diff options
Diffstat (limited to 'source')
| -rw-r--r-- | source/elna/ast.elna | 21 | ||||
| -rw-r--r-- | source/elna/lexer.elna | 270 | ||||
| -rw-r--r-- | source/elna/parser.elna | 29 | ||||
| -rw-r--r-- | source/main.elna | 14 | ||||
| -rw-r--r-- | source/sys/posix/stdio.elna | 20 | ||||
| -rw-r--r-- | source/sys/posix/stdlib.elna | 15 | ||||
| -rw-r--r-- | source/sys/posix/string.elna | 6 |
7 files changed, 218 insertions, 157 deletions
diff --git a/source/elna/ast.elna b/source/elna/ast.elna new file mode 100644 index 0000000..2fca087 --- /dev/null +++ b/source/elna/ast.elna @@ -0,0 +1,21 @@ +(* This Source Code Form is subject to the terms of the Mozilla Public License, + v. 2.0. If a copy of the MPL was not distributed with this file, You can + obtain one at https://mozilla.org/MPL/2.0/. *) + +(* Abstract syntax tree representation. *) +module + +import elna.source + +type + ElnaAstKind* = ( + unit + ) + ElnaAstNode* = record + kind: ElnaAstKind; + position: ElnaSourceSpan + end + ElnaAstUnit* = record(ElnaAstNode) + end + +end. diff --git a/source/elna/lexer.elna b/source/elna/lexer.elna index f7bb4f7..05dd1b2 100644 --- a/source/elna/lexer.elna +++ b/source/elna/lexer.elna @@ -6,7 +6,7 @@ module import - std, sys.posix.stdlib, sys.posix.string, sys.posix.stdio, sys.posix.errno, + std, sys.ffi, sys.posix.stdlib, sys.posix.string, sys.posix.stdio, sys.posix.errno, elna.source type @@ -925,7 +925,7 @@ begin hex_digits := [0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8] begin memcpy(hex_digits.ptr, digits_start, cast(current_position - digits_start: Word)); - codepoint := cast(strtoul(hex_digits.ptr, cast(0: ^^Word8), 16): Word32); + codepoint := cast(strtoul(hex_digits.ptr, cast(0: ^^CChar), cast(16: CInt)): Word32); if (codepoint >= ELNA_LEXER_UTF8_SURROGATE_RANGE_START & codepoint <= ELNA_LEXER_UTF8_SURROGATE_RANGE_END) or codepoint > ELNA_LEXER_UTF8_MAX_UNICODE_CODEPOINT then @@ -948,7 +948,7 @@ begin if elna_lexer_is_hex(current_position^) & elna_lexer_is_hex((current_position + 1)^) then var hex_digits: const := [current_position^, (current_position + 1)^, 0u8] - character: const := cast(strtol(hex_digits.ptr, cast(0: ^^Word8), 16): Word8) + character: const := cast(strtol(hex_digits.ptr, cast(0: ^^CChar), cast(16: CInt)): Word8) begin output^^ := character; output^ := output^ + 1; @@ -1121,14 +1121,14 @@ begin if number_string.length > 2u & number_string[1] = '0' & (number_string[2] = 'b' or number_string[2] = 'B') then start := start + 2; - integer_token.magnitude := strtoull(start, @number_stop, 2); + integer_token.magnitude := strtoull(start, @number_stop, cast(2: CInt)); is_prefixed := true elsif number_string.length > 2u & number_string[1] = '0' & (number_string[2] = 'x' or number_string[2] = 'X') then start := start + 2; - integer_token.magnitude := strtoull(start, @number_stop, 16); + integer_token.magnitude := strtoull(start, @number_stop, cast(16: CInt)); is_prefixed := true else - integer_token.magnitude := strtoull(start, @number_stop, 10) + integer_token.magnitude := strtoull(start, @number_stop, cast(10: CInt)) end; if number_stop > start & errno()^ <> ERANGE then if cursor = number_stop then @@ -1194,13 +1194,13 @@ return Creates a token of the given