static enum rules_token
lex_include_string(struct scanner *s, struct xkb_compose_table *table,
union lvalue *val_out)
{
while (is_space(peek(s)))
if (next(s) == '\n')
return TOK_END_OF_LINE;
s->token_line = s->line;
s->token_column = s->column;
s->buf_pos = 0;
if (!chr(s, '\"')) {
scanner_err(s, "include statement must be followed by a path");
return TOK_ERROR;
}
while (!eof(s) && !eol(s) && peek(s) != '\"') {
if (chr(s, '%')) {
if (chr(s, '%')) {
buf_append(s, '%');
}
else if (chr(s, 'H')) {
const char *home = secure_getenv("HOME");
if (!home) {
scanner_err(s, "%%H was used in an include statement, but the HOME environment variable is not set");
return TOK_ERROR;
}
if (!buf_appends(s, home)) {
scanner_err(s, "include path after expanding %%H is too long");
return TOK_ERROR;
}
}
else if (chr(s, 'L')) {
char *path = get_locale_compose_file_path(table->locale);
if (!path) {
scanner_err(s, "failed to expand %%L to the locale Compose file");
return TOK_ERROR;
}
if (!buf_appends(s, path)) {
free(path);
scanner_err(s, "include path after expanding %%L is too long");
return TOK_ERROR;
}
free(path);
}
else if (chr(s, 'S')) {
const char *xlocaledir = get_xlocaledir_path();
if (!buf_appends(s, xlocaledir)) {
scanner_err(s, "include path after expanding %%S is too long");
return TOK_ERROR;
}
}
else {
scanner_err(s, "unknown %% format (%c) in include statement", peek(s));
return TOK_ERROR;
}
} else {
buf_append(s, next(s));
}
}
if (!chr(s, '\"')) {
scanner_err(s, "unterminated include statement");
return TOK_ERROR;
}
if (!buf_append(s, '\0')) {
scanner_err(s, "include path is too long");
return TOK_ERROR;
}
val_out->string.str = s->buf;
val_out->string.len = s->buf_pos;
return TOK_INCLUDE_STRING;
}
static enum rules_token
lex(struct scanner *s, union lvalue *val)
{
skip_more_whitespace_and_comments:
/* Skip spaces. */
while (chr(s, ' ') || chr(s, '\t'));
/* Skip comments. */
if (lit(s, "//")) {
while (!eof(s) && !eol(s)) next(s);
}
/* New line. */
if (eol(s)) {
while (eol(s)) next(s);
return TOK_END_OF_LINE;
}
/* Escaped line continuation. */
if (chr(s, '\\')) {
if (!eol(s)) {
scanner_err(s, "illegal new line escape; must appear at end of line");
return TOK_ERROR;
}
next(s);
goto skip_more_whitespace_and_comments;
}
/* See if we're done. */
if (eof(s)) return TOK_END_OF_FILE;
/* New token. */
s->token_line = s->line;
s->token_column = s->column;
/* Operators and punctuation. */
if (chr(s, '!')) return TOK_BANG;
if (chr(s, '=')) return TOK_EQUALS;
if (chr(s, '*')) return TOK_STAR;
/* Group name. */
if (chr(s, '$')) {
val->string.start = s->s + s->pos;
val->string.len = 0;
while (is_ident(peek(s))) {
next(s);
val->string.len++;
}
if (val->string.len == 0) {
scanner_err(s, "unexpected character after \'$\'; expected name");
return TOK_ERROR;
}
return TOK_GROUP_NAME;
}
/* Identifier. */
if (is_ident(peek(s))) {
val->string.start = s->s + s->pos;
val->string.len = 0;
while (is_ident(peek(s))) {
next(s);
val->string.len++;
}
return TOK_IDENTIFIER;
}
scanner_err(s, "unrecognized token");
return TOK_ERROR;
}
static enum rules_token
lex(struct scanner *s, union lvalue *val)
{
skip_more_whitespace_and_comments:
/* Skip spaces. */
while (is_space(peek(s)))
if (next(s) == '\n')
return TOK_END_OF_LINE;
/* Skip comments. */
if (chr(s, '#')) {
skip_to_eol(s);
goto skip_more_whitespace_and_comments;
}
/* See if we're done. */
if (eof(s)) return TOK_END_OF_FILE;
/* New token. */
s->token_line = s->line;
s->token_column = s->column;
s->buf_pos = 0;
/* LHS Keysym. */
if (chr(s, '<')) {
while (peek(s) != '>' && !eol(s))
buf_append(s, next(s));
if (!chr(s, '>')) {
scanner_err(s, "unterminated keysym literal");
return TOK_ERROR;
}
if (!buf_append(s, '\0')) {
scanner_err(s, "keysym literal is too long");
return TOK_ERROR;
}
val->string.str = s->buf;
val->string.len = s->buf_pos;
return TOK_LHS_KEYSYM;
}
/* Colon. */
if (chr(s, ':'))
return TOK_COLON;
if (chr(s, '!'))
