在robotframework中使用 databaselibrary時，有個引數令我很疑惑，它就是 sanstran，這個引數在很多地方出現，它的解釋是
set optional input
`sanstran` to true to run command without an explicit transaction
commit or rollback.
execute_sql_string為例：def execute_sql_string(self, sqlstring, sanstran=false):
"""executes the sqlstring as sql commands and returns number of rows
affected. useful to pass arguments to your sql. set optional input
`sanstran` to true to run command without an explicit transaction
commit or rollback.
sql commands are expected to be delimited by a semi-colon (';').
for example:
| execute sql string | delete from person_employee_table; delete from person_table |
for example with an argument:
| execute sql string | select * from person where first_name = $ |
using optional `sanstran` to run command without an explicit transaction commit or rollback:
| execute sql string | delete from person_employee_table; delete from person_table | true |
"""cur = none
result = 0
try:
cur = self._dbconnection.cursor()
logger.info('executing : execute sql string | %s ' % sqlstring)
result = self.__execute_sql(cur, sqlstring)
if not sanstran:
self._dbconnection.commit()
finally:
if cur:
if not sanstran:
self._dbconnection.rollback()
return result
是我理解錯了，還是文件寫的有問題呢？？？？
乙個小疑惑
auto ply new playerchannel 看到這個我想到和new playerchannel 到底有啥區別 預設建構函式 只有當不提供任何建構函式，編譯器才會合成乙個預設建構函式，引數初始化取決於定義物件的位置 當提供乙個函式時就不會有預設建構函式 不管new playerchannel...
Metric Learning的乙個疑惑
我看的是liu yang的survey distance metric learning a comprehensive survey.裡面總結道，eric xing等把metric learning formulate成這樣乙個凸優化問題 begin min limits sum limits i...
乙個關於kylin的疑惑
自從研究了kylin這個系統之後，就一直有乙個疑惑，那就是kylin針對某一特定查詢條件建立的cube，只能在這個條件下查詢預計算的結果，如果條件一旦換了，cube就廢了。帶著這個疑問我在很多技術群裡和別人討論和爭辯，都沒有結果。看了很多網上分析的文章，都沒有關於這個疑問的詳細解答，於是今天自己試著...