};
static struct lex_source *lex_source__ (const struct lexer *);
+static struct substring lex_source_get_syntax__ (const struct lex_source *,
+ int n0, int n1);
static const struct lex_token *lex_next__ (const struct lexer *, int n);
static void lex_source_push_endcmd__ (struct lex_source *);
src->tokens = deque_expand (&src->deque, src->tokens, sizeof *src->tokens);
token = &src->tokens[deque_push_front (&src->deque)];
- token_init (&token->token);
+ token->token = (struct token) { .type = T_STOP };
return token;
}
return lex_source_next__ (src, n);
else
{
- static const struct lex_token stop_token =
- { TOKEN_INITIALIZER (T_STOP, 0.0, ""), 0, 0, 0, 0 };
-
+ static const struct lex_token stop_token = { .token = { .type = T_STOP } };
return &stop_token;
}
}
return lex_next (lexer, n)->string;
}
+/* Returns the text of the syntax in tokens N0 ahead of the current one,
+ through N1 ahead of the current one, inclusive. (For example, if N0 and N1
+ are both zero, this requests the syntax for the current token.) The caller
+ must not modify or free the returned string. The syntax is encoded in UTF-8
+ and in the original form supplied to the lexer so that, for example, it may
+ include comments, spaces, and new-lines if it spans multiple tokens. */
+struct substring
+lex_next_representation (const struct lexer *lexer, int n0, int n1)
+{
+ return lex_source_get_syntax__ (lex_source__ (lexer), n0, n1);
+}
+
static bool
lex_tokens_match (const struct token *actual, const struct token *expected)
{
int i;
i = 0;
- string_lexer_init (&slex, s, strlen (s), SEG_MODE_INTERACTIVE);
+ string_lexer_init (&slex, s, strlen (s), SEG_MODE_INTERACTIVE, true);
while (string_lexer_next (&slex, &token))
if (token.type != SCAN_SKIP)
{
return src == NULL ? NULL : src->reader->file_name;
}
+/* Returns a newly allocated msg_location for the syntax that represents tokens
+ with 0-based offsets N0...N1, inclusive, from the current token. The caller
+ must eventually free the location (with msg_location_destroy()). */
+struct msg_location *
+lex_get_location (const struct lexer *lexer, int n0, int n1)
+{
+ struct msg_location *loc = lex_get_lines (lexer, n0, n1);
+ loc->first_column = lex_get_first_column (lexer, n0);
+ loc->last_column = lex_get_last_column (lexer, n1);
+ return loc;
+}
+
+/* Returns a newly allocated msg_location for the syntax that represents tokens
+ with 0-based offsets N0...N1, inclusive, from the current token. The
+ location only covers the tokens' lines, not the columns. The caller must
+ eventually free the location (with msg_location_destroy()). */
+struct msg_location *
+lex_get_lines (const struct lexer *lexer, int n0, int n1)
+{
+ struct msg_location *loc = xmalloc (sizeof *loc);
+ *loc = (struct msg_location) {
+ .file_name = xstrdup_if_nonnull (lex_get_file_name (lexer)),
+ .first_line = lex_get_first_line_number (lexer, n0),
+ .last_line = lex_get_last_line_number (lexer, n1),
+ };
+ return loc;
+}
+
const char *
lex_get_encoding (const struct lexer *lexer)
{
src->journal_pos = src->seg_pos = src->line_pos = 0;
src->n_newlines = 0;
src->suppress_next_newline = false;
- segmenter_init (&src->segmenter, segmenter_get_mode (&src->segmenter));
+ src->segmenter = segmenter_init (segmenter_get_mode (&src->segmenter),
+ false);
while (!deque_is_empty (&src->deque))
lex_source_pop__ (src);
lex_source_push_endcmd__ (src);
}
static struct substring
-lex_source_get_syntax__ (const struct lex_source *src, int n0, int n1)
+lex_tokens_get_syntax__ (const struct lex_source *src,
+ const struct lex_token *token0,
+ const struct lex_token *token1)
{
- const struct lex_token *token0 = lex_source_next__ (src, n0);
- const struct lex_token *token1 = lex_source_next__ (src, MAX (n0, n1));
size_t start = token0->token_pos;
size_t end = token1->token_pos + token1->token_len;
return ss_buffer (&src->buffer[start - src->tail], end - start);
}
-static void
-lex_ellipsize__ (struct substring in, char *out, size_t out_size)
+static struct substring
+lex_source_get_syntax__ (const struct lex_source *src, int n0, int n1)
{
- size_t out_maxlen;
- size_t out_len;
- int mblen;
-
- assert (out_size >= 16);
- out_maxlen = out_size - 1;
- if (in.length > out_maxlen - 3)
- out_maxlen -= 3;
-
- for (out_len = 0; out_len < in.length; out_len += mblen)
- {
- if (in.string[out_len] == '\n'
- || in.string[out_len] == '\0'
- || (in.string[out_len] == '\r'
- && out_len + 1 < in.length
- && in.string[out_len + 1] == '\n'))
- break;
-
- mblen = u8_mblen (CHAR_CAST (const uint8_t *, in.string + out_len),
- in.length - out_len);
-
- if (mblen < 0)
- break;
-
- if (out_len + mblen > out_maxlen)
- break;
- }
-
- memcpy (out, in.string, out_len);
- strcpy (&out[out_len], out_len < in.length ? "..." : "");
+ return lex_tokens_get_syntax__ (src,
+ lex_source_next__ (src, n0),
+ lex_source_next__ (src, MAX (n0, n1)));
}
static void
{
char syntax_cstr[64];
- lex_ellipsize__ (syntax, syntax_cstr, sizeof syntax_cstr);
+ str_ellipsize (syntax, syntax_cstr, sizeof syntax_cstr);
ds_put_format (&s, _("Syntax error at `%s'"), syntax_cstr);
}
else
if (ds_last (&s) != '.')
