From 91454e911e680d6e408547341f41f0a50fef0cc4 Mon Sep 17 00:00:00 2001 From: Nigel Tao Date: Thu, 13 Feb 2020 13:52:41 +1100 Subject: [PATCH] Have std/json decode '"' --- release/c/wuffs-unsupported-snapshot.c | 71 +++++++++++++++++++++++++- std/json/decode_json.wuffs | 69 ++++++++++++++++++++++++- 2 files changed, 136 insertions(+), 4 deletions(-) diff --git a/release/c/wuffs-unsupported-snapshot.c b/release/c/wuffs-unsupported-snapshot.c index fb86d266..be5b3b94 100644 --- a/release/c/wuffs-unsupported-snapshot.c +++ b/release/c/wuffs-unsupported-snapshot.c @@ -18223,6 +18223,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, self->private_impl.active_coroutine = 0; wuffs_base__status status = wuffs_base__make_status(NULL); + uint32_t v_string_length = 0; uint32_t v_whitespace_length = 0; uint32_t v_depth = 0; uint32_t v_stack_byte = 0; @@ -18322,8 +18323,74 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } if (v_c <= 57) { if (v_c == 34) { - status = wuffs_base__make_status(wuffs_json__error__bad_input); - goto exit; + if (0 == (v_expect & 2)) { + status = wuffs_base__make_status(wuffs_json__error__bad_input); + goto exit; + } + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + (iop_a_src += 1, wuffs_base__make_empty_struct()); + label_2_continue:; + while (true) { + if (((uint64_t)(io2_a_dst - iop_a_dst)) <= 0) { + status = + wuffs_base__make_status(wuffs_base__suspension__short_write); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(3); + goto label_2_continue; + } + v_string_length = 0; + while (true) { + if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { + if (v_string_length > 0) { + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(6291457)) + << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(((uint64_t)(v_string_length)))) + << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + v_string_length = 0; + } + goto label_2_continue; + } + v_c = wuffs_base__load_u8be(iop_a_src); + (iop_a_src += 1, wuffs_base__make_empty_struct()); + if (v_c == 34) { + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(6291456)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(((uint64_t)(v_string_length)))) + << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + v_string_length = 0; + goto label_2_break; + } else if (v_c == 92) { + status = wuffs_base__make_status(wuffs_json__error__bad_input); + goto exit; + } else if (v_string_length >= 65534) { + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(6291457)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(65535)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + v_string_length = 0; + goto label_2_continue; + } + v_string_length += 1; + } + } + label_2_break:; + while (true) { + if (((uint64_t)(io2_a_dst - iop_a_dst)) > 0) { + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + goto label_3_break; + } + status = + wuffs_base__make_status(wuffs_base__suspension__short_write); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(4); + } + label_3_break:; + if (0 == (v_expect & 1)) { + v_expect = 8; + goto label_0_continue; + } } else { status = wuffs_base__make_status(wuffs_json__error__bad_input); goto exit; diff --git a/std/json/decode_json.wuffs b/std/json/decode_json.wuffs index 38ebaeb1..fd5c3c09 100644 --- a/std/json/decode_json.wuffs +++ b/std/json/decode_json.wuffs @@ -50,6 +50,7 @@ pub struct decoder?( ) pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { + var string_length : base.u32[..= 0xFFFE] var whitespace_length : base.u32[..= 0xFFFE] var depth : base.u32[..= 1024] var stack_byte : base.u32[..= (1024 / 32) - 1] @@ -135,9 +136,72 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if c <= 0x39 { // 0x39 is '9'. if c == 0x22 { // 0x22 is '"'. - // TODO: strings. + // -------- BEGIN parse strings. + if 0 == (expect & 0x02) { // 0x02 is EXPECT_STRING_VALUE. + return "#bad input" + } - return "#bad input" + // The leading '"' is filler. + args.dst.write_fast_token!(value: 0, length: 1) + args.src.skip32_fast!(actual: 1, worst_case: 1) + + while.string_loop true { + if args.dst.available() <= 0 { + yield? base."$short write" + continue.string_loop + } + + string_length = 0 + while true, + pre args.dst.available() > 0, + { + if args.src.available() <= 0 { + if string_length > 0 { + args.dst.write_fast_token!(value: 0x60_0001, length: string_length as base.u64) + string_length = 0 + } + continue.string_loop + } + c = args.src.peek_u8() + args.src.skip32_fast!(actual: 1, worst_case: 1) + + if c == 0x22 { // 0x22 is '"'. + args.dst.write_fast_token!(value: 0x60_0000, length: string_length as base.u64) + string_length = 0 + break.string_loop + + } else if c == 0x5C { // 0x5C is '\\'. + // TODO: escapes. + return "#bad input" + + } else if string_length >= 0xFFFE { + args.dst.write_fast_token!(value: 0x60_0001, length: 0xFFFF) + string_length = 0 + continue.string_loop + } + + string_length += 1 + } + } + + // The trailing '"' is filler. + while true { + if args.dst.available() > 0 { + args.dst.write_fast_token!(value: 0, length: 1) + break + } + yield? base."$short write" + } + + // As above, expect must have contained EXPECT_STRING_VALUE. If + // it didn't also contain EXPECT_NON_STRING_VALUE then we were + // parsing an object key and the next token should be ':'. + // Otherwise, fall below to "We've just parsed a value". + if 0 == (expect & 0x01) { // 0x01 is EXPECT_NON_STRING_VALUE. + expect = 0x08 // 0x08 is EXPECT_COLON. + continue.outer + } + // -------- END parse strings. } else { // TODO: numbers. @@ -181,6 +245,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { return "#bad input" } + // We've just parsed a value. if depth == 0 { break.outer }