Update party and remove internal parsing lib
This commit is contained in:
parent
9a3bceff68
commit
bde07bb394
4 changed files with 176 additions and 616 deletions
|
@ -10,7 +10,6 @@ links = [
|
||||||
|
|
||||||
[dependencies]
|
[dependencies]
|
||||||
gleam_stdlib = "~> 0.34 or ~> 1.0"
|
gleam_stdlib = "~> 0.34 or ~> 1.0"
|
||||||
pears = "~> 0.3"
|
|
||||||
party = "~> 1.0"
|
party = "~> 1.0"
|
||||||
|
|
||||||
[dev-dependencies]
|
[dev-dependencies]
|
||||||
|
|
|
@ -3,13 +3,11 @@
|
||||||
|
|
||||||
packages = [
|
packages = [
|
||||||
{ name = "gleam_stdlib", version = "0.36.0", build_tools = ["gleam"], requirements = [], otp_app = "gleam_stdlib", source = "hex", outer_checksum = "C0D14D807FEC6F8A08A7C9EF8DFDE6AE5C10E40E21325B2B29365965D82EB3D4" },
|
{ name = "gleam_stdlib", version = "0.36.0", build_tools = ["gleam"], requirements = [], otp_app = "gleam_stdlib", source = "hex", outer_checksum = "C0D14D807FEC6F8A08A7C9EF8DFDE6AE5C10E40E21325B2B29365965D82EB3D4" },
|
||||||
{ name = "gleeunit", version = "1.0.2", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleeunit", source = "hex", outer_checksum = "D364C87AFEB26BDB4FB8A5ABDE67D635DC9FA52D6AB68416044C35B096C6882D" },
|
{ name = "gleeunit", version = "1.1.2", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "gleeunit", source = "hex", outer_checksum = "72CDC3D3F719478F26C4E2C5FED3E657AC81EC14A47D2D2DEBB8693CA3220C3B" },
|
||||||
{ name = "party", version = "1.0.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "party", source = "hex", outer_checksum = "13D70704E008F1AB163E6E5AED717014473898100B26852E158BD6E9BAE583BC" },
|
{ name = "party", version = "1.0.2", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "party", source = "hex", outer_checksum = "6169240A11DE6793FDFB77E7375E79441B6D7FB3D3EF8274B3C109ABAFF1A8FD" },
|
||||||
{ name = "pears", version = "0.3.0", build_tools = ["gleam"], requirements = ["gleam_stdlib"], otp_app = "pears", source = "hex", outer_checksum = "F823EDF5C7F8606A0B7764C071B6EE8515FC341D6FC69902E81120633DF7E983" },
|
|
||||||
]
|
]
|
||||||
|
|
||||||
[requirements]
|
[requirements]
|
||||||
gleam_stdlib = { version = "~> 0.34 or ~> 1.0" }
|
gleam_stdlib = { version = "~> 0.34 or ~> 1.0" }
|
||||||
gleeunit = { version = "~> 1.0" }
|
gleeunit = { version = "~> 1.0" }
|
||||||
party = { version = "~> 1.0"}
|
party = { version = "~> 1.0" }
|
||||||
pears = { version = "~> 0.3" }
|
|
||||||
|
|
366
src/jasper.gleam
366
src/jasper.gleam
|
@ -4,21 +4,16 @@ import gleam/string
|
||||||
import gleam/result
|
import gleam/result
|
||||||
import gleam/dict.{type Dict}
|
import gleam/dict.{type Dict}
|
||||||
import gleam/list
|
import gleam/list
|
||||||
import jasper/internal/parsing.{
|
import party.{
|
||||||
type ParseError, type Parser, Parser, Position, Unexpected, between, char,
|
type ParseError, type Parser, char, choice, digit, do, either, end, go, lazy,
|
||||||
choice, choice_char, concat, digit, do, either, end, go, lazy, left, letter,
|
letter, many1_concat, many_concat, map, perhaps, return, satisfy, sep, string,
|
||||||
many, many1_concat, many_concat, map, pair, perhaps, perhaps_default,
|
until,
|
||||||
perhaps_empty, return, right, satisfy, sep, string, to,
|
|
||||||
}
|
}
|
||||||
|
|
||||||
pub type JsonParseError =
|
pub fn to(p: Parser(a, e), v: b) -> Parser(b, e) {
|
||||||
Nil
|
use _ <- do(p)
|
||||||
|
return(v)
|
||||||
pub type JsonParserError =
|
}
|
||||||
ParseError(JsonParseError)
|
|
||||||
|
|
||||||
pub type JsonParser(a) =
|
|
||||||
Parser(a, JsonParseError)
|
|
||||||
|
|
||||||
pub type JsonObject =
|
pub type JsonObject =
|
||||||
Dict(String, JsonValue)
|
Dict(String, JsonValue)
|
||||||
|
@ -35,8 +30,26 @@ pub type JsonValue {
|
||||||
Null
|
Null
|
||||||
}
|
}
|
||||||
|
|
||||||
fn ws(options: JsonParseOptions) -> JsonParser(String) {
|
fn multiline_comment() -> Parser(String, e) {
|
||||||
many_concat(
|
use _ <- do(string("/*"))
|
||||||
|
use comment <- do(until(satisfy(fn(_) { True }), string("*/")))
|
||||||
|
return(string.concat(comment))
|
||||||
|
}
|
||||||
|
|
||||||
|
fn singleline_comment() -> Parser(String, e) {
|
||||||
|
use _ <- do(string("//"))
|
||||||
|
use comment <- do(
|
||||||
|
many_concat(satisfy(fn(c) { !string.contains("\r\n\u{2028}\u{2029}", c) })),
|
||||||
|
)
|
||||||
|
return(comment)
|
||||||
|
}
|
||||||
|
|
||||||
|
fn comment() -> Parser(String, e) {
|
||||||
|