kind. The created token is a simple token without any payload. *) -proc elna_lexer_token_create(kind: ElnaLexerTokenKind) -> ^ElnaLexerToken +proc elna_lexer_token_create(kind: ElnaLexerTokenKind) -> StdOptional#[ElnaLexerToken] var - result: ^ElnaLexerToken + plain_token: ^ElnaLexerToken begin - result := malloc#[ElnaLexerToken](#size(ElnaLexerToken)); - result^.kind := kind -return result + plain_token := malloc#[ElnaLexerToken](#size(ElnaLexerToken)); + plain_token^.kind := kind +return std_optional_make(plain_token) (* Frees the memory allocated for the token itself and for its internal storage. @@ -1229,9 +1229,9 @@ proc elna_lexer_token_destroy*(token: ^ElnaLexerToken) free(token) return -proc elna_lexer_identify_keyword(start, cursor: ^const Word8) -> ^ElnaLexerToken +proc elna_lexer_identify_keyword(start, cursor: ^const Word8) -> StdOptional#[ElnaLexerToken] var - result: ^ElnaLexerToken + result: StdOptional#[ElnaLexerToken] identifier: StdString := start[1 to cursor - start] begin case identifier of @@ -1278,11 +1278,11 @@ begin identifier_token^.kind := ElnaLexerTokenKind.identifier; identifier_token^.identifier := identifier_pointer[1 to identifier.length]; - result := identifier_token + result := std_optional_make#[ElnaLexerToken](identifier_token) end return result -proc elna_lexer_identify_trait(start, cursor: ^const Word8) -> ^ElnaLexerToken +proc elna_lexer_identify_trait(start, cursor: ^const Word8) -> StdOptional#[ElnaLexerToken] var identifier: StdString := start[2 to cursor - start] result := malloc#[ElnaLexerIdentifierToken](#size(ElnaLexerIdentifierToken)) @@ -1292,11 +1292,11 @@ begin result^.kind := ElnaLexerTokenKind.trait; result^.identifier := identifier_pointer[1 to identifier.length] -return result +return std_optional_make#[ElnaLexerToken](result) -proc elna_lexer_stop_single(start: ^const Word8) -> ^ElnaLexerToken +proc elna_lexer_stop_single(start: ^const Word8) -> StdOptional#[ElnaLexerToken] var - result: ^ElnaLexerToken + result: StdOptional#[ElnaLexerToken] begin case start^ of '&': result := elna_lexer_token_create(ElnaLexerTokenKind.and) | @@ -1322,7 +1322,7 @@ begin end return result -proc elna_lexer_stop_comment(kind: ElnaLexerTokenKind; start, cursor: ^const Word8) -> ^ElnaLexerCommentToken +proc elna_lexer_stop_comment(kind: ElnaLexerTokenKind; start, cursor: ^const Word8) -> StdOptional#[ElnaLexerToken] var comment_token := malloc#[ElnaLexerCommentToken](#size(ElnaLexerCommentToken)) comment_start: ^const Word8 := start + 2u @@ -1332,17 +1332,14 @@ begin comment_token^.kind := kind; memcpy(comment_pointer, comment_start, comment_length); comment_token^.content := comment_pointer[1 to comment_length] -return comment_token +return std_optional_make#[ElnaLexerToken](comment_token) proc elna_lexer_delimit_string(start, cursor: ^const Word8) -> StdOptional#[ElnaLexerToken] var result := { StdOptional#[ElnaLexerToken] } begin if start^ = '(' then - var - comment_token := elna_lexer_stop_comment(ElnaLexerTokenKind.docblock, start, cursor - 2) - begin - result := std_optional_make#[ElnaLexerToken](comment_token) + result := elna_lexer_stop_comment(ElnaLexerTokenKind.docblock, start, cursor - 2) elsif start^ = '\'' then result := elna_lexer_parse_byte_literal(start) elsif