diff --git a/release/c/wuffs-unsupported-snapshot.c b/release/c/wuffs-unsupported-snapshot.c index db52d6ef..a6767555 100644 --- a/release/c/wuffs-unsupported-snapshot.c +++ b/release/c/wuffs-unsupported-snapshot.c @@ -18150,6 +18150,11 @@ static uint32_t // wuffs_json__decoder__decode_number(wuffs_json__decoder* self, wuffs_base__io_buffer* a_src); +static uint32_t // +wuffs_json__decoder__decode_digits(wuffs_json__decoder* self, + wuffs_base__io_buffer* a_src, + uint32_t a_n); + // ---------------- VTables // ---------------- Initializer Implementations @@ -18450,7 +18455,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, v_expect = 3; } goto label_0_continue; - } else if (v_c >= 48) { + } else if ((v_c >= 48) || (v_c == 45)) { while (true) { if (a_src) { a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); @@ -18678,7 +18683,127 @@ wuffs_json__decoder__decode_number(wuffs_json__decoder* self, io2_a_src = io0_a_src + a_src->meta.wi; } - v_n = 0; + while (true) { + v_n = 0; + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (!(a_src && a_src->meta.closed)) { + v_n |= 768; + } + goto label_0_break; + } + v_c = wuffs_base__load_u8be(iop_a_src); + if (v_c != 45) { + } else { + v_n += 1; + (iop_a_src += 1, wuffs_base__make_empty_struct()); + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (!(a_src && a_src->meta.closed)) { + v_n |= 768; + } + v_n |= 256; + goto label_0_break; + } + v_c = wuffs_base__load_u8be(iop_a_src); + } + if (v_c == 48) { + v_n += 1; + (iop_a_src += 1, wuffs_base__make_empty_struct()); + } else { + if (a_src) { + a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); + } + v_n = wuffs_json__decoder__decode_digits(self, a_src, v_n); + if (a_src) { + iop_a_src = a_src->data.ptr + a_src->meta.ri; + } + if (v_n > 253) { + goto label_0_break; + } + } + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (!(a_src && a_src->meta.closed)) { + v_n |= 768; + } + goto label_0_break; + } + v_c = wuffs_base__load_u8be(iop_a_src); + if (v_c != 46) { + } else { + v_n += 1; + (iop_a_src += 1, wuffs_base__make_empty_struct()); + if (a_src) { + a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); + } + v_n = wuffs_json__decoder__decode_digits(self, a_src, v_n); + if (a_src) { + iop_a_src = a_src->data.ptr + a_src->meta.ri; + } + if (v_n > 253) { + goto label_0_break; + } + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (!(a_src && a_src->meta.closed)) { + v_n |= 768; + } + goto label_0_break; + } + v_c = wuffs_base__load_u8be(iop_a_src); + } + if ((v_c != 69) && (v_c != 101)) { + goto label_0_break; + } + v_n += 1; + (iop_a_src += 1, wuffs_base__make_empty_struct()); + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (!(a_src && a_src->meta.closed)) { + v_n |= 768; + } + v_n |= 256; + goto label_0_break; + } + v_c = wuffs_base__load_u8be(iop_a_src); + if ((v_c != 43) && (v_c != 45)) { + } else { + v_n += 1; + (iop_a_src += 1, wuffs_base__make_empty_struct()); + } + if (a_src) { + a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); + } + v_n = wuffs_json__decoder__decode_digits(self, a_src, v_n); + if (a_src) { + iop_a_src = a_src->data.ptr + a_src->meta.ri; + } + goto label_0_break; + } +label_0_break:; + if (a_src) { + a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); + } + return v_n; +} + +// -------- func json.decoder.decode_digits + +static uint32_t // +wuffs_json__decoder__decode_digits(wuffs_json__decoder* self, + wuffs_base__io_buffer* a_src, + uint32_t a_n) { + uint8_t v_c = 0; + uint32_t v_n = 0; + + uint8_t* iop_a_src = NULL; + uint8_t* io0_a_src WUFFS_BASE__POTENTIALLY_UNUSED = NULL; + uint8_t* io1_a_src WUFFS_BASE__POTENTIALLY_UNUSED = NULL; + uint8_t* io2_a_src WUFFS_BASE__POTENTIALLY_UNUSED = NULL; + if (a_src) { + io0_a_src = a_src->data.ptr; + io1_a_src = io0_a_src + a_src->meta.ri; + iop_a_src = io1_a_src; + io2_a_src = io0_a_src + a_src->meta.wi; + } + + v_n = a_n; while (true) { if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { if (!