"and"#))), 0, 1))))) ) , t2rules!(t2_text!("TRANSF2"), t2_funct!("endl"), t2_byname!("transf2"), t2_text!("EOTRANSF2"), t2_funct!("endl"), t2_text!("AND"), t2_funct!("endl"), t2_byname!("andline"), t2_text!("CLOSE_MEXPR"), t2_funct!("endl"), t2_byname!("and"), ) )), and!(ref_rule!(r#"andline"#), rep!(or!(and!(transf2!( and!( and!(rep!(or!(and!(lit!(" ")), and!(ref_rule!(r#"comment"#))), 0), rep!(ref_rule!(r#"eol"#), 1), ref_rule!(r#"_"#)) ) , t2rules!(t2_funct!("none"), ) ), not!(or!(and!(ref_rule!(r#"rule_name"#), ref_rule!(r#"_"#), or!(and!(lit!("=")), and!(lit!("{")))))), ref_rule!(r#"and"#))), 0, 1)))
, r#"_"# => or!(and!(transf2!( and!( and!(or!(and!(rep!(or!(and!(lit!(" ")), and!(ref_rule!(r#"eol"#)), and!(ref_rule!(r#"comment"#))), 0)))) ) , t2rules!(t2_funct!("none"), ) )))
, r#"or"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"_"#), ref_rule!(r#"and"#)) ) , t2rules!(t2_text!("AND"), t2_funct!("endl"), t2_byname!("and"), t2_text!("CLOSE_MEXPR"), t2_funct!("endl"), ) ), transf2!( and!( and!(rep!(or!(and!(ref_rule!(r#"_"#), lit!("/"), ref_rule!(r#"_"#), ref_rule!(r#"or"#))), 0, 1)) ) , t2rules!(t2_byname!("or"), ) )))
, r#"_1"# => or!(and!(lit!(" ")), and!(transf2!( and!( and!(ref_rule!(r#"eol"#)) ) , t2rules!(t2_funct!("none"), ) )))
, r#"lit_esc"# => or!(and!(transf2!( and!( and!(or!(and!(ref_rule!(r#"_""#), named!("l", rep!(or!(and!(ref_rule!(r#"esc_char"#)), and!(ref_rule!(r#"hex_char"#)), and!(not!(ref_rule!(r#"_""#)), dot!())), 0)), ref_rule!(r#"_""#)))) ) , t2rules!(t2_byname!("l"), ) )))
, r#"rep_symbol"# => or!(and!(transf2!( and!( and!(lit!("*")) ) , t2rules!(t2_text!("REPEAT"), t2_funct!("endl"), t2_text!("0"), t2_funct!("endl"), t2_text!("inf"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(lit!("+")) ) , t2rules!(t2_text!("REPEAT"), t2_funct!("endl"), t2_text!("1"), t2_funct!("endl"), t2_text!("inf"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(lit!("?")) ) , t2rules!(t2_text!("REPEAT"), t2_funct!("endl"), t2_text!("0"), t2_funct!("endl"), t2_text!("1"), t2_funct!("endl"), ) )))
, r#"parenth"# => or!(and!(transf2!( and!( and!(lit!("("), ref_rule!(r#"_"#), ref_rule!(r#"expr"#), ref_rule!(r#"_"#)) ) , t2rules!(t2_byname!("expr"), ) ), or!(and!(transf2!( and!( and!(lit!(")")) ) , t2rules!(t2_funct!("none"), ) )), and!(error!("unbalanced parethesis: missing ')'")))))
, r#"comment"# => or!(and!(transf2!( and!( and!(or!(and!(ref_rule!(r#"line_comment"#)), and!(ref_rule!(r#"mline_comment"#)))) ) , t2rules!(t2_funct!("none"), ) )))
, r#"_eol"# => or!(and!(rep!(or!(and!(lit!(" ")), and!(ref_rule!(r#"comment"#))), 0), ref_rule!(r#"eol"#)))
, r#"esc_char"# => or!(and!(lit!("\r")), and!(lit!("\n")), and!(lit!("\t")), and!(lit!("\\")), and!(lit!("\\\"")))
, r#"match"# => or!(and!(transf2!( and!( and!(lit!("[")) ) , t2rules!(t2_funct!("none"), ) ), or!(and!(transf2!( and!( and!(ref_rule!(r#"mchars"#), named!("b", or!(and!(rep!(ref_rule!(r#"mbetween"#), 0))))) ) , t2rules!(t2_text!("CHARS"), t2_funct!("endl"), t2_byname!("mchars"), t2_funct!("endl"), t2_text!("BETW"), t2_funct!("endl"), t2_byname!("b"), t2_text!("EOBETW"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(named!("b", or!(and!(rep!(ref_rule!(r#"mbetween"#), 1))))) ) , t2rules!(t2_text!("BETW"), t2_funct!("endl"), t2_byname!("b"), t2_text!("EOBETW"), t2_funct!("endl"), ) ))), transf2!( and!( and!(lit!("]")) ) , t2rules!(t2_funct!("none"), ) )))
, r#"rep_or_unary"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"atom_or_par"#), rep!(ref_rule!(r#"rep_symbol"#), 0, 1)) ) , t2rules!(t2_byname_opt!("rep_symbol"), t2_byname!("atom_or_par"), ) )), and!(transf2!( and!( and!(lit!("!"), ref_rule!(r#"atom_or_par"#)) ) , t2rules!(t2_text!("NEGATE"), t2_funct!("endl"), t2_byname!("atom_or_par"), ) )), and!(transf2!( and!( and!(lit!("&"), ref_rule!(r#"atom_or_par"#)) ) , t2rules!(t2_text!("PEEK"), t2_funct!("endl"), t2_byname!("atom_or_par"), ) )))
