[ https://issues.jboss.org/browse/TEIID-2383?page=com.atlassian.jira.plugin... ] Hisanobu Okuda commented on TEIID-2383: --------------------------------------- File translator uses buffered reader to read file. In the sample, you can see the Japanese character '道'={0xe9 0x81 0x93} is buffered separately, i.e., {0xe9} and {0x81 0x93}. {code} [hokuda@localhost 00769999]$ rlwrap -l jdb.log -pRED -C jdb_teiid /opt/jdk1.6.0_37/bin/jdb -attach localhost:8787 Set uncaught java.lang.Throwable Set deferred uncaught java.lang.Throwable Initializing jdb ... Set breakpoint org.teiid.core.util.InputStreamReader:93 Breakpoint hit: "thread=Worker2_QueryProcessorQueue16", org.teiid.core.util.InputStreamReader.read(), line=93 bci=178 Worker2_QueryProcessorQueue16[1] cont Breakpoint hit: "thread=Worker2_QueryProcessorQueue16", org.teiid.core.util.InputStreamReader.read(), line=93 bci=178 Worker2_QueryProcessorQueue16[1] cont Breakpoint hit: "thread=Worker2_QueryProcessorQueue16", org.teiid.core.util.InputStreamReader.read(), line=93 bci=178 Worker2_QueryProcessorQueue16[1] cont Breakpoint hit: "thread=Worker2_QueryProcessorQueue16", org.teiid.core.util.InputStreamReader.read(), line=93 bci=178 Worker2_QueryProcessorQueue16[1] cont Breakpoint hit: "thread=Worker3_QueryProcessorQueue18", org.teiid.core.util.InputStreamReader.read(), line=93 bci=178 Worker3_QueryProcessorQueue18[1] print this.bb this.bb = "java.nio.HeapByteBuffer[pos=8191 lim=8192 cap=8192]" Worker3_QueryProcessorQueue18[1] print this.bb.hb[8191] this.bb.hb[8191] = -23 (=0xe9) {code} Before executing line 93, position=8191 and limit=8192 in this.bb. This means the last byte remained to be decoded at next turn. The value of the last byte is -23 = 0xe9. It is the first byte of Japanese character '道'={0xe9 0x81 0x93}. {code} Worker3_QueryProcessorQueue18[1] locals Method arguments: cbuf = instance of char[8192] (id=10433) off = 0 len = 8192 Local variables: read = 8191 cr = instance of java.nio.charset.CoderResult(id=10434) Since bb.position()=8191 and read=8191, bb.clear() is invoked. Actually, bb.compact() should be invoked though. Worker3_QueryProcessorQueue18[1] next Step completed: "thread=Worker3_QueryProcessorQueue18", org.teiid.core.util.InputStreamReader.read(), line=96 bci=201 Worker3_QueryProcessorQueue18[1] next Step completed: "thread=Worker3_QueryProcessorQueue18", org.teiid.core.util.InputStreamReader.read(), line=98 bci=209 Worker3_QueryProcessorQueue18[1] print this.bb this.bb = "java.nio.HeapByteBuffer[pos=0 lim=8192 cap=8192]" {code} Therefore read=8191 and bb.position()=8191, bb.clear() is invoked, the position of bb is reset to 0, and finally the last byte this.bb.hb[8191] = -23 (=0xe9) is discarded. In next blocking read, 0x81 and 0x93 are stored in bb. {code} Worker3_QueryProcessorQueue18[1] stop at org.teiid.core.util.InputStreamReader:82 Set breakpoint org.teiid.core.util.InputStreamReader:82 Worker3_QueryProcessorQueue18[1] cont Breakpoint hit: "thread=Worker3_QueryProcessorQueue18", org.teiid.core.util.InputStreamReader.read(), line=82 bci=100 Worker3_QueryProcessorQueue18[1] print this.bb this.bb = "java.nio.HeapByteBuffer[pos=0 lim=4 cap=8192]" Worker3_QueryProcessorQueue18[1] print this.bb.hb[0] this.bb.hb[0] = -127 (=0x81) Worker3_QueryProcessorQueue18[1] print this.bb.hb[1] this.bb.hb[1] = -109 (=0x93) Worker3_QueryProcessorQueue18[1] print this.bb.hb[2] this.bb.hb[2] = 39 Worker3_QueryProcessorQueue18[1] print this.bb.hb[3] this.bb.hb[3] = 10 Worker3_QueryProcessorQueue18[1] {code} The first byte 0xe9 is missing, decoding fails. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira