mirror of
https://github.com/ddidderr/wuffs.git
synced 2026-11-18 08:32:11 +01:00
Redo Wuffs structs' quirk bool fields
This commit is contained in:
1 parent
532e18c4a5
commit
e39f3cbf42
7 files changed
+162
-290
No files matched your search
@@ -66,14 +66,15 @@ pri const DECODER_NUMBER_LENGTH_MAX_INCL : base.u64 = 99
|
||||
//
|
||||
// If the element is non-zero (but the 0x80 bit is not set) then "\i"'s
|
||||
// validity depends on the relevant quirk. The element's value is an enum:
|
||||
// - 1: "\a", U+0007, quirk_allow_backslash_a.
|
||||
// - 2: "\e", U+001B, quirk_allow_backslash_e.
|
||||
// - 3: "backslash new_line(not_n)", U+000A, quirk_allow_backslash_new_line.
|
||||
// - 4: "\?", U+003F, quirk_allow_backslash_question_mark.
|
||||
// - 5: "\'", U+0027, quirk_allow_backslash_single_quote.
|
||||
// - 6: "\v", U+000B, quirk_allow_backslash_v.
|
||||
// - 7: "\0", U+0000, quirk_allow_backslash_zero.
|
||||
// The U+1234 values are held in LUT_QUIRKY_BACKSLASHES, below.
|
||||
// - 1: "\a", U+0007, QUIRK_ALLOW_BACKSLASH_A.
|
||||
// - 2: "\e", U+001B, QUIRK_ALLOW_BACKSLASH_E.
|
||||
// - 3: "backslash new_line(not_n)", U+000A, QUIRK_ALLOW_BACKSLASH_NEW_LINE.
|
||||
// - 4: "\?", U+003F, QUIRK_ALLOW_BACKSLASH_QUESTION_MARK.
|
||||
// - 5: "\'", U+0027, QUIRK_ALLOW_BACKSLASH_SINGLE_QUOTE.
|
||||
// - 6: "\v", U+000B, QUIRK_ALLOW_BACKSLASH_V.
|
||||
// - 7: "\0", U+0000, QUIRK_ALLOW_BACKSLASH_ZERO.
|
||||
// The quirk and U+1234 values are held in LUT_QUIRKY_BACKSLASHES_QUIRKS and
|
||||
// LUT_QUIRKY_BACKSLASHES_CHARS, below.
|
||||
//
|
||||
// If the element is zero then "\i" is invalid, or it is a special case, the
|
||||
// start of "\x12", "\u1234" or "\U12345678".
|
||||
@@ -119,10 +120,24 @@ pri const LUT_BACKSLASHES : array[256] base.u8 = [
|
||||
// 8 9 A B C D E F
|
||||
]
|
||||
|
||||
// LUT_QUIRKY_BACKSLASHES is discussed in the LUT_BACKSLASHES comment. The
|
||||
// first element (index 0) is not used, but 8 is a round power of 2, so
|
||||
// LUT_QUIRKY_BACKSLASHES_QUIRKS is discussed in the LUT_BACKSLASHES comment.
|
||||
// The first element (index 0) is not used, but 8 is a round power of 2, so
|
||||
// enforcing index-in-bounds is a simple "&7" operation.