(a_src && a_src->meta.closed)) { @@ -18698,6 +18823,9 @@ wuffs_json__decoder__decode_number(wuffs_json__decoder* self, (iop_a_src += 1, wuffs_base__make_empty_struct()); } label_0_break:; + if (v_n == a_n) { + v_n |= 256; + } if (a_src) { a_src->meta.ri = ((size_t)(iop_a_src - a_src->data.ptr)); } diff --git a/std/json/decode_json.wuffs b/std/json/decode_json.wuffs index 830048a6..0739b823 100644 --- a/std/json/decode_json.wuffs +++ b/std/json/decode_json.wuffs @@ -246,7 +246,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { } continue.outer - } else if c >= 0x30 { // 0x30 is '0'. + } else if (c >= 0x30) or (c == 0x2D) { // 0x30 is '0', 0x2D is '-'. // -------- BEGIN parse numbers. while true, pre args.dst.available() > 0, @@ -399,10 +399,128 @@ pri func decoder.decode_number!(src: base.io_reader) base.u32[..= 0x3FF] { var c : base.u8 var n : base.u32[..= 0x3FF] - // TODO: fractions, exponents, leading '-' sign, special casing when the - // first digit is '0'. + while.goto_done true { + n = 0 - n = 0 + // Peek. + if args.src.available() <= 0 { + if not args.src.is_closed() { + n |= 0x300 + } + break.goto_done + } + c = args.src.peek_u8() + + // Scan the optional minus sign. + if c <> 0x2D { // 0x2D is '-'. + assert args.src.available() > 0 + assert n <= 1 + } else { + n += 1 + args.src.skip32_fast!(actual: 1, worst_case: 1) + + // Peek. + if args.src.available() <= 0 { + if not args.src.is_closed() { + n |= 0x300 + } + n |= 0x100 // A '-' without digits is invalid. + break.goto_done + } + c = args.src.peek_u8() + + assert args.src.available() > 0 + assert n <= 1 + } + + // Scan the opening digits. + if c == 0x30 { // 0x30 is '0'. + n += 1 + args.src.skip32_fast!(actual: 1, worst_case: 1) + assert n <= 0xFD + } else { + n = this.decode_digits!(src: args.src, n: n) + if n > 0xFD { + break.goto_done + } + assert n <= 0xFD + } + + // Peek. + if args.src.available() <= 0 { + if not args.src.is_closed() { + n |= 0x300 + } + break.goto_done + } + c = args.src.peek_u8() + + // Scan the optional fraction. + if c <> 0x2E { // 0x2E is '.'. + assert args.src.available() > 0 + assert n <= 0xFD + } else { + n += 1 + args.src.skip32_fast!(actual: 1, worst_case: 1) + + n = this.decode_digits!(src: args.src, n: n) + if n > 0xFD { + break.goto_done + } + + // Peek. + if args.src.available() <= 0 { + if not args.src.is_closed() { + n |= 0x300 + } + break.goto_done + } + c = args.src.peek_u8() + + assert args.src.available() > 0 + assert n <= 0xFD + } + + // Scan the optional 'E' or 'e'. + if (c <> 0x45) and (c <> 0x65) { // 0x45 and 0x65 are 'E' and 'e'. + break.goto_done + } + n += 1 + args.src.skip32_fast!(actual: 1, worst_case: 1) + assert n <= 0xFE + + // Peek. + if args.src.available() <= 0 { + if not args.src.is_closed() { + n |= 0x300 + } + n |= 0x100 // An 'E' or 'e' without digits is invalid. + break.goto_done + } + c = args.src.peek_u8() + + // Scan the optional '+' or '-'. + if (c <> 0x2B) and (c <> 0x2D) { // 0x2B and 0x2D are '+' and '-'. + assert n <= 255 + } else { + n += 1 + args.src.skip32_fast!(actual: 1, worst_case: 1) + assert n <= 255 + } + + // Scan the exponent digits. + n = this.decode_digits!(src: args.src, n: n) + + break.goto_done + } // while.goto_done + return n +} + +pri func decoder.decode_digits!(src: base.io_reader, n: base.u32[..= 0xFF]) base.u32[..= 0x3FF] { + var c : base.u8 + var n : base.u32[..= 0x3FF] + + n = args.n while true { if args.src.available() <= 0 { if not args.src.is_closed() { @@ -421,5 +539,8 @@ pri func decoder.decode_number!(src: base.io_reader) base.u32[..= 0x3FF] { n += 1 args.src.skip32_fast!(actual: 1, worst_case: 1) } + if n == args.n { + n |= 0x100 + } return n }