diff --git a/example/jsonptr/jsonptr.cc b/example/jsonptr/jsonptr.cc index c3531cab..c4b28ce6 100644 --- a/example/jsonptr/jsonptr.cc +++ b/example/jsonptr/jsonptr.cc @@ -99,8 +99,6 @@ wuffs_base__token_buffer tok; // token. An invariant is that (curr_token_end_src_index <= src.meta.ri). size_t curr_token_end_src_index; -bool prev_token_incomplete; - uint64_t depth; enum class context { @@ -132,8 +130,6 @@ initialize_globals(int argc, char** argv) { curr_token_end_src_index = 0; - prev_token_incomplete = false; - depth = 0; ctx = context::none; @@ -319,19 +315,20 @@ handle_token(wuffs_base__token t) { goto after_value; } - // Write preceding whitespace and punctuation, if it wasn't ']' or '}'. - if (!prev_token_incomplete) { - if (ctx == context::in_dict_after_key) { - TRY(write_dst(": ", 2)); - } else if (ctx != context::none) { - if ((ctx != context::in_list_after_bracket) && - (ctx != context::in_dict_after_brace)) { - TRY(write_dst(",", 1)); - } - TRY(write_dst("\n", 1)); - for (size_t i = 0; i < depth; i++) { - TRY(write_dst(INDENT_STRING, indent)); - } + // Write preceding whitespace and punctuation, if it wasn't ']', '}' or a + // continuation of a multi-token chain. + if (t.link_prev()) { + // No-op. + } else if (ctx == context::in_dict_after_key) { + TRY(write_dst(": ", 2)); + } else if (ctx != context::none) { + if ((ctx != context::in_list_after_bracket) && + (ctx != context::in_dict_after_brace)) { + TRY(write_dst(",", 1)); + } + TRY(write_dst("\n", 1)); + for (size_t i = 0; i < depth; i++) { + TRY(write_dst(INDENT_STRING, indent)); } } @@ -339,7 +336,7 @@ handle_token(wuffs_base__token t) { // value (number, string or literal). switch (vbc) { case WUFFS_BASE__TOKEN__VBC__STRING: - if (!prev_token_incomplete) { + if (!t.link_prev()) { TRY(write_dst("\"", 1)); } @@ -352,9 +349,7 @@ handle_token(wuffs_base__token t) { return "main: internal error: unexpected string-token conversion"; } - prev_token_incomplete = - vbd & WUFFS_BASE__TOKEN__VBD__STRING__INCOMPLETE; - if (prev_token_incomplete) { + if (t.link_next()) { return nullptr; } TRY(write_dst("\"", 1)); diff --git a/internal/cgen/base/token-public.h b/internal/cgen/base/token-public.h index 389cab65..b34f073a 100644 --- a/internal/cgen/base/token-public.h +++ b/internal/cgen/base/token-public.h @@ -17,22 +17,52 @@ // ---------------- Tokens typedef struct { - // The repr is divided as: - // - Bits 63 .. 40 (24 bits) is the major value. - // - Bits 39 .. 16 (24 bits) is the minor value. + // The repr's 64 bits are divided as: + // + // +-----+-------------+-------+-------------+-----+-----+-----------+ + // | 1 | 21 | 3 | 21 | 1 | 1 | 16 | + // +-----+-------------+-------+-------------+-----+-----+-----------+ + // [..................value..................] LP LN length + // [..0..|.value_major.|.....value_minor.....] + // [..0..|.........VBC.........|.....VBD.....] + // + // The broad divisions are: + // - Bits 63 .. 18 (46 bits) is the value. + // - Bits 17 .. 16 ( 2 bits) is LP and LN (link_prev and link_next). // - Bits 15 .. 0 (16 bits) is the length. // - // The major value is a [Base38](doc/note/base38-and-fourcc.md) value. If - // zero (special cased for Wuffs' built-in "base" package) then the minor - // value is further sub-divided: - // - Bits 39 .. 37 ( 3 bits) is the value_base_category. - // - Bits 36 .. 16 (21 bits) is the value_base_detail. + // The value bits can be sub-divided in multiple ways: + // - Bits 63 .. 63 ( 1 bits) is reserved (a zero bit). + // - Bits 62 .. 42 (21 bits) is the value_major. + // - Bits 41 .. 18 (24 bits) is the value_minor. + // - Bits 62 .. 39 (24 bits) is the VBC (value_base_category). + // - Bits 38 .. 18 (21 bits) is the VBD (value_base_detail). // - // In particular, at 21 bits, the value_base_detail can hold every valid - // Unicode code point. + // The value_major is a 21-bit [Base38](doc/note/base38-and-fourcc.md) value. + // If all of its bits are zero (special cased for Wuffs' built-in "base" + // package) then the value_minor is further sub-divided: + // - Bits 41 .. 39 ( 3 bits) is the VBC (value_base_category). + // - Bits 38 .. 18 (21 bits) is the VBD (value_base_detail). // - // If the major value is non-zero then the minor value has whatever arbitrary - // meaning the tokenizer's package assigns to it. + // The high 46 bits (bits 63 .. 18) only have VBC and VBD semantics when the + // high 22 bits (the value_major) are all zero. An equivalent test is that + // the high 25 bits (the notional VBC) has a numerical value less than 8. + // + // At 21 bits, the VBD can hold every valid Unicode code point. + // + // If value_major is non-zero then value_minor has whatever arbitrary meaning + // the tokenizer's package assigns to it. + // + // Multiple consecutive tokens can form a larger conceptual unit. For + // example, an "abc\tz" string is a single higher level concept but at the + // lower level, it could consist of multiple tokens: the quotes '"', the + // ASCII texts "abc" and "z" and the backslash-escaped tab '\t'. The LP and + // LN (link_prev and link_next) bits denote tokens that are part of a + // multi-token chain: + // - LP means that this token is not the first (there is a previous token). + // - LN means that this token is not the last (there is a next token). + // + // In particular, a stand-alone token will have both link bits set to zero. uint64_t repr; #ifdef __cplusplus @@ -41,6 +71,8 @@ typedef struct { inline uint64_t value_minor() const; inline uint64_t value_base_category() const; inline uint64_t value_base_detail() const; + inline bool link_prev() const; + inline bool link_next() const; inline uint64_t length() const; #endif // __cplusplus @@ -55,20 +87,25 @@ wuffs_base__make_token(uint64_t repr) { // -------- -#define WUFFS_BASE__TOKEN__VALUE__MASK 0xFFFFFFFFFFFF -#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0xFFFFFF +#define WUFFS_BASE__TOKEN__VALUE__MASK 0x3FFFFFFFFFFF +#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0x3FFFFF #define WUFFS_BASE__TOKEN__VALUE_MINOR__MASK 0xFFFFFF -#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x7FFFFFF +#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x1FFFFFF #define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK 0x1FFFFF +#define WUFFS_BASE__TOKEN__LINK__MASK 0x3 #define WUFFS_BASE__TOKEN__LENGTH__MASK 0xFFFF -#define WUFFS_BASE__TOKEN__VALUE__SHIFT 16 -#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 40 -#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 16 -#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 37 -#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 16 +#define WUFFS_BASE__TOKEN__VALUE__SHIFT 18 +#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 42 +#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 18 +#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 39 +#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 18 +#define WUFFS_BASE__TOKEN__LINK__SHIFT 16 #define WUFFS_BASE__TOKEN__LENGTH__SHIFT 0 +#define WUFFS_BASE__TOKEN__LINK_PREV 0x20000 +#define WUFFS_BASE__TOKEN__LINK_NEXT 0x10000 + // -------- #define WUFFS_BASE__TOKEN__VBC__FILLER 0 @@ -77,41 +114,19 @@ wuffs_base__make_token(uint64_t repr) { #define WUFFS_BASE__TOKEN__VBC__NUMBER 3 #define WUFFS_BASE__TOKEN__VBC__STRUCTURE 4 -// -------- + // -------- -// INCOMPLETE means that this token combines with the following token. -// -// For example, tokenizing a comment that is longer than the source buffer can -// result in multiple VBC__FILLER tokens. All but the last one is INCOMPLETE. -// -// By convention, whitespace is not marked incomplete. Two whitespace tokens of -// length 30 are equivalent to one whitespace token of length 60. -// -// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between -// VBD_FILLER and VBD_STRING. -#define WUFFS_BASE__TOKEN__VBD__FILLER__INCOMPLETE 0x00001 - -#define WUFFS_BASE__TOKEN__VBD__FILLER__END_OF_CONSECUTIVE_COMMENTS 0x00010 -#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00020 -#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00040 +#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00001 +#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00002 // -------- -// INCOMPLETE means that this token combines with the following token. -// -// For example, tokenizing a string that is longer than the source buffer can -// result in multiple VBC__STRING tokens. All but the last one is INCOMPLETE. -// -// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between -// VBD_FILLER and VBD_STRING. -#define WUFFS_BASE__TOKEN__VBD__STRING__INCOMPLETE 0x00001 - -// "DEFINITELY_FOO" means that the source bytes (and also the destination -// bytes, assuming 1_DST_1_SRC_COPY) are in the FOO format. Definitely means -// that the lack of the bit is conservative: it is valid for all-ASCII strings -// to have neither DEFINITELY_ASCII or DEFINITELY_UTF_8 bits set. -#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00010 -#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00020 +// "DEFINITELY_FOO" means that the destination bytes (and also the source +// bytes, for 1_DST_1_SRC_COPY) are in the FOO format. Definitely means that +// the lack of the bit is conservative: it is valid for all-ASCII strings to +// have neither DEFINITELY_UTF_8 or DEFINITELY_ASCII bits set. +#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00001 +#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00002 // "CONVERT_D_DST_S_SRC" means that multiples of S source bytes (possibly // padded) produces multiples of D