start^ = '`' then @@ -1352,17 +1349,15 @@ begin end return result --- TODO: Return type should be "error or a pointer". --- TODO: Save the position in the token. -proc elna_lexer_lex*(cursor: ^ElnaLexerCursor) -> ^ElnaLexerToken +proc elna_lexer_lex*(cursor: ^ElnaLexerCursor) -> StdOptional#[ElnaLexerToken] var - result: ^ElnaLexerToken := cast(0: ^ElnaLexerToken) + result := { StdOptional#[ElnaLexerToken] } begin elna_lexer_skip_trivia(cursor); case cursor^.state of ElnaLexerState.stop: - result := elna_lexer_delimit_string(cursor^.start, cursor^.cursor).value + result := elna_lexer_delimit_string(cursor^.start, cursor^.cursor) | ElnaLexerState.comment: result := elna_lexer_stop_comment(ElnaLexerTokenKind.comment, cursor^.start, cursor^.cursor) @@ -1370,12 +1365,12 @@ begin ElnaLexerState.leading_zero, ElnaLexerState.integer, ElnaLexerState.integer_tail: - result := elna_lexer_parse_integer(cursor^.start, cursor^.cursor).value + result := elna_lexer_parse_integer(cursor^.start, cursor^.cursor) | ElnaLexerState.float_tail, ElnaLexerState.exponent, ElnaLexerState.fraction: - result := elna_lexer_parse_float(cursor^.start, cursor^.cursor).value + result := elna_lexer_parse_float(cursor^.start, cursor^.cursor) | ElnaLexerState.identifier: result := elna_lexer_identify_keyword(cursor^.start, cursor^.cursor) | ElnaLexerState.single: result := elna_lexer_stop_single(cursor^.start) | @@ -1395,166 +1390,167 @@ begin ElnaLexerState.less: result := elna_lexer_token_create(ElnaLexerTokenKind.less_than) | ElnaLexerState.trait: result := elna_lexer_identify_trait(cursor^.start, cursor^.cursor) | ElnaLexerState.eof: result := elna_lexer_token_create(ElnaLexerTokenKind.eof) - else - result := malloc#[ElnaLexerToken](#size(ElnaLexerToken)); - memset(result, 0, #size(ElnaLexerToken)) + end; + if result.has_value then + result.value^.position := { ElnaSourceSpan; + start_offset: cursor^.position, + end_offset: cursor^.position + cast(cursor^.cursor - cursor^.start: Word) + } end return result -proc elna_lexer_token_dump_comment(token: ^ElnaLexerCommentToken) +proc elna_lexer_token_dump_comment(token: ^ElnaLexerCommentToken; stream: ^FILE) if token^.content.length > 6u then - printf("--%.*s…%.*s\0".ptr, 3, token^.content.ptr, 3, token^.content.ptr + token^.content.length - 3u) + fprintf(stream, "--%.*s…%.*s\0".ptr, 3, token^.content.ptr, 3, token^.content.ptr + token^.content.length - 3u) else - printf("--%.*s\0".ptr, token^.content.length, token^.content.ptr) + fprintf(stream, "--%.*s\0".ptr, token^.content.length, token^.content.ptr) end return -proc elna_lexer_token_dump_docblock(token: ^ElnaLexerCommentToken) +proc elna_lexer_token_dump_docblock(token: ^ElnaLexerCommentToken; stream: ^FILE) if token^.content.length > 6u then - printf("(*%.*s…%.*s*)\0".ptr, 3, token^.content.ptr, + fprintf(stream, "(*%.*s…%.*s*)\0".ptr, 3, token^.content.ptr, 3, token^.content.ptr + token^.content.length - 3u) else - printf("(*%.*s*)\0".ptr, token^.content.length, token^.content.ptr) + fprintf(stream, "(*%.*s*)\0".ptr, token^.content.length, token^.content.ptr) end return -proc elna_lexer_dump_string(token: ^ElnaLexerStringToken) - putchar(cast('"': Int)); +proc elna_lexer_dump_string(token: ^ElnaLexerStringToken; stream: ^FILE) + fputc(cast('"': CInt), stream); for c of token^.string do if c^ = '\f' then - printf("\\f\0".ptr) + fprintf(stream, "\\f\0".ptr) elsif c^ = '\v' then - printf("\\v\0".ptr) + fprintf(stream, "\\v\0".ptr) elsif c^ = '\\' then - printf("\\\\\0".ptr) + fprintf(stream, "\\\\\0".ptr) elsif c^ = '"' then - printf("\\\"\0".ptr) + fprintf(stream, "\\\"\0".ptr) elsif c^ = '\0' then - printf("\\0\0".ptr) + fprintf(stream, "\\0\0".ptr) elsif c^ = '\n' or c^ = '\t' or c^ = '\r' then - putchar(cast(c^: Int)) + fputc(cast(c^: CInt), stream) elsif c^ < 32u8 or c^ = 127u8 then - printf("\\x%02X\0".ptr, c^) + fprintf(stream, "\\x%02X\0".ptr, c^) else - putchar(cast(c^: Int)) + fputc(cast(c^: CInt), stream) end end; - putchar(cast('"': Int)) + fputc(cast('"': CInt), stream) return -proc elna_lexer_dump_integer(token: ^ElnaLexerIntegerToken) +proc elna_lexer_dump_integer(token: ^ElnaLexerIntegerToken; stream: ^FILE) if token^.is_signed then if token^.size = 0u8 then - printf("«%lld»\0".ptr, token^.magnitude) + fprintf(stream, "«%lld»\0".ptr, token^.magnitude) else - printf("«%lldi%d»\0".ptr, token^.magnitude, token^.size) + fprintf(stream, "«%lldi%d»\0".ptr, token^.magnitude, token^.size) end else if token^.size = 0u8 then - printf("«0x%Xu»\0".ptr, token^.magnitude) + fprintf(stream, "«0x%Xu»\0".ptr, token^.magnitude) else - printf("«0x%Xu%d»\0".ptr, token^.magnitude, token^.size) + fprintf(stream, "«0x%Xu%d»\0".ptr, token^.magnitude, token^.size) end end return --- TODO: Replace printf and other writing functions with a writer callback. --- TODO: Cursor shouldn't be a parameter, only token. -proc elna_lexer_token_dump*(token: ^ElnaLexerToken; cursor: ^ElnaLexerCursor) - printf("⟨\0".ptr); +proc elna_lexer_token_dump*(token: ^ElnaLexerToken; stream: ^FILE) + fprintf(stream, "⟨\0".ptr); case token^.kind of - ElnaLexerTokenKind.comment: elna_lexer_token_dump_comment(cast(token: ^ElnaLexerCommentToken)) | - ElnaLexerTokenKind.docblock: elna_lexer_token_dump_docblock(cast(token: ^ElnaLexerCommentToken)) | - ElnaLexerTokenKind.if_: printf("«if»\0".ptr) | - ElnaLexerTokenKind.then_: printf("«then»\0".ptr) | - ElnaLexerTokenKind.else_: printf("«else»\0".ptr) | - ElnaLexerTokenKind.elsif_: printf("«elsif»\0".ptr) | - ElnaLexerTokenKind.while_: printf("«while»\0".ptr) | - ElnaLexerTokenKind.for_: printf("«for»\0".ptr) | - ElnaLexerTokenKind.repeat_: printf("«repeat»\0".ptr) | - ElnaLexerTokenKind.until_: printf("«until»\0".ptr) | - ElnaLexerTokenKind.with_: printf("«with»\0".ptr) | - ElnaLexerTokenKind.do_: printf("«do»\0".ptr) | - ElnaLexerTokenKind.proc_: printf("«proc»\0".ptr) | - ElnaLexerTokenKind.begin_: printf("«begin»\0".ptr) | - ElnaLexerTokenKind.end_: printf("«end»\0".ptr) | - ElnaLexerTokenKind.extern_: printf("«extern»\0".ptr) | - ElnaLexerTokenKind.const_: printf("«const»\0".ptr) | - ElnaLexerTokenKind.var_: printf("«var»\0".ptr) | - ElnaLexerTokenKind.threadvar_: printf("«threadvar»\0".ptr) | - ElnaLexerTokenKind.type_: printf("«type»\0".ptr) | - ElnaLexerTokenKind.record_: printf("«record»\0".ptr) | - ElnaLexerTokenKind.true_: printf("«true»\0".ptr) | - ElnaLexerTokenKind.false_: printf("«false»\0".ptr) | - ElnaLexerTokenKind.and: printf("«&»\0".ptr) | - ElnaLexerTokenKind.xor_: