Skip to content

Commit

Permalink
fix: 21944, 21977 in oss-fuzz
Browse files Browse the repository at this point in the history
  • Loading branch information
@kkos
kkos committed May 3, 2020
1 parent efa4e3a commit 8b8b37e
Show file tree
Hide file tree
Showing 8 changed files with 98 additions and 35 deletions.
13 changes: 10 additions & 3 deletions src/big5.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
big5.c - Oniguruma (regular expression library)
**********************************************************************/
/*-
* Copyright (c) 2002-2019 K.Kosako
* Copyright (c) 2002-2020 K.Kosako
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
Expand Down Expand Up @@ -58,8 +58,15 @@ static int
big5_code_to_mbclen(OnigCodePoint code)
{
if ((code & (~0xffff)) != 0) return ONIGERR_INVALID_CODE_POINT_VALUE;
if ((code & 0xff00) != 0) return 2;
if (EncLen_BIG5[(int )(code & 0xff)] == 1) return 1;

if ((code & 0xff00) != 0) {
if (EncLen_BIG5[(int )(code >> 8) & 0xff] == 2)
return 2;
}
else {
if (EncLen_BIG5[(int )(code & 0xff)] == 1)
return 1;
}

return ONIGERR_INVALID_CODE_POINT_VALUE;
}
Expand Down
20 changes: 15 additions & 5 deletions src/euc_jp.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
euc_jp.c - Oniguruma (regular expression library)
**********************************************************************/
/*-
* Copyright (c) 2002-2019 K.Kosako
* Copyright (c) 2002-2020 K.Kosako
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
Expand Down Expand Up @@ -114,10 +114,20 @@ static int
code_to_mbclen(OnigCodePoint code)
{
if (ONIGENC_IS_CODE_ASCII(code)) return 1;
else if ((code & 0xff0000) != 0) return 3;
else if ((code & 0xff00) != 0) return 2;
else
return ONIGERR_INVALID_CODE_POINT_VALUE;
else if ((code & 0xff0000) != 0) {
if (EncLen_EUCJP[(int )(code >> 16) & 0xff] == 3)
return 3;
}
else if ((code & 0xff00) != 0) {
if (EncLen_EUCJP[(int )(code >> 8) & 0xff] == 2)
return 2;
}
else if (code < 256) {
if (EncLen_EUCJP[(int )(code & 0xff)] == 1)
return 1;
}

return ONIGERR_INVALID_CODE_POINT_VALUE;
}

static int
Expand Down
13 changes: 10 additions & 3 deletions src/euc_kr.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
euc_kr.c - Oniguruma (regular expression library)
**********************************************************************/
/*-
* Copyright (c) 2002-2019 K.Kosako
* Copyright (c) 2002-2020 K.Kosako
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
Expand Down Expand Up @@ -58,8 +58,15 @@ static int
euckr_code_to_mbclen(OnigCodePoint code)
{
if ((code & (~0xffff)) != 0) return ONIGERR_INVALID_CODE_POINT_VALUE;
if ((code & 0xff00) != 0) return 2;
if (EncLen_EUCKR[(int )(code & 0xff)] == 1) return 1;

if ((code & 0xff00) != 0) {
if (EncLen_EUCKR[(int )(code >> 8) & 0xff] == 2)
return 2;
}
else {
if (EncLen_EUCKR[(int )(code & 0xff)] == 1)
return 1;
}

return ONIGERR_INVALID_CODE_POINT_VALUE;
}
Expand Down
19 changes: 13 additions & 6 deletions src/euc_tw.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
euc_tw.c - Oniguruma (regular expression library)
**********************************************************************/
/*-
* Copyright (c) 2002-2019 K.Kosako
* Copyright (c) 2002-2020 K.Kosako
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
Expand Down Expand Up @@ -57,15 +57,22 @@ euctw_mbc_enc_len(const UChar* p)
static int
euctw_code_to_mbclen(OnigCodePoint code)
{
if ((code & 0xff000000) != 0) return 4;
else if ((code & 0xff0000) != 0) return ONIGERR_INVALID_CODE_POINT_VALUE;
else if ((code & 0xff00) != 0) return 2;
if ((code & 0xff000000) != 0) {
if (EncLen_EUCTW[(int )(code >> 24) & 0xff] == 4)
return 4;
}
else if ((code & 0xff0000) != 0)
return ONIGERR_INVALID_CODE_POINT_VALUE;
else if ((code & 0xff00) != 0) {
if (EncLen_EUCTW[(int )(code >> 8) & 0xff] == 2)
return 2;
}
else {
if (EncLen_EUCTW[(int )(code & 0xff)] == 1)
return 1;

