Skip to content

M
mruby
Commit 11b0dda6 authored by Yukihiro "Matz" Matsumoto's avatar Yukihiro "Matz" Matsumoto
Browse files
Options

Merge pull request #1509 from FUKUZAWA-Tadashi/heredoc-bugfix 

fix bugs on Heredocument
parents 91940477 8d34e001
Hide whitespace changes
Inline Side-by-side
Showing with 170 additions and 36 deletions
include/mruby/compile.h
View file @ 11b0dda6
......@@ -128,9 +128,10 @@ struct mrb_parser_state {
char buf[MRB_PARSER_BUF_SIZE];
int bidx;
mrb_ast_node *heredocs; /* list of mrb_parser_heredoc_info* */
mrb_ast_node *all_heredocs; /* list of mrb_parser_heredoc_info* */
mrb_ast_node *heredocs_from_nextline;
mrb_ast_node *parsing_heredoc;
mrb_bool heredoc_starts_nextline:1;
mrb_ast_node *lex_strterm_before_heredoc;
mrb_bool heredoc_end_now:1; /* for mirb */
void *ylval;
......
src/parse.y
View file @ 11b0dda6
......@@ -915,6 +915,48 @@ parsing_heredoc_inf(parser_state *p)
return (parser_heredoc_info*)nd->car->cdr;
}
static void
heredoc_treat_nextline(parser_state *p)
{
if (p->heredocs_from_nextline == NULL)
return;
if (p->parsing_heredoc == NULL) {
node *n;
p->parsing_heredoc = p->heredocs_from_nextline;
p->lex_strterm_before_heredoc = p->lex_strterm;
p->lex_strterm = new_strterm(p, parsing_heredoc_inf(p)->type, 0, 0);
n = p->all_heredocs;
if (n) {
while (n->cdr)
n = n->cdr;
n->cdr = p->parsing_heredoc;
} else {
p->all_heredocs = p->parsing_heredoc;
}
} else {
node *n, *m;
m = p->heredocs_from_nextline;
while (m->cdr)
m = m->cdr;
n = p->all_heredocs;
mrb_assert(n != NULL);
if (n == p->parsing_heredoc) {
m->cdr = n;
p->all_heredocs = p->heredocs_from_nextline;
p->parsing_heredoc = p->heredocs_from_nextline;
} else {
while (n->cdr != p->parsing_heredoc) {
n = n->cdr;
mrb_assert(n != NULL);
}
m->cdr = n->cdr;
n->cdr = p->heredocs_from_nextline;
p->parsing_heredoc = p->heredocs_from_nextline;
}
}
p->heredocs_from_nextline = NULL;
}
static void
heredoc_end(parser_state *p)
{
......@@ -923,6 +965,8 @@ heredoc_end(parser_state *p)
p->lstate = EXPR_BEG;
p->cmd_start = TRUE;
end_strterm(p);
p->lex_strterm = p->lex_strterm_before_heredoc;
p->lex_strterm_before_heredoc = NULL;
p->heredoc_end_now = TRUE;
} else {
/* next heredoc */
......@@ -1054,7 +1098,8 @@ heredoc_end(parser_state *p)
%token tSYMBEG tREGEXP_BEG tWORDS_BEG tSYMBOLS_BEG
%token tSTRING_BEG tXSTRING_BEG tSTRING_DVAR tLAMBEG
%token <nd> tHEREDOC_BEG /* <<, <<- */
%token tHEREDOC_END tLITERAL_DELIM
%token tHEREDOC_END tLITERAL_DELIM tHD_LITERAL_DELIM
%token <nd> tHD_STRING_PART tHD_STRING_MID
/*
* precedence table
......@@ -1938,6 +1983,14 @@ args : arg_value
{
$$ = push($1, new_splat(p, $4));
}
| args ',' heredoc_bodies arg_value
{
$$ = push($1, $4);
}
| args ',' heredoc_bodies tSTAR arg_value
{
$$ = push($1, new_splat(p, $5));
}
;
mrhs : args ',' arg_value
......@@ -2604,6 +2657,10 @@ string_interp : tSTRING_MID
{
$$ = list1(new_literal_delim(p));
}
| tHD_LITERAL_DELIM heredoc_bodies
{