|
||||
pri const LUT_QUIRKY_BACKSLASHES : array[8] base.u8 = [
|
||||
pri const LUT_QUIRKY_BACKSLASHES_QUIRKS : array[8] base.u8[..= 0x09] = [
|
||||
0,
|
||||
0x01, // (QUIRK_ALLOW_BACKSLASH_A - QUIRKS_BASE) as base.u8,
|
||||
0x03, // (QUIRK_ALLOW_BACKSLASH_E - QUIRKS_BASE) as base.u8,
|
||||
0x04, // (QUIRK_ALLOW_BACKSLASH_NEW_LINE - QUIRKS_BASE) as base.u8,
|
||||
0x05, // (QUIRK_ALLOW_BACKSLASH_QUESTION_MARK - QUIRKS_BASE) as base.u8,
|
||||
0x06, // (QUIRK_ALLOW_BACKSLASH_SINGLE_QUOTE - QUIRKS_BASE) as base.u8,
|
||||
0x07, // (QUIRK_ALLOW_BACKSLASH_V - QUIRKS_BASE) as base.u8,
|
||||
0x09, // (QUIRK_ALLOW_BACKSLASH_ZERO - QUIRKS_BASE) as base.u8,
|
||||
]
|
||||
|
||||
// LUT_QUIRKY_BACKSLASHES_CHARS is discussed in the LUT_BACKSLASHES comment.
|
||||
// The first element (index 0) is not used, but 8 is a round power of 2, so
|
||||
// enforcing index-in-bounds is a simple "&7" operation.
|
||||
pri const LUT_QUIRKY_BACKSLASHES_CHARS : array[8] base.u8 = [
|
||||
0x00, 0x07, 0x1B, 0x0A, 0x3F, 0x27, 0x0B, 0x00,
|
||||
]
|
||||
|
||||
|
||||
+33
-75
@@ -13,21 +13,7 @@
|
||||
// limitations under the License.
|
||||
|
||||
pub struct decoder? implements base.token_decoder(
|
||||
// quirk_enabled_allow_backslash_etc, an 8-element array, is indexed by the
|
||||
// same enum as LUT_QUIRKY_BACKSLASHES.
|
||||
quirk_enabled_allow_backslash_etc : array[8] base.bool,
|
||||
|
||||
quirk_enabled_allow_ascii_control_codes : base.bool,
|
||||
quirk_enabled_allow_backslash_capital_u : base.bool,
|
||||
quirk_enabled_allow_backslash_x : base.bool,
|
||||
quirk_enabled_allow_comment_block : base.bool,
|
||||
quirk_enabled_allow_comment_line : base.bool,
|
||||
quirk_enabled_allow_extra_comma : base.bool,
|
||||
quirk_enabled_allow_inf_nan_numbers : base.bool,
|
||||
quirk_enabled_allow_leading_ascii_record_separator : base.bool,
|
||||
quirk_enabled_allow_leading_unicode_byte_order_mark : base.bool,
|
||||
quirk_enabled_allow_trailing_new_line : base.bool,
|
||||
quirk_enabled_replace_invalid_unicode : base.bool,
|
||||
quirks : array[QUIRKS_COUNT] base.bool,
|
||||
|
||||
allow_leading_ars : base.bool,
|
||||
allow_leading_ubom : base.bool,
|
||||
@@ -68,42 +54,11 @@ pub struct decoder? implements base.token_decoder(
|
||||
)
|
||||
|
||||
pub func decoder.set_quirk_enabled!(quirk: base.u32, enabled: base.bool) {
|
||||
if args.quirk == QUIRK_ALLOW_ASCII_CONTROL_CODES {
|
||||
this.quirk_enabled_allow_ascii_control_codes = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_A {
|
||||
this.quirk_enabled_allow_backslash_etc[1] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_CAPITAL_U {
|
||||
this.quirk_enabled_allow_backslash_capital_u = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_E {
|
||||
this.quirk_enabled_allow_backslash_etc[2] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_NEW_LINE {
|
||||
this.quirk_enabled_allow_backslash_etc[3] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_QUESTION_MARK {
|
||||
this.quirk_enabled_allow_backslash_etc[4] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_SINGLE_QUOTE {
|
||||
this.quirk_enabled_allow_backslash_etc[5] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_V {
|
||||
this.quirk_enabled_allow_backslash_etc[6] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_X {
|
||||
this.quirk_enabled_allow_backslash_x = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_BACKSLASH_ZERO {
|
||||
this.quirk_enabled_allow_backslash_etc[7] = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_COMMENT_BLOCK {
|
||||
this.quirk_enabled_allow_comment_block = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_COMMENT_LINE {