printf("«xor»\0".ptr) | - ElnaLexerTokenKind.or_: printf("«or»\0".ptr) | - ElnaLexerTokenKind.pipe: printf("«|»\0".ptr) | - ElnaLexerTokenKind.not: printf("«~»\0".ptr) | - ElnaLexerTokenKind.return_: printf("«return»\0".ptr) | - ElnaLexerTokenKind.import_: printf("«import»\0".ptr) | - ElnaLexerTokenKind.cast_: printf("«cast»\0".ptr) | - ElnaLexerTokenKind.defer_: printf("«defer»\0".ptr) | - ElnaLexerTokenKind.block_: printf("«block»\0".ptr) | - ElnaLexerTokenKind.break_: printf("«break»\0".ptr) | - ElnaLexerTokenKind.case_: printf("«case»\0".ptr) | - ElnaLexerTokenKind.of_: printf("«of»\0".ptr) | - ElnaLexerTokenKind.to_: printf("«to»\0".ptr) | - ElnaLexerTokenKind.program_: printf("«program»\0".ptr) | - ElnaLexerTokenKind.module_: printf("«module»\0".ptr) | + ElnaLexerTokenKind.comment: elna_lexer_token_dump_comment(cast(token: ^ElnaLexerCommentToken), stream) | + ElnaLexerTokenKind.docblock: elna_lexer_token_dump_docblock(cast(token: ^ElnaLexerCommentToken), stream) | + ElnaLexerTokenKind.if_: fprintf(stream, "«if»\0".ptr) | + ElnaLexerTokenKind.then_: fprintf(stream, "«then»\0".ptr) | + ElnaLexerTokenKind.else_: fprintf(stream, "«else»\0".ptr) | + ElnaLexerTokenKind.elsif_: fprintf(stream, "«elsif»\0".ptr) | + ElnaLexerTokenKind.while_: fprintf(stream, "«while»\0".ptr) | + ElnaLexerTokenKind.for_: fprintf(stream, "«for»\0".ptr) | + ElnaLexerTokenKind.repeat_: fprintf(stream, "«repeat»\0".ptr) | + ElnaLexerTokenKind.until_: fprintf(stream, "«until»\0".ptr) | + ElnaLexerTokenKind.with_: fprintf(stream, "«with»\0".ptr) | + ElnaLexerTokenKind.do_: fprintf(stream, "«do»\0".ptr) | + ElnaLexerTokenKind.proc_: fprintf(stream, "«proc»\0".ptr) | + ElnaLexerTokenKind.begin_: fprintf(stream, "«begin»\0".ptr) | + ElnaLexerTokenKind.end_: fprintf(stream, "«end»\0".ptr) | + ElnaLexerTokenKind.extern_: fprintf(stream, "«extern»\0".ptr) | + ElnaLexerTokenKind.const_: fprintf(stream, "«const»\0".ptr) | + ElnaLexerTokenKind.var_: fprintf(stream, "«var»\0".ptr) | + ElnaLexerTokenKind.threadvar_: fprintf(stream, "«threadvar»\0".ptr) | + ElnaLexerTokenKind.type_: fprintf(stream, "«type»\0".ptr) | + ElnaLexerTokenKind.record_: fprintf(stream, "«record»\0".ptr) | + ElnaLexerTokenKind.true_: fprintf(stream, "«true»\0".ptr) | + ElnaLexerTokenKind.false_: fprintf(stream, "«false»\0".ptr) | + ElnaLexerTokenKind.and: fprintf(stream, "«&»\0".ptr) | + ElnaLexerTokenKind.xor_: fprintf(stream, "«xor»\0".ptr) | + ElnaLexerTokenKind.or_: fprintf(stream, "«or»\0".ptr) | + ElnaLexerTokenKind.pipe: fprintf(stream, "«|»\0".ptr) | + ElnaLexerTokenKind.not: fprintf(stream, "«~»\0".ptr) | + ElnaLexerTokenKind.return_: fprintf(stream, "«return»\0".ptr) | + ElnaLexerTokenKind.import_: fprintf(stream, "«import»\0".ptr) | + ElnaLexerTokenKind.cast_: fprintf(stream, "«cast»\0".ptr) | + ElnaLexerTokenKind.defer_: fprintf(stream, "«defer»\0".ptr) | + ElnaLexerTokenKind.block_: fprintf(stream, "«block»\0".ptr) | + ElnaLexerTokenKind.break_: fprintf(stream, "«break»\0".ptr) | + ElnaLexerTokenKind.case_: fprintf(stream, "«case»\0".ptr) | + ElnaLexerTokenKind.of_: fprintf(stream, "«of»\0".ptr) | + ElnaLexerTokenKind.to_: fprintf(stream, "«to»\0".ptr) | + ElnaLexerTokenKind.program_: fprintf(stream, "«program»\0".ptr) | + ElnaLexerTokenKind.module_: fprintf(stream, "«module»\0".ptr) | ElnaLexerTokenKind.identifier: var identifier := cast(token: ^ElnaLexerIdentifierToken)^.identifier begin - printf("Identifier «%.