[Linux C]利用libxml2解析xml文件

[Linux C]利用libxml2解析xml文件

分类： Linux_C 2012-03-09 15:42  382人阅读  评论(0)  收藏  举报

为了解析xml，可以使用Linux下默认安装的libxml2。

 /*
     a.c
     功能：利用libxml2解析xml文件
 */

 #include
 #include
 #include
 #include
 #include
 #include
 #include
 #include

 int GetCurFilePath(char *lpOut)     // get full path of the executable file
 {
     char chPath[BUFSIZ] = {0};
     int nRetVal = readlink("/proc/self/exe", chPath, sizeof(chPath)); // get full path of the current-executable file
     if(nRetVal < 0)
     {
         strcpy(lpOut, ".");
         return -1;
     }
     else
     {
         strcpy(lpOut, chPath);
         return 0;
     }
 }

 int GetCurDir(char *lpOut)                                  // get directory-path of current executable-file
 {
     char    chPath[BUFSIZ] = { 0 };
     if( GetCurFilePath(chPath) < 0 )
         return - 1;
     dirname(chPath);                                        // dirname will change value of "chPath"(contain result)
     strcpy(lpOut, chPath);                                  // copy result to out-param

     return 0;
 }

 xmlDocPtr getdoc(char *docname)                         // 根据文件名得到文档指针
 {
     xmlDocPtr doc;
     doc = xmlParseFile(docname);
     if(doc == NULL)
     {
         fprintf(stderr, "Document not parsed successfully.\n");
         return NULL;
     }
     return doc;
 }

 // 在文档doc中解析xpath表达式，返回结果集指针
 xmlXPathObjectPtr getnodeset(xmlDocPtr doc, xmlChar *xpath)
 {
     xmlXPathContextPtr context;
     xmlXPathObjectPtr result;
     context = xmlXPathNewContext(doc);
     if(context == NULL)
     {
         printf("Error in xmlXPathNewContent\n");
         return NULL;
     }
     result = xmlXPathEvalExpression(xpath, context);        // 在context中解析表达式xpath
     xmlXPathFreeContext(context);                           // 释放context
     if(result == NULL)
     {
         printf("Error in xmlXPathEvalExpression\n");
         return NULL;
     }
     if(xmlXPathNodeSetIsEmpty(result->nodesetval))           // 解析表达式的结果集为空
     {
         xmlXPathFreeObject(result);
         printf("No result\n");
         return NULL;
     }
     return result;
 }

 // 解析xmlPath路径的结点
 void testReadXmlDoc(char *filepath, char *xmlPath)
 {
     xmlDocPtr doc = getdoc(filepath);
     if(NULL == doc)
         return ;

     xmlChar *xpath = (xmlChar*) xmlPath;
     xmlXPathObjectPtr result = getnodeset(doc, xpath);          // 获取结果集
     if(result)
     {
         xmlNodeSetPtr nodeset = result->nodesetval;
         xmlChar *name, *value;
         printf("nodeset->nodeNr = %d\n", nodeset->nodeNr);        // 打印结果集中结点个数
         for(int i = 0; i < nodeset->nodeNr; i++)
         {
             xmlNodePtr cur = nodeset->nodeTab[i];                // products
             printf("cur->name = %s\n", cur->name);
             cur = cur->xmlChildrenNode;
             while(cur)
             {
                 if(xmlStrcmp(cur->name, (const xmlChar*) "text"))        // cur->name不为"text"
                 {
                     printf("cur->name = %s\t", cur->name);
                     name = xmlGetProp(cur, (const xmlChar*) "name");    // 获取属性值
                     value = xmlGetProp(cur, (const xmlChar*) "value");
                     printf("name = %s, value = %s\n", name, value);
                     xmlFree(name);
                     xmlFree(value);
                 }
                 cur = cur->next;
             }
             printf("\n");
         }
         xmlXPathFreeObject(result);
     }
     xmlFreeDoc(doc);
     xmlCleanupParser();
 }

 int main(void)
 {
     char curDir[100] = {0};
     char docname[100] = {0};
     GetCurDir(curDir);
     strcpy(docname, curDir);
     strcat(docname, "/dprod.xml");
     testReadXmlDoc(docname, "/allproducts/products");

     return EXIT_SUCCESS;
 }

makefile文件：

 CC=gcc
 CFLAGS=
 BIN=a
 INC=/usr/include/libxml2

 $(BIN): $(BIN).c
     $(CC) $(CFLAGS) -o $(BIN) $(BIN).c -I$(INC) -lxml2 -std=c99

 clean:
     rm -f *.o $(BIN)

xml文件(dprod.xml)内容：

 xml version="1.0"?>
 <allproducts>
   <products>
     <product name="name11" value="value11" />
     <product name="name12" value="value12" />
     <product name="name13" value="value13" />
     <product name="name14" value="value14" />
   products>
   <products>
     <product name="name21" value="value21" />
     <product name="name22" value="value22" />
     <product name="name23" value="value23" />
   products>
   <products>
     <product name="name31" value="value31" />
     <product name="name32" value="value32" />
   products>
 allproducts>

