在SAS中读取XML数据时设置数据类型

Mar*_*und 2 xml types sas

在SAS中读取XML数据时,我需要控制数据类型.使用SAS中的XML libname引擎编写和访问XML数据.

SAS似乎根据列的内容猜测数据类型:如果我在字符列中将"20081002"写入我的XML数据,它将作为数字变量读回.

一个例子:

filename my_xml '/tmp/my.xml'; * Yes, I use SAS on Unix *;
libname my_xml XML;

data my_xml.data_type_test;
  text_char="This is obviously text";
  date_char="20081002";
  num_char="42";
  genuine_num=42;
run;

proc copy inlib=my_xml outlib=WORK;
run;

libname my_xml;
filename my_xml CLEAR;
Run Code Online (Sandbox Code Playgroud)

只有最后一列被定义为XML数据中的数值数据类型,但是当我将其复制到我的WORK库中时,只有text_char列是字符.其他3个现在是数字.

在SAS中读取XML数据时如何控制数据类型?

小智 6

看看SAS XML Mapper.它允许您创建一个映射来读取(和9.2中的wrte)XML文件并指定列属性.

如果这是您的XML文件:

这显然是文本20081002 42 42您可以像这样创建一个MAP:

<!-- ############################################################ -->
<TABLE name="DATA_TYPE_TEST">
    <TABLE-PATH syntax="XPath">/TABLE/DATA_TYPE_TEST</TABLE-PATH>

    <COLUMN name="text_char">
        <PATH syntax="XPath">/TABLE/DATA_TYPE_TEST/text_char</PATH>
        <TYPE>character</TYPE>
        <DATATYPE>string</DATATYPE>
        <LENGTH>22</LENGTH>
    </COLUMN>

    <COLUMN name="date_char">
        <PATH syntax="XPath">/TABLE/DATA_TYPE_TEST/date_char</PATH>
        <TYPE>numeric</TYPE>
        <DATATYPE>integer</DATATYPE>
        <FORMAT width="9">DATE</FORMAT>
        <INFORMAT width="8">ND8601DA</INFORMAT>
    </COLUMN>

    <COLUMN name="num_char">
        <PATH syntax="XPath">/TABLE/DATA_TYPE_TEST/num_char</PATH>
        <TYPE>character</TYPE>
        <DATATYPE>string</DATATYPE>
        <LENGTH>2</LENGTH>
    </COLUMN>

    <COLUMN name="genuine_num">
        <PATH syntax="XPath">/TABLE/DATA_TYPE_TEST/genuine_num</PATH>
        <TYPE>numeric</TYPE>
        <DATATYPE>integer</DATATYPE>
    </COLUMN>

</TABLE>
Run Code Online (Sandbox Code Playgroud)

然后阅读XML文件:

filename  my 'C:\temp\my.xml';
filename  SXLEMAP 'C:\temp\MyMap.map';
libname   my xml xmlmap=SXLEMAP access=READONLY;

title 'Table DATA_TYPE_TEST';
proc contents data=my.DATA_TYPE_TEST varnum; 
run;
proc print data=my.DATA_TYPE_TEST(obs=10); 
run;
Run Code Online (Sandbox Code Playgroud)

结果:

Table DATA_TYPE_TEST

The CONTENTS Procedure

Data Set Name        MY.DATA_TYPE_TEST    Observations            
Member Type          DATA                 Variables             4 
Engine               XML                  Indexes               0 
Created              .                    Observation Length    0 
Last Modified        .                    Deleted Observations  0 
Protection                                Compressed            NO
Data Set Type                             Sorted                NO
Label                                                             
Data Representation  Default                                      
Encoding             Default                                      


Variables in Creation Order

#    Variable       Type    Len    Format    Informat      Label

1    text_char      Char     22    $22.      $22.          text_char  
2    date_char      Num       8    DATE9.    ND8601DA8.    date_char  
3    num_char       Char      2    $2.       $2.           num_char   
4    genuine_num    Num       8    F8.       F8.           genuine_num

Table DATA_TYPE_TEST

                                                               genuine_
     Obs    text_char                 date_char    num_char      num

       1    This is obviously text    02OCT2008       42             42
Run Code Online (Sandbox Code Playgroud)