DomDoc/SimpleXML/XSLT：解析以向元素的每个唯一元素子元素添加自动递增 id 属性

Question

我已经解决这个问题有一段时间了，而且我对编程有点陌生。即使我发现错误，也很难弄清楚如何纠正它。现在，我正在试图弄清楚我是如何错误地使用xpath的，因为有人告诉我我使用了错误的xpath。我希望有人可以告诉我我做错了什么，特别是迭代，如果我做错了什么，可以给我一个快速的开始。这是我参与这个项目的最后一晚，如果可以的话，我真的很想完成它。所以，我真的需要帮助。这是我正在使用的代码，带有注释：

$xml = @simplexml_load_file("original.xml"); //Loading the original file, dubbed original.xml.
$array_key_target_parent = count($xml->xpath('/doc/*'); //Puts all of the children of <doc> into an _iterable_ array.
$key_targets = foreach($array_key_target_parent;){
  foreach($array_key_target_parent as $single_target){ // I tried foreach($array_key_target_parent[$i]).  It doesn't work, so don't even go there.
    $current_target = current($single_target);
    count($xml->xpath('/doc/$current_target/*');
  }
} */ ////Puts the targets for keying into iterable arrays.  =>1 makes the array start from 1, so the id's will be right.


/* At this point, we have multiple elements that we want to key, each having a unique name.  There's <element_type1a> and <element_type1b>, etc.  We want each one to have its own id set.  So, we have to embed iteration within iteration. */
foreach($key_target){ //This will ensure that every unique element that we want to key gets its key set.
  {
  $id = current($key_target=>1); //This allows us to reset the id to 1 (=>1), each time the key algorithm starts for a new element.
  foreach($key_target as $id){ //I tried for($i=0, $key_target[$i]; $i>$key_target; $i++), and it didn't work, so don't even go there.
    addAttribute('id', '$id');
  }
}  //Adds an 'id' attribute and a unique number to each target.

$xml->asXML("new.xml"); //saves the output as a new xml document, new.xml

我还有一个通用的 XML 文件：

<doc>
    <info_type1>
        <element_type1a>not_unique_data</element_type1a>
        <element_type1b>unique_data</element_type1b>
        <element_type2a>not_unique_data</element_type2a>
        <element_type2b>not_unique_data</element_type2b>
        <element_type2c lang="fr">not_unique_data</element_type2c>
        <!-- ... --->
        <element_typeNxM>unique_data</element_typeNxM>
    </info_type1>
    <info_type2>
        <element_type1a>repeat_data(info_type1_element1a)</element_type1a>
        <element_type2a>not_unique_data</element_type2a>
    </info_type2>
    <!-- ... --->
    <info_typeN>
        <descendants></descendants>
    </info_typeN>
</doc>

所需的输出：

<datatables>
    <table id="element_type1">
        <element_type1a id="1">unique_data</element_type1a>
        <element_type1b id="2">unique_data</element_type1b>
        <!-- ... --->
        <element_type1N id="M">unique_data</element_type1N>
    </table>
    <table id="element_type2">
        <element_type2a id="1">unique_data</element_type2a>
        <element_type2b id="2">unique_data</element_type2b>
        <!-- ... --->
        <element_type2N id="M">unique_data</element_type2N>
    </table>
    <table id="element_type2_fr">
        <element_type2a lang="fr" id="1">unique_data</element_type2a>
        <element_type2b lang="fr" id="2">unique_data</element_type2>
        <!-- ... (there are five languages) --->
        <element_type2N lang="fr" id="M">unique_data</element_type2N>
    </table>
    <!-- ... --->
    <table id="element_typeN">
        <descendants></descendants>
    </table>
</datatables>

我

<intermediary_tables>
    <table id="intermediary_table_type1xtype2">
        <element id="1">
            <type1ID>1</type1ID>
            <type2ID>1</type2ID>
        </element>
        <element id="2">
            <type1ID>1</type1ID>
            <type2ID>2</type2ID>
        </element>
        <element id="3">
            <type1ID>2</type1ID>
            <type2ID>1</type2ID>
        </element>
        <element id="4">
            <type1ID>2</type1ID>
            <type2ID>2</type2ID>
        </element>
        <!-- ... --->
        <element id="N">
            <type1ID>M</type1ID>
            <type2ID>Z</type2ID>
        </element_type2N>
    </table>