either(multiline_comment(), singleline_comment())
|
||||||
|
}
|
||||||
|
|
||||||
|
fn ws(options: JsonParseOptions) -> Parser(String, e) {
|
||||||
|
let ws =
|
||||||
satisfy(fn(c) {
|
satisfy(fn(c) {
|
||||||
string.contains(
|
string.contains(
|
||||||
case options.ecma_whitespace {
|
case options.ecma_whitespace {
|
||||||
|
@ -46,52 +59,46 @@ fn ws(options: JsonParseOptions) -> JsonParser(String) {
|
||||||
},
|
},
|
||||||
c,
|
c,
|
||||||
)
|
)
|
||||||
}),
|
})
|
||||||
)
|
many_concat(case options.comments {
|
||||||
|
True -> either(ws, comment())
|
||||||
|
False -> ws
|
||||||
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
fn padded(p: JsonParser(a), options: JsonParseOptions) -> JsonParser(a) {
|
fn symbol(s: String, options: JsonParseOptions) -> Parser(String, e) {
|
||||||
left(p, ws(options))
|
use _ <- do(ws(options))
|
||||||
|
use c <- do(char(s))
|
||||||
|
return(c)
|
||||||
}
|
}
|
||||||
|
|
||||||
fn symbol(s: String, options: JsonParseOptions) -> JsonParser(String) {
|
fn json_null() -> Parser(Nil, e) {
|
||||||
char(s)
|
to(string("null"), Nil)
|
||||||
|> padded(options)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_null_parser() -> JsonParser(Nil) {
|
fn json_boolean() -> Parser(Bool, e) {
|
||||||
string("null")
|
|
||||||
|> to(Nil)
|
|
||||||
}
|
|
||||||
|
|
||||||
fn json_boolean_parser() -> JsonParser(Bool) {
|
|
||||||
either(to(string("true"), True), to(string("false"), False))
|
either(to(string("true"), True), to(string("false"), False))
|
||||||
}
|
}
|
||||||
|
|
||||||
fn hex_digit_parser() -> JsonParser(String) {
|
fn hex_digit() -> Parser(String, e) {
|
||||||
choice_char([
|
satisfy(fn(c) { string.contains("0123456789abcdefABCDEF", c) })
|
||||||
"0", "1", "2", "3", "4", "5", "6", "7", "8", "9", "a", "b", "c", "d", "e",
|
|
||||||
"f", "A", "B", "C", "D", "E", "F",
|
|
||||||
])
|
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_number_parser(options: JsonParseOptions) -> JsonParser(Float) {
|
fn json_number(options: JsonParseOptions) -> Parser(Float, e) {
|
||||||
let sign_parser =
|
let sign_parser = perhaps(either(char("+"), char("-")))
|
||||||
either(char("+"), char("-"))
|
|
||||||
|> perhaps_empty
|
|
||||||
let int_parser =
|
let int_parser =
|
||||||
either(
|
either(char("0"), {
|
||||||
char("0"),
|
use d <- do(satisfy(fn(c) { string.contains("123456789", c) }))
|
||||||
concat(
|
use ds <- do(many_concat(digit()))
|
||||||
choice_char(["1", "2", "3", "4", "5", "6", "7", "8", "9"]),
|
return(d <> ds)
|
||||||
many_concat(digit()),
|
})
|
||||||
),
|
|
||||||
)
|
|
||||||
let exp_parser =
|
let exp_parser =
|
||||||
either(char("e"), char("E"))
|
perhaps({
|
||||||
|> concat(sign_parser)
|
use e <- do(either(char("e"), char("E")))
|
||||||
|> concat(many1_concat(digit()))
|
use sign <- do(sign_parser)
|
||||||
|> perhaps_empty
|
use digits <- do(many1_concat(digit()))
|
||||||
|
return(e <> result.unwrap(sign, "") <> digits)
|
||||||
|
})
|
||||||
let min_double = -1.7976931348623158e308
|
let min_double = -1.7976931348623158e308
|
||||||
let max_double = 1.7976931348623158e308
|
let max_double = 1.7976931348623158e308
|
||||||
|
|
||||||
|
@ -99,38 +106,41 @@ fn json_number_parser(options: JsonParseOptions) -> JsonParser(Float) {
|
||||||
True -> {
|
True -> {
|
||||||
use sign <- do(sign_parser)
|
use sign <- do(sign_parser)
|
||||||
let inf = case sign {
|
let inf = case sign {
|
||||||
"-" -> min_double
|
Ok("-") -> min_double
|
||||||
_ -> max_double
|
_ -> max_double
|
||||||
}
|
}
|
||||||
choice([
|
choice([
|
||||||
to(string("Infinity"), inf),
|
to(string("Infinity"), inf),
|
||||||
to(string("NaN"), 0.0),
|
to(string("NaN"), 0.0),
|
||||||
{
|
{
|
||||||
use hex <- do(
|
use _ <- do(either(string("0x"), string("0X")))
|
||||||
either(string("0x"), string("0X"))
|
use hex <- do(many1_concat(hex_digit()))
|
||||||
|> right(many1_concat(hex_digit_parser())),
|
let assert Ok(n) = int.base_parse(result.unwrap(sign, "") <> hex, 16)
|
||||||
)
|
|
||||||
let assert Ok(n) = int.base_parse(sign <> hex, 16)
|
|
||||||
return(int.to_float(n))
|
return(int.to_float(n))
|
||||||
},
|
},
|
||||||
{
|
{
|
||||||
use n <- do(either(
|
use n <- do(
|
||||||
{
|
either(
|
||||||
use ns <- do(int_parser)
|
{
|
||||||
use ds <- do(
|
use ns <- do(int_parser)
|
||||||
right(
|
use ds <- do(
|
||||||
char("."),
|
perhaps({
|
||||||
many1_concat(digit())
|
use _ <- do(char("."))