destination bytes. For example, @@ -121,21 +136,13 @@ wuffs_base__make_token(uint64_t repr) { // When src is the empty string, multiple conversion algorithms are applicable // (so these bits are not necessarily mutually exclusive), all producing the // same empty dst string. -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00100 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00200 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00400 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00800 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x01000 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x02000 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x04000 - - // -------- - - // UNICODE_CODE_POINT tokens have no room in their VBD for an INCOMPLETE bit. - // All 21 bits are used to hold the Unicode code point. Such tokens preserve - // the INCOMPLETEness of the previous token (if a FILLER or STRING). - -#define WUFFS_BASE__TOKEN__VBD__UNICODE_CODE_POINT__MAX_INCL 0x10FFFF +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00010 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00020 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00040 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00080 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x00100 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x00200 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x00400 // -------- @@ -201,6 +208,16 @@ wuffs_base__token__value_base_detail(const wuffs_base__token* t) { WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK; } +static inline bool // +wuffs_base__token__link_prev(const wuffs_base__token* t) { + return t->repr & WUFFS_BASE__TOKEN__LINK_PREV; +} + +static inline bool // +wuffs_base__token__link_next(const wuffs_base__token* t) { + return t->repr & WUFFS_BASE__TOKEN__LINK_NEXT; +} + static inline uint64_t // wuffs_base__token__length(const wuffs_base__token* t) { return (t->repr >> WUFFS_BASE__TOKEN__LENGTH__SHIFT) & @@ -234,6 +251,16 @@ wuffs_base__token::value_base_detail() const { return wuffs_base__token__value_base_detail(this); } +inline bool // +wuffs_base__token::link_prev() const { + return wuffs_base__token__link_prev(this); +} + +inline bool // +wuffs_base__token::link_next() const { + return wuffs_base__token__link_next(this); +} + inline uint64_t // wuffs_base__token::length() const { return wuffs_base__token__length(this); diff --git a/internal/cgen/builtin.go b/internal/cgen/builtin.go index 55562a4c..a6676fc4 100644 --- a/internal/cgen/builtin.go +++ b/internal/cgen/builtin.go @@ -325,11 +325,27 @@ func (g *gen) writeBuiltinIOWriter(b *buffer, recv *a.Expr, method t.ID, args [] func (g *gen) writeBuiltinTokenWriter(b *buffer, recv *a.Expr, method t.ID, args []*a.Node, depth uint32) error { if method == t.IDWriteFastToken { b.printf("*iop_a_dst++ = wuffs_base__make_token((((uint64_t)(") - if err := g.writeExpr(b, args[0].AsArg().Value(), depth); err != nil { + + if cv := args[0].AsArg().Value().ConstValue(); (cv == nil) || (cv.Sign() != 0) { + if err := g.writeExpr(b, args[0].AsArg().Value(), depth); err != nil { + return err + } + b.writes(")) << WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT) | (((uint64_t)(") + } + + if err := g.writeExpr(b, args[1].AsArg().Value(), depth); err != nil { return err } - b.writes(")) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | (((uint64_t)(") - if err := g.writeExpr(b, args[1].AsArg().Value(), depth); err != nil { + b.writes(")) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(") + + if cv := args[2].AsArg().Value().ConstValue(); (cv == nil) || (cv.Sign() != 0) { + if err := g.writeExpr(b, args[2].AsArg().Value(), depth); err != nil { + return err + } + b.writes(")) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(") + } + + if err := g.writeExpr(b, args[3].AsArg().Value(), depth); err != nil { return err } b.writes(")) << WUFFS_BASE__TOKEN__LENGTH__SHIFT))") @@ -699,7 +715,7 @@ func (g *gen) writeBuiltinQuestionCall(b *buffer, n *a.Expr, depth uint32) error if err := g.writeExpr(b, n.Args()[0].AsArg().Value(), depth); err != nil { return err } - b.writes(")) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | (((uint64_t)(") + b.writes(")) << WUFFS_BASE__TOKEN__VLL__SHIFT) | (((uint64_t)(") if err := g.writeExpr(b, n.Args()[1].AsArg().Value(), depth); err != nil { return err } diff --git a/internal/cgen/data.go b/internal/cgen/data.go index ebb7fa9a..fee98742 100644 --- a/internal/cgen/data.go +++ b/internal/cgen/data.go @@ -308,26 +308,26 @@ const baseTokenPrivateH = "" + "" const baseTokenPublicH = "" + - "// ---------------- Tokens\n\ntypedef struct {\n // The repr is divided as:\n // - Bits 63 .. 40 (24 bits) is the major value.\n // - Bits 39 .. 16 (24 bits) is the minor value.\n // - Bits 15 .. 0 (16 bits) is the length.\n //\n // The major value is a [Base38](doc/note/base38-and-fourcc.md) value. If\n // zero (special cased for Wuffs' built-in \"base\" package) then the minor\n // value is further sub-divided:\n // - Bits 39 .. 37 ( 3 bits) is the value_base_category.\n // - Bits 36 .. 16 (21 bits) is the value_base_detail.\n //\n // In particular, at 21 bits, the value_base_detail can hold every valid\n // Unicode code point.\n //\n // If the major value is non-zero then the minor value has whatever arbitrary\n // meaning the tokenizer's package assigns to it.\n uint64_t repr;\n\n#ifdef __cplusplus\n inline uint64_t value() const;\n inline uint64_t value_major() const;\n inline uint64_t value_minor() const;\n inline uint64_t value_base_category() const;\n inline uint64_t value_base_detail() const;\n inlin" + - "e uint64_t length() const;\n#endif // __cplusplus\n\n} wuffs_base__token;\n\nstatic inline wuffs_base__token //\nwuffs_base__make_token(uint64_t repr) {\n wuffs_base__token ret;\n ret.repr = repr;\n return ret;\n}\n\n " + + "// ---------------- Tokens\n\ntypedef struct {\n // The repr's 64 bits are divided as:\n //\n // +-----+-------------+-------+-------------+-----+-----+-----------+\n // | 1 | 21 | 3 | 21 | 1 | 1 | 16 |\n // +-----+-------------+-------+-------------+-----+-----+-----------+\n // [..................value..................] LP LN length\n // [..0..|.value_major.|.....value_minor.....]\n // [..0..|.........VBC.........|.....VBD.....]\n //\n // The broad divisions are:\n // - Bits 63 .. 18 (46 bits) is the value.\n // - Bits 17 .. 16 ( 2 bits) is LP and LN (link_prev and link_next).\n // - Bits 15 .. 0 (16 bits) is the length.\n //\n // The value bits can be sub-divided in multiple ways:\n // - Bits 63 .. 63 ( 1 bits) is reserved (a zero bit).\n // - Bits 62 .. 42 (21 bits) is the value_major.\n // - Bits 41 .. 18 (24 bits) is the value_minor.\n // - Bits 62 .. 39 (24 bits) is the VBC (value_base_category).\n // - Bits 38 .. 18 (21 bits) is the VBD (value_base_de" + + "tail).\n //\n // The value_major is a 21-bit [Base38](doc/note/base38-and-fourcc.md) value.\n // If all of its bits are zero (special cased for Wuffs' built-in \"base\"\n // package) then the value_minor is further sub-divided:\n // - Bits 41 .. 39 ( 3 bits) is the VBC (value_base_category).\n // - Bits 38 .. 18 (21 bits) is the VBD (value_base_detail).\n //\n // The high 46 bits (bits 63 .. 18) only have VBC and VBD semantics when the\n // high 22 bits (the value_major) are all zero. An equivalent test is that\n // the high 25 bits (the notional VBC) has a numerical value less than 8.\n //\n // At 21 bits, the VBD can hold every valid Unicode code point.\n //\n // If value_major is non-zero then value_minor has whatever arbitrary meaning\n // the tokenizer's package assigns to it.\n //\n // Multiple consecutive tokens can form a larger conceptual unit. For\n // example, an \"abc\\tz\" string is a single higher level concept but at the\n // lower level, it could consist of multiple tokens: the quotes '\"', the\n " + + "// ASCII texts \"abc\" and \"z\" and the backslash-escaped tab '\\t'. The LP and\n // LN (link_prev and link_next) bits denote tokens that are part of a\n // multi-token chain:\n // - LP means that this token is not the first (there is a previous token).\n // - LN means that this token is not the last (there is a next token).\n //\n // In particular, a stand-alone token will have both link bits set to zero.