, r#"tmpl_rule"# => or!(and!(transf2!( and!( and!(lit!("$(")) ) , t2rules!(t2_funct!("none"), ) ), or!(and!(transf2!( and!( and!(lit!("?"), ref_rule!(r#"symbol"#)) ) , t2rules!(t2_text!("NAMED_OPT"), t2_funct!("endl"), t2_byname!("symbol"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(ref_rule!(r#"symbol"#)) ) , t2rules!(t2_text!("NAMED"), t2_funct!("endl"), t2_byname!("symbol"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(lit!("."), named!("pos", or!(and!(rep!(ematch!(chlist r#""# , from '0', to '9' ), 1))))) ) , t2rules!(t2_text!("POS"), t2_funct!("endl"), t2_byname!("symbol"), t2_byname!("pos"), t2_funct!("endl"), ) )), and!(transf2!( and!( and!(lit!(":")) ) , t2rules!(t2_funct!("none"), ) ), transf2!( and!( and!(named!("fn", or!(and!(rep!(or!(and!(not!(or!(and!(lit!(")")), and!(ref_rule!(r#"eol"#)))), dot!())), 1))))) ) , t2rules!(t2_text!("FUNCT"), t2_funct!("endl"), t2_byname!("fn"), t2_funct!("endl"), ) ))), transf2!( and!( and!(lit!(")")) ) , t2rules!(t2_funct!("none"), ) )))
, r#"main"# => or!(and!(transf2!( and!( and!(ref_rule!(r#"grammar"#)) ) , t2rules!(t2_byname!("grammar"), t2_text!("EOP"), ) )))
, r#"rule_name"# => or!(and!(ref_rule!(r#"symbol"#)))
)
}
Writting it by hand, it's dificult.
Isn't this program desineg to receive a text peg grammar and an text input and produce a text output?
IR
IR is from Intermediate Representation
Why???
Once we parse the input, we have an AST.
We could process the AST but...
The AST is strongly coupled to the grammar. Most of the times we modify the grammar, we will need to modify the code to process the AST.
Some times the grammar modification will be a syntax modif, or adding some feature that requiere some syntax modification, therefore a different AST but all, or almost all of the concepts remain the same.
Imagine if we wanted to add de function sqrt to the math expresion compiler. We will need to modify the rules generator in order to process the new AST
To decouple the peg grammar from parsing the AST, we will create the IR (Intermediate Representation)
How to get the IR will be defined in the own peg grammar as transformation rules.
An interpreter of the IR will produce the rules in memory. Later, we can generate de rust code from the rules produced, or we could have a specific interpreter to generate them, but it's nice to get it from rust data structures
To develop this feature... we need a parser, and a code generator... Hey!!! I do it. dpr does that!!!
How to generate the IR
peg_grammar()
.parse(peg_grammar())
.gen_rules()
.replace()
The peg_grammar will have in transformation rules the intructions to generate the IR
Thanks to the IR it's easy to modify this program, and we don't need to deal with the AST coupled to the peg-grammar
Let's see step by step
Creating rules...
extern crate dpr;
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
"
main = char+
char = 'a' -> A
/ 'b' -> B
/ .
",
)
.gen_rules()?
// .parse("aaacbbabdef")?
// .replace()?
// ...
;
println!("{:#?}", result);
Ok(())
}
Produce a set of rules like...
SetOfRules(
{
"main": And(
MultiExpr(
[
Repeat(
RepInfo {
expression: RuleName(
"char",
),
min: NRep(
1,
),
max: None,
},
),
],
),
),
"char": Or(
MultiExpr(
[
And(
MultiExpr(
[
MetaExpr(
Transf2(
Transf2Expr {
mexpr: MultiExpr(
[
Simple(
Literal(
"a",
),
),
],
),
transf2_rules: "A",
},
),
),
],
),
),
And(
MultiExpr(
[
MetaExpr(
Transf2(
Transf2Expr {
mexpr: MultiExpr(
[
Simple(
Literal(
"b",
),
),
],
),
transf2_rules: "B",
},
),
),
],
),
),
And(
MultiExpr(
[
Simple(
Dot,
),
],
),
),
],
),
),
},
)
This set of rules will let us to parse and generate the AST for any input
Next step, parsing the input with generated rules...
Creating rules...
(With a simplified input in order to reduce the output size)
extern crate dpr;
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
"
main = char+
char = 'a' -> A
/ 'b' -> B
/ .
",
)
.gen_rules()?
.parse("acb")?
// .replace()?
// ...
;
println!("{:#?}", result);
Ok(())
}
Now you can see de produced AST
Rule(
(
"main",
[
Rule(
(
"char",
[
Transf2(
(
"A",
[
Val(
"a",
),
],
),
),
],
),
),
Rule(
(
"char",
[
Val(
"c",
),
],
),
),
Rule(
(
"char",
[
Transf2(
(
"B",
[
Val(
"b",
),
],
),
),
],
),
),
],
),
)
And running the transformations...
extern crate dpr;
fn main() -> Result<(), dpr::Error> {
let result = dpr::Peg::new(
"
main = char+
char = 'a' -> A
/ 'b' -> B
/ .
",
)
.gen_rules()?
.parse("acb")?
.replace()?
// ...
;
println!("{:#?}", result);
Ok(())
}
"AcB"
Dependencies
~1MB
~25K SLoC