|
||||||
|> perhaps_default("0"),
|
use ds <- do(perhaps(many1_concat(digit())))
|
||||||
|
return(result.unwrap(ds, "0"))
|
||||||
|
}),
|
||||||
)
|
)
|
||||||
|> perhaps_default("0"),
|
return(ns <> "." <> result.unwrap(ds, "0"))
|
||||||
)
|
},
|
||||||
return(ns <> "." <> ds)
|
{
|
||||||
},
|
use _ <- do(char("."))
|
||||||
concat(to(char("."), "0."), many1_concat(digit())),
|
use ds <- do(many1_concat(digit()))
|
||||||
))
|
return("0." <> ds)
|
||||||
|
},
|
||||||
|
),
|
||||||
|
)
|
||||||
use ex <- do(exp_parser)
|
use ex <- do(exp_parser)
|
||||||
{ sign <> n <> ex }
|
{ result.unwrap(sign, "") <> n <> result.unwrap(ex, "") }
|
||||||
|> float.parse
|
|> float.parse
|
||||||
|> result.unwrap(inf)
|
|> result.unwrap(inf)
|
||||||
|> return
|
|> return
|
||||||
|
@ -138,21 +148,27 @@ fn json_number_parser(options: JsonParseOptions) -> JsonParser(Float) {
|
||||||
])
|
])
|
||||||
}
|
}
|
||||||
False -> {
|
False -> {
|
||||||
use sign <- do(
|
use sign <- do(perhaps(char("-")))
|
||||||
char("-")
|
|
||||||
|> perhaps_empty,
|
|
||||||
)
|
|
||||||
let inf = case sign {
|
let inf = case sign {
|
||||||
"-" -> min_double
|
Ok("-") -> min_double
|
||||||
_ -> max_double
|
_ -> max_double
|
||||||
}
|
}
|
||||||
use ns <- do(int_parser)
|
use ns <- do(int_parser)
|
||||||
use ds <- do(
|
use ds <- do(
|
||||||
right(char("."), many1_concat(digit()))
|
perhaps({
|
||||||
|> perhaps_default("0"),
|
use _ <- do(char("."))
|
||||||
|
use ds <- do(many1_concat(digit()))
|
||||||
|
return(ds)
|
||||||
|
}),
|
||||||
)
|
)
|
||||||
use ex <- do(exp_parser)
|
use ex <- do(exp_parser)
|
||||||
{ sign <> ns <> "." <> ds <> ex }
|
{
|
||||||
|
result.unwrap(sign, "")
|
||||||
|
<> ns
|
||||||
|
<> "."
|
||||||
|
<> result.unwrap(ds, "0")
|
||||||
|
<> result.unwrap(ex, "")
|
||||||
|
}
|
||||||
|> float.parse
|
|> float.parse
|
||||||
|> result.unwrap(inf)
|
|> result.unwrap(inf)
|
||||||
|> return
|
|> return
|
||||||
|
@ -172,29 +188,27 @@ fn valid_string_char(c: String) -> Bool {
|
||||||
i >= 0x20 && i <= 0x10FFFF
|
i >= 0x20 && i <= 0x10FFFF
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_string_parser(options: JsonParseOptions) -> JsonParser(String) {
|
fn json_string(options: JsonParseOptions) -> Parser(String, e) {
|
||||||
let hex_digit = hex_digit_parser()
|
let unicode_escape = {
|
||||||
let unicode_escape =
|
use _ <- do(char("u"))
|
||||||
map(
|
use a <- do(hex_digit())
|
||||||
right(
|
use b <- do(hex_digit())
|
||||||
char("u"),
|
use c <- do(hex_digit())
|
||||||
hex_digit
|
use d <- do(hex_digit())
|
||||||
|> concat(hex_digit)
|
return(string_to_codepoint(a <> b <> c <> d))
|
||||||
|> concat(hex_digit)
|
}
|
||||||
|> concat(hex_digit),
|
|
||||||
),
|
|
||||||
string_to_codepoint,
|
|
||||||
)
|
|
||||||
|
|
||||||
let escape =
|
let escape = {
|
||||||
char("\\")
|
use _ <- do(char("\\"))
|
||||||
|> right(
|
use c <- do(
|
||||||
choice(case options.ecma_strings {
|
choice(case options.ecma_strings {
|
||||||
True -> [
|
True -> [
|
||||||
map(
|
{
|
||||||
right(char("x"), concat(hex_digit, hex_digit)),
|
use _ <- do(char("x"))
|
||||||
string_to_codepoint,
|
use a <- do(hex_digit())
|
||||||
),
|
use b <- do(hex_digit())
|
||||||
|
return(string_to_codepoint(a <> b))
|
||||||
|
},
|
||||||
unicode_escape,
|
unicode_escape,
|
||||||
char("\\"),
|
char("\\"),
|
||||||
char("/"),
|
char("/"),
|
||||||
|
@ -232,12 +246,19 @@ fn json_string_parser(options: JsonParseOptions) -> JsonParser(String) {
|
||||||
]
|
]
|
||||||
}),
|
}),
|
||||||
)
|
)
|
||||||
|
return(c)
|
||||||
|
}
|
||||||
|
|
||||||
let str = fn(q) {
|
let str = fn(q) {
|
||||||
satisfy(fn(c) { c != q && c != "\\" && valid_string_char(c) })
|
use _ <- do(char(q))
|
||||||
|> either(escape)
|
use str <- do(
|
||||||
|> many_concat()
|
many_concat(either(
|
||||||
|> between(char(q), char(q))
|
escape,
|
||||||
|
satisfy(fn(c) { c != q && c != "\\" && valid_string_char(c) }),
|
||||||
|
)),
|
||||||
|
)
|
||||||
|
use _ <- do(char(q))
|
||||||
|
return(str)
|
||||||
}
|
}
|
||||||
|
|
||||||
case options.ecma_strings {
|
case options.ecma_strings {
|
||||||
|