编译运行：

 [zcm@tmp #115]$make
 gcc  -o a a.c -I/usr/include/libxml2 -lxml2 -std=c99
 a.c: 在函数‘GetCurFilePath’中:
 a.c:18: 警告：隐式声明函数‘readlink’
 [zcm@tmp #116]$./a
 nodeset->nodeNr = 3
 cur->name = products
 cur->name = product  name = name11, value = value11
 cur->name = product  name = name12, value = value12
 cur->name = product  name = name13, value = value13
 cur->name = product  name = name14, value = value14

 cur->name = products
 cur->name = product  name = name21, value = value21
 cur->name = product  name = name22, value = value22
 cur->name = product  name = name23, value = value23

 cur->name = products
 cur->name = product  name = name31, value = value31
 cur->name = product  name = name32, value = value32

 [zcm@tmp #117]$

说明：对于编译中出现的“a.c:18: 警告：隐式声明函数‘readlink’”错误，实在不能明白。我查了下手册，这个函数在unistd.h中，而且我也已经#include了，为什么还会出现这个错误呢？

后来突然想到，可能是-std=c99的原因，将它改为-std=gnu99后，这个警告就没有了!

--------------------------------------------------------------------------------------------------------------------------------------------------

修改了xml文件和上面源码中的testReadXmlDoc()后，发现结果相当神奇，看来对libxml2的理解还是比较缺乏。

1. 修改xml文件内容：

 xml version="1.0"?>
 <allproducts>
   <products>h1
     <product name="name11" value="value11" />h2
     <product name="name12" value="value12" />h3
     <product name="name13" value="value13" />h4
     <product name="name14" value="value14" />h5
   products>
   <products>
     <product name="name21" value="value21" />
     <product name="name22" value="value22" />
     <product name="name23" value="value23" />
   products>
   <products>
     <product name="name31" value="value31" />g1
     <product name="name32" value="value32" />
 g2products>
 allproducts>

2. 修改testReadXmlDoc()

 // 解析xmlPath路径的结点
 void testReadXmlDoc(char *filepath, char *xmlPath)
 {
     xmlDocPtr doc = getdoc(filepath);
     if(NULL == doc)
         return ;

     xmlChar *xpath = (xmlChar*) xmlPath;
     xmlXPathObjectPtr result = getnodeset(doc, xpath);          // 获取结果集
     if(result)
     {
         xmlNodeSetPtr nodeset = result->nodesetval;
         xmlChar *name, *value;
         printf("nodeset->nodeNr = %d\n", nodeset->nodeNr);        // 打印结果集中结点个数
         for(int i = 0; i < nodeset->nodeNr; i++)
         {
             xmlNodePtr cur = nodeset->nodeTab[i];                // products
             printf("cur->name = %s\n", cur->name);
             cur = cur->xmlChildrenNode;
             int ctext = 0;
             while(cur)
             {
                 if(xmlStrcmp(cur->name, (const xmlChar*) "text"))        // cur->name不为"text"
                 {
                     printf("cur->name = %s\t", cur->name);
                     name = xmlGetProp(cur, (const xmlChar*) "name");    // 获取属性值
                     value = xmlGetProp(cur, (const xmlChar*) "value");
                     printf("name = %s, value = %s\n", name, value);
                     xmlFree(name);
                     xmlFree(value);
                 }
                 else
                 {
                     ctext++;
                     xmlChar *v = xmlNodeListGetString(doc, cur, 1);
                     printf("cur->content = [%s], v = [%s]", cur->content, v); // cur->content获取cur的内容
                     xmlFree(v);
                 }
                 cur = cur->next;
             }
             printf("ctext = %d\n", ctext);
             printf("\n");
         }
         xmlXPathFreeObject(result);
     }
     xmlFreeDoc(doc);
     xmlCleanupParser();
 }

运行结果：

 [zcm@tmp #168]$make
 gcc  -o a a.c -I/usr/include/libxml2 -lxml2 -std=gnu99
 [zcm@tmp #169]$./a
 nodeset->nodeNr = 3
 cur->name = products
 cur->content = [h1
     ], v = [h1
     h2
     h3
     h4
     h5
   ]cur->name = product   name = name11, value = value11
 cur->content = [h2
     ], v = [h2
     h3
     h4
     h5
   ]cur->name = product   name = name12, value = value12
 cur->content = [h3
     ], v = [h3
     h4
     h5
   ]cur->name = product   name = name13, value = value13
 cur->content = [h4
     ], v = [h4
     h5
   ]cur->name = product   name = name14, value = value14
 cur->content = [h5
   ], v = [h5
   ]ctext = 5

 cur->name = products
 cur->content = [
     ], v = [



   ]cur->name = product   name = name21, value = value21
 cur->content = [
     ], v = [


   ]cur->name = product   name = name22, value = value22
 cur->content = [
     ], v = [

   ]cur->name = product   name = name23, value = value23
 cur->content = [
   ], v = [
   ]ctext = 4

 cur->name = products
 cur->content = [
     ], v = [
     g1

 g2]cur->name = product   name = name31, value = value31
 cur->content = [g1
     ], v = [g1

 g2]cur->name = product   name = name32, value = value32
 cur->content = [
 g2], v = [
 g2]ctext = 3

 [zcm@tmp #170]$

由此可见，一般情况下，我们用的比较多的可能会是cur->content这个东西了!

补充：

在网上看到一个人的写法，可以在解析xml文件时，直接忽略掉结点之间的无效空白。对于本文，就是将：

doc = xmlParseFile(docname);　－－－＞修改为：doc = xmlParseFile(docname, "UTF-8", XML_PARSE_NOBLANKS);  // 第3个参数是关键

http://blog.csdn.net/joeblackzqq/article/details/7337093