\n uint64_t repr;\n\n#ifdef __cplusplus\n inline uint64_t value() const;\n inline uint64_t value_major() const;\n inline uint64_t value_minor() const;\n inline uint64_t value_base_category() const;\n inline uint64_t value_base_detail() const;\n inline bool link_prev() const;\n inline bool link_next() const;\n inline uint64_t length() const;\n#endif // __cplusplus\n\n} wuffs_base__token;\n\nstatic inline wuffs_base__token //\nwuffs_base__make_token(uint64_t repr) {\n wuffs_base__token ret;\n ret.repr = repr;\n return ret;\n}\n\n " + "" + - "// --------\n\n#define WUFFS_BASE__TOKEN__VALUE__MASK 0xFFFFFFFFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0xFFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_MINOR__MASK 0xFFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x7FFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK 0x1FFFFF\n#define WUFFS_BASE__TOKEN__LENGTH__MASK 0xFFFF\n\n#define WUFFS_BASE__TOKEN__VALUE__SHIFT 16\n#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 40\n#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 16\n#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 37\n#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 16\n#define WUFFS_BASE__TOKEN__LENGTH__SHIFT 0\n\n " + + "// --------\n\n#define WUFFS_BASE__TOKEN__VALUE__MASK 0x3FFFFFFFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0x3FFFFF\n#define WUFFS_BASE__TOKEN__VALUE_MINOR__MASK 0xFFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x1FFFFFF\n#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK 0x1FFFFF\n#define WUFFS_BASE__TOKEN__LINK__MASK 0x3\n#define WUFFS_BASE__TOKEN__LENGTH__MASK 0xFFFF\n\n#define WUFFS_BASE__TOKEN__VALUE__SHIFT 18\n#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 42\n#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 18\n#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 39\n#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 18\n#define WUFFS_BASE__TOKEN__LINK__SHIFT 16\n#define WUFFS_BASE__TOKEN__LENGTH__SHIFT 0\n\n#define WUFFS_BASE__TOKEN__LINK_PREV 0x20000\n#define WUFFS_BASE__TOKEN__LINK_NEXT 0x10000\n\n " + "" + - "// --------\n\n#define WUFFS_BASE__TOKEN__VBC__FILLER 0\n#define WUFFS_BASE__TOKEN__VBC__STRING 1\n#define WUFFS_BASE__TOKEN__VBC__UNICODE_CODE_POINT 2\n#define WUFFS_BASE__TOKEN__VBC__NUMBER 3\n#define WUFFS_BASE__TOKEN__VBC__STRUCTURE 4\n\n" + + "// --------\n\n#define WUFFS_BASE__TOKEN__VBC__FILLER 0\n#define WUFFS_BASE__TOKEN__VBC__STRING 1\n#define WUFFS_BASE__TOKEN__VBC__UNICODE_CODE_POINT 2\n#define WUFFS_BASE__TOKEN__VBC__NUMBER 3\n#define WUFFS_BASE__TOKEN__VBC__STRUCTURE 4\n\n " + "" + - "// --------\n\n// INCOMPLETE means that this token combines with the following token.\n//\n// For example, tokenizing a comment that is longer than the source buffer can\n// result in multiple VBC__FILLER tokens. All but the last one is INCOMPLETE.\n//\n// By convention, whitespace is not marked incomplete. Two whitespace tokens of\n// length 30 are equivalent to one whitespace token of length 60.\n//\n// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between\n// VBD_FILLER and VBD_STRING.\n#define WUFFS_BASE__TOKEN__VBD__FILLER__INCOMPLETE 0x00001\n\n#define WUFFS_BASE__TOKEN__VBD__FILLER__END_OF_CONSECUTIVE_COMMENTS 0x00010\n#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00020\n#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00040\n\n" + + "// --------\n\n#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00001\n#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00002\n\n" + "" + - "// --------\n\n// INCOMPLETE means that this token combines with the following token.\n//\n// For example, tokenizing a string that is longer than the source buffer can\n// result in multiple VBC__STRING tokens. All but the last one is INCOMPLETE.\n//\n// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between\n// VBD_FILLER and VBD_STRING.\n#define WUFFS_BASE__TOKEN__VBD__STRING__INCOMPLETE 0x00001\n\n// \"DEFINITELY_FOO\" means that the source bytes (and also the destination\n// bytes, assuming 1_DST_1_SRC_COPY) are in the FOO format. Definitely means\n// that the lack of the bit is conservative: it is valid for all-ASCII strings\n// to have neither DEFINITELY_ASCII or DEFINITELY_UTF_8 bits set.\n#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00010\n#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00020\n\n// \"CONVERT_D_DST_S_SRC\" means that multiples of S source bytes (possibly\n// padded) produces multiples of D destination bytes. For example,\n// CONVERT_1_DST_4_SRC_BACKSLASH_X means a sourc" + - "e like \"\\\\x23\\\\x67\\\\xAB\", where\n// 12 src bytes encode 3 dst bytes.\n//\n// When src is the empty string, multiple conversion algorithms are applicable\n// (so these bits are not necessarily mutually exclusive), all producing the\n// same empty dst string.\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00100\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00200\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00400\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00800\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x01000\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x02000\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x04000\n\n " + - "" + - "// --------\n\n // UNICODE_CODE_POINT tokens have no room in their VBD for an INCOMPLETE bit.\n // All 21 bits are used to hold the Unicode code point. Such tokens preserve\n // the INCOMPLETEness of the previous token (if a FILLER or STRING).\n\n#define WUFFS_BASE__TOKEN__VBD__UNICODE_CODE_POINT__MAX_INCL 0x10FFFF\n\n " + + "// --------\n\n// \"DEFINITELY_FOO\" means that the destination bytes (and also the source\n// bytes, for 1_DST_1_SRC_COPY) are in the FOO format. Definitely means that\n// the lack of the bit is conservative: it is valid for all-ASCII strings to\n// have neither DEFINITELY_UTF_8 or DEFINITELY_ASCII bits set.\n#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00001\n#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00002\n\n// \"CONVERT_D_DST_S_SRC\" means that multiples of S source bytes (possibly\n// padded) produces multiples of D destination bytes. For example,\n// CONVERT_1_DST_4_SRC_BACKSLASH_X means a source like \"\\\\x23\\\\x67\\\\xAB\", where\n// 12 src bytes encode 3 dst bytes.\n//\n// When src is the empty string, multiple conversion algorithms are applicable\n// (so these bits are not necessarily mutually exclusive), all producing the\n// same empty dst string.\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00010\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00020\n#" + + "define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00040\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00080\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x00100\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x00200\n#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x00400\n\n " + "" + "// --------\n\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__LITERAL 0x00001\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__LITERAL__UNDEFINED 0x00101\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__LITERAL__NULL 0x00201\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__LITERAL__FALSE 0x00401\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__LITERAL__TRUE 0x00801\n\n// For a source string of \"123\" or \"0x9A\", it is valid for a tokenizer to\n// return any one of:\n// - WUFFS_BASE__TOKEN__VBD__NUMBER__FLOATING_POINT.\n// - WUFFS_BASE__TOKEN__VBD__NUMBER__INTEGER_SIGNED.\n// - WUFFS_BASE__TOKEN__VBD__NUMBER__INTEGER_UNSIGNED.\n//\n// For a source string of \"+123\" or \"-0x9A\", only the first two are valid.\n//\n// For a source string of \"123.\", only the first one is valid.\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__FLOATING_POINT 0x00002\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__INTEGER_SIGNED 0x00004\n#define WUFFS_BASE__TOKEN__VBD__NUMBER__INTEGER_UNSIGNED 0x00008\n\n " + "" + "// --------\n\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__PUSH 0x00001\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__POP 0x00002\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__FROM_NONE 0x00010\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__FROM_LIST 0x00020\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__FROM_DICT 0x00040\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__TO_NONE 0x01000\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__TO_LIST 0x02000\n#define WUFFS_BASE__TOKEN__VBD__STRUCTURE__TO_DICT 0x04000\n\n" + "" + "// --------\n\nstatic inline uint64_t //\nwuffs_base__token__value(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__VALUE__SHIFT) &\n WUFFS_BASE__TOKEN__VALUE__MASK;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__value_major(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT) &\n WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__value_minor(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) &\n WUFFS_BASE__TOKEN__VALUE_MINOR__MASK;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__value_base_category(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT) &\n WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__value_base_detail(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT) &\n WUFFS_BASE__TOKEN__VALUE_BASE_DETA" + - "IL__MASK;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__length(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__LENGTH__SHIFT) &\n WUFFS_BASE__TOKEN__LENGTH__MASK;\n}\n\n#ifdef __cplusplus\n\ninline uint64_t //\nwuffs_base__token::value() const {\n return wuffs_base__token__value(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_major() const {\n return wuffs_base__token__value_major(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_minor() const {\n return wuffs_base__token__value_minor(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_base_category() const {\n return wuffs_base__token__value_base_category(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_base_detail() const {\n return wuffs_base__token__value_base_detail(this);\n}\n\ninline uint64_t //\nwuffs_base__token::length() const {\n return wuffs_base__token__length(this);\n}\n\n#endif // __cplusplus\n\n" + + "IL__MASK;\n}\n\nstatic inline bool //\nwuffs_base__token__link_prev(const wuffs_base__token* t) {\n return t->repr & WUFFS_BASE__TOKEN__LINK_PREV;\n}\n\nstatic inline bool //\nwuffs_base__token__link_next(const wuffs_base__token* t) {\n return t->repr & WUFFS_BASE__TOKEN__LINK_NEXT;\n}\n\nstatic inline uint64_t //\nwuffs_base__token__length(const wuffs_base__token* t) {\n return (t->repr >> WUFFS_BASE__TOKEN__LENGTH__SHIFT) &\n WUFFS_BASE__TOKEN__LENGTH__MASK;\n}\n\n#ifdef __cplusplus\n\ninline