@ -246,106 +267,67 @@ fn json_string_parser(options: JsonParseOptions) -> JsonParser(String) {
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
fn parse_multiline_comment() -> JsonParser(String) {
|
fn allow_trailing_comma(p: Parser(a, e), options: JsonParseOptions) {
|
||||||
Parser(fn(source, pos) {
|
|
||||||
let assert Position(row, col) = pos
|
|
||||||
case source {
|
|
||||||
[h, ..t] ->
|
|
||||||
case h {
|
|
||||||
"\n" -> Ok(#(h, t, Position(row + 1, 0)))
|
|
||||||
"*" ->
|
|
||||||
case t {
|
|
||||||
["/", ..] -> Error(Unexpected(pos, h))
|
|
||||||
_ -> Ok(#(h, t, Position(row, col + 1)))
|
|
||||||
}
|
|
||||||
_ -> Ok(#(h, t, Position(row, col + 1)))
|
|
||||||
}
|
|
||||||
[] -> Error(Unexpected(pos, "EOF"))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
|> many_concat
|
|
||||||
|> between(string("/*"), string("*/"))
|
|
||||||
}
|
|
||||||
|
|
||||||
fn parse_singleline_comment() -> JsonParser(String) {
|
|
||||||
right(
|
|
||||||
string("//"),
|
|
||||||
many_concat(satisfy(fn(c) { !string.contains("\r\n\u{2028}\u{2029}", c) })),
|
|
||||||
)
|
|
||||||
}
|
|
||||||
|
|
||||||
fn comment_parser(options: JsonParseOptions) -> JsonParser(String) {
|
|
||||||
either(parse_multiline_comment(), parse_singleline_comment())
|
|
||||||
|> padded(options)
|
|
||||||
}
|
|
||||||
|
|
||||||
fn allow_comments(options: JsonParseOptions, p: Parser(_, _)) {
|
|
||||||
case options.comments {
|
|
||||||
True -> left(p, many(comment_parser(options)))
|
|
||||||
False -> p
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
fn allow_trailing_comma(options: JsonParseOptions, p: Parser(_, _)) {
|
|
||||||
case options.trailing_comma {
|
case options.trailing_comma {
|
||||||
True -> right(perhaps(symbol(",", options)), p)
|
True -> {
|
||||||
|
use _ <- do(perhaps(symbol(",", options)))
|
||||||
|
use v <- do(p)
|
||||||
|
return(v)
|
||||||
|
}
|
||||||
False -> p
|
False -> p
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_array_parser(options: JsonParseOptions) -> JsonParser(JsonArray) {
|
fn json_array(options: JsonParseOptions) -> Parser(JsonArray, e) {
|
||||||
let json_value_parser = lazy(fn() { json_value_parser(options) })
|
use _ <- do(symbol("[", options))
|
||||||
let allow_comments = allow_comments(options, _)
|
use values <- do(sep(lazy(fn() { json_value(options) }), symbol(",", options)))
|
||||||
sep(json_value_parser, allow_comments(symbol(",", options)))
|
use _ <- do(allow_trailing_comma(symbol("]", options), options))
|
||||||
|> between(
|
return(values)
|
||||||
allow_comments(symbol("[", options)),
|
|
||||||
allow_trailing_comma(options, symbol("]", options)),
|
|
||||||
)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// todo
|
// todo
|
||||||
fn ecmascript_identifier_parser() -> JsonParser(String) {
|
fn ecmascript_identifier() -> Parser(String, e) {
|
||||||
letter()
|
use c <- do(letter())
|
||||||
|> concat(
|
use cs <- do(many_concat(either(letter(), digit())))
|
||||||
either(letter(), digit())
|
return(c <> cs)
|
||||||
|> many_concat,
|
|
||||||
)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_object_parser(options: JsonParseOptions) -> JsonParser(JsonObject) {
|
fn json_object(options: JsonParseOptions) -> Parser(JsonObject, e) {
|
||||||
let json_value_parser = lazy(fn() { json_value_parser(options) })
|
use _ <- do(symbol("{", options))
|
||||||
let allow_comments = allow_comments(options, _)
|
use key_value_pairs <- do(sep(
|
||||||
let key_parser = case options.ecma_object_keys {
|
{
|
||||||
True -> either(json_string_parser(options), ecmascript_identifier_parser())
|
use _ <- do(ws(options))
|
||||||
False -> json_string_parser(options)
|
use key <- do(case options.ecma_object_keys {
|
||||||
}
|
True -> either(json_string(options), ecmascript_identifier())
|
||||||
allow_comments(key_parser)
|
False -> json_string(options)
|
||||||
|> left(allow_comments(symbol(":", options)))
|
})
|
||||||
|> pair(json_value_parser)
|
use _ <- do(symbol(":", options))
|
||||||
|> sep(allow_comments(symbol(",", options)))
|
use value <- do(lazy(fn() { json_value(options) }))
|
||||||
|> map(dict.from_list)
|
return(#(key, value))
|
||||||
|> between(
|
},
|