return ONIGERR_INVALID_CODE_POINT_VALUE;
}

return ONIGERR_INVALID_CODE_POINT_VALUE;
}

static int
Expand Down
24 changes: 17 additions & 7 deletions src/gb18030.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -89,15 +89,25 @@ gb18030_mbc_enc_len(const UChar* p)
static int
gb18030_code_to_mbclen(OnigCodePoint code)
{
if ((code & 0xff000000) != 0) return 4;
else if ((code & 0xff0000) != 0) return ONIGERR_INVALID_CODE_POINT_VALUE;
else if ((code & 0xff00) != 0) return 2;
if ((code & 0xff000000) != 0) {
if (GB18030_MAP[(int )(code >> 24) & 0xff] == CM)
if (GB18030_MAP[(int )(code >> 16) & 0xff] == C4)
return 4;
}
else if ((code & 0xff0000) != 0) return ONIGERR_INVALID_CODE_POINT_VALUE;
else if ((code & 0xff00) != 0) {
if (GB18030_MAP[(int )(code >> 8) & 0xff] == CM) {
char c = GB18030_MAP[(int )code & 0xff];
if (c == CM || c == C2)
return 2;
}
}
else {
if (GB18030_MAP[(int )(code & 0xff)] == CM)
return ONIGERR_INVALID_CODE_POINT_VALUE;

return 1;
if (GB18030_MAP[(int )(code & 0xff)] != CM)
return 1;
}

return ONIGERR_INVALID_CODE_POINT_VALUE;
}

static int
Expand Down
4 changes: 4 additions & 0 deletions src/regexec.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2362,6 +2362,10 @@ static int string_cmp_ic(OnigEncoding enc, int case_fold_flag,
p1++;
p2++;
}
if (s2 >= end2) {
if (s1 < end1) return 0;
else break;
}
}

*ps2 = s2;
Expand Down
26 changes: 21 additions & 5 deletions src/regparse.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4564,7 +4564,7 @@ fetch_interval(UChar** src, UChar* end, PToken* tok, ScanEnv* env)

/* \M-, \C-, \c, or \... */
static int
fetch_escaped_value(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
fetch_escaped_value_raw(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
{
int v;
OnigCodePoint c;
Expand All @@ -4583,7 +4583,7 @@ fetch_escaped_value(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
if (PEND) return ONIGERR_END_PATTERN_AT_META;
PFETCH_S(c);
if (c == MC_ESC(env->syntax)) {
v = fetch_escaped_value(&p, end, env, &c);
v = fetch_escaped_value_raw(&p, end, env, &c);
if (v < 0) return v;
}
c = ((c & 0xff) | 0x80);
Expand Down Expand Up @@ -4612,7 +4612,7 @@ fetch_escaped_value(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
}
else {
if (c == MC_ESC(env->syntax)) {
v = fetch_escaped_value(&p, end, env, &c);
v = fetch_escaped_value_raw(&p, end, env, &c);
if (v < 0) return v;
}
c &= 0x9f;
Expand All @@ -4634,6 +4634,21 @@ fetch_escaped_value(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
return 0;
}

static int
fetch_escaped_value(UChar** src, UChar* end, ScanEnv* env, OnigCodePoint* val)
{
int r;
int len;

r = fetch_escaped_value_raw(src, end, env, val);
if (r != 0) return r;

len = ONIGENC_CODE_TO_MBCLEN(env->enc, *val);
if (len < 0) return len;

return 0;
}

static int fetch_token(PToken* tok, UChar** src, UChar* end, ScanEnv* env);

static OnigCodePoint
Expand Down Expand Up @@ -5192,7 +5207,7 @@ fetch_token_cc(PToken* tok, UChar** src, UChar* end, ScanEnv* env, int state)
else {
int curr_state;

curr_state = (state == CS_RANGE) ? CPS_EMPTY : CPS_START;
curr_state = (state == CS_RANGE) ? CPS_EMPTY : CPS_START;
r = check_code_point_sequence_cc(p, end, tok->base_num, enc,
curr_state);
if (r < 0) return r;
Expand Down Expand Up @@ -8419,8 +8434,9 @@ parse_exp(Node** np, PToken* tok, int term, UChar** src, UChar* end,
case TK_CODE_POINT:
{
UChar buf[ONIGENC_CODE_TO_MBC_MAXLEN];
len = ONIGENC_CODE_TO_MBC(env->enc, tok->u.code, buf);
len = ONIGENC_CODE_TO_MBCLEN(env->enc, tok->u.code);
if (len < 0) return len;
len = ONIGENC_CODE_TO_MBC(env->enc, tok->u.code, buf);
#ifdef NUMBERED_CHAR_IS_NOT_CASE_AMBIG
*np = node_new_str_crude(buf, buf + len, env->options);
#else
Expand Down
14 changes: 8 additions & 6 deletions src/sjis.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
sjis.c - Oniguruma (regular expression library)
**********************************************************************/
/*-
* Copyright (c) 2002-2019 K.Kosako
* Copyright (c) 2002-2020 K.Kosako
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
Expand Down Expand Up @@ -113,13 +113,15 @@ static int
code_to_mbclen(OnigCodePoint code)
{
if (code < 256) {
return EncLen_SJIS[(int )code] == 1;
if (EncLen_SJIS[(int )code] == 1)
return 1;
}
else if (code <= 0xffff) {
return 2;
else if (code < 0x10000) {
if (EncLen_SJIS[(int )(code >> 8) & 0xff] == 2)
return 2;
}
else
return ONIGERR_INVALID_CODE_POINT_VALUE;

return ONIGERR_INVALID_CODE_POINT_VALUE;
}

static OnigCodePoint
Expand Down

0 comments on commit 8b8b37e

Please sign in to comment.