$$ = list1(new_literal_delim(p));
}
;
xstring : tXSTRING_BEG tXSTRING
......@@ -2629,7 +2686,7 @@ regexp : tREGEXP_BEG tREGEXP
heredoc : tHEREDOC_BEG
;
opt_heredoc_bodies : none
opt_heredoc_bodies : /* none */
| heredoc_bodies
;
......@@ -2639,16 +2696,40 @@ heredoc_bodies : heredoc_body
heredoc_body : tHEREDOC_END
{
parsing_heredoc_inf(p)->doc = list1(new_str(p, "", 0));
parser_heredoc_info * inf = parsing_heredoc_inf(p);
inf->doc = push(inf->doc, new_str(p, "", 0));
heredoc_end(p);
}
| string_rep tHEREDOC_END
| heredoc_string_rep tHEREDOC_END
{
parsing_heredoc_inf(p)->doc = $1;
heredoc_end(p);
}
;
heredoc_string_rep : heredoc_string_interp
| heredoc_string_rep heredoc_string_interp
;
heredoc_string_interp : tHD_STRING_MID
{
parser_heredoc_info * inf = parsing_heredoc_inf(p);
inf->doc = push(inf->doc, $1);
heredoc_treat_nextline(p);
}
| tHD_STRING_PART
{
$<nd>$ = p->lex_strterm;
p->lex_strterm = NULL;
}
compstmt
'}'
{
parser_heredoc_info * inf = parsing_heredoc_inf(p);
p->lex_strterm = $<nd>2;
inf->doc = push(push(inf->doc, $1), $3);
}
;
words : tWORDS_BEG tSTRING
{
$$ = new_words(p, list1($2));
......@@ -3477,12 +3558,12 @@ read_escape(parser_state *p)
buf[0] = c;
for (i=1; i<3; i++) {
buf[i] = nextc(p);
if (buf[i] == -1) goto eof;
if (buf[i] < '0' || '7' < buf[i]) {
pushback(p, buf[i]);
break;
}
buf[i] = nextc(p);
if (buf[i] == -1) goto eof;
if (buf[i] < '0' || '7' < buf[i]) {
pushback(p, buf[i]);
break;
}
}
c = scan_oct(buf, i, &i);
}
......@@ -3591,12 +3672,12 @@ parse_string(parser_state *p)
}
if (c == -1) {
char buf[256];
snprintf(buf, sizeof(buf), "can't find string \"%s\" anywhere before EOF", hinf->term);
snprintf(buf, sizeof(buf), "can't find heredoc delimiter \"%s\" anywhere before EOF", hinf->term);
yyerror(p, buf);
return 0;
}
yylval.nd = new_str(p, tok(p), toklen(p));
return tSTRING_MID;
return tHD_STRING_MID;
}
if (c == -1) {
yyerror(p, "unterminated string meets end of file");
......@@ -3660,8 +3741,10 @@ parse_string(parser_state *p)
p->lstate = EXPR_BEG;
p->cmd_start = TRUE;
yylval.nd = new_str(p, tok(p), toklen(p));
if (hinf)
if (hinf) {
hinf->line_head = FALSE;
return tHD_STRING_PART;
}
return tSTRING_PART;
}
tokadd(p, '#');
......@@ -3674,6 +3757,10 @@ parse_string(parser_state *p)
if (c == '\n') {
p->lineno++;
p->column = 0;
heredoc_treat_nextline(p);
if (p->parsing_heredoc != NULL) {
return tHD_LITERAL_DELIM;
}
}
} while (ISSPACE(c = nextc(p)));
pushback(p, c);
......@@ -3801,14 +3888,7 @@ heredoc_identifier(parser_state *p)
info->allow_indent = indent;
info->line_head = TRUE;
info->doc = NULL;
p->heredocs = push(p->heredocs, newnode);
if (p->parsing_heredoc == NULL) {
node *n = p->heredocs;
while (n->cdr)
n = n->cdr;
p->parsing_heredoc = n;
}
p->heredoc_starts_nextline = TRUE;
p->heredocs_from_nextline = push(p->heredocs_from_nextline, newnode);
p->lstate = EXPR_END;
yylval.nd = newnode;
......@@ -3835,7 +3915,7 @@ parser_yylex(parser_state *p)
