diff options
author | David Robillard <d@drobilla.net> | 2018-11-25 18:08:54 +0100 |
---|---|---|
committer | David Robillard <d@drobilla.net> | 2018-11-25 22:12:48 +0100 |
commit | 2a1a2522b23122e2428019489df9ddb5b0ec30ea (patch) | |
tree | a3edb59a31ca127f145cc0ca572a0d3eaa2f72e4 | |
parent | 61f44d803f7b197c83547927ef86329d9a13c5d2 (diff) | |
download | serd-2a1a2522b23122e2428019489df9ddb5b0ec30ea.tar.gz serd-2a1a2522b23122e2428019489df9ddb5b0ec30ea.tar.bz2 serd-2a1a2522b23122e2428019489df9ddb5b0ec30ea.zip |
Remove reader status and handle errors more precisely
-rw-r--r-- | src/byte_source.h | 5 | ||||
-rw-r--r-- | src/n3.c | 55 | ||||
-rw-r--r-- | src/reader.c | 13 | ||||
-rw-r--r-- | src/reader.h | 8 |
4 files changed, 37 insertions, 44 deletions
diff --git a/src/byte_source.h b/src/byte_source.h index 34df4fa3..bd4ec988 100644 --- a/src/byte_source.h +++ b/src/byte_source.h @@ -88,13 +88,14 @@ serd_byte_source_advance(SerdByteSource* source) } if (source->from_stream) { - source->eof = false; if (source->page_size > 1) { if (++source->read_head == source->page_size) { + source->eof = false; st = serd_byte_source_page(source); } } else { if (!source->read_func(&source->read_byte, 1, 1, source->stream)) { + source->eof = true; st = source->error_func(source->stream) ? SERD_ERR_UNKNOWN : SERD_FAILURE; } @@ -103,7 +104,7 @@ serd_byte_source_advance(SerdByteSource* source) ++source->read_head; // Move to next character in string } - return source->eof ? SERD_FAILURE : st; + return st ? st : source->eof ? SERD_FAILURE : SERD_SUCCESS; } #endif // SERD_BYTE_SOURCE_H @@ -195,9 +195,10 @@ read_utf8_character(SerdReader* reader, SerdNode* dest, uint8_t c) uint32_t size; uint8_t bytes[4]; SerdStatus st = read_utf8_bytes(reader, bytes, &size, c); - if (st) { - push_bytes(reader, dest, replacement_char, 3); + if (st && reader->strict) { return st; + } else if (st) { + return push_bytes(reader, dest, replacement_char, 3); } return push_bytes(reader, dest, bytes, size); @@ -209,9 +210,10 @@ read_utf8_code(SerdReader* reader, SerdNode* dest, uint32_t* code, uint8_t c) uint32_t size; uint8_t bytes[4]; SerdStatus st = read_utf8_bytes(reader, bytes, &size, c); - if (st) { - push_bytes(reader, dest, replacement_char, 3); + if (st && reader->strict) { return st; + } else if (st) { + return push_bytes(reader, dest, replacement_char, 3); } if (!(st = push_bytes(reader, dest, bytes, size))) { @@ -299,7 +301,7 @@ static SerdStatus read_STRING_LITERAL_LONG(SerdReader* reader, SerdNode* dest, uint8_t q) { SerdStatus st = SERD_SUCCESS; - while (!reader->status) { + while (!st) { const uint8_t c = peek_byte(reader); uint32_t code; switch (c) { @@ -315,7 +317,10 @@ read_STRING_LITERAL_LONG(SerdReader* reader, SerdNode* dest, uint8_t q) } break; default: - if (c == q) { + if (reader->source.eof) { + st = r_err(reader, SERD_ERR_BAD_SYNTAX, + "unexpected end of file\n"); + } else if (c == q) { eat_byte_safe(reader, q); const uint8_t q2 = eat_byte_safe(reader, peek_byte(reader)); const uint8_t q3 = peek_byte(reader); @@ -332,6 +337,7 @@ read_STRING_LITERAL_LONG(SerdReader* reader, SerdNode* dest, uint8_t q) } } } + return st; } @@ -341,24 +347,25 @@ static SerdStatus read_STRING_LITERAL(SerdReader* reader, SerdNode* dest, uint8_t q) { SerdStatus st = SERD_SUCCESS; - while (!reader->status) { + while (!st) { const uint8_t c = peek_byte(reader); uint32_t code = 0; switch (c) { case '\n': case '\r': - r_err(reader, SERD_ERR_BAD_SYNTAX, "line end in short string\n"); - return SERD_ERR_BAD_SYNTAX; + st = r_err(reader, SERD_ERR_BAD_SYNTAX, + "line end in short string\n"); + break; case '\\': eat_byte_safe(reader, c); if ((st = read_ECHAR(reader, dest, &dest->flags)) && (st = read_UCHAR(reader, dest, &code))) { - return