*s»\0".ptr, identifier.length, identifier.ptr) + fprintf(stream, "Identifier «%.*s»\0".ptr, identifier.length, identifier.ptr) | ElnaLexerTokenKind.trait: var identifier := cast(token: ^ElnaLexerIdentifierToken)^.identifier begin - printf("#%.*s\0".ptr, identifier.length, identifier.ptr) + fprintf(stream, "#%.*s\0".ptr, identifier.length, identifier.ptr) | - ElnaLexerTokenKind.generic_open: printf("«#[»\0".ptr) | - ElnaLexerTokenKind.attribute_open: printf("«#{»\0".ptr) | - ElnaLexerTokenKind.integer: elna_lexer_dump_integer(cast(token: ^ElnaLexerIntegerToken)) | + ElnaLexerTokenKind.generic_open: fprintf(stream, "«#[»\0".ptr) | + ElnaLexerTokenKind.attribute_open: fprintf(stream, "«#{»\0".ptr) | + ElnaLexerTokenKind.integer: elna_lexer_dump_integer(cast(token: ^ElnaLexerIntegerToken), stream) | ElnaLexerTokenKind.double: - printf("«%g»\0".ptr, cast(token: ^ElnaLexerFloatToken)^.value) | + fprintf(stream, "«%g»\0".ptr, cast(token: ^ElnaLexerFloatToken)^.value) | ElnaLexerTokenKind.single: - printf("«%gf»\0".ptr, cast(token: ^ElnaLexerFloatToken)^.value) | + fprintf(stream, "«%gf»\0".ptr, cast(token: ^ElnaLexerFloatToken)^.value) | ElnaLexerTokenKind.character: - printf("\\{U+%X}\0".ptr, cast(token: ^ElnaLexerCharacterToken)^.codepoint) | - ElnaLexerTokenKind.string: elna_lexer_dump_string(cast(token: ^ElnaLexerStringToken)) | - ElnaLexerTokenKind.left_paren: printf("«(»\0".ptr) | - ElnaLexerTokenKind.right_paren: printf("«)»\0".ptr) | - ElnaLexerTokenKind.left_square: printf("«[»\0".ptr) | - ElnaLexerTokenKind.right_square: printf("«]»\0".ptr) | - ElnaLexerTokenKind.left_brace: printf("«{»\0".ptr) | - ElnaLexerTokenKind.right_brace: printf("«}»\0".ptr) | - ElnaLexerTokenKind.shift_left: printf("«<<»\0".ptr) | - ElnaLexerTokenKind.shift_right: printf("«>>»\0".ptr) | - ElnaLexerTokenKind.greater_equal: printf("«>=»\0".ptr) | - ElnaLexerTokenKind.less_equal: printf("«<=»\0".ptr) | - ElnaLexerTokenKind.greater_than: printf("«>»\0".ptr) | - ElnaLexerTokenKind.less_than: printf("«<»\0".ptr) | - ElnaLexerTokenKind.not_equal: printf("«<>»\0".ptr) | - ElnaLexerTokenKind.equals: printf("«=»\0".ptr) | - ElnaLexerTokenKind.semicolon: printf("«;»\0".ptr) | - ElnaLexerTokenKind.dot: printf("«.»\0".ptr) | - ElnaLexerTokenKind.comma: printf("«,»\0".ptr) | - ElnaLexerTokenKind.plus: printf("«+»\0".ptr) | - ElnaLexerTokenKind.minus: printf("«-»\0".ptr) | - ElnaLexerTokenKind.multiplication: printf("«*»\0".ptr) | - ElnaLexerTokenKind.division: printf("«/»\0".ptr) | - ElnaLexerTokenKind.remainder: printf("«%»\0".ptr) | - ElnaLexerTokenKind.assignment: printf("«:=»\0".ptr) | - ElnaLexerTokenKind.colon: printf("«:»\0".ptr) | - ElnaLexerTokenKind.hat: printf("«^»\0".ptr) | - ElnaLexerTokenKind.at: printf("«@»\0".ptr) | - ElnaLexerTokenKind.arrow: printf("«->»\0".ptr) | - ElnaLexerTokenKind.eof: printf("EOF\0".ptr) + fprintf(stream, "\\{U+%X}\0".ptr, cast(token: ^ElnaLexerCharacterToken)^.codepoint) | + ElnaLexerTokenKind.string: elna_lexer_dump_string(cast(token: ^ElnaLexerStringToken), stream) | + ElnaLexerTokenKind.left_paren: fprintf(stream, "«(»\0".ptr) | + ElnaLexerTokenKind.right_paren: fprintf(stream, "«)»\0".ptr) | + ElnaLexerTokenKind.left_square: fprintf(stream, "«[»\0".ptr) | + ElnaLexerTokenKind.right_square: fprintf(stream, "«]»\0".ptr) | + ElnaLexerTokenKind.left_brace: fprintf(stream, "«{»\0".ptr) | + ElnaLexerTokenKind.right_brace: fprintf(stream, "«}»\0".ptr) | + ElnaLexerTokenKind.shift_left: fprintf(stream, "«<<»\0".ptr) | + ElnaLexerTokenKind.shift_right: fprintf(stream, "«>>»\0".ptr) | + ElnaLexerTokenKind.greater_equal: fprintf(stream, "«>=»\0".ptr) | + ElnaLexerTokenKind.less_equal: fprintf(stream, "«<=»\0".ptr) | + ElnaLexerTokenKind.greater_than: fprintf(stream, "«>»\0".ptr) | + ElnaLexerTokenKind.less_than: fprintf(stream, "«<»\0".ptr) | + ElnaLexerTokenKind.not_equal: fprintf(stream, "«<>»\0".ptr) | + ElnaLexerTokenKind.equals: fprintf(stream, "«=»\0".ptr) | + ElnaLexerTokenKind.semicolon: fprintf(stream, "«;»\0".ptr) | + ElnaLexerTokenKind.dot: fprintf(stream, "«.»\0".ptr) | + ElnaLexerTokenKind.comma: fprintf(stream, "«,»\0".ptr) | + ElnaLexerTokenKind.plus: fprintf(stream, "«+»\0".ptr) | + ElnaLexerTokenKind.minus: fprintf(stream, "«-»\0".ptr) | + ElnaLexerTokenKind.multiplication: fprintf(stream, "«*»\0".ptr) | + ElnaLexerTokenKind.division: fprintf(stream, "«/»\0".ptr) | + ElnaLexerTokenKind.remainder: fprintf(stream, "«%»\0".ptr) | + ElnaLexerTokenKind.assignment: fprintf(stream, "«:=»\0".ptr) | + ElnaLexerTokenKind.colon: fprintf(stream, "«:»\0".ptr) | + ElnaLexerTokenKind.hat: fprintf(stream, "«^»\0".ptr) | + ElnaLexerTokenKind.at: fprintf(stream, "«@»\0".ptr) | + ElnaLexerTokenKind.arrow: fprintf(stream, "«->»\0".ptr) | + ElnaLexerTokenKind.eof: fprintf(stream, "EOF\0".ptr) end; - printf("; Position: %lu⟩\n\0".ptr, cursor^.position) + fprintf(stream, "; Position: [%lu, %lu)⟩\n\0".ptr, token^.position.start_offset, token^.position.end_offset) return end. diff --git a/source/elna/parser.elna b/source/elna/parser.elna new file mode 100644 index 0000000..d1448fa --- /dev/null +++ b/source/elna/parser.elna @@ -0,0 +1,29 @@ +(* This Source Code Form is subject to the terms of the Mozilla Public License, + v. 2.0. If a copy of the MPL was not distributed with this file, You can + obtain one at https://mozilla.org/MPL/2.0/. *) + +(* Syntactic analysis. *) +module + +import sys.posix.stdlib, sys.posix.stdio, std, elna.lexer, elna.ast + +proc elna_parser_program*(lexer_cursor: ^ElnaLexerCursor) -> StdOptional#[ElnaAstUnit] +var + token: ^ElnaLexerToken + unit: ^ElnaAstUnit +begin + token := elna_lexer_lex(lexer_cursor).value; + elna_lexer_token_dump(token, stderr_()); + + while token^.kind <> ElnaLexerTokenKind.eof do + elna_lexer_token_destroy(token); + + token := elna_lexer_lex(lexer_cursor).value; + elna_lexer_token_dump(token, stderr_()) + end; + elna_lexer_token_destroy(token); + + unit := malloc#[ElnaAstUnit](#size(ElnaAstUnit)) +return std_optional_make(unit) + +end. diff --git a/source/main.elna b/source/main.elna index fb1ae86..ebdeee1 100644 --- a/source/main.elna +++ b/source/main.elna @@ -2,12 +2,13 @@ v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at https://mozilla.org/MPL/2.0/. *) program() -import sys.posix.stdio, elna.lexer +import sys.posix.stdio, std, elna.lexer, elna.ast, elna.parser