return TOK_BANG;
if (chr(s, '~'))
return TOK_TILDE;
/* String literal. */
if (chr(s, '\"')) {
while (!eof(s) && !eol(s) && peek(s) != '\"') {
if (chr(s, '\\')) {
uint8_t o;
if (chr(s, '\\')) {
buf_append(s, '\\');
}
else if (chr(s, '"')) {
buf_append(s, '"');
}
else if (chr(s, 'x') || chr(s, 'X')) {
if (hex(s, &o))
buf_append(s, (char) o);
else
scanner_warn(s, "illegal hexadecimal escape sequence in string literal");
}
else if (oct(s, &o)) {
buf_append(s, (char) o);
}
else {
scanner_warn(s, "unknown escape sequence (%c) in string literal", peek(s));
/* Ignore. */
}
} else {
buf_append(s, next(s));
}
}
if (!chr(s, '\"')) {
scanner_err(s, "unterminated string literal");
return TOK_ERROR;
}
if (!buf_append(s, '\0')) {
scanner_err(s, "string literal is too long");
return TOK_ERROR;
}
if (!is_valid_utf8(s->buf, s->buf_pos - 1)) {
scanner_err(s, "string literal is not a valid UTF-8 string");
return TOK_ERROR;
}
val->string.str = s->buf;
val->string.len = s->buf_pos;
return TOK_STRING;
}
/* Identifier or include. */
if (is_alpha(peek(s)) || peek(s) == '_') {
s->buf_pos = 0;
while (is_alnum(peek(s)) || peek(s) == '_')
buf_append(s, next(s));
if (!buf_append(s, '\0')) {
scanner_err(s, "identifier is too long");
return TOK_ERROR;
}
if (streq(s->buf, "include"))
return TOK_INCLUDE;
val->string.str = s->buf;
val->string.len = s->buf_pos;
return TOK_IDENT;
}
/* Discard rest of line. */
skip_to_eol(s);
scanner_err(s, "unrecognized token");
return TOK_ERROR;
}
int
_xkbcommon_lex(YYSTYPE *yylval, struct scanner *s)
{
int tok;
skip_more_whitespace_and_comments:
/* Skip spaces. */
while (is_space(peek(s))) next(s);
/* Skip comments. */
if (lit(s, "//") || chr(s, '#')) {
while (!eof(s) && !eol(s)) next(s);
goto skip_more_whitespace_and_comments;
}
/* See if we're done. */
if (eof(s)) return END_OF_FILE;
/* New token. */
s->token_line = s->line;
s->token_column = s->column;
s->buf_pos = 0;
/* String literal. */
if (chr(s, '\"')) {
while (!eof(s) && !eol(s) && peek(s) != '\"') {
if (chr(s, '\\')) {
uint8_t o;
if (chr(s, '\\')) buf_append(s, '\\');
else if (chr(s, 'n')) buf_append(s, '\n');
else if (chr(s, 't')) buf_append(s, '\t');
else if (chr(s, 'r')) buf_append(s, '\r');
else if (chr(s, 'b')) buf_append(s, '\b');
else if (chr(s, 'f')) buf_append(s, '\f');
else if (chr(s, 'v')) buf_append(s, '\v');
else if (chr(s, 'e')) buf_append(s, '\033');
else if (oct(s, &o)) buf_append(s, (char) o);
else {
scanner_warn(s, "unknown escape sequence in string literal");
/* Ignore. */
}
} else {
buf_append(s, next(s));
}
}
if (!buf_append(s, '\0') || !chr(s, '\"')) {
scanner_err(s, "unterminated string literal");
return ERROR_TOK;
}
yylval->str = strdup(s->buf);
if (!yylval->str)
return ERROR_TOK;
return STRING;
}
/* Key name literal. */
if (chr(s, '<')) {
while (is_graph(peek(s)) && peek(s) != '>')
buf_append(s, next(s));
if (!buf_append(s, '\0') || !chr(s, '>')) {
scanner_err(s, "unterminated key name literal");
return ERROR_TOK;
}
/* Empty key name literals are allowed. */
yylval->sval = xkb_atom_intern(s->ctx, s->buf, s->buf_pos - 1);
return KEYNAME;
}
/* Operators and punctuation. */
if (chr(s, ';')) return SEMI;
if (chr(s, '{')) return OBRACE;
if (chr(s, '}')) return CBRACE;
if (chr(s, '=')) return EQUALS;
if (chr(s, '[')) return OBRACKET;
if (chr(s, ']')) return CBRACKET;
if (chr(s, '(')) return OPAREN;
if (chr(s, ')')) return CPAREN;
if (chr(s, '.')) return DOT;
if (chr(s, ',')) return COMMA;
if (chr(s, '+')) return PLUS;
if (chr(s, '-')) return MINUS;
if (chr(s, '*')) return TIMES;
if (chr(s, '/')) return DIVIDE;
if (chr(s, '!')) return EXCLAM;
if (chr(s, '~')) return INVERT;
/* Identifier. */
if (is_alpha(peek(s)) || peek(s) == '_') {
s->buf_pos = 0;
while (is_alnum(peek(s)) || peek(s) == '_')
buf_append(s, next(s));
if (!buf_append(s, '\0')) {
scanner_err(s, "identifier too long");
return ERROR_TOK;
}
/* Keyword. */
tok = keyword_to_token(s->buf, s->buf_pos - 1);
if (tok != -1) return tok;
yylval->str = strdup(s->buf);
if (!yylval->str)
return ERROR_TOK;
return IDENT;
}
/* Number literal (hexadecimal / decimal / float). */
if (number(s, &yylval->num, &tok)) {
if (tok == ERROR_TOK) {
scanner_err(s, "malformed number literal");
return ERROR_TOK;
}
return tok;
}
scanner_err(s, "unrecognized token");
return ERROR_TOK;
}