r_err(reader, st, - "invalid escape `\\%c'\n", peek_byte(reader)); + st = r_err(reader, st, + "invalid escape `\\%c'\n", peek_byte(reader)); } break; default: if (c == q) { - eat_byte_check(reader, q); + eat_byte_safe(reader, q); return SERD_SUCCESS; } else { st = read_character( @@ -367,8 +374,7 @@ read_STRING_LITERAL(SerdReader* reader, SerdNode* dest, uint8_t q) } } - return st ? st - : eat_byte_check(reader, q) ? SERD_SUCCESS : SERD_ERR_BAD_SYNTAX; + return st; } static SerdStatus @@ -671,10 +677,7 @@ read_IRIREF(SerdReader* reader, SerdNode** dest) } else if (!(c & 0x80)) { push_byte(reader, *dest, c); } else if (read_utf8_character(reader, *dest, c)) { - if (reader->strict) { - return SERD_ERR_BAD_SYNTAX; - } - st = SERD_FAILURE; + st = reader->strict ? SERD_ERR_BAD_SYNTAX : SERD_FAILURE; } } } @@ -1457,8 +1460,6 @@ read_n3_statement(SerdReader* reader) ctx.subject = subj = 0; TRY(st, read_wrappedGraph(reader, &ctx)); read_ws_star(reader); - } else if (!ctx.subject) { - st = r_err(reader, SERD_ERR_BAD_SYNTAX, "bad subject\n"); } else if ((st = read_triples(reader, ctx, &ate_dot))) { if (st == SERD_FAILURE && s_type == '[') { return SERD_SUCCESS; @@ -1493,23 +1494,22 @@ read_turtleTrigDoc(SerdReader* reader) const size_t orig_stack_size = reader->stack.size; const SerdStatus st = read_n3_statement(reader); if (st > SERD_FAILURE) { - if (reader->strict) { + if (reader->strict || reader->source.eof) { serd_stack_pop_to(&reader->stack, orig_stack_size); return st; } skip_until(reader, '\n'); - reader->status = SERD_SUCCESS; } serd_stack_pop_to(&reader->stack, orig_stack_size); } - return reader->status; + return SERD_SUCCESS; } SerdStatus read_nquadsDoc(SerdReader* reader) { SerdStatus st = SERD_SUCCESS; - while (!reader->source.eof) { + while (!st && !reader->source.eof) { const size_t orig_stack_size = reader->stack.size; SerdStatementFlags flags = 0; @@ -1554,11 +1554,8 @@ read_nquadsDoc(SerdReader* reader) } } - if (emit_statement(reader, ctx, ctx.object)) { - break; - } - + st = emit_statement(reader, ctx, ctx.object); serd_stack_pop_to(&reader->stack, orig_stack_size); } - return reader->status; + return st; } diff --git a/src/reader.c b/src/reader.c index e286fd4e..ace6ef7d 100644 --- a/src/reader.c +++ b/src/reader.c @@ -83,7 +83,6 @@ push_node_padded(SerdReader* reader, size_t maxlen, &reader->stack, sizeof(SerdNode) + maxlen + 1, sizeof(SerdNode)); if (!mem) { - reader->status = SERD_ERR_OVERFLOW; return NULL; } @@ -278,15 +277,15 @@ serd_reader_start_string(SerdReader* reader, static SerdStatus serd_reader_prepare(SerdReader* reader) { - reader->status = serd_byte_source_prepare(&reader->source); - if (reader->status == SERD_SUCCESS) { - reader->status = skip_bom(reader); - } else if (reader->status == SERD_FAILURE) { + SerdStatus st = serd_byte_source_prepare(&reader->source); + if (st == SERD_SUCCESS) { + st = skip_bom(reader); + } else if (st == SERD_FAILURE) { reader->source.eof = true; } else { - r_err(reader, reader->status, "read error: %s\n", strerror(errno)); + r_err(reader, st, "read error: %s\n", strerror(errno)); } - return reader->status; + return st; } SerdStatus diff --git a/src/reader.h b/src/reader.h index 138c59b0..6e946f60 100644 --- a/src/reader.h +++ b/src/reader.h @@ -55,7 +55,6 @@ struct SerdReaderImpl { SerdStack stack; SerdSyntax syntax; unsigned next_id; - SerdStatus status; uint8_t* buf; char* bprefix; size_t bprefix_len; @@ -99,11 +98,8 @@ peek_byte(SerdReader* reader) static inline uint8_t eat_byte(SerdReader* reader) { - const uint8_t c = peek_byte(reader); - const SerdStatus st = serd_byte_source_advance(&reader->source); - if (st) { - reader->status = st; - } + const uint8_t c = peek_byte(reader); + serd_byte_source_advance(&reader->source); return c; } |