if (p->lex_strterm) {
if (is_strterm_type(p, STR_FUNC_HEREDOC)) {
if ((p->parsing_heredoc != NULL) && (! p->heredoc_starts_nextline))
if (p->parsing_heredoc != NULL)
return parse_string(p);
}
else
......@@ -3862,11 +3942,7 @@ parser_yylex(parser_state *p)
skip(p, '\n');
/* fall through */
case '\n':
p->heredoc_starts_nextline = FALSE;
if (p->parsing_heredoc != NULL) {
p->lex_strterm = new_strterm(p, parsing_heredoc_inf(p)->type, 0, 0);
goto normal_newline;
}
heredoc_treat_nextline(p);
switch (p->lstate) {
case EXPR_BEG:
case EXPR_FNAME:
......@@ -3875,10 +3951,16 @@ parser_yylex(parser_state *p)
case EXPR_VALUE:
p->lineno++;
p->column = 0;
if (p->parsing_heredoc != NULL) {
return parse_string(p);
}
goto retry;
default:
break;
}
if (p->parsing_heredoc != NULL) {
return '\n';
}
while ((c = nextc(p))) {
switch (c) {
case ' ': case '\t': case '\f': case '\r':
......@@ -4061,9 +4143,9 @@ parser_yylex(parser_state *p)
}
if (p->lstate == EXPR_DOT) {
if (cmd_state)
p->lstate = EXPR_CMDARG;
p->lstate = EXPR_CMDARG;
else
p->lstate = EXPR_ARG;
p->lstate = EXPR_ARG;
return '`';
}
p->lex_strterm = new_strterm(p, str_xquote, '`', 0);
......@@ -4762,9 +4844,9 @@ parser_yylex(parser_state *p)
case '_': /* $_: last read line string */
c = nextc(p);
if (c != -1 && identchar(c)) { /* if there is more after _ it is a variable */
tokadd(p, '$');
tokadd(p, c);
break;
tokadd(p, '$');
tokadd(p, c);
break;
}
pushback(p, c);
c = '_';
......@@ -5134,7 +5216,8 @@ mrb_parser_new(mrb_state *mrb)
#endif
p->lex_strterm = NULL;
p->heredocs = p->parsing_heredoc = NULL;
p->all_heredocs = p->parsing_heredoc = NULL;
p->lex_strterm_before_heredoc = NULL;
p->current_filename_index = -1;
p->filename_table = NULL;
......
test/t/literals.rb
View file @ 11b0dda6
......@@ -138,6 +138,46 @@ FFF
123
KKK
m = [<<MM1, <<MM2]
x#{m2 = {x:<<MM3}}y
mm3
MM3
mm1
MM1
mm2
MM2
n = [1, "#{<<NN1}", 3,
nn1
NN1
4]
qqq = Proc.new {|*x| x.join(' $ ')}
q1 = qqq.call("a", <<QQ1, "c",
q
QQ1
"d")
q2 = qqq.call("l", "m#{<<QQ2}n",
qq
QQ2
"o")
w = %W( 1 #{<<WWW} 3
www
WWW
4 5 )
x = [1, <<XXX1,
foo #{<<XXX2} bar
222 #{<<XXX3} 444
333
XXX3
5
XXX2
6
XXX1
9]
z = <<'ZZZ'
ZZZ
......@@ -152,9 +192,18 @@ ZZZ
assert_equal " iii\n", i
assert_equal [" j1j\n", " j2j\n", " j\#{3}j\n"], j
assert_equal 123, k
assert_equal ["x{:x=>\"mm3\\n\"}y\nmm1\n", "mm2\n"], m
assert_equal ({:x=>"mm3\n"}), m2
assert_equal [1, "nn1\n", 3, 4], n
assert_equal "a $ q\n $ c $ d", q1
assert_equal "l $ mqq\nn $ o", q2
assert_equal ["1", "www\n", "3", "4", "5"], w
assert_equal [1, "foo 222 333\n 444\n5\n bar\n6\n", 9], x
assert_equal "", z
end
assert('Literals Array', '8.7.6.4') do
a = %W{abc#{1+2}def \}g}
b = %W(abc #{2+3} def \(g)
......@@ -209,6 +258,7 @@ d
assert_equal ["a\\nb", "test abc", "c\nd", "x\\y", "x\\y", "x\\\\y"], h
end
assert('Literals Array of symbols') do
a = %I{abc#{1+2}def \}g}
b = %I(abc #{2+3} def \(g)
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment