Hi.Definitely i'll do this functionality, but not as a filter - as a part of HTable. Little problem with submission, I'm waiting to approval from the company (that I'm working under its policy) to permit me to contribute my work to community. As soon as this approval will be ready, i'll submit.
Best Regards. 2009/2/17 stack <[email protected]> > Yeah, it certainly looks that way Slava. Want to try writing one? Make a > patch and we'll commit it? > St.Ack > > > On Tue, Feb 17, 2009 at 7:28 AM, Slava Gorelik <[email protected] > >wrote: > > > Hi.I tried to understand this filter and from method > > > > public boolean filterColumn(final byte [] rowKey, final byte [] colKey, > > final byte[] data) { > > if (filterRowKey(rowKey)) { > > return true; > > } > > if (filtersByColumnValue()) { > > byte[] filterValue = equalsMap.get(colKey); > > if (null != filterValue) { > > return !Arrays.equals(filterValue, data); > > } > > } > > if (nullColumns.contains(colKey)) { > > if (data != null && !HLogEdit.isDeleted(data)) { > > return true; > > } > > } > > return false; > > } > > > > I see that filtering is done on column value and not column name and as i > > see it from code the comparison not even done > > by RegEx (the regex is applied only on rowkey). > > > > Correct me if I wrong. > > > > Thank You and Best Regards. > > Slava. > > > > > > On Tue, Feb 17, 2009 at 4:33 PM, Slava Gorelik <[email protected] > > >wrote: > > > > > Thank You !I'll look on this closer. > > > > > > > > > On Tue, Feb 17, 2009 at 4:15 PM, Rasit OZDAS <[email protected]> > > wrote: > > > > > >> In that API documentation, filterColumn function has the definition: > > >> "Filters on row key, column name, and column value." > > >> > > >> > > > http://hadoop.apache.org/hbase/docs/current/api/org/apache/hadoop/hbase/filter/RegExpRowFilter.html#filterColumn(byte[],%20byte[],%20byte[]) > < > http://hadoop.apache.org/hbase/docs/current/api/org/apache/hadoop/hbase/filter/RegExpRowFilter.html#filterColumn%28byte%5B%5D,%20byte%5B%5D,%20byte%5B%5D%29 > > > > >> < > > >> > > > http://hadoop.apache.org/hbase/docs/current/api/org/apache/hadoop/hbase/filter/RegExpRowFilter.html#filterColumn%28byte%5B%5D,%20byte%5B%5D,%20byte%5B%5D%29 > > >> > > > >> > > >> Sorry if you tried this and didn't help. > > >> > > >> Rasit > > >> > > >> > > >> 2009/2/17 Slava Gorelik <[email protected]> > > >> > > >> > HiThank You for the answer but as i understand this filter works on > > >> column > > >> > value and i need column name. > > >> > > > >> > Best Regards. > > >> > > > >> > > > >> > On Tue, Feb 17, 2009 at 3:17 PM, Rasit OZDAS <[email protected]> > > >> wrote: > > >> > > > >> > > Slava, there is a regex filter in HBase, (I didn't try yet). > > >> > > > > >> > > > > >> > > > >> > > > http://hadoop.apache.org/hbase/docs/current/api/org/apache/hadoop/hbase/filter/RegExpRowFilter.html > > >> > > > > >> > > and there is an extension of this: > > >> > > > > >> > > > > >> > > > >> > > > http://hadoop.apache.org/hbase/docs/current/api/org/apache/hadoop/hbase/filter/WhileMatchRowFilter.html > > >> > > > > >> > > Rasit > > >> > > > > >> > > 2009/2/16 Slava Gorelik <[email protected]>: > > >> > > > Do you have any example of this ? > > >> > > > > > >> > > > Thank You. > > >> > > > Slava. > > >> > > > > > >> > > > On Mon, Feb 16, 2009 at 8:48 PM, stack <[email protected]> > wrote: > > >> > > > > > >> > > >> A regex filter? > > >> > > >> St.Ack > > >> > > >> > > >> > > >> On Mon, Feb 16, 2009 at 10:46 AM, Slava Gorelik < > > >> > > [email protected] > > >> > > >> >wrote: > > >> > > >> > > >> > > >> > Hi. > > >> > > >> > I'm looking for a way retrieve columns by regular expression. > > >> > > >> > Is in 0.19.0 such functionality ? If not, what do you think > the > > >> best > > >> > > way > > >> > > >> to > > >> > > >> > do it ? > > >> > > >> > > > >> > > >> > Thank You and Best Regards. > > >> > > >> > Slava. > > >> > > >> > > > >> > > >> > > >> > > > > > >> > > > > >> > > > > >> > > > > >> > > -- > > >> > > M. Raşit ÖZDAŞ > > >> > > > > >> > > > >> > > >> > > >> > > >> -- > > >> M. Raşit ÖZDAŞ > > >> > > > > > > > > >