||||||
allow_comments(symbol("{", options)),
|
symbol(",", options),
|
||||||
allow_trailing_comma(options, symbol("}", options)),
|
))
|
||||||
)
|
use _ <- do(allow_trailing_comma(symbol("}", options), options))
|
||||||
|
return(dict.from_list(key_value_pairs))
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_value_parser(options: JsonParseOptions) -> JsonParser(JsonValue) {
|
fn json_value(options: JsonParseOptions) -> Parser(JsonValue, e) {
|
||||||
|
use _ <- do(ws(options))
|
||||||
choice([
|
choice([
|
||||||
to(json_null_parser(), Null),
|
to(json_null(), Null),
|
||||||
map(json_boolean_parser(), Boolean),
|
map(json_boolean(), Boolean),
|
||||||
map(json_number_parser(options), Number),
|
map(json_number(options), Number),
|
||||||
map(json_string_parser(options), String),
|
map(json_string(options), String),
|
||||||
map(json_array_parser(options), Array),
|
map(json_array(options), Array),
|
||||||
map(json_object_parser(options), Object),
|
map(json_object(options), Object),
|
||||||
])
|
])
|
||||||
|> padded(options)
|
|
||||||
|> allow_comments(options, _)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
fn json_parser(options: JsonParseOptions) -> JsonParser(JsonValue) {
|
fn json_parser(options: JsonParseOptions) -> Parser(JsonValue, e) {
|
||||||
json_value_parser(options)
|
use value <- do(json_value(options))
|
||||||
|> between(allow_comments(options, ws(options)), end())
|
use _ <- do(ws(options))
|
||||||
|
use _ <- do(end())
|
||||||
|
return(value)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub type JsonParseOptions {
|
pub type JsonParseOptions {
|
||||||
|
@ -391,19 +373,19 @@ const jsonl_options = json_options
|
||||||
pub fn parse_json_custom(
|
pub fn parse_json_custom(
|
||||||
value: String,
|
value: String,
|
||||||
options: JsonParseOptions,
|
options: JsonParseOptions,
|
||||||
) -> Result(JsonValue, JsonParserError) {
|
) -> Result(JsonValue, ParseError(e)) {
|
||||||
go(json_parser(options), value)
|
go(json_parser(options), value)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn parse_json(value: String) -> Result(JsonValue, JsonParserError) {
|
pub fn parse_json(value: String) -> Result(JsonValue, ParseError(e)) {
|
||||||
parse_json_custom(value, json_options)
|
parse_json_custom(value, json_options)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn parse_jsonc(value: String) -> Result(JsonValue, JsonParserError) {
|
pub fn parse_jsonc(value: String) -> Result(JsonValue, ParseError(e)) {
|
||||||
parse_json_custom(value, jsonc_options)
|
parse_json_custom(value, jsonc_options)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn parse_json5(value: String) -> Result(JsonValue, JsonParserError) {
|
pub fn parse_json5(value: String) -> Result(JsonValue, ParseError(e)) {
|
||||||
parse_json_custom(value, json5_options)
|
parse_json_custom(value, json5_options)
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -415,14 +397,12 @@ fn split_jsonl(value: String) -> List(String) {
|
||||||
|> string.split("\n")
|
|> string.split("\n")
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn parse_jsonl(value: String) -> Result(List(JsonValue), JsonParserError) {
|
pub fn parse_jsonl(value: String) -> Result(List(JsonValue), ParseError(e)) {
|
||||||
let parse = go(json_parser(jsonl_options), _)
|
let parse = go(json_parser(jsonl_options), _)
|
||||||
list.try_map(split_jsonl(value), parse)
|
list.try_map(split_jsonl(value), parse)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn parse_jsonl_all(
|
pub fn parse_jsonl_all(value: String) -> List(Result(JsonValue, ParseError(e))) {
|
||||||
value: String,
|
|
||||||
) -> List(Result(JsonValue, JsonParserError)) {
|
|
||||||
let parse = go(json_parser(jsonl_options), _)
|
let parse = go(json_parser(jsonl_options), _)
|
||||||
list.map(split_jsonl(value), parse)
|
list.map(split_jsonl(value), parse)
|
||||||
}
|
}
|
||||||
|
|
|
@ -1,417 +0,0 @@
|
||||||
import gleam/string
|
|
||||||
import gleam/result
|
|
||||||
import gleam/list
|
|
||||||
|
|
||||||
/// The custom error type for the parser,
|
|
||||||
/// which can itself be parameterized by a user-defined error type.