    <table id="intermediary_table_typeMxtypeN">
        <descendants></descendants>
    </table>
</intermediary_tables>

也看到了许多非常相似的问题，我从他们那里收集了一些资源并阅读：

• http://www.ibm.com/developerworks/xml/library/x-xmlphp1/index.html
• http://www.php.net/manual/en/simplexmlelement.addattribute.php
• http://www.learn-xslt-tutorial.com/
• http://msdn.microsoft.com/en-us/library/ms256103.aspx
• http://php.net/manual/en/class.domdocument.php

这些是最有用的链接：

• http://www.capcourse.com/Library/NormalizingXML/Part1.html
• http://forums.tizag.com/showthread.php?t=17821

我发现这些问题的应用程序都无法产生我正在尝试的结果来实现。不过，capcourse.com 链接是个例外。它面向已毕业的 CS 受众，他们似乎在做同样的事情，只不过他们使用的 ID 不是自动增量的。他们使用的算法极其复杂，而且他们根本没有对他们的代码进行注释。由于某种原因，他们在其名称空间内使用名称空间，即使它是我能找到的最接近的名称空间，我也无法复制它。

---

更新

从 XML 文档中提取的真实世界摘录，我想解析该文档以更改数据结构：

<?xml version="1.0"?>
<!DOCTYPE catalog [
<!ELEMENT catalog (entry*)>
<!ELEMENT entry (ent_seq, country*, arist+, info?, title+)><!-- Entries consist of the name of the album, artist, and more information about the CD.  Each entry must contain an artist and an album title. -->
<!ELEMENT ent_seq (#PCDATA)><!-- A unique numeric sequence, showing the entry number -->
<!ELEMENT title (#PCDATA)><!-- The title of the album/the album name. -->
<!ELEMENT artist (band+, name, nickname*)><!-- The name of the band, and if there was a famous artist, his name and nickname.  Must contain a band element. -->
<!ELEMENT band (#PCDATA)><!-- The name of the band. -->
<!ELEMENT name (#PCDATA)><!-- The name of any famous artist in the band. -->
<!ELEMENT nickname (#PCDATA)><!-- The nickname of the popular artist that precedes the nickname element, from the band. -->
<!ELEMENT country (#PCDATA)><!-- Specifies countries where the album was released -->
<!ELEMENT company (name, country)><!-- Company/producer info.  The company's name is in the name element, and the country where the company originated is in the country element. -->
<!ELEMENT name (#PCDATA)><!-- The name of the producer -->
<!ELEMENT country (#PCDATA)><!-- The country where the company does its primary business -->
<!ELEMENT year (#PCDATA)><!-- The year of the album's release -->
<!ELEMENT info (link*, bibl*)><!-- Additional info, including links and bibliography information -->
<!ELEMENT link (#PCDATA)><!-- Links where people can read more about the album -->
<!ELEMENT bibl (#PCDATA)><!-- Bibliography text about the artist -->
]>
<catalog>
  <cd>
    <ent_seq>1</ent_seq>
    <title>For Your Love</title>
    <artist>
      <name>The Yardbirds</name>
      <name>Eric Clapton</name>
      <nickname>Slowhand</nickname>
    </artist>
    <country>USA</country>
    <country>UK</country>
    <company>
      <name>Sweet Music</name>
      <country>USA</country>
    </company>
    <year>1965</year>
    <info>
      <link>http://en.wikipedia.org/wiki/For_Your_Love</link>
    </info>
  </cd>
  <cd>
    <ent_seq>2</ent_seq>
    <title>Splish Splash</title>
    <artist>
      <name>Roberto Carlos</name>
      <nickname>The King</nickname>
    </artist>
    <country>USA</country>
    <country>Brazil</country>
    <country>Italy</country>
    <company>
      <name>Sweet Music</name>
    <country>Brazil</country>
    </company>
    <year>1965</year>
  </cd>
  <cd>
    <ent_seq>3</ent_seq>
    <title>How Great Thuo Art</title>
    <artist>
      <name>Elvis Presley</name>
      <nickname>The King</nickname>
      <nickname>The King of Rock 'n Roll</nickname>
    </artist>
    <country>USA</country>
    <country>Canada</country>
    <country>UK</country>
    <company>
      <name>Felton Jarvis</name>
      <country>USA</country>
    </company>
    <year>1965</year>
  </cd>
  <cd>
    <ent_seq>4</ent_seq>
    <title>Big Willie style</title>
    <artist>
      <band>Will Smith</band>
      <name>Will Smith</name>
    </artist>
    <country>USA</country>
    <company>Columbia</company>
    <year>1997</year>
  </cd>
  <cd>
    <ent_seq>5</ent_seq>
    <title>Empire Burlesque</title>
    <artist>
      <band>Bob Dylan and Boby Rockhammer</band>
      <name>Bob Dylan</name>
      <name>Boby Rockhammer</name>
    </artist>
    <country>USA</country>
    <country>India</country>
    <company>Columbia</company>
    <year>1985</year>
  </cd>
  <cd>  <!-- Update part 1: New Entry -->
    <ent_seq>6</ent_seq>
    <title>Merry Christmas</title>
    <title>White Christmas</title>
    <artist>
      <name>Bing Crosby</name>
    <artist>
    <country>USA</country>
    <company>MCA Records</company>
    <year>1995</year>
  </cd> <!-- End update part 1-->
</catalog>

所需输出示例的真实世界示例：

<datatable>
  <table id="album title">
    <title id="1">For your Love</title>
    <title id="2">Splish Splash</title>
    <title id="3">How Great Thuo Art</title>
    <title id="4">Big Willie style</title>
    <title id="5">Empire Burlesque</title>
    <title id="6">Merry Christmas</title> <!-- Update part 2: New output -->
    <title id="7">White Christmas</title> <!-- Update part 2: New output -->
  </table>
  <table id="Band Name">
    <artist id="1">The Yardbirds</artist>
    <artist id="2">Roberto Carlos</artist>
    <artist id="3">Elvis Presley</artist>
    <artist id="4">Will Smith</artist>
    <artist id="5">Bob Dylan and Boby Rockhammer</artist>
    <artist id="6"> <!-- Update part 2: New output -->
  </table>
  <table id="artist name">
    <artist id="1">Eric Clapton</artist>
    <artist id="2">Roberto Carlos</artist>
    <artist id="3">Elvis Presley</artist>
    <artist id="4">Will Smith</artist>
    <artist id="5">Bob Dylan</artist>
    <artist id="6">Boby Rockhammer</artist>
    <artist id="7">Bing Crosby</artist> <!-- Update part 2: New output -->
  </table>
  <table id="nickname">
    <nickname id="1">Slowhand</nickname>
    <nickname id="2">The King</nickname>
    <nickname id="3">The King of Rock 'n Roll</nickname>
  </table>
</datatable>

以及

<intermediarytable>
  <table id="artist by band name">
    <entry id="1">
      <band_id>1</band_id>
      <artist_id>1</artist_id>
    </entry>
    <entry id="2">
      <band_id>2</band_id>
      <artist_id>2</artist_id>
    </entry>
    <entry id="3">
      <band_id>3</band_id>
      <artist_id>3</artist_id>
    </entry>
    <entry id="4">
      <band_id>4</band_id>
      <artist_id>4</artist_id>
    </entry>
    <entry id="5">
      <band_id>5</band_id>
      <artist_id>5</artist_id>
    </entry>
    <entry id="6">
      <band_id>5</band_id>
      <artist_id>6</artist_id>
    </entry>
    <entry id="7">
      <band_id>6</band_id>
      <artist_id>7</artist_id>
    </entry>
  </table>
  <table id="artist by nickname">
    <entry id="1">
      <artist_id>1</artist_id>
      <nickname_id>1</artist_id>
    </entry>
    <entry id="2">
      <artist_id>2</artist_id>
      <nickname_id>2</nickname_id>
    </entry>
    <entry id="3">
      <artist_id>2</artist_id>
      <nickname_id>3</nickname_id>
    </entry>
    <entry id="4">
      <artist_id>3</artist_id>
      <nickname_id>3</nickname_id>
    </entry>
  </table>
</intermediarytable>

--更新-- 存在问题哪两个元素共享相同的条目 ID

在我的另一个 XML 文档中，

<entry id="1">
  <word>blue</word>
  <word>beryl</word>
  <word lang="SP">azul</word>
</entry>

我希望输出为

数据表：

<table id="en">
  <word lang="en" id="0">blue</word>
  <word lang="en" id="1">beryl</word>
</table>
<table id="sp">
  