On 10/02/2015 08:22, Guido Wimmel wrote:
Hi,
I noticed an inefficiency in the code for loading content.xml:
before every insert, the whole table content seems to be read just to
get the metadata - see ContentLoaderHandler.getParameters():
57 Map<String, Integer> colTypes = jdbcTemplate.query("SELECT * FROM "
+ tableName,
58 new ResultSetExtractor<Map<String, Integer>>() {
...
This usually shouldn't be a big problem, but could be awkward if one e.g. wants
to use this
mechanism to insert large amounts of test data, or if it is used during migration
from 1.1 -> 1.2
to insert records into tables which are already filled.
I suggest adding "WHERE 0=1". I tried this and it does not affect the build and
integration test.
If it's ok, I can make the corresponding change - but I'd need to know if I
should open a JIRA,
and if there's anything special to consider when merging from 1.2 because of
the code reorganization
(SYNCOPE-620).
WDYT?
Cool stuff, I like optimizations! :-)
Please open an issue on JIRA with fix-for-version 1.2.3 (1_2_X) & 1.3.0
(master): I'll take care of merging into the 2_0_X feature branch.
Regards.
--
Francesco Chicchiriccò
Tirasa - Open Source Excellence
http://www.tirasa.net/
Involved at The Apache Software Foundation:
member, Syncope PMC chair, Cocoon PMC, Olingo PMC
http://people.apache.org/~ilgrosso/
