2021-11-20 20:48:44 +03:00
|
|
|
// Copyright (c) 2021 Lars Pontoppidan. All rights reserved.
|
|
|
|
// Use of this source code is governed by an MIT license
|
|
|
|
// that can be found in the LICENSE file.
|
|
|
|
module decoder
|
|
|
|
|
|
|
|
import toml.ast
|
|
|
|
import toml.ast.walker
|
|
|
|
import toml.token
|
|
|
|
import toml.scanner
|
|
|
|
import strconv
|
|
|
|
|
2021-11-23 12:02:43 +03:00
|
|
|
const (
|
|
|
|
// utf8_max is the largest inclusive value of the Unicodes scalar value ranges.
|
|
|
|
utf8_max = 0x10FFFF
|
|
|
|
)
|
|
|
|
|
2021-11-20 20:48:44 +03:00
|
|
|
// Decoder decode special sequences in a tree of TOML `ast.Value`'s.
|
|
|
|
pub struct Decoder {
|
|
|
|
scanner &scanner.Scanner
|
|
|
|
}
|
|
|
|
|
|
|
|
// decode decodes certain `ast.Value`'s and all it's children.
|
|
|
|
pub fn (d Decoder) decode(mut n ast.Value) ? {
|
|
|
|
walker.walk_and_modify(d, mut n) ?
|
|
|
|
}
|
|
|
|
|
|
|
|
fn (d Decoder) modify(mut value ast.Value) ? {
|
|
|
|
match value {
|
|
|
|
ast.Quoted {
|
|
|
|
mut v := &(value as ast.Quoted)
|
|
|
|
d.decode_quoted(mut v) ?
|
|
|
|
}
|
|
|
|
else {}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// excerpt returns a string of the token's surroundings
|
|
|
|
fn (d Decoder) excerpt(tp token.Position) string {
|
|
|
|
return d.scanner.excerpt(tp.pos, 10)
|
|
|
|
}
|
|
|
|
|
|
|
|
// decode_quoted returns an error if `q` is not a valid quoted TOML string.
|
|
|
|
fn (d Decoder) decode_quoted(mut q ast.Quoted) ? {
|
2021-11-23 12:02:43 +03:00
|
|
|
decode_quoted_escapes(mut q) ?
|
2021-11-20 20:48:44 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
// decode_quoted_escapes returns an error for any disallowed escape sequences.
|
|
|
|
// Delimiters in TOML has significant meaning:
|
|
|
|
// '/''' delimits *literal* strings (WYSIWYG / What-you-see-is-what-you-get)
|
|
|
|
// "/""" delimits *basic* strings
|
|
|
|
// Allowed escapes in *basic* strings are:
|
|
|
|
// \b - backspace (U+0008)
|
|
|
|
// \t - tab (U+0009)
|
|
|
|
// \n - linefeed (U+000A)
|
|
|
|
// \f - form feed (U+000C)
|
|
|
|
// \r - carriage return (U+000D)
|
|
|
|
// \" - quote (U+0022)
|
|
|
|
// \\ - backslash (U+005C)
|
|
|
|
// \uXXXX - Unicode (U+XXXX)
|
|
|
|
// \UXXXXXXXX - Unicode (U+XXXXXXXX)
|
2021-11-23 12:02:43 +03:00
|
|
|
pub fn decode_quoted_escapes(mut q ast.Quoted) ? {
|
2021-11-20 20:48:44 +03:00
|
|
|
// Setup a scanner in stack memory for easier navigation.
|
2021-11-23 12:02:43 +03:00
|
|
|
mut eat_whitespace := false
|
2021-11-20 20:48:44 +03:00
|
|
|
// TODO use string builder
|
|
|
|
mut decoded_s := ''
|
|
|
|
// See https://toml.io/en/v1.0.0#string for more info on string types.
|
|
|
|
is_basic := q.quote == `\"`
|
|
|
|
if !is_basic {
|
|
|
|
return
|
|
|
|
}
|
2021-11-23 12:02:43 +03:00
|
|
|
|
|
|
|
mut s := scanner.new_simple(q.text) ?