uint64_t //\nwuffs_base__token::value() const {\n return wuffs_base__token__value(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_major() const {\n return wuffs_base__token__value_major(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_minor() const {\n return wuffs_base__token__value_minor(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_base_category() const {\n return wuffs_base__token__value_base_category(this);\n}\n\ninline uint64_t //\nwuffs_base__token::value_base_detail() const {\n return wuffs_base__to" + + "ken__value_base_detail(this);\n}\n\ninline bool //\nwuffs_base__token::link_prev() const {\n return wuffs_base__token__link_prev(this);\n}\n\ninline bool //\nwuffs_base__token::link_next() const {\n return wuffs_base__token__link_next(this);\n}\n\ninline uint64_t //\nwuffs_base__token::length() const {\n return wuffs_base__token__length(this);\n}\n\n#endif // __cplusplus\n\n" + "" + "// --------\n\ntypedef WUFFS_BASE__SLICE(wuffs_base__token) wuffs_base__slice_token;\n\nstatic inline wuffs_base__slice_token //\nwuffs_base__make_slice_token(wuffs_base__token* ptr, size_t len) {\n wuffs_base__slice_token ret;\n ret.ptr = ptr;\n ret.len = len;\n return ret;\n}\n\n" + "" + diff --git a/lang/builtin/builtin.go b/lang/builtin/builtin.go index 22e2dd68..2ec22216 100644 --- a/lang/builtin/builtin.go +++ b/lang/builtin/builtin.go @@ -326,9 +326,13 @@ var Funcs = []string{ // ---- token_writer - "token_writer.write_token?(value: u64[..= 0xFFFFFFFFFFFF], length: u64[..= 0xFFFF])", + "token_writer.write_token?(" + + "value_major: u32[..= 0x1F_FFFF], value_minor: u32[..= 0xFF_FFFF]," + + "link: u32[..= 0x3], length: u32[..= 0xFFFF])", - "token_writer.write_fast_token!(value: u64[..= 0xFFFFFFFFFFFF], length: u64[..= 0xFFFF])", + "token_writer.write_fast_token!(" + + "value_major: u32[..= 0x1F_FFFF], value_minor: u32[..= 0xFF_FFFF]," + + "link: u32[..= 0x3], length: u32[..= 0xFFFF])", "token_writer.available() u64", diff --git a/release/c/wuffs-unsupported-snapshot.c b/release/c/wuffs-unsupported-snapshot.c index 6b3a64a4..75ca1012 100644 --- a/release/c/wuffs-unsupported-snapshot.c +++ b/release/c/wuffs-unsupported-snapshot.c @@ -1733,22 +1733,52 @@ wuffs_base__io_buffer::writer_io_position() const { // ---------------- Tokens typedef struct { - // The repr is divided as: - // - Bits 63 .. 40 (24 bits) is the major value. - // - Bits 39 .. 16 (24 bits) is the minor value. + // The repr's 64 bits are divided as: + // + // +-----+-------------+-------+-------------+-----+-----+-----------+ + // | 1 | 21 | 3 | 21 | 1 | 1 | 16 | + // +-----+-------------+-------+-------------+-----+-----+-----------+ + // [..................value..................] LP LN length + // [..0..|.value_major.|.....value_minor.....] + // [..0..|.........VBC.........|.....VBD.....] + // + // The broad divisions are: + // - Bits 63 .. 18 (46 bits) is the value. + // - Bits 17 .. 16 ( 2 bits) is LP and LN (link_prev and link_next). // - Bits 15 .. 0 (16 bits) is the length. // - // The major value is a [Base38](doc/note/base38-and-fourcc.md) value. If - // zero (special cased for Wuffs' built-in "base" package) then the minor - // value is further sub-divided: - // - Bits 39 .. 37 ( 3 bits) is the value_base_category. - // - Bits 36 .. 16 (21 bits) is the value_base_detail. + // The value bits can be sub-divided in multiple ways: + // - Bits 63 .. 63 ( 1 bits) is reserved (a zero bit). + // - Bits 62 .. 42 (21 bits) is the value_major. + // - Bits 41 .. 18 (24 bits) is the value_minor. + // - Bits 62 .. 39 (24 bits) is the VBC (value_base_category). + // - Bits 38 .. 18 (21 bits) is the VBD (value_base_detail). // - // In particular, at 21 bits, the value_base_detail can hold every valid - // Unicode code point. + // The value_major is a 21-bit [Base38](doc/note/base38-and-fourcc.md) value. + // If all of its bits are zero (special cased for Wuffs' built-in "base" + // package) then the value_minor is further sub-divided: + // - Bits 41 .. 39 ( 3 bits) is the VBC (value_base_category). + // - Bits 38 .. 18 (21 bits) is the VBD (value_base_detail). // - // If the major value is non-zero then the minor value has whatever arbitrary - // meaning the tokenizer's package assigns to it. + // The high 46 bits (bits 63 .. 18) only have VBC and VBD semantics when the + // high 22 bits (the value_major) are all zero. An equivalent test is that + // the high 25 bits (the notional VBC) has a numerical value less than 8. + // + // At 21 bits, the VBD can hold every valid Unicode code point. + // + // If value_major is non-zero then value_minor has whatever arbitrary meaning + // the tokenizer's package assigns to it. + // + // Multiple consecutive tokens can form a larger conceptual unit. For + // example, an "abc\tz" string is a single higher level concept but at the + // lower level, it could consist of multiple tokens: the quotes '"', the + // ASCII texts "abc" and "z" and the backslash-escaped tab '\t'. The LP and + // LN (link_prev and link_next) bits denote tokens that are part of a + // multi-token chain: + // - LP means that this token is not the first (there is a previous token). + // - LN means that this token is not the last (there is a next token). + // + // In particular, a stand-alone token will have both link bits set to zero. uint64_t repr; #ifdef __cplusplus @@ -1757,6 +1787,8 @@ typedef struct { inline uint64_t value_minor() const; inline uint64_t value_base_category() const; inline uint64_t value_base_detail() const; + inline bool link_prev() const; + inline bool link_next() const; inline uint64_t length() const; #endif // __cplusplus @@ -1771,20 +1803,25 @@ wuffs_base__make_token(uint64_t repr) { // -------- -#define WUFFS_BASE__TOKEN__VALUE__MASK 0xFFFFFFFFFFFF -#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0xFFFFFF +#define WUFFS_BASE__TOKEN__VALUE__MASK 0x3FFFFFFFFFFF +#define WUFFS_BASE__TOKEN__VALUE_MAJOR__MASK 0x3FFFFF #define WUFFS_BASE__TOKEN__VALUE_MINOR__MASK 0xFFFFFF -#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x7FFFFFF +#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__MASK 0x1FFFFFF #define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK 0x1FFFFF +#define WUFFS_BASE__TOKEN__LINK__MASK 0x3 #define WUFFS_BASE__TOKEN__LENGTH__MASK 0xFFFF -#define WUFFS_BASE__TOKEN__VALUE__SHIFT 16 -#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 40 -#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 16 -#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 37 -#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 16 +#define WUFFS_BASE__TOKEN__VALUE__SHIFT 18 +#define WUFFS_BASE__TOKEN__VALUE_MAJOR__SHIFT 42 +#define WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT 18 +#define WUFFS_BASE__TOKEN__VALUE_BASE_CATEGORY__SHIFT 39 +#define WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__SHIFT 18 +#define WUFFS_BASE__TOKEN__LINK__SHIFT 16 #define WUFFS_BASE__TOKEN__LENGTH__SHIFT 0 +#define WUFFS_BASE__TOKEN__LINK_PREV 0x20000 +#define WUFFS_BASE__TOKEN__LINK_NEXT 0x10000 + // -------- #define WUFFS_BASE__TOKEN__VBC__FILLER 0 @@ -1793,41 +1830,19 @@ wuffs_base__make_token(uint64_t repr) { #define WUFFS_BASE__TOKEN__VBC__NUMBER 3 #define WUFFS_BASE__TOKEN__VBC__STRUCTURE 4 -// -------- + // -------- -// INCOMPLETE means that this token combines with the following token. -// -// For example, tokenizing a comment that is longer than the source buffer can -// result in multiple VBC__FILLER tokens. All but the last one is INCOMPLETE. -// -// By convention, whitespace is not marked incomplete. Two whitespace tokens of -// length 30 are equivalent to one whitespace token of length 60. -// -// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between -// VBD_FILLER and VBD_STRING. -#define WUFFS_BASE__TOKEN__VBD__FILLER__INCOMPLETE 0x00001 - -#define WUFFS_BASE__TOKEN__VBD__FILLER__END_OF_CONSECUTIVE_COMMENTS 0x00010 -#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00020 -#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00040 +#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_LINE 0x00001 +#define WUFFS_BASE__TOKEN__VBD__FILLER__COMMENT_BLOCK 0x00002 // -------- -// INCOMPLETE means that this token combines with the following token. -// -// For example, tokenizing a string that is longer than the source buffer can -// result in multiple VBC__STRING tokens. All but the last one is INCOMPLETE. -// -// Bits 0x2, 0x4 and 0x8 are reserved for flags that are common between -// VBD_FILLER and VBD_STRING. -#define WUFFS_BASE__TOKEN__VBD__STRING__INCOMPLETE 0x00001 - -// "DEFINITELY_FOO" means that the source bytes (and also the destination -// bytes, assuming 1_DST_1_SRC_COPY) are in the FOO format. Definitely means -// that the lack of the bit is conservative: it is valid for all-ASCII strings -// to have neither DEFINITELY_ASCII or DEFINITELY_UTF_8 bits set. -#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00010 -#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00020 +// "DEFINITELY_FOO" means that the destination bytes (and also the source +// bytes, for 1_DST_1_SRC_COPY) are in the FOO format. Definitely means that +// the lack of the bit is conservative: it is valid for all-ASCII strings to +// have neither DEFINITELY_UTF_8 or DEFINITELY_ASCII bits set. +#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_UTF_8 0x00001 +#define WUFFS_BASE__TOKEN__VBD__STRING__DEFINITELY_ASCII 0x00002 // "CONVERT_D_DST_S_SRC" means that multiples of S source bytes (possibly // padded) produces multiples of D