|
||||
this.quirk_enabled_allow_comment_line = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_EXTRA_COMMA {
|
||||
this.quirk_enabled_allow_extra_comma = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_INF_NAN_NUMBERS {
|
||||
this.quirk_enabled_allow_inf_nan_numbers = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_LEADING_ASCII_RECORD_SEPARATOR {
|
||||
this.quirk_enabled_allow_leading_ascii_record_separator = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_LEADING_UNICODE_BYTE_ORDER_MARK {
|
||||
this.quirk_enabled_allow_leading_unicode_byte_order_mark = args.enabled
|
||||
} else if args.quirk == QUIRK_ALLOW_TRAILING_NEW_LINE {
|
||||
this.quirk_enabled_allow_trailing_new_line = args.enabled
|
||||
} else if args.quirk == QUIRK_REPLACE_INVALID_UNICODE {
|
||||
this.quirk_enabled_replace_invalid_unicode = args.enabled
|
||||
if args.quirk >= QUIRKS_BASE {
|
||||
args.quirk -= QUIRKS_BASE
|
||||
if args.quirk < QUIRKS_COUNT {
|
||||
this.quirks[args.quirk] = args.enabled
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@@ -161,8 +116,8 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
return base."@end of data"
|
||||
}
|
||||
|
||||
if this.quirk_enabled_allow_leading_ascii_record_separator or
|
||||
this.quirk_enabled_allow_leading_unicode_byte_order_mark {
|
||||
if this.quirks[QUIRK_ALLOW_LEADING_ASCII_RECORD_SEPARATOR - QUIRKS_BASE] or
|
||||
this.quirks[QUIRK_ALLOW_LEADING_UNICODE_BYTE_ORDER_MARK - QUIRKS_BASE] {
|
||||
this.decode_leading?(dst: args.dst, src: args.src)
|
||||
}
|
||||
|
||||
@@ -370,11 +325,11 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
continue.string_loop_outer
|
||||
|
||||
} else if backslash <> 0 {
|
||||
if this.quirk_enabled_allow_backslash_etc[backslash & 7] {
|
||||
if this.quirks[LUT_QUIRKY_BACKSLASHES_QUIRKS[backslash & 7]] {
|
||||
args.src.skip32_fast!(actual: 2, worst_case: 2)
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_0000 | (LUT_QUIRKY_BACKSLASHES[backslash & 7] as base.u32),
|
||||
value_minor: 0x60_0000 | (LUT_QUIRKY_BACKSLASHES_CHARS[backslash & 7] as base.u32),
|
||||
continued: 1,
|
||||
length: 2)
|
||||
continue.string_loop_outer
|
||||
@@ -435,7 +390,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
// offset of 4.
|
||||
if args.src.available() < 12 {
|
||||
if args.src.is_closed() {
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.src.skip32_fast!(actual: 6, worst_case: 6)
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
@@ -497,7 +452,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
}
|
||||
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
if args.src.available() < 6 {
|
||||
return "#internal error: inconsistent I/O"
|
||||
}
|
||||
@@ -512,7 +467,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
// -------- END backslash-u.
|
||||
|
||||
} else if (c == 'U') and
|
||||
this.quirk_enabled_allow_backslash_capital_u {
|
||||
this.quirks[QUIRK_ALLOW_BACKSLASH_CAPITAL_U - QUIRKS_BASE] {
|
||||
// -------- BEGIN backslash-capital-u.
|
||||
if args.src.available() < 10 {
|
||||
if args.src.is_closed() {
|
||||
@@ -566,7 +521,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
continued: 1,
|
||||
length: 10)
|
||||
continue.string_loop_outer
|
||||
} else if this.quirk_enabled_replace_invalid_unicode {
|
||||
} else if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.src.skip32_fast!(actual: 10, worst_case: 10)
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
@@ -578,7 +533,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
// -------- END backslash-capital-u.