|
|
||||||
/// The user-defined error type is useful for, for example,
|
|
||||||
/// adding a `int.parse` call into your parser pipeline.
|
|
||||||
/// See `try` for using this feature.
|
|
||||||
pub type ParseError(e) {
|
|
||||||
Unexpected(pos: Position, error: String)
|
|
||||||
UserError(pos: Position, error: e)
|
|
||||||
}
|
|
||||||
|
|
||||||
/// The type for positions within a string.
|
|
||||||
pub type Position {
|
|
||||||
Position(row: Int, col: Int)
|
|
||||||
}
|
|
||||||
|
|
||||||
/// The parser type, parameterized by the type it parses and
|
|
||||||
/// the user-defined error type it can return.
|
|
||||||
pub type Parser(a, e) {
|
|
||||||
Parser(
|
|
||||||
parse: fn(List(String), Position) ->
|
|
||||||
Result(#(a, List(String), Position), ParseError(e)),
|
|
||||||
)
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Apply a parser to a list of graphemes (holding on to extra result info that is hidden from the library user).
|
|
||||||
fn run(
|
|
||||||
p: Parser(a, e),
|
|
||||||
src: List(String),
|
|
||||||
pos: Position,
|
|
||||||
) -> Result(#(a, List(String), Position), ParseError(e)) {
|
|
||||||
case p {
|
|
||||||
Parser(f) -> f(src, pos)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Apply a parser to a string.
|
|
||||||
pub fn go(p: Parser(a, e), src: String) -> Result(a, ParseError(e)) {
|
|
||||||
case run(p, string.to_graphemes(src), Position(1, 1)) {
|
|
||||||
Ok(#(x, _, _)) -> Ok(x)
|
|
||||||
Error(e) -> Error(e)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Get the current parser position.
|
|
||||||
pub fn pos() -> Parser(Position, e) {
|
|
||||||
Parser(fn(source, p) { Ok(#(p, source, p)) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a character if it matches the predicate.
|
|
||||||
pub fn satisfy(when pred: fn(String) -> Bool) -> Parser(String, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
let assert Position(row, col) = pos
|
|
||||||
case source {
|
|
||||||
[h, ..t] ->
|
|
||||||
case pred(h) {
|
|
||||||
True ->
|
|
||||||
case h {
|
|
||||||
"\n" -> Ok(#(h, t, Position(row + 1, 0)))
|
|
||||||
_ -> Ok(#(h, t, Position(row, col + 1)))
|
|
||||||
}
|
|
||||||
False -> Error(Unexpected(pos, h))
|
|
||||||
}
|
|
||||||
[] -> Error(Unexpected(pos, "EOF"))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a lowercase letter.
|
|
||||||
pub fn lowercase_letter() -> Parser(String, e) {
|
|
||||||
satisfy(when: fn(c) { string.contains("abcdefghijklmnopqrstuvwxyz", c) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse an uppercase letter.
|
|
||||||
pub fn uppercase_letter() -> Parser(String, e) {
|
|
||||||
satisfy(when: fn(c) { string.contains("ABCDEFGHIJKLMNOPQRSTUVWXYZ", c) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a lowercase or uppercase letter.
|
|
||||||
pub fn letter() -> Parser(String, e) {
|
|
||||||
either(lowercase_letter(), uppercase_letter())
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a specific character.
|
|
||||||
pub fn char(c) -> Parser(String, e) {
|
|
||||||
satisfy(when: fn(c2) { c == c2 })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a digit.
|
|
||||||
pub fn digit() -> Parser(String, e) {
|
|
||||||
satisfy(fn(c) { string.contains("0123456789", c) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a sequence of digits.
|
|
||||||
pub fn digits() -> Parser(String, e) {
|
|
||||||
many1_concat(digit())
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse the first parser, or the second if the first fails.
|
|
||||||
pub fn either(p: Parser(a, e), q: Parser(a, e)) -> Parser(a, e) {
|
|
||||||
Parser(fn(source, pos) { result.or(run(p, source, pos), run(q, source, pos)) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse with the first parser in the list that doesn't fail.
|
|
||||||
pub fn choice(ps: List(Parser(a, e))) -> Parser(a, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case ps {
|
|
||||||
[] -> panic as "choice doesn't accept an empty list of parsers"
|
|
||||||
// TODO: should this be an Unexpected instead?
|
|
||||||
[p] -> run(p, source, pos)
|
|
||||||
[p, ..t] ->
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(#(x, r, pos2)) -> Ok(#(x, r, pos2))
|
|
||||||
Error(_) -> run(choice(t), source, pos)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse an alphanumeric character.
|
|
||||||
pub fn alphanum() -> Parser(String, e) {
|
|
||||||
either(digit(), letter())
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse zero or more whitespace characters.
|
|
||||||
pub fn whitespace() -> Parser(String, e) {
|
|
||||||
many_concat(choice([char(" "), char("\t"), char("\n")]))
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse one or more whitespace characters.
|
|
||||||
pub fn whitespace1() -> Parser(String, e) {
|
|
||||||
many1_concat(choice([char(" "), char("\t"), char("\n")]))
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Keep trying the parser until it fails, and return the array of parsed results.
