From b43e2493a2f644e1f1424e42fff2e00d9fdc4c83 Mon Sep 17 00:00:00 2001 From: Nigel Tao Date: Sun, 22 Mar 2020 21:41:04 +1100 Subject: [PATCH] Add std/json quirk_allow_final_comma --- release/c/wuffs-unsupported-snapshot.c | 12 ++++- std/json/decode_json.wuffs | 12 ++++- test/c/std/json.c | 69 ++++++++++++++++++++++++++ 3 files changed, 89 insertions(+), 4 deletions(-) diff --git a/release/c/wuffs-unsupported-snapshot.c b/release/c/wuffs-unsupported-snapshot.c index fe922b05..b615ecc3 100644 --- a/release/c/wuffs-unsupported-snapshot.c +++ b/release/c/wuffs-unsupported-snapshot.c @@ -21320,9 +21320,17 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); if (0 == (v_expect & 256)) { - v_expect = 4098; + if (self->private_impl.f_quirk_enabled_allow_final_comma) { + v_expect = 4162; + } else { + v_expect = 4098; + } } else { - v_expect = 7858; + if (self->private_impl.f_quirk_enabled_allow_final_comma) { + v_expect = 8114; + } else { + v_expect = 7858; + } } goto label__outer__continue; } else if (v_class == 3) { diff --git a/std/json/decode_json.wuffs b/std/json/decode_json.wuffs index 176fd095..566811d6 100644 --- a/std/json/decode_json.wuffs +++ b/std/json/decode_json.wuffs @@ -832,9 +832,17 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { // What's valid after a comma depends on whether or not we're // in an array or an object. if 0 == (expect & 0x0100) { // 0x0100 is (1 << CLASS_CLOSE_SQUARE_BRACKET). - expect = 0x1002 // 0x1002 is EXPECT_STRING. + if this.quirk_enabled_allow_final_comma { + expect = 0x1042 // 0x1042 is EXPECT_STRING | EXPECT_CLOSE_CURLY_BRACE. + } else { + expect = 0x1002 // 0x1002 is EXPECT_STRING. + } } else { - expect = 0x1EB2 // 0x0EB2 is EXPECT_VALUE. + if this.quirk_enabled_allow_final_comma { + expect = 0x1FB2 // 0x0FB2 is EXPECT_VALUE | EXPECT_CLOSE_SQUARE_BRACKET. + } else { + expect = 0x1EB2 // 0x0EB2 is EXPECT_VALUE. + } } continue.outer diff --git a/test/c/std/json.c b/test/c/std/json.c index e345344b..f1bfc1ab 100644 --- a/test/c/std/json.c +++ b/test/c/std/json.c @@ -1329,6 +1329,74 @@ test_wuffs_json_decode_quirk_allow_backslash_x() { return NULL; } +const char* // +test_wuffs_json_decode_quirk_allow_final_comma() { + CHECK_FOCUS(__func__); + + struct { + // want has 2 bytes, one for each possible q: + // - q&1 sets WUFFS_JSON__QUIRK_ALLOW_FINAL_COMMA. + // An 'X', '+' or '-' means that decoding should succeed (and consume the + // entire input), succeed (without consuming the entire input) or fail. + const char* want; + const char* str; + } test_cases[] = { + {.want = "-X", .str = "[0,]"}, + {.want = "-X", .str = "[[], {},{\"k\":\"v\",\n}\n,\n]"}, + {.want = "--", .str = "[,]"}, + {.want = "--", .str = "{,}"}, + }; + + int tc; + for (tc = 0; tc < WUFFS_TESTLIB_ARRAY_SIZE(test_cases); tc++) { + int q; + for (q = 0; q < 2; q++) { + wuffs_json__decoder dec; + CHECK_STATUS("initialize", wuffs_json__decoder__initialize( + &dec, sizeof dec, WUFFS_VERSION, + WUFFS_INITIALIZE__DEFAULT_OPTIONS)); + wuffs_json__decoder__set_quirk_enabled( + &dec, WUFFS_JSON__QUIRK_ALLOW_FINAL_COMMA, q & 1); + + wuffs_base__token_buffer tok = + wuffs_base__make_token_buffer_writer(global_have_token_slice); + wuffs_base__io_buffer src = wuffs_base__make_io_buffer_reader( + wuffs_base__make_slice_u8((void*)(test_cases[tc].str), + strlen(test_cases[tc].str)), + true); + const char* have = + wuffs_json__decoder__decode_tokens(&dec, &tok, &src).repr; + const char* want = + (test_cases[tc].want[q] != '-') ? NULL : wuffs_json__error__bad_input; + if (have != want) { + RETURN_FAIL("tc=%d, q=%d: decode_tokens: have \"%s\", want \"%s\"", tc, + q, have, want); + } + + size_t total_length = 0; + while (tok.meta.ri < tok.meta.wi) { + total_length += wuffs_base__token__length(&tok.data.ptr[tok.meta.ri++]); + } + if (total_length != src.meta.ri) { + RETURN_FAIL("tc=%d, q=%d: total_length: have %zu, want %zu", tc, q, + total_length, src.meta.ri); + } + if (test_cases[tc].want[q] == 'X') { + if (total_length != src.data.len) { + RETURN_FAIL("tc=%d, q=%d: total_length: have %zu, want %zu", tc, q, + total_length, src.data.len); + } + } else if (test_cases[tc].want[q] == '+') { + if (total_length >= src.data.len) { + RETURN_FAIL("tc=%d, q=%d: total_length: have %zu, want < %zu", tc, q, + total_length, src.data.len); + } + } + } + } + return NULL; +} + const char* // test_wuffs_json_decode_quirk_allow_comment_etc() { CHECK_FOCUS(__func__); @@ -2046,6 +2114,7 @@ proc tests[] = { test_wuffs_json_decode_quirk_allow_backslash_etc, test_wuffs_json_decode_quirk_allow_backslash_x, test_wuffs_json_decode_quirk_allow_comment_etc, + test_wuffs_json_decode_quirk_allow_final_comma, test_wuffs_json_decode_quirk_allow_leading_etc, test_wuffs_json_decode_quirk_allow_trailing_etc, test_wuffs_json_decode_quirk_replace_invalid_utf_8,