ds_put_byte (&s, '.');
- struct msg m = {
- .category = MSG_C_SYNTAX,
- .severity = MSG_S_ERROR,
- .file_name = src->reader->file_name,
+ struct msg_location *location = xmalloc (sizeof *location);
+ *location = (struct msg_location) {
+ .file_name = xstrdup_if_nonnull (src->reader->file_name),
.first_line = lex_source_get_first_line_number (src, n0),
.last_line = lex_source_get_last_line_number (src, n1),
.first_column = lex_source_get_first_column (src, n0),
.last_column = lex_source_get_last_column (src, n1),
+ };
+ struct msg *m = xmalloc (sizeof *m);
+ *m = (struct msg) {
+ .category = MSG_C_SYNTAX,
+ .severity = MSG_S_ERROR,
+ .location = location,
.text = ds_steal_cstr (&s),
};
- msg_emit (&m);
+ msg_emit (m);
}
-static void PRINTF_FORMAT (2, 3)
-lex_get_error (struct lex_source *src, const char *format, ...)
+static void PRINTF_FORMAT (4, 5)
+lex_source_error (struct lex_source *src, int n0, int n1,
+ const char *format, ...)
{
va_list args;
- int n;
-
va_start (args, format);
+ lex_source_error_valist (src, n0, n1, format, args);
+ va_end (args);
+}
- n = deque_count (&src->deque) - 1;
- lex_source_error_valist (src, n, n, format, args);
+static void
+lex_get_error (struct lex_source *src, const char *s)
+{
+ int n = deque_count (&src->deque) - 1;
+ lex_source_error (src, n, n, "%s", s);
lex_source_pop_front (src);
-
- va_end (args);
}
/* Attempts to append an additional token into SRC's deque, reading more from
break;
case SCAN_BAD_HEX_LENGTH:
- lex_get_error (src, _("String of hex digits has %d characters, which "
- "is not a multiple of 2"),
- (int) token->token.number);
- break;
-
case SCAN_BAD_HEX_DIGIT:
case SCAN_BAD_UNICODE_DIGIT:
- lex_get_error (src, _("`%c' is not a valid hex digit"),
- (int) token->token.number);
- break;
-
case SCAN_BAD_UNICODE_LENGTH:
- lex_get_error (src, _("Unicode string contains %d bytes, which is "
- "not in the valid range of 1 to 8 bytes"),
- (int) token->token.number);
- break;
-
case SCAN_BAD_UNICODE_CODE_POINT:
- lex_get_error (src, _("U+%04X is not a valid Unicode code point"),
- (int) token->token.number);
- break;
-
case SCAN_EXPECTED_QUOTE:
- lex_get_error (src, _("Unterminated string constant"));
- break;
-
case SCAN_EXPECTED_EXPONENT:
- lex_get_error (src, _("Missing exponent following `%s'"),
- token->token.string.string);
- break;
-
case SCAN_UNEXPECTED_CHAR:
- {
- char c_name[16];
- lex_get_error (src, _("Bad character %s in input"),
- uc_name (token->token.number, c_name));
- }
+ char *msg = scan_token_to_error (&token->token);
+ lex_get_error (src, msg);
+ free (msg);
break;
case SCAN_SKIP:
static struct lex_source *
lex_source_create (struct lex_reader *reader)
{
- struct lex_source *src;
-
- src = xzalloc (sizeof *src);
- src->reader = reader;
- segmenter_init (&src->segmenter, reader->syntax);
- src->tokens = deque_init (&src->deque, 4, sizeof *src->tokens);
+ struct lex_source *src = xmalloc (sizeof *src);
+ *src = (struct lex_source) {
+ .reader = reader,
+ .segmenter = segmenter_init (reader->syntax, false),
+ .tokens = deque_init (&src->deque, 4, sizeof *src->tokens),
+ };
lex_source_push_endcmd__ (src);
projects / pspp / blobdiff