|
|
||||||
/// This cannot fail because it parses zero or more times!
|
|
||||||
pub fn many(p: Parser(a, e)) -> Parser(List(a), e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Error(_) -> Ok(#([], source, pos))
|
|
||||||
Ok(#(x, r, pos2)) ->
|
|
||||||
result.map(run(many(p), r, pos2), fn(res) {
|
|
||||||
#([x, ..res.0], res.1, res.2)
|
|
||||||
})
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a certain string as many times as possible, returning everything that was parsed.
|
|
||||||
/// This cannot fail because it parses zero or more times!
|
|
||||||
pub fn many_concat(p: Parser(String, e)) -> Parser(String, e) {
|
|
||||||
many(p)
|
|
||||||
|> map(string.concat)
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Keep trying the parser until it fails, and return the array of parsed results.
|
|
||||||
/// This can fail, because it must parse successfully at least once!
|
|
||||||
pub fn many1(p: Parser(a, e)) -> Parser(List(a), e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Error(e) -> Error(e)
|
|
||||||
Ok(#(x, r, pos2)) ->
|
|
||||||
result.map(run(many(p), r, pos2), fn(res) {
|
|
||||||
#([x, ..res.0], res.1, res.2)
|
|
||||||
})
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a certain string as many times as possible, returning everything that was parsed.
|
|
||||||
/// This can fail, because it must parse successfully at least once!
|
|
||||||
pub fn many1_concat(p: Parser(String, e)) -> Parser(String, e) {
|
|
||||||
many1(p)
|
|
||||||
|> map(string.concat)
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Do the first parser, ignore its result, then do the second parser.
|
|
||||||
pub fn seq(p: Parser(a, e), q: Parser(b, e)) -> Parser(b, e) {
|
|
||||||
use _ <- do(p)
|
|
||||||
q
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a sequence separated by the given separator parser.
|
|
||||||
pub fn sep(parser: Parser(a, e), by s: Parser(b, e)) -> Parser(List(a), e) {
|
|
||||||
use mb_a <- do(perhaps(parser))
|
|
||||||
case mb_a {
|
|
||||||
Ok(a) -> {
|
|
||||||
use rest <- do(many(right(s, parser)))
|
|
||||||
return([a, ..rest])
|
|
||||||
}
|
|
||||||
Error(Nil) -> return([])
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse a sequence separated by the given separator parser.
|
|
||||||
/// This only succeeds if at least one element of the sequence was parsed.
|
|
||||||
pub fn sep1(parser: Parser(a, e), by s: Parser(b, e)) -> Parser(List(a), e) {
|
|
||||||
use sequence <- do(sep(parser, by: s))
|
|
||||||
case sequence {
|
|
||||||
[] -> fail()
|
|
||||||
_ -> return(sequence)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Do `p`, then apply `f` to the result if it succeeded.
|
|
||||||
pub fn map(p: Parser(a, e), f: fn(a) -> b) -> Parser(b, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(#(x, r, pos2)) -> Ok(#(f(x), r, pos2))
|
|
||||||
Error(e) -> Error(e)
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Do `p`, the apply `f` to the result if it succeeded.
|
|
||||||
/// `f` itself can fail with the user-defined error type,
|
|
||||||
/// and if it does the result is a `UserError` with the error.
|
|
||||||
pub fn try(p: Parser(a, e), f: fn(a) -> Result(b, e)) -> Parser(b, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(#(x, r, pos2)) ->
|
|
||||||
case f(x) {
|
|
||||||
Ok(a) -> Ok(#(a, r, pos2))
|
|
||||||
Error(e) -> Error(UserError(pos2, e))
|
|
||||||
}
|
|
||||||
Error(e) -> Error(e)
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Transform the user-defined error type
|
|
||||||
/// with a user-provided conversion function.
|
|
||||||
pub fn error_map(p: Parser(a, e), f: fn(e) -> f) -> Parser(a, f) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(res) -> Ok(res)
|
|
||||||
Error(e) ->
|
|
||||||
case e {
|
|
||||||
UserError(pos, e) -> Error(UserError(pos, f(e)))
|
|
||||||
Unexpected(pos, s) -> Error(Unexpected(pos, s))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Try running a parser, but still succeed (with `Error(Nil)`) if it failed.
|
|
||||||
pub fn perhaps(p: Parser(a, e)) -> Parser(Result(a, Nil), e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(#(x, r, pos2)) -> Ok(#(Ok(x), r, pos2))
|
|
||||||
Error(_) -> Ok(#(Error(Nil), source, pos))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Do each parser in the list, returning the result of the last parser.
|
|
||||||
pub fn all(ps: List(Parser(a, e))) -> Parser(a, e) {
|
|
||||||
case ps {
|
|
||||||
[p] -> p
|
|
||||||
[h, ..t] -> {
|
|
||||||
use _ <- do(h)
|
|
||||||
all(t)
|
|
||||||
}
|
|
||||||
_ -> panic as "all(parsers) doesn't accept an empty list of parsers"
|
|
||||||
}
|
|
||||||
// TODO: should this be an Unexpected instead?
