初始化cobol中的索引变量

Question

我们有一个COBOL程序，我们在其中将值填充到COBOL内部表中，然后搜索此表以找出某个值。在此搜索之前，我们初始化表索引变量。

SET PAF-IDX                  TO 1.   

任何人都可以澄清一下，如果在COBOL中允许初始化索引变量，则可以这样做。

INITIALIZE                   PAF-IDX.

Answer 1

没有。你为什么要“初始化”呢？

这来自IBM Enterprise Cobol手册：

identifier-1
Receiving areas.
identifier-1 must reference one of the following:
v An alphanumeric group item
v A national group item
v An elementary data item of one of the following categories:
Alphabetic
Alphanumeric
Alphanumeric-edited
DBCS
External floating-point
Internal floating-point
National
National-edited
Numeric
Numeric-edited
v A special register that is valid as a receiving operand in a MOVE
statement with identifer-2 or literal-1 as the sending operand.
When identifier-1 re

编辑：

OpenCobol程序员指南专门记录了它：

The list of data items eligible to be set to new values by this statement is:
Every elementary item specified as identifier-1 ..., PLUS...
Every elementary item defined subordinate to every group item specified as dentifier-1 
..., with the following exceptions:
USAGE INDEX items are excluded.
Items with a REDEFINES as part of their definition are excluded; this means that
tems subordinate to them are excluded as well. 

Cobol标准草案不太明确，但在INITIALIZE中使用时这些项目是通过生成SET而不是MOVE来处理的：DATA-POINTER，FUNCTION-POINTER，OBJECT-REFERENCE或PROGRAM-POINTER。

编辑：

看到OpenCobol引用并不像我想象的那样“具体”：在IBM Cobol中，目前，作为MOVE目标无法操纵的任何东西都可以是INITIALIZEd。对于当前的OpenCobol，这是相同的。 Cobol草案列出了一些例外，但既不包括INDEXED BY（它们不是表本身的一部分，而是编译器本身定义存储的单独项目），也不包括USAGE INDEX。