yhirose
commented
5 years ago @sapgan, thanks for the report. But I didn't see the problem in my test. Could you give me a minimum CSV file that can reproduce the crash problem?
Closed sapgan closed 5 years ago
yhirose
commented
5 years ago @sapgan, thanks for the report. But I didn't see the problem in my test. Could you give me a minimum CSV file that can reproduce the crash problem?
sapgan
commented
5 years ago the calc3 binary in the example folder crashes, but the other 2 binaries run fine. For the csv example, try the following from the lint folder:
./peglint --ast --opt --source a,b --trace ../../grammar/csv.peg compiler version is gcc 6.1 and distro is centos 7.2
yhirose
commented
5 years ago Thanks for the report. Both the calc3 and peglint examples worked on my Mac. The followings are my results.
Since I don't have linux machine right now, I'll try to setup the virtual box with centos to see if the problem occurs on my machine when I have time.
$ ./calc3 1+2*3
+ EXPRESSION
- TERM/0[NUMBER] (1)
- TERM_OPERATOR (+)
+ TERM
- FACTOR/0[NUMBER] (2)
- FACTOR_OPERATOR (*)
- FACTOR/0[NUMBER] (3)
1+2*3 = 7$ ./peglint --ast --opt --source a,b --trace ../../grammar/csv.peg | pbcopy
pos:lev rule/ope
------- --------
0:0 file
0:1 Sequence
0:2 Option
0:3 Sequence
0:4 header
0:5 Sequence
0:6 name
0:7 field
0:8 PrioritizedChoice
0:9 escaped
0:10 Sequence
0:11 DQUOTE
0:12 LiteralString
0:9 non_escaped
0:10 ZeroOrMore
0:11 TEXTDATA
0:12 Sequence
0:13 NotPredicate
0:14 PrioritizedChoice
0:15 CharacterClass
0:15 NL
0:16 PrioritizedChoice
0:17 Sequence
0:18 CR
0:19 LiteralString
0:17 CR
0:18 LiteralString
0:17 LF
0:18 LiteralString
0:13 AnyCharacter
1:11 TEXTDATA
1:12 Sequence
1:13 NotPredicate
1:14 PrioritizedChoice
1:15 CharacterClass
1:6 ZeroOrMore
1:7 Sequence
1:8 COMMA
1:9 LiteralString
2:8 name
2:9 field
2:10 PrioritizedChoice
2:11 escaped
2:12 Sequence
2:13 DQUOTE
2:14 LiteralString
2:11 non_escaped
2:12 ZeroOrMore
2:13 TEXTDATA
2:14 Sequence
2:15 NotPredicate
2:16 PrioritizedChoice
2:17 CharacterClass
2:17 NL
2:18 PrioritizedChoice
2:19 Sequence
2:20 CR
2:21 LiteralString
2:19 CR
2:20 LiteralString
2:19 LF
2:20 LiteralString
2:15 AnyCharacter
3:4 NL
3:5 PrioritizedChoice
3:6 Sequence
3:7 CR
3:8 LiteralString
3:6 CR
3:7 LiteralString
3:6 LF
3:7 LiteralString
0:2* record
0:3 Sequence
0:4 field
0:5 PrioritizedChoice
0:6 escaped
0:7 Sequence
0:8 DQUOTE
0:9 LiteralString
0:6 non_escaped
0:7 ZeroOrMore
0:8 TEXTDATA
0:9 Sequence
0:10 NotPredicate
0:11 PrioritizedChoice
0:12 CharacterClass
0:12 NL
0:13 PrioritizedChoice
0:14 Sequence
0:15 CR
0:16 LiteralString
0:14 CR
0:15 LiteralString
0:14 LF
0:15 LiteralString
0:10 AnyCharacter
1:8 TEXTDATA
1:9 Sequence
1:10 NotPredicate
1:11 PrioritizedChoice
1:12 CharacterClass
1:4 ZeroOrMore
1:5 Sequence
1:6 COMMA
1:7 LiteralString
2:6 field
2:7 PrioritizedChoice
2:8 escaped
2:9 Sequence
2:10 DQUOTE
2:11 LiteralString
2:8 non_escaped
2:9 ZeroOrMore
2:10 TEXTDATA
2:11 Sequence
2:12 NotPredicate
2:13 PrioritizedChoice
2:14 CharacterClass
2:14 NL
2:15 PrioritizedChoice
2:16 Sequence
2:17 CR
2:18 LiteralString
2:16 CR
2:17 LiteralString
2:16 LF
2:17 LiteralString
2:12 AnyCharacter
3:2 ZeroOrMore
3:2 Option
3:3 NL
3:4 PrioritizedChoice
3:5 Sequence
3:6 CR
3:7 LiteralString
3:5 CR
3:6 LiteralString
3:5 LF
3:6 LiteralString
+ file/0[record]
- field/1[TEXTDATA] (a)
- COMMA (,)
- field/1[TEXTDATA] (b)@sapgan, I just wonder if it happens due to the same reason as #23 and #46.
Could you try to link pthread to see if it fixes the crash? If it fixes the problem, I'll update the 'IMPORTANT NOTE' section on README to include 'CentOS' as well.
sapgan
commented
5 years ago Yes that solved the issue. Thanks
For the csv grammar in the given example it core dumps when forming an AST of it
#
CSV grammar based on RFC 4180 (http://www.ietf.org/rfc/rfc4180.txt)
#
file <- (header NL)? record (NL record) NL? header <- name (COMMA name) record <- field (COMMA field) name <- field field <- escaped / non_escaped escaped <- DQUOTE (TEXTDATA / COMMA / CR / LF / D_DQUOTE) DQUOTE non_escaped <- TEXTDATA* COMMA <- ',' CR <- '\r' DQUOTE <- '"' LF <- '\n' NL <- CR LF / CR / LF TEXTDATA <- !([",] / NL) . D_DQUOTE <- '"' '"'
0 0x000000000043267a in std::_Hashtable<std::string, std::pair<std::string const, peg::Definition>, std::allocator<std::pair<std::string const, peg::Definition> >, std::detail::_Select1st, std::equal_to, std::hash, std:: detail::_Mod_range_hashing, std::detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::detail::_Hashtable_traits<true, false, true> >::_M_begin (this=0x0)
1 0x0000000000429eae in std::_Hashtable<std::string, std::pair<std::string const, peg::Definition>, std::allocator<std::pair<std::string const, peg::Definition> >, std::detail::_Select1st, std::equal_to, std::hash, std:: detail::_Mod_range_hashing, std::detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::detail::_Hashtable_traits<true, false, true> >::begin (this=0x0)
2 0x00000000004216f6 in std::unordered_map<std::string, peg::Definition, std::hash, std::equal_to, std::allocator<std::pair<std::string const, peg::Definition> > >::begin (this=0x0)
3 0x00000000004237e9 in peg::parser::enable_ast<peg::AstBase > (this=0x7ffe1c236ca0) at cpp-peglib/peglib.h:3254
If the (?) operator is not used, it doesn't crash. Any reason to not use the (?) operator