destination bytes. For example, @@ -1837,21 +1852,13 @@ wuffs_base__make_token(uint64_t repr) { // When src is the empty string, multiple conversion algorithms are applicable // (so these bits are not necessarily mutually exclusive), all producing the // same empty dst string. -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00100 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00200 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00400 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00800 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x01000 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x02000 -#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x04000 - - // -------- - - // UNICODE_CODE_POINT tokens have no room in their VBD for an INCOMPLETE bit. - // All 21 bits are used to hold the Unicode code point. Such tokens preserve - // the INCOMPLETEness of the previous token (if a FILLER or STRING). - -#define WUFFS_BASE__TOKEN__VBD__UNICODE_CODE_POINT__MAX_INCL 0x10FFFF +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_0_DST_1_SRC_DROP 0x00010 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_1_SRC_COPY 0x00020 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_2_SRC_HEXADECIMAL 0x00040 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_1_DST_4_SRC_BACKSLASH_X 0x00080 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_STD 0x00100 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_3_DST_4_SRC_BASE_64_URL 0x00200 +#define WUFFS_BASE__TOKEN__VBD__STRING__CONVERT_4_DST_5_SRC_ASCII_85 0x00400 // -------- @@ -1917,6 +1924,16 @@ wuffs_base__token__value_base_detail(const wuffs_base__token* t) { WUFFS_BASE__TOKEN__VALUE_BASE_DETAIL__MASK; } +static inline bool // +wuffs_base__token__link_prev(const wuffs_base__token* t) { + return t->repr & WUFFS_BASE__TOKEN__LINK_PREV; +} + +static inline bool // +wuffs_base__token__link_next(const wuffs_base__token* t) { + return t->repr & WUFFS_BASE__TOKEN__LINK_NEXT; +} + static inline uint64_t // wuffs_base__token__length(const wuffs_base__token* t) { return (t->repr >> WUFFS_BASE__TOKEN__LENGTH__SHIFT) & @@ -1950,6 +1967,16 @@ wuffs_base__token::value_base_detail() const { return wuffs_base__token__value_base_detail(this); } +inline bool // +wuffs_base__token::link_prev() const { + return wuffs_base__token__link_prev(this); +} + +inline bool // +wuffs_base__token::link_next() const { + return wuffs_base__token__link_next(this); +} + inline uint64_t // wuffs_base__token::length() const { return wuffs_base__token__length(this); @@ -19257,7 +19284,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, self->private_impl.active_coroutine = 0; wuffs_base__status status = wuffs_base__make_status(NULL); - uint64_t v_token_value = 0; + uint32_t v_vminor = 0; uint32_t v_number_length = 0; uint32_t v_number_status = 0; uint32_t v_string_length = 0; @@ -19329,8 +19356,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { if (v_whitespace_length > 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_whitespace_length)))) + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(v_whitespace_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_whitespace_length = 0; } @@ -19352,7 +19379,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 1, wuffs_base__make_empty_struct()); if (v_whitespace_length >= 65534) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(65535)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_whitespace_length = 0; goto label__outer__continue; @@ -19362,8 +19389,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, label__ws__break:; if (v_whitespace_length > 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_whitespace_length)))) + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(v_whitespace_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_whitespace_length = 0; if (((uint64_t)(io2_a_dst - iop_a_dst)) <= 0) { @@ -19376,53 +19403,28 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } if (v_class == 1) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(2097171)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); (iop_a_src += 1, wuffs_base__make_empty_struct()); - label__0__continue:; - while (true) { - if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { - if (a_src && a_src->meta.closed) { - status = wuffs_base__make_status(wuffs_json__error__bad_input); - goto exit; - } - status = - wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(3); - goto label__0__continue; - } - if (wuffs_base__load_u8be__no_bounds_check(iop_a_src) != 92) { - goto label__0__break; - } - 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(4); - goto label__0__continue; - } - *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2129713)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); - goto label__0__break; - } - label__0__break:; label__string_loop__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(5); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(3); goto label__string_loop__continue; } v_string_length = 0; - label__1__continue:; + label__0__continue:; 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)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_string_length)))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)(v_string_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; } @@ -19433,7 +19435,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(6); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(4); v_string_length = 0; goto label__string_loop__continue; } @@ -19443,28 +19445,34 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 1, wuffs_base__make_empty_struct()); if (v_string_length >= 65531) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(65532)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; goto label__string_loop__continue; } v_string_length += 1; - goto label__1__continue; + goto label__0__continue; } else if (v_char == 1) { - *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097696)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_string_length)))) - << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); - v_string_length = 0; + if (v_string_length != 0) { + *iop_a_dst++ = wuffs_base__make_token( + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)(v_string_length)) + << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); + v_string_length = 0; + } goto label__string_loop__break; } else if (v_char == 2) { if (v_string_length > 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_string_length)))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)(v_string_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; if (((uint64_t)(io2_a_dst - iop_a_dst)) <= 0) { @@ -19479,7 +19487,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(7); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(5); v_string_length = 0; v_char = 0; goto label__string_loop__continue; @@ -19491,8 +19499,9 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 2, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( (((uint64_t)( - (4194304 | ((uint64_t)((v_backslash & 127)))))) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (4194304 | ((uint32_t)((v_backslash & 127)))))) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(2)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); goto label__string_loop__continue; } else if (v_c == 117) { @@ -19504,7 +19513,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(8); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(6); v_string_length = 0; v_char = 0; goto label__string_loop__continue; @@ -19535,8 +19544,9 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if ((v_uni4_value < 55296) || (57343 < v_uni4_value)) { (iop_a_src += 6, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)((4194304 | ((uint64_t)(v_uni4_value))))) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)((4194304 | v_uni4_value))) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(6)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); goto label__string_loop__continue; } else if (v_uni4_value >= 56320) { @@ -19549,7 +19559,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(9); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(7); v_string_length = 0; v_uni4_value = 0; v_char = 0; @@ -19592,10 +19602,10 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, v_uni4_value -= 56320; (iop_a_src += 8, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)( - (4194304 | ((uint64_t)((v_uni4_high_surrogate + - v_uni4_value)))))) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)((4194304 | v_uni4_high_surrogate | + v_uni4_value))) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(12)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); goto label__string_loop__continue; @@ -19626,7 +19636,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(10); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(8); v_string_length = 0; v_char = 0; goto label__string_loop__continue; @@ -19639,15 +19649,16 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 2, wuffs_base__make_empty_struct()); if (v_string_length >= 65528) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)((v_string_length + 2))))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)((v_string_length + 2))) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; goto label__string_loop__continue; } v_string_length += 2; - goto label__1__continue; + goto label__0__continue; } } else if (v_char == 4) { if (((uint64_t)(io2_a_src - iop_a_src)) < 3) { @@ -19658,7 +19669,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(11); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(9); v_string_length = 0; v_char = 0; goto label__string_loop__continue; @@ -19675,15 +19686,16 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 3, wuffs_base__make_empty_struct()); if (v_string_length >= 65528) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)((v_string_length + 3))))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)((v_string_length + 3))) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; goto label__string_loop__continue; } v_string_length += 3; - goto label__1__continue; + goto label__0__continue; } } } else if (v_char == 5) { @@ -19695,7 +19707,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status( wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(12); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(10); v_string_length = 0; v_char = 0; goto label__string_loop__continue; @@ -19712,22 +19724,25 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 4, wuffs_base__make_empty_struct()); if (v_string_length >= 65528) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)((v_string_length + 4))))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)((v_string_length + 4))) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; goto label__string_loop__continue; } v_string_length += 4; - goto label__1__continue; + goto label__0__continue; } } } if (v_string_length > 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(2097697)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_string_length)))) + (((uint64_t)(2097185)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(3)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | + (((uint64_t)(v_string_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_string_length = 0; } @@ -19741,7 +19756,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } } label__string_loop__break:; - label__2__continue:; + label__1__continue:; while (true) { if (((uint64_t)(io2_a_src - iop_a_src)) <= 0) { if (a_src && a_src->meta.closed) { @@ -19750,22 +19765,24 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(13); - goto label__2__continue; + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(11); + goto label__1__continue; } 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(14); - goto label__2__continue; + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(12); + goto label__1__continue; } (iop_a_src += 1, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(2097171)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(2)) << WUFFS_BASE__TOKEN__LINK__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); - goto label__2__break; + goto label__1__break; } - label__2__break:; + label__1__break:; if (0 == (v_expect & 16)) { v_expect = 8; goto label__outer__continue; @@ -19774,7 +19791,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if (v_class == 2) { (iop_a_src += 1, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); if (0 == (v_expect & 256)) { v_expect = 2; @@ -19785,7 +19802,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if (v_class == 3) { (iop_a_src += 1, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(0)) << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 3762; goto label__outer__continue; @@ -19799,18 +19816,18 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, iop_a_src = a_src->data.ptr + a_src->meta.ri; } v_number_status = (v_number_length >> 8); - v_token_value = 6291460; + v_vminor = 6291460; if ((v_number_length & 128) != 0) { - v_token_value = 6291458; + v_vminor = 6291458; } v_number_length = (v_number_length & 127); if (v_number_status == 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(v_token_value)) - << WUFFS_BASE__TOKEN__VALUE__SHIFT) | - (((uint64_t)(((uint64_t)(v_number_length)))) + (((uint64_t)(v_vminor)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | + (((uint64_t)(v_number_length)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); - goto label__3__break; + goto label__2__break; } while (v_number_length > 0) { v_number_length -= 1; @@ -19832,23 +19849,23 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else { status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(15); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(13); while (((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(16); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(14); } } } - label__3__break:; + label__2__break:; goto label__goto_parsed_a_leaf_value__break; } else if (v_class == 5) { - v_token_value = 8405009; + v_vminor = 8405009; if (v_depth == 0) { } else if (0 != (v_expect_after_value & 64)) { - v_token_value = 8405057; + v_vminor = 8405057; } else { - v_token_value = 8405025; + v_vminor = 8405025; } if (v_depth >= 1024) { status = wuffs_base__make_status( @@ -19862,7 +19879,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, v_depth += 1; (iop_a_src += 1, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(v_token_value)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(v_vminor)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 66; v_expect_after_value = 68; @@ -19871,7 +19889,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 1, wuffs_base__make_empty_struct()); if (v_depth <= 1) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8392770)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8392770)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); goto label__outer__break; } @@ -19881,25 +19900,27 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, if (0 == (self->private_data.f_stack[v_stack_byte] & (((uint32_t)(1)) << v_stack_bit))) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8396866)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8396866)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 260; v_expect_after_value = 260; } else { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8405058)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8405058)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 68; v_expect_after_value = 68; } goto label__outer__continue; } else if (v_class == 7) { - v_token_value = 8396817; + v_vminor = 8396817; if (v_depth == 0) { } else if (0 != (v_expect_after_value & 64)) { - v_token_value = 8396865; + v_vminor = 8396865; } else { - v_token_value = 8396833; + v_vminor = 8396833; } if (v_depth >= 1024) { status = wuffs_base__make_status( @@ -19913,7 +19934,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, v_depth += 1; (iop_a_src += 1, wuffs_base__make_empty_struct()); *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(v_token_value)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(v_vminor)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 4018; v_expect_after_value = 260; @@ -19922,7 +19944,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, (iop_a_src += 1, wuffs_base__make_empty_struct()); if (v_depth <= 1) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8392738)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8392738)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); goto label__outer__break; } @@ -19932,13 +19955,15 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, if (0 == (self->private_data.f_stack[v_stack_byte] & (((uint32_t)(1)) << v_stack_bit))) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8396834)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8396834)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 260; v_expect_after_value = 260; } else { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(8405026)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(8405026)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(1)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); v_expect = 68; v_expect_after_value = 68; @@ -19949,7 +19974,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, 111546413966853); if (v_match == 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(6292481)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(6292481)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(5)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); if (((uint64_t)(io2_a_src - iop_a_src)) < 5) { status = wuffs_base__make_status( @@ -19961,7 +19987,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if (v_match == 1) { status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(17); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(15); goto label__outer__continue; } } else if (v_class == 10) { @@ -19969,7 +19995,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, 435762131972); if (v_match == 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(6293505)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(6293505)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(4)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); if (((uint64_t)(io2_a_src - iop_a_src)) < 4) { status = wuffs_base__make_status( @@ -19981,7 +20008,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if (v_match == 1) { status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(18); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(16); goto label__outer__continue; } } else if (v_class == 11) { @@ -19989,7 +20016,8 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, 465676103172); if (v_match == 0) { *iop_a_dst++ = wuffs_base__make_token( - (((uint64_t)(6291969)) << WUFFS_BASE__TOKEN__VALUE__SHIFT) | + (((uint64_t)(6291969)) + << WUFFS_BASE__TOKEN__VALUE_MINOR__SHIFT) | (((uint64_t)(4)) << WUFFS_BASE__TOKEN__LENGTH__SHIFT)); if (((uint64_t)(io2_a_src - iop_a_src)) < 4) { status = wuffs_base__make_status( @@ -20001,7 +20029,7 @@ wuffs_json__decoder__decode_tokens(wuffs_json__decoder* self, } else if (v_match == 1) { status = wuffs_base__make_status(wuffs_base__suspension__short_read); - WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(19); + WUFFS_BASE__COROUTINE_SUSPENSION_POINT_MAYBE_SUSPEND(17); goto label__outer__continue; } } diff --git a/script/print-json-token-debug-format.c b/script/print-json-token-debug-format.c index af38a958..5ddc73fa 100644 --- a/script/print-json-token-debug-format.c +++ b/script/print-json-token-debug-format.c @@ -17,23 +17,38 @@ // print-json-token-debug-format.c parses JSON from stdin and prints the // resulting token stream, eliding any non-essential (e.g. whitespace) tokens. // -// The output format is only for debugging, and certainly not for long term -// storage. It isn't guaranteed to be stable between versions of this program -// and of the Wuffs standard library. +// The output format is only for debugging or regression testing, and certainly +// not for long term storage. It isn't guaranteed to be stable between versions +// of this program and of the Wuffs standard library. +// +// It prints 16 bytes (128 bits) per token, containing big-endian numbers: +// +// +---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+ +// | POS | LEN | LP| LN| MAJOR | MINOR | +// | | | | | |VBC| VBD | +// +---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+---+ +// +// - POS (4 bytes) is the position: the sum of all previous tokens' lengths, +// including elided tokens. +// - LEN (2 bytes) is the length. +// - LP (1 bytes) is the link_prev bit. +// - LN (1 bytes) is the link_next bit +// - MAJOR (4 bytes) is the value_major. +// +// The final 4 bytes are either value_minor (when the value_major is non-zero) +// or 1 + 3 bytes for value_base_category and value_base_detail (otherwise). // -// It prints 16 bytes (4 big-endian uint32's) per token: the token's position -// (total length of all previous tokens), length, major value and minor value. // Together with the hexadecimal WUFFS_BASE__TOKEN__ETC constants defined in // token-public.h, this format is somewhat human-readable when piped through a -// hex-dump program (e.g. /usr/bin/hd). +// hex-dump program (such as /usr/bin/hd), printing one token per line. // // If the input or output is larger than the program's buffers (64 MiB and // 131072 tokens by default), there may be multiple valid tokenizations of any // given input. For example, if a source string "abcde" straddles an I/O -// boundary, it may be tokenized as a 5-length complete string or as a 3-length -// incomplete string followed by a 2-length complete string. +// boundary, it may be tokenized as single (no-link) 5-length string or as a +// 3-length link_next'ed string followed by a 2-length link_prev'ed string. // -// A Wuffs token stream, in general, can support inputs more than (1 << 32) +// A Wuffs token stream, in general, can support inputs more than 0xFFFF_FFFF // bytes long, but this program can not, as it tracks the tokens' cumulative // position as a uint32. @@ -134,17 +149,28 @@ main1(int argc, char** argv) { while (tok.meta.ri < tok.meta.wi) { wuffs_base__token* t = &tok.data.ptr[tok.meta.ri++]; - uint64_t len = wuffs_base__token__length(t); + uint16_t len = wuffs_base__token__length(t); if (wuffs_base__token__value(t) != 0) { - uint64_t maj = wuffs_base__token__value_major(t); - uint64_t min = wuffs_base__token__value_minor(t); + uint8_t lp = wuffs_base__token__link_prev(t) ? 1 : 0; + uint8_t ln = wuffs_base__token__link_next(t) ? 1 : 0; + uint32_t vmajor = wuffs_base__token__value_major(t); uint8_t buf[16]; - wuffs_base__store_u32be__no_bounds_check(&buf[0 * 4], (uint32_t)(pos)); - wuffs_base__store_u32be__no_bounds_check(&buf[1 * 4], (uint32_t)(len)); - wuffs_base__store_u32be__no_bounds_check(&buf[2 * 4], (uint32_t)(maj)); - wuffs_base__store_u32be__no_bounds_check(&buf[3 * 4], (uint32_t)(min)); + wuffs_base__store_u32be__no_bounds_check(&buf[0x0], (uint32_t)(pos)); + wuffs_base__store_u16be__no_bounds_check(&buf[0x4], len); + wuffs_base__store_u8__no_bounds_check(&buf[0x0006], lp); + wuffs_base__store_u8__no_bounds_check(&buf[0x0007], ln); + wuffs_base__store_u32be__no_bounds_check(&buf[0x8], vmajor); + if (vmajor) { + uint32_t vminor = wuffs_base__token__value_minor(t); + wuffs_base__store_u32be__no_bounds_check(&buf[0xC], vminor); + } else { + uint8_t vbc = wuffs_base__token__value_base_category(t); + uint32_t vbd = wuffs_base__token__value_base_detail(t); + wuffs_base__store_u8__no_bounds_check(&buf[0x000C], vbc); + wuffs_base__store_u24be__no_bounds_check(&buf[0xD], vbd); + } const int stdout_fd = 1; write(stdout_fd, &buf[0], 16); } diff --git a/std/json/decode_json.wuffs b/std/json/decode_json.wuffs index 959589e8..beb23a42 100644 --- a/std/json/decode_json.wuffs +++ b/std/json/decode_json.wuffs @@ -45,7 +45,7 @@ pub struct decoder? implements base.token_decoder( ) pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { - var token_value : base.u64[..= 0xFF_FFFF] + var vminor : base.u32[..= 0xFF_FFFF] var number_length : base.u32[..= 0x3FF] var number_status : base.u32[..= 0x3] var string_length : base.u32[..= 0xFFFB] @@ -95,8 +95,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if args.src.available() <= 0 { if whitespace_length > 0 { args.dst.write_fast_token!( - value: 0, - length: whitespace_length as base.u64) + value_major: 0, + value_minor: 0, + link: 0x0, + length: whitespace_length) whitespace_length = 0 } if args.src.is_closed() { @@ -116,7 +118,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if whitespace_length >= 0xFFFE { args.dst.write_fast_token!( - value: 0, + value_major: 0, + value_minor: 0, + link: 0x0, length: 0xFFFF) whitespace_length = 0 continue.outer @@ -127,8 +131,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { // Emit whitespace. if whitespace_length > 0 { args.dst.write_fast_token!( - value: 0, - length: whitespace_length as base.u64) + value_major: 0, + value_minor: 0, + link: 0x0, + length: whitespace_length) whitespace_length = 0 if args.dst.available() <= 0 { continue.outer @@ -149,37 +155,14 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if class == 0x01 { // 0x01 is CLASS_STRING. // -------- BEGIN parse strings. - // The leading '"' is filler. + // Emit the leading '"'. args.dst.write_fast_token!( - value: 0, + value_major: 0, + value_minor: 0x20_0013, + link: 0x1, length: 1) args.src.skip32_fast!(actual: 1, worst_case: 1) - // If the string contents starts with an escape code, issue an - // incomplete zero-length string token (with - // WUFFS_BASE__TOKEN__VBC__STRING) before issuing a token with - // WUFFS_BASE__TOKEN__VBC__UNICODE_CODE_POINT. - while true { - if args.src.available() <= 0 { - if args.src.is_closed() { - return "#bad input" - } - yield? base."$short read" - continue - } - if args.src.peek_u8() <> 0x5C { // 0x5C is '\\'. - break - } - if args.dst.available() <= 0 { - yield? base."$short write" - continue - } - args.dst.write_fast_token!( - value: 0x20_7F31, - length: 0) - break - } - while.string_loop true { if args.dst.available() <= 0 { yield? base."$short write" @@ -193,8 +176,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if args.src.available() <= 0 { if string_length > 0 { args.dst.write_fast_token!( - value: 0x20_0221, - length: string_length as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length) string_length = 0 } if args.src.is_closed() { @@ -211,7 +196,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) if string_length >= 0xFFFB { args.dst.write_fast_token!( - value: 0x20_0221, + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, length: 0xFFFC) string_length = 0 continue.string_loop @@ -220,17 +207,23 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { continue } else if char == 0x01 { // '"' - args.dst.write_fast_token!( - value: 0x20_0220, - length: string_length as base.u64) - string_length = 0 + if string_length <> 0 { + args.dst.write_fast_token!( + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length) + string_length = 0 + } break.string_loop } else if char == 0x02 { // '\\'. if string_length > 0 { args.dst.write_fast_token!( - value: 0x20_0221, - length: string_length as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length) string_length = 0 if args.dst.available() <= 0 { continue.string_loop @@ -252,7 +245,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if backslash > 0 { args.src.skip32_fast!(actual: 2, worst_case: 2) args.dst.write_fast_token!( - value: 0x40_0000 | ((backslash & 0x7F) as base.u64), + value_major: 0, + value_minor: 0x40_0000 | ((backslash & 0x7F) as base.u32), + link: 0x3, length: 2) continue.string_loop @@ -294,7 +289,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { // Not a Unicode surrogate. We're good. args.src.skip32_fast!(actual: 6, worst_case: 6) args.dst.write_fast_token!( - value: 0x40_0000 | (uni4_value as base.u64), + value_major: 0, + value_minor: 0x40_0000 | uni4_value, + link: 0x3, length: 6) continue.string_loop @@ -358,7 +355,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { uni4_value -= 0xDC00 args.src.skip32_fast!(actual: 8, worst_case: 8) args.dst.write_fast_token!( - value: 0x40_0000 | ((uni4_high_surrogate + uni4_value) as base.u64), + value_major: 0, + value_minor: 0x40_0000 | uni4_high_surrogate | uni4_value, + link: 0x3, length: 12) continue.string_loop @@ -399,8 +398,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 2, worst_case: 2) if string_length >= 0xFFF8 { args.dst.write_fast_token!( - value: 0x20_0221, - length: (string_length + 2) as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length + 2) string_length = 0 continue.string_loop } @@ -429,8 +430,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 3, worst_case: 3) if string_length >= 0xFFF8 { args.dst.write_fast_token!( - value: 0x20_0221, - length: (string_length + 3) as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length + 3) string_length = 0 continue.string_loop } @@ -459,8 +462,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 4, worst_case: 4) if string_length >= 0xFFF8 { args.dst.write_fast_token!( - value: 0x20_0221, - length: (string_length + 4) as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length + 4) string_length = 0 continue.string_loop } @@ -472,8 +477,10 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { if string_length > 0 { args.dst.write_fast_token!( - value: 0x20_0221, - length: string_length as base.u64) + value_major: 0, + value_minor: 0x20_0021, + link: 0x3, + length: string_length) string_length = 0 } if char == 0x80 { @@ -483,7 +490,7 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { } } endwhile.string_loop - // The trailing '"' is filler. + // Emit the trailing '"'. while true { if args.src.available() <= 0 { if args.src.is_closed() { @@ -498,7 +505,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { } args.src.skip32_fast!(actual: 1, worst_case: 1) args.dst.write_fast_token!( - value: 0, + value_major: 0, + value_minor: 0x20_0013, + link: 0x2, length: 1) break } @@ -517,7 +526,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) // The ',' is filler. args.dst.write_fast_token!( - value: 0, + value_major: 0, + value_minor: 0, + link: 0x0, length: 1) // What's valid after a comma depends on whether or not we're // in an array or an object. @@ -532,7 +543,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) // The ':' is filler. args.dst.write_fast_token!( - value: 0, + value_major: 0, + value_minor: 0, + link: 0x0, length: 1) expect = 0x0EB2 // 0x0EB2 is EXPECT_VALUE. continue.outer @@ -544,15 +557,17 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { { number_length = this.decode_number!(src: args.src) number_status = number_length >> 8 - token_value = 0x60_0004 + vminor = 0x60_0004 if (number_length & 0x80) <> 0 { - token_value = 0x60_0002 + vminor = 0x60_0002 } number_length = number_length & 0x7F if number_status == 0 { args.dst.write_fast_token!( - value: token_value, - length: number_length as base.u64) + value_major: 0, + value_minor: vminor, + link: 0x0, + length: number_length) break } @@ -582,13 +597,13 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { // -------- END parse numbers. } else if class == 0x05 { // 0x05 is CLASS_OPEN_CURLY_BRACE. - token_value = 0x80_4011 + vminor = 0x80_4011 if depth == 0 { // No-op. } else if 0 <> (expect_after_value & 0x0040) { // 0x0040 is EXPECT_CLOSE_CURLY_BRACE. - token_value = 0x80_4041 + vminor = 0x80_4041 } else { - token_value = 0x80_4021 + vminor = 0x80_4021 } if depth >= 1024 { return "#unsupported recursion depth" @@ -600,7 +615,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) args.dst.write_fast_token!( - value: token_value, + value_major: 0, + value_minor: vminor, + link: 0x0, length: 1) expect = 0x0042 // 0x0042 is (EXPECT_CLOSE_CURLY_BRACE | EXPECT_STRING). expect_after_value = 0x0044 // 0x0044 is (EXPECT_CURLY_CLOSE_BRACE | EXPECT_COMMA). @@ -610,7 +627,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) if depth <= 1 { args.dst.write_fast_token!( - value: 0x80_1042, + value_major: 0, + value_minor: 0x80_1042, + link: 0x0, length: 1) break.outer } @@ -619,14 +638,18 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { stack_bit = (depth - 1) & 31 if 0 == (this.stack[stack_byte] & ((1 as base.u32) << stack_bit)) { args.dst.write_fast_token!( - value: 0x80_2042, + value_major: 0, + value_minor: 0x80_2042, + link: 0x0, length: 1) // 0x0104 is (EXPECT_SQUARE_CLOSE_BRACKET | EXPECT_COMMA). expect = 0x0104 expect_after_value = 0x0104 } else { args.dst.write_fast_token!( - value: 0x80_4042, + value_major: 0, + value_minor: 0x80_4042, + link: 0x0, length: 1) // 0x0044 is (EXPECT_CURLY_CLOSE_BRACE | EXPECT_COMMA). expect = 0x0044 @@ -635,13 +658,13 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { continue.outer } else if class == 0x07 { // 0x07 is CLASS_OPEN_SQUARE_BRACKET. - token_value = 0x80_2011 + vminor = 0x80_2011 if depth == 0 { // No-op. } else if 0 <> (expect_after_value & 0x0040) { // 0x0040 is EXPECT_CLOSE_CURLY_BRACE. - token_value = 0x80_2041 + vminor = 0x80_2041 } else { - token_value = 0x80_2021 + vminor = 0x80_2021 } if depth >= 1024 { return "#unsupported recursion depth" @@ -653,7 +676,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) args.dst.write_fast_token!( - value: token_value, + value_major: 0, + value_minor: vminor, + link: 0x0, length: 1) expect = 0x0FB2 // 0x0FB2 is (EXPECT_CLOSE_SQUARE_BRACKET | EXPECT_VALUE). expect_after_value = 0x0104 // 0x0104 is (EXPECT_CLOSE_SQUARE_BRACKET | EXPECT_COMMA). @@ -663,7 +688,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { args.src.skip32_fast!(actual: 1, worst_case: 1) if depth <= 1 { args.dst.write_fast_token!( - value: 0x80_1022, + value_major: 0, + value_minor: 0x80_1022, + link: 0x0, length: 1) break.outer } @@ -672,14 +699,18 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { stack_bit = (depth - 1) & 31 if 0 == (this.stack[stack_byte] & ((1 as base.u32) << stack_bit)) { args.dst.write_fast_token!( - value: 0x80_2022, + value_major: 0, + value_minor: 0x80_2022, + link: 0x0, length: 1) // 0x0104 is (EXPECT_CLOSE_SQUARE_BRACKET | EXPECT_COMMA). expect = 0x0104 expect_after_value = 0x0104 } else { args.dst.write_fast_token!( - value: 0x80_4022, + value_major: 0, + value_minor: 0x80_4022, + link: 0x0, length: 1) // 0x0044 is (EXPECT_CLOSE_CURLY_BRACE | EXPECT_COMMA). expect = 0x0044 @@ -691,7 +722,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { match = args.src.match7(a: 0x6573_6C61_6605) // 5 bytes "false". if match == 0 { args.dst.write_fast_token!( - value: 0x60_0401, + value_major: 0, + value_minor: 0x60_0401, + link: 0x0, length: 5) if args.src.available() < 5 { return "#internal error: inconsistent I/O" @@ -707,7 +740,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { match = args.src.match7(a: 0x65_7572_7404) // 4 bytes "true". if match == 0 { args.dst.write_fast_token!( - value: 0x60_0801, + value_major: 0, + value_minor: 0x60_0801, + link: 0x0, length: 4) if args.src.available() < 4 { return "#internal error: inconsistent I/O" @@ -723,7 +758,9 @@ pub func decoder.decode_tokens?(dst: base.token_writer, src: base.io_reader) { match = args.src.match7(a: 0x6C_6C75_6E04) // 4 bytes "null". if match == 0 { args.dst.write_fast_token!( - value: 0x60_0201, + value_major: 0, + value_minor: 0x60_0201, + link: 0x0, length: 4) if args.src.available() < 4 { return "#internal error: inconsistent I/O" diff --git a/test/c/testlib/testlib.c b/test/c/testlib/testlib.c index 91327942..fe61b707 100644 --- a/test/c/testlib/testlib.c +++ b/test/c/testlib/testlib.c @@ -1089,11 +1089,12 @@ do_test__wuffs_base__token_decoder(wuffs_base__token_decoder* b, uint64_t pos = 0; while (tok.meta.ri < tok.meta.wi) { wuffs_base__token* t = &tok.data.ptr[tok.meta.ri++]; - uint64_t len = wuffs_base__token__length(t); + uint16_t len = wuffs_base__token__length(t); if (wuffs_base__token__value(t) != 0) { - uint64_t maj = wuffs_base__token__value_major(t); - uint64_t min = wuffs_base__token__value_minor(t); + uint8_t lp = wuffs_base__token__link_prev(t) ? 1 : 0; + uint8_t ln = wuffs_base__token__link_next(t) ? 1 : 0; + uint32_t vmajor = wuffs_base__token__value_major(t); if ((have.data.len - have.meta.wi) < 16) { return "testlib: output is too long"; @@ -1101,10 +1102,21 @@ do_test__wuffs_base__token_decoder(wuffs_base__token_decoder* b, // This 16-bytes-per-token debug format is the same one used by // `script/print-json-token-debug-format.c`. uint8_t* ptr = have.data.ptr + have.meta.wi; - wuffs_base__store_u32be__no_bounds_check(ptr + (0 * 4), (uint32_t)(pos)); - wuffs_base__store_u32be__no_bounds_check(ptr + (1 * 4), (uint32_t)(len)); - wuffs_base__store_u32be__no_bounds_check(ptr + (2 * 4), (uint32_t)(maj)); - wuffs_base__store_u32be__no_bounds_check(ptr + (3 * 4), (uint32_t)(min)); + + wuffs_base__store_u32be__no_bounds_check(ptr + 0x0, (uint32_t)(pos)); + wuffs_base__store_u16be__no_bounds_check(ptr + 0x4, len); + wuffs_base__store_u8__no_bounds_check(ptr + 0x0006, lp); + wuffs_base__store_u8__no_bounds_check(ptr + 0x0007, ln); + wuffs_base__store_u32be__no_bounds_check(ptr + 0x8, vmajor); + if (vmajor) { + uint32_t vminor = wuffs_base__token__value_minor(t); + wuffs_base__store_u32be__no_bounds_check(ptr + 0xC, vminor); + } else { + uint8_t vbc = wuffs_base__token__value_base_category(t); + uint32_t vbd = wuffs_base__token__value_base_detail(t); + wuffs_base__store_u8__no_bounds_check(ptr + 0x000C, vbc); + wuffs_base__store_u24be__no_bounds_check(ptr + 0xD, vbd); + } have.meta.wi += 16; } diff --git a/test/data/australian-abc-local-stations.tokens b/test/data/australian-abc-local-stations.tokens index 97ffa61e..f8b40ee6 100644 Binary files a/test/data/australian-abc-local-stations.tokens and b/test/data/australian-abc-local-stations.tokens differ diff --git a/test/data/json-things.unformatted.tokens b/test/data/json-things.unformatted.tokens index 4e930942..856ea18e 100644 Binary files a/test/data/json-things.unformatted.tokens and b/test/data/json-things.unformatted.tokens differ