DO NOT REPLY [Bug 33239] New: - date encoding limitation removing

[bu...@apache.org]

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG�
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=33239>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND�
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=33239

           Summary: date encoding limitation removing
           Product: Lucene
           Version: 1.4
          Platform: PC
        OS/Version: Windows XP
            Status: NEW
          Severity: enhancement
          Priority: P5
         Component: Index
        AssignedTo: lucene-dev@jakarta.apache.org
        ReportedBy: vbychkoviak@i-hypergrid.com


currently there is some limitation to date encoding in lucene. I think it's 
because dates should preserve lexicografical ordering, i.e. if one date precedes 
another date then encoded values should keep same ordering.

I know that it can be difficult to integrate it into existing version but there 
is way to remove this limitation.
Date milliseconds can be encoded as unsigned values with prefix that indicates 
positive or negative value.

In more details:
I used hex encoding and prefix &#8216;p&#8217; and &#8216;n&#8217; for positive and negative values. I 
got following results:

Value -10000 is encoded with nffffffffffffd8f0, 
-100	- nffffffffffffff9c
0	- p0000000000000000
100	- p0000000000000064
10000	- p0000000000002710

This preserves ordering between values and theirs encoding.

Also hex encoding can be replaced with Character.MAX_RADIX encoding.

Part of code that do this work:
   final static char[] digits = {
	'0' , '1' , '2' , '3' , '4' , '5' ,
	'6' , '7' , '8' , '9' , 'a' , 'b' ,
	'c' , 'd' , 'e' , 'f' , 'g' , 'h' ,
	'i' , 'j' , 'k' , 'l' , 'm' , 'n' ,
	'o' , 'p' , 'q' , 'r' , 's' , 't' ,
	'u' , 'v' , 'w' , 'x' , 'y' , 'z'
    };


    char prefix;
    if (time >= 0) {
      prefix = 'p';
    } else {
      prefix = 'n';
    }

    char[] chars = new char[DATE_LEN + 1];
    int index = DATE_LEN;
    while (time != 0) {
      int b = (int) (time & 0x0F);
      chars[index--] = digits[b];
      time = time >>> 4;
    }

    while (index >= 0) {
      chars[index--] = '0';
    }
    chars[0] = prefix;

    return new String(chars);

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

---------------------------------------------------------------------
To unsubscribe, e-mail: lucene-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: lucene-dev-help@jakarta.apache.org