程序间数据共享与传递（2）：EXPORT/IMPORT、SAP/ABAP Memory

本文详细介绍了ABAP中使用IMPORT与EXPORT语句进行数据导出与导入的方法，包括不同类型数据表的操作方式及参数说明，并展示了如何利用IMPORT DIRECTORY语句获取存储信息。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

IMPORT语句

IMPORT {p1 = dobj1 p2 = dobj2 ...} | {p1 TO dobj1 p2 TO dobj2 ...} | (ptab)
FROM | { MEMORY ID id }
  | { DATABASE      dbtab(ar) [TO wa] [CLIENT cl] ID id }
  | { SHARED MEMORY dbtab(ar) [TO wa] [CLIENT cl] ID id }
  | { SHARED BUFFER dbtab(ar) [TO wa] [CLIENT cl] ID id }

从簇数据表中读取数据，各项参数与EXPORT是一样的，请参考EXPORT各项解释

TYPES:BEGIN OF tab,
    col1 TYPE i,
    col2 TYPE i,
  END OF tab.
DATA:wa_indx TYPE demo_indx_table,
  wa_itab TYPE tab,
  itab    TYPE STANDARD TABLE OF tab.
IMPORT tab = itab FROM DATABASE demo_indx_table(sq) TO wa_indx ID 'TABLE'.
WRITE: wa_indx-timestamp, wa_indx-userid.
ULINE.
LOOP AT itab INTO wa_itab.
  WRITE: / wa_itab-col1, wa_itab-col2.
ENDLOOP.

实例二：

TABLES indx.
DATA: BEGIN OF jtab OCCURS 100,
  col1 TYPE i,
  col2 TYPE i,
END OF jtab.

"注意：i_tab的名称不能是其他的，必须与EXPORT语句中的分类
"存储标签名一样。如果i_tab本身又是一个jtab类型的内表，则
"TO后面的jtab可以省略
IMPORT i_tab TO jtab FROM DATABASE indx(hk) ID 'Key'.
"注：在该程序中并没有明显的为indx工作区设置值，但由于使用了
"TABLES indx.语句定义了与indx簇数据库表同名的结构变量，所以
"上面IMPORT会默认加上使用 TO indx 选项
WRITE: / 'AEDAT:', indx-aedat,
      / 'USERA:', indx-usera,
      / 'PGMID:', indx-pgmid.
SKIP.
WRITE 'JTAB:'.
LOOP AT jtab FROM 1 TO 5.
  WRITE: / jtab-col1, jtab-col2.
ENDLOOP.

AEDAT: 2011.10.12

USERA: ZHENGJUN

PGMID: YJZJ_TEST2

JTAB:

1 1

2 4

3 9

4 16

5 25

IMPORT DIRECTORY语句

IMPORT DIRECTORY INTO itab FROM DATABASE dbtab(ar) [TO wa] [CLIENT cl] ID id.

该语句的作用用来读取存储到簇数据表中的变量信息，如存储名是什么、以什么样的类型存储的、数据长度等（有点类似反射）。itab 为cdir词典类型的内表，其结构与各字段作用：

Component	Type	Meaning
NAME	CHAR(30)	Name of the parameter under which a data object was saved.
OTYPE:	CHAR(1)	General type of the data object saved. The following values are permitted: "F" for elementary, flat data objects, "G" for strings, "R" for flat structures, "S" for deep structures, "T" for internal tables with flat row type and "C" for tables with a deep row type.
FTYPE	CHAR(1)	More specific type of the data object saved. For elementary data objects and internal tables with a elementary row type, the data or row type is returned in accordance with the table of return values fromDESCRIBE FIELD ... TYPE ("a", "b", "C", "D", "e", "F", "g", "I", "N", "P", "s", "T", "X", "y"). In the case of flat structures and internal tables with flat structured row types, "C" is returned. In the case of deepstructures and internal tables with deep structured row types, "v" is returned. In the case of a table that has an internal table as a row type, "h" is returned.
TFILL	INT4	Length filled of the saved data object. For strings the length of the content in bytes is returned and for internal tables the number of rows is returned. The value 0 is returned for other data objects.
FLENG	INT2	Length of saved data object or saved table row in bytes. The value 8 is returned for strings.

DATA: f1 TYPE decfloat16,
      f2 TYPE TABLE OF i,
      f3 TYPE spfli.
DATA itab TYPE STANDARD TABLE OF cdir.
DO 10 TIMES.
  APPEND sy-index TO f2.
ENDDO.
EXPORT par1 = f1 par2 = f2 par3 = f3 TO DATABASE demo_indx_table(hk) ID 'HK'.
IMPORT DIRECTORY INTO itab FROM DATABASE demo_indx_table(hk) ID 'HK'.

NAME	OTYPE:	FTYPE	TFILL	FLENG
"PAR1"	"F"	"a"	0	8
"PAR2"	"T"	"I"	10	4
"PAR3"	"R"	"C"	0	168

DELETE FROM语句

DELETE FROM { {MEMORY ID id}
            | {DATABASE      dbtab(ar) [CLIENT cl] ID id}
            | {SHARED MEMORY dbtab(ar) [CLIENT cl] ID id}
            | {SHARED BUFFER dbtab(ar) [CLIENT cl] ID id} }.

用来清理EXPORT语句的存储的数据

其中DELETE FROM MEMORY ID id.与FREE MEMORY ID id.作用一样

DATA: id    TYPE c LENGTH 4 VALUE 'TEXT',
      text1 TYPE string     VALUE 'Tina',
      text2 TYPE string     VALUE 'Mike'.

EXPORT p1 = text1 p2 = text2 TO SHARED BUFFER demo_indx_table(xy) ID id.

IMPORT p1 = text2 p2 = text1 FROM SHARED BUFFER demo_indx_table(xy) ID id.

...

DELETE FROM SHARED BUFFER demo_indx_table(xy) ID id.
"此语句会执行后，sy-subrc返回4
IMPORT p1 = text2 p2 = text1 FROM SHARED BUFFER demo_indx_table(xy) ID id.