var global_cursor: ElnaLexerCursor sample_source_text := "Here.a few (* documentation block *) or #size if '\\n' `У` \"некоторая строка\" 23 -- End line comment\0" token: ^ElnaLexerToken + tree: StdOptional#[ElnaAstUnit] begin elna_lexer_transitions(); @@ -17,14 +18,7 @@ begin position: 1u }; - token := elna_lexer_lex(@global_cursor); - elna_lexer_token_dump(token, @global_cursor); + tree := elna_parser_program(@global_cursor); - while token^.kind <> ElnaLexerTokenKind.eof do - elna_lexer_token_destroy(token); - - token := elna_lexer_lex(@global_cursor); - elna_lexer_token_dump(token, @global_cursor) - end; - elna_lexer_token_destroy(token) + puts("void main(){}\0".ptr) end. diff --git a/source/sys/posix/stdio.elna b/source/sys/posix/stdio.elna index 39756a8..c667f43 100644 --- a/source/sys/posix/stdio.elna +++ b/source/sys/posix/stdio.elna @@ -3,13 +3,27 @@ obtain one at https://mozilla.org/MPL/2.0/. *) module -proc printf*(format: ^const Word8) #{varargs} -> Int +import sys.ffi, bits.stdio* + +proc fprintf*(stream: ^FILE; format: ^const CChar) #{varargs} -> CInt +extern + +proc printf*(format: ^const CChar) #{varargs} -> CInt +extern + +proc fputs*(s: ^const CChar; stream: ^FILE) -> CInt +extern + +proc puts*(s: ^const CChar) -> CInt extern -proc puts*(s: ^const Word8) -> Int +proc fputc*(c: CInt; stream: ^FILE) -> CInt extern -proc putchar*(c: Int) -> Int +proc putc*(c: CInt; stream: ^FILE) -> CInt +return fputc(c, stream) + +proc putchar*(c: CInt) -> CInt extern end. diff --git a/source/sys/posix/stdlib.elna b/source/sys/posix/stdlib.elna index 497e87d..272400a 100644 --- a/source/sys/posix/stdlib.elna +++ b/source/sys/posix/stdlib.elna @@ -3,6 +3,8 @@ obtain one at https://mozilla.org/MPL/2.0/. *) module +import sys.ffi + proc malloc#[T]*(size: Word) -> ^T extern @@ -12,19 +14,22 @@ extern proc realloc#[T]*(ptr: ^T; size: Word) -> ^T extern -proc strtol*(str: ^const Word8; endptr: ^^Word8; base: Int) -> Int +proc strtol*(str: ^const CChar; endptr: ^^CChar; base: CInt) -> CLong +extern + +proc strtoll*(str: ^const CChar; endptr: ^^CChar; base: CInt) -> CLongLong extern -proc strtoul*(str: ^const Word8; endptr: ^^Word8; base: Int) -> Word +proc strtoul*(str: ^const CChar; endptr: ^^CChar; base: CInt) -> CULong extern -proc strtoull*(str: ^const Word8; endptr: ^^Word8; base: Int) -> Word64 +proc strtoull*(str: ^const CChar; endptr: ^^CChar; base: CInt) -> CULongLong extern -proc strtof*(nptr: ^const Word8; endptr: ^^Word8) -> Single +proc strtof*(nptr: ^const CChar; endptr: ^^CChar) -> Single extern -proc strtod*(nptr: ^const Word8; endptr: ^^Word8) -> Double +proc strtod*(nptr: ^const CChar; endptr: ^^CChar) -> Double extern end. diff --git a/source/sys/posix/string.elna b/source/sys/posix/string.elna index 561d975..a9a1ce9 100644 --- a/source/sys/posix/string.elna +++ b/source/sys/posix/string.elna @@ -3,13 +3,15 @@ obtain one at https://mozilla.org/MPL/2.0/. *) module -proc memset#[T]*(b: ^T; c: Int; len: Word) -> ^T +import sys.ffi + +proc memset#[T]*(b: ^T; c: CInt; len: Word) -> ^T extern proc memcpy#[T]*(dst: ^T; src: ^const T; n: Word) -> ^T extern -proc memchr#[T]*(s: ^T; c: Int; n: Word) -> ^T +proc memchr#[T]*(s: ^T; c: CInt; n: Word) -> ^T extern end. |