|
||||
|
||||
} else if (c == 'x') and
|
||||
this.quirk_enabled_allow_backslash_x {
|
||||
this.quirks[QUIRK_ALLOW_BACKSLASH_X - QUIRKS_BASE] {
|
||||
// -------- BEGIN backslash-x
|
||||
if args.src.available() < 4 {
|
||||
if args.src.is_closed() {
|
||||
@@ -640,7 +595,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
}
|
||||
if args.src.is_closed() {
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_FFFD,
|
||||
@@ -688,7 +643,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
}
|
||||
if args.src.is_closed() {
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_FFFD,
|
||||
@@ -741,7 +696,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
}
|
||||
if args.src.is_closed() {
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_FFFD,
|
||||
@@ -792,7 +747,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
}
|
||||
if (char & 0x80) <> 0 {
|
||||
if this.quirk_enabled_allow_ascii_control_codes {
|
||||
if this.quirks[QUIRK_ALLOW_ASCII_CONTROL_CODES - QUIRKS_BASE] {
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_0000 | ((char & 0x7F) as base.u32),
|
||||
@@ -803,7 +758,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
}
|
||||
return "#bad C0 control code"
|
||||
}
|
||||
if this.quirk_enabled_replace_invalid_unicode {
|
||||
if this.quirks[QUIRK_REPLACE_INVALID_UNICODE - QUIRKS_BASE] {
|
||||
args.dst.write_simple_token_fast!(
|
||||
value_major: 0,
|
||||
value_minor: 0x60_FFFD,
|
||||
@@ -859,13 +814,13 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
// What's valid after a comma depends on whether or not we're in an
|
||||
// array or an object.
|
||||
if 0 == (expect & ((1 as base.u32) << CLASS_CLOSE_SQUARE_BRACKET)) {
|
||||
if this.quirk_enabled_allow_extra_comma {
|
||||
if this.quirks[QUIRK_ALLOW_EXTRA_COMMA - QUIRKS_BASE] {
|
||||
expect = 0x1042 // 0x1042 is EXPECT_STRING | EXPECT_CLOSE_CURLY_BRACE.
|
||||
} else {
|
||||
expect = 0x1002 // 0x1002 is EXPECT_STRING.
|
||||
}
|
||||
} else {
|
||||
if this.quirk_enabled_allow_extra_comma {
|
||||
if this.quirks[QUIRK_ALLOW_EXTRA_COMMA - QUIRKS_BASE] {
|
||||
expect = 0x1FB2 // 0x0FB2 is EXPECT_VALUE | EXPECT_CLOSE_SQUARE_BRACKET.
|
||||
} else {
|
||||
expect = 0x1EB2 // 0x0EB2 is EXPECT_VALUE.