|
|
|
|
q.text = q.text.replace('\\"', '"')
|
|
|
|
|
2021-11-20 20:48:44 +03:00
|
|
|
for {
|
|
|
|
ch := s.next()
|
|
|
|
if ch == scanner.end_of_text {
|
|
|
|
break
|
|
|
|
}
|
|
|
|
ch_byte := byte(ch)
|
|
|
|
|
2021-11-23 12:02:43 +03:00
|
|
|
if eat_whitespace && ch_byte.is_space() {
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
eat_whitespace = false
|
|
|
|
|
2021-11-20 20:48:44 +03:00
|
|
|
if ch == `\\` {
|
2021-11-23 12:02:43 +03:00
|
|
|
ch_next := s.at()
|
|
|
|
ch_next_byte := byte(ch_next)
|
2021-11-20 20:48:44 +03:00
|
|
|
|
|
|
|
if ch_next == `\\` {
|
2021-11-23 12:02:43 +03:00
|
|
|
decoded_s += ch_next_byte.ascii_str()
|
2021-11-20 20:48:44 +03:00
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
2021-11-23 12:02:43 +03:00
|
|
|
if q.is_multiline {
|
|
|
|
if ch_next_byte.is_space() {
|
|
|
|
eat_whitespace = true
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2021-11-20 20:48:44 +03:00
|
|
|
if ch_next == `"` {
|
|
|
|
decoded_s += '"'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
if ch_next == `n` {
|
|
|
|
decoded_s += '\n'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
2021-11-23 12:02:43 +03:00
|
|
|
if ch_next == `t` {
|
|
|
|
decoded_s += '\t'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
if ch_next == `b` {
|
|
|
|
decoded_s += '\b'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
if ch_next == `r` {
|
|
|
|
decoded_s += '\r'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
if ch_next == `f` {
|
|
|
|
decoded_s += '\f'
|
|
|
|
s.next()
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
escape := ch_byte.ascii_str() + ch_next_byte.ascii_str()
|
2021-11-20 20:48:44 +03:00
|
|
|
// Decode unicode escapes
|
|
|
|
if escape.to_lower() == '\\u' {
|
2021-11-23 12:02:43 +03:00
|
|
|
is_valid_short := byte(s.peek(1)).is_hex_digit() && byte(s.peek(2)).is_hex_digit()
|
|
|
|
&& byte(s.peek(3)).is_hex_digit() && byte(s.peek(4)).is_hex_digit()
|
|
|
|
|
|
|
|
if is_valid_short {
|
|
|
|
// is_valid_long := byte(s.peek(5)).is_hex_digit() && byte(s.peek(6)).is_hex_digit() && byte(s.peek(7)).is_hex_digit() && byte(s.peek(8)).is_hex_digit()
|
|
|
|
// Long type Unicode (\UXXXXXXXX) is a maximum of 10 chars: '\' + 'U' + 8 hex characters
|
|
|
|
// we pass in 10 characters from the `u`/`U` which is the longest possible sequence
|
|
|
|
// of 9 chars plus one extra.
|
|
|
|
mut decoded := ''
|
|
|
|
mut sequence_length := 0
|
|
|
|
mut unicode_val := 0
|
|
|
|
if s.remaining() >= 10 {
|
|
|
|
pos := s.state().pos
|
|
|
|
sequence := s.text[pos..pos + 11]
|
|
|
|
|
|
|
|
decoded, unicode_val, sequence_length = decode_unicode_escape(sequence) or {
|
|
|
|
decoded_s += escape
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
if unicode_val > decoder.utf8_max || unicode_val < 0 {
|
|
|
|
decoded_s += escape
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
// Check if the Unicode value is actually in the valid Unicode scalar value ranges.
|
|
|
|
if !((unicode_val >= 0x0000 && unicode_val <= 0xD7FF)
|
|
|
|
|| (unicode_val >= 0xE000 && unicode_val <= decoder.utf8_max)) {
|
|
|
|
decoded_s += escape
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
if unicode_val in [0x7F, 0x1F, 0x5C, 0x75] {
|
|
|
|
sequence_length -= 2
|
|
|
|
}
|
|
|
|
decoded_s += decoded
|
|
|
|
s.skip_n(s.text[pos..pos + 2 + sequence_length + 1].len)
|
|
|
|
continue
|
|
|
|
} else {
|
|
|
|
pos := s.state().pos
|
|
|
|
sequence := s.text[pos..]
|
|
|
|
decoded, _, _ = decode_unicode_escape(sequence) or {
|
|
|
|
decoded_s += escape
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
decoded_s += decoded
|
|
|
|
s.skip_n(s.text[pos..].len)
|
|
|
|
continue
|
2021-11-20 20:48:44 +03:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
decoded_s += ch_byte.ascii_str()
|
|
|
|
}
|
|
|
|
q.text = decoded_s
|
|
|
|
}
|
|
|
|
|
|
|
|
// decode_unicode_escape returns an error if `esc_unicode` is not
|
|
|
|
// a valid Unicode escape sequence. `esc_unicode` is expected to be
|
|
|
|
// prefixed with either `u` or `U`.
|
2021-11-23 12:02:43 +03:00
|
|
|
fn decode_unicode_escape(esc_unicode string) ?(string, int, int) {
|
2021-11-20 20:48:44 +03:00
|
|
|
is_long_esc_type := esc_unicode.starts_with('U')
|
|
|
|
mut sequence := esc_unicode[1..]
|
|
|
|
hex_digits_len := if is_long_esc_type { 8 } else { 4 }
|
2021-11-23 12:02:43 +03:00
|
|
|
mut sequence_len := hex_digits_len
|
2021-11-20 20:48:44 +03:00
|
|
|
|
|
|
|
sequence = sequence[..hex_digits_len]
|
|
|
|
|
|
|
|
mut unicode_point := sequence
|
|
|
|
if unicode_point.len < 8 {
|
|
|
|
unicode_point = '0'.repeat(8 - unicode_point.len) + unicode_point
|
|
|
|
}
|
2021-11-23 12:02:43 +03:00
|
|
|
i64_val := strconv.parse_int(unicode_point, 16, 0) ?
|
|
|
|
rn := rune(i64_val)
|
|
|
|
return '$rn', int(i64_val), sequence_len
|
2021-11-20 20:48:44 +03:00
|
|
|
}
|