起因: 使用happybase 访问hbase 时
def scan(self, row_start=None, row_stop=None, row_prefix=None,
columns=None, filter=None, timestamp=None,
include_timestamp=False, batch_size=1000, scan_batching=None,
limit=None, sorted_columns=False):
scan 函数中有1个row_prefix 参数，而这个参数在java client 对应函数并没有出现，它到底有甚么作用呢
查看源码，我们能看到
if row_prefix is not None:
if row_start is not None or row_stop is not None:
raise TypeError(
"'row_prefix' cannot be combined with 'row_start' "
"or 'row_stop'")
row_start = row_prefix
row_stop = str_increment(row_prefix)
str_increment 的具体代码
def str_increment(s):
"""Increment and truncate a byte string (for sorting purposes)
This functions returns the shortest string that sorts after the given
string when compared using regular string comparison semantics.
This function increments the last byte that is smaller than ``0xFF``, and
drops everything after it. If the string only contains ``0xFF`` bytes,
`None` is returned.
"""
for i in xrange(len(s) - 1, ⑴, ⑴):
if s[i] != 'xff':
return s[:i] + chr(ord(s[i]) + 1)
return None
看完代码大家应当很明白了,row_prefix 被转换成了row_start 和row_stop。
当有以下场景
微博表
用户ID_微博ID
假定我们想获得此用户的所有微博，在scan时就没有必要设定scan范围 '用户ID_0' ~ '用户ID_a'
而可以直接使用row_prefix = '用户ID'
PS：回头我会提供str_increment 的java 实现