Hi, I'm posting here to let you know that I've pushed the following patches to the 3.1 repository: a4d7533d9c826dcf405d94aa6644587feacf9e96 [patterndb] fixed parser node first char match handling 9e7a729f6b450c7ef5bf0b22943771f040fd110e [patterndb] Consider the value of the parameter when checking if two parser nodes are equal 73fe61c0dccd0453cb18d41e3bdca54a2525fba1 [patterndb] use strcmp instead of prefix match when looking up parser name 8c9e4099477866829b9de8dbdbcca97806856f00 [patterndb] added FLOAT radix parser ae1b7fd8f386c141056247264ccd06da78614136 On error r_insert_node displays more information about the node that caused the problem in logpatterns.c for rules, the rule_id is di 6be05a55fcfd55b4efa106e5c507ad3fad9cbdbc [patterndb] adapted db_parser configuration loader to new syntax with rulesets b0869709f98566d39b5b3d8bd793cdf12ccc1996 [patterndb] db_parser NUMBER parser now handles hexadecimal numbers in 0xXXXXX form 60dc446418bcb9e41436d5ab287f66a604172a6f [unit tests] white space cleanup 32e346803c04c8454ae7f9313ebf5f9fdac83553 [patterndb] added IPv6 and IPvANY parsers to db_parser a2b156f67a4cb7c5ee9438864c6d93870e668ac1 fixed a memory leak around parser match storing, check for parser node name length 53c07dc8bba32dab918d6ab782e0e1dfc6b210ff fixed a memory leak in db-parser() dynamic value handling 1c31f647245df4e6f8773d49c04ca32deb9d9bfc added testcase for real log radix parsing This contains a couple of patterndb bugfixes, some new parsers and the support for the v2 patterndb format. This code currently only supports v2, I'm going to publish a script that converts v1 to v2 automatically. Otherwise this also means that from now it is possible to use the syslog-ng Store Box patterns compiled by BalaBit directly from syslog-ng OSE: http://www.balabit.com/downloads/files/patterndb/1.0-20081117/patterndb/ If anyone asks, I'm willing to create an "alpha" release from the current state of affairs. The alpha sign only means that we still haven't reached all goals for 3.1 (as per the syslog-ng roadmap published on the website). I don't expect to have stability problems with the above changes, those are mostly trivial and have associated automatic testcases. -- Bazsi