|
||||
@@ -915,7 +870,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
} endwhile
|
||||
|
||||
if number_status == 1 {
|
||||
if this.quirk_enabled_allow_inf_nan_numbers {
|
||||
if this.quirks[QUIRK_ALLOW_INF_NAN_NUMBERS - QUIRKS_BASE] {
|
||||
this.decode_inf_nan?(dst: args.dst, src: args.src)
|
||||
break
|
||||
}
|
||||
@@ -1110,13 +1065,14 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
continue.outer
|
||||
}
|
||||
|
||||
if this.quirk_enabled_allow_inf_nan_numbers {
|
||||
if this.quirks[QUIRK_ALLOW_INF_NAN_NUMBERS - QUIRKS_BASE] {
|
||||
this.decode_inf_nan?(dst: args.dst, src: args.src)
|
||||
break.goto_parsed_a_leaf_value
|
||||
}
|
||||
|
||||
} else if class == CLASS_COMMENT {
|
||||
if this.quirk_enabled_allow_comment_block or this.quirk_enabled_allow_comment_line {
|
||||
if this.quirks[QUIRK_ALLOW_COMMENT_BLOCK - QUIRKS_BASE] or
|
||||
this.quirks[QUIRK_ALLOW_COMMENT_LINE - QUIRKS_BASE] {
|
||||
this.decode_comment?(dst: args.dst, src: args.src)
|
||||
continue.outer
|
||||
}
|
||||
@@ -1133,7 +1089,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader, wor
|
||||
expect = expect_after_value
|
||||
} endwhile.outer
|
||||
|
||||
if this.quirk_enabled_allow_trailing_new_line {
|
||||
if this.quirks[QUIRK_ALLOW_TRAILING_NEW_LINE - QUIRKS_BASE] {
|
||||
this.decode_trailing_new_line?(dst: args.dst, src: args.src)
|
||||
}
|
||||
|
||||
@@ -1319,8 +1275,10 @@ pri func decoder.decode_leading?(dst: base.token_writer, src: base.io_reader) {
|
||||
var c : base.u8
|
||||
var u : base.u32
|
||||
|
||||
this.allow_leading_ars = this.quirk_enabled_allow_leading_ascii_record_separator
|
||||
this.allow_leading_ubom = this.quirk_enabled_allow_leading_unicode_byte_order_mark
|
||||
this.allow_leading_ars =
|
||||
this.quirks[QUIRK_ALLOW_LEADING_ASCII_RECORD_SEPARATOR - QUIRKS_BASE]
|
||||
this.allow_leading_ubom =
|
||||
this.quirks[QUIRK_ALLOW_LEADING_UNICODE_BYTE_ORDER_MARK - QUIRKS_BASE]
|
||||
|
||||
while this.allow_leading_ars or this.allow_leading_ubom {
|
||||
if args.dst.available() <= 0 {
|
||||
@@ -1382,7 +1340,7 @@ pri func decoder.decode_comment?(dst: base.token_writer, src: base.io_reader) {
|
||||
} endwhile
|
||||
c2 = args.src.peek_u16le()
|
||||
|
||||
if (c2 == '/*'le) and this.quirk_enabled_allow_comment_block {
|
||||
if (c2 == '/*'le) and this.quirks[QUIRK_ALLOW_COMMENT_BLOCK - QUIRKS_BASE] {
|
||||
args.src.skip32_fast!(actual: 2, worst_case: 2)
|
||||
length = 2
|
||||
|
||||
@@ -1437,7 +1395,7 @@ pri func decoder.decode_comment?(dst: base.token_writer, src: base.io_reader) {
|
||||
} endwhile
|
||||
} endwhile.comment_block
|
||||
|
||||
} else if (c2 == '//'le) and this.quirk_enabled_allow_comment_line {
|
||||
} else if (c2 == '//'le) and this.quirks[QUIRK_ALLOW_COMMENT_LINE - QUIRKS_BASE] {
|
||||
args.src.skip32_fast!(actual: 2, worst_case: 2)
|
||||
length = 2
|
||||
|
||||
|
||||
@@ -18,6 +18,7 @@
|
||||
//
|
||||
// The base38 encoding of "json" is 0x12_4265. Left shifting by 10 gives
|
||||
// 0x4909_9400.
|
||||
pri const QUIRKS_BASE : base.u32 = 0x4909_9400
|
||||
|
||||
// --------
|
||||
|
||||
@@ -204,3 +205,5 @@ pub const QUIRK_ALLOW_TRAILING_NEW_LINE : base.u32 = 0x4909_9400 | 0x10
|
||||
// unit that is an invalid Unicode code point (i.e. in the range U+D800 ..=
|
||||
// U+DFFF or above U+10FFFF) is similarly replaced with U+FFFD.
|
||||
pub const QUIRK_REPLACE_INVALID_UNICODE : base.u32 = 0x4909_9400 | 0x11
|
||||
|
||||
pri const QUIRKS_COUNT : base.u32 = 0x12
|
||||
Reference in new issue
Block a user