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parse an exact string of characters.
|
|
||||||
pub fn string(s: String) -> Parser(String, e) {
|
|
||||||
case string.pop_grapheme(s) {
|
|
||||||
Ok(#(h, t)) -> {
|
|
||||||
use c <- do(char(h))
|
|
||||||
use rest <- do(string(t))
|
|
||||||
return(c <> rest)
|
|
||||||
}
|
|
||||||
Error(_) -> return("")
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Negate a parser: if it succeeds, this fails, and vice versa.
|
|
||||||
/// Example: `seq(string("if"), not(either(alphanum(), char("_"))))`
|
|
||||||
pub fn not(p: Parser(a, e)) -> Parser(Nil, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(_) -> Error(Unexpected(pos, ""))
|
|
||||||
// todo: better error message here (add a label system)
|
|
||||||
Error(_) -> Ok(#(Nil, source, pos))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Parses successfully only when at the end of the input string.
|
|
||||||
pub fn end() -> Parser(Nil, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case source {
|
|
||||||
[] -> Ok(#(Nil, source, pos))
|
|
||||||
[h, ..] -> Error(Unexpected(pos, h))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Run a parser as normal, but the parser itself isn't evaluated until it is used.
|
|
||||||
/// This is needed for recursive grammars, such as `E := n | E + E` where `n` is a number.
|
|
||||||
/// Example: `lazy(digit)` instead of `digit()`.
|
|
||||||
pub fn lazy(p: fn() -> Parser(a, e)) -> Parser(a, e) {
|
|
||||||
Parser(fn(source, pos) { run(p(), source, pos) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// A monadic bind for pleasant interplay with gleam's `use` syntax.
|
|
||||||
/// example:
|
|
||||||
/// ```
|
|
||||||
/// fn identifier() -> Parser(String, e) {
|
|
||||||
/// use pos <- do(pos())
|
|
||||||
/// use first <- do(lowercase_letter())
|
|
||||||
/// use rest <- do(many(alt(alphanum(), char("_"))))
|
|
||||||
/// return(Ident(pos, first <> string.concat(rest)))
|
|
||||||
/// }
|
|
||||||
/// ```
|
|
||||||
pub fn do(p: Parser(a, e), f: fn(a) -> Parser(b, e)) -> Parser(b, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case run(p, source, pos) {
|
|
||||||
Ok(#(x, r, pos2)) -> run(f(x), r, pos2)
|
|
||||||
Error(e) -> Error(e)
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
/// A monadic return for pleasant interplay with gleam's `use` syntax.
|
|
||||||
/// see `do` for more details and an example.
|
|
||||||
/// This is redundant if the last `do` is a `map` instead.
|
|
||||||
/// But I prefer using it, stylistically.
|
|
||||||
pub fn return(x) {
|
|
||||||
Parser(fn(source, pos) { Ok(#(x, source, pos)) })
|
|
||||||
}
|
|
||||||
|
|
||||||
/// Immediately fail regardless of the next input
|
|
||||||
pub fn fail() -> Parser(a, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
case source {
|
|
||||||
[] -> Error(Unexpected(pos, "EOF"))
|
|
||||||
[h, ..] -> Error(Unexpected(pos, h))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn left(p: Parser(a, e), q: Parser(b, e)) -> Parser(a, e) {
|
|
||||||
use x <- do(p)
|
|
||||||
use _ <- do(q)
|
|
||||||
return(x)
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn right(p: Parser(a, e), q: Parser(b, e)) -> Parser(b, e) {
|
|
||||||
use _ <- do(p)
|
|
||||||
use x <- do(q)
|
|
||||||
return(x)
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn pair(p: Parser(a, e), q: Parser(b, e)) -> Parser(#(a, b), e) {
|
|
||||||
use x <- do(p)
|
|
||||||
use y <- do(q)
|
|
||||||
return(#(x, y))
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn between(
|
|
||||||
p: Parser(a, e),
|
|
||||||
q: Parser(b, e),
|
|
||||||
r: Parser(c, e),
|
|
||||||
) -> Parser(a, e) {
|
|
||||||
use _ <- do(q)
|
|
||||||
use x <- do(p)
|
|
||||||
use _ <- do(r)
|
|
||||||
return(x)
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn concat(p: Parser(String, e), q: Parser(String, e)) -> Parser(String, e) {
|
|
||||||
use x <- do(p)
|
|
||||||
use y <- do(q)
|
|
||||||
return(x <> y)
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn choice_char(vs: List(String)) -> Parser(String, e) {
|
|
||||||
vs
|
|
||||||
|> list.map(char)
|
|
||||||
|> choice
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn perhaps_default(p: Parser(a, e), d: a) -> Parser(a, e) {
|
|
||||||
use x <- do(perhaps(p))
|
|
||||||
return(result.unwrap(x, d))
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn perhaps_empty(p: Parser(String, e)) -> Parser(String, e) {
|
|
||||||
perhaps_default(p, "")
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn to(p: Parser(a, e), v: b) -> Parser(b, e) {
|
|
||||||
use _ <- do(p)
|
|
||||||
return(v)
|
|
||||||
}
|
|
||||||
|
|
||||||
pub fn any() -> Parser(String, e) {
|
|
||||||
Parser(fn(source, pos) {
|
|
||||||
let assert Position(row, col) = pos
|
|
||||||
case source {
|
|
||||||
[h, ..t] ->
|
|
||||||
case h {
|
|
||||||
"\n" -> Ok(#(h, t, Position(row + 1, 0)))
|
|
||||||
_ -> Ok(#(h, t, Position(row, col + 1)))
|
|
||||||
}
|
|
||||||
[] -> Error(Unexpected(pos, "EOF"))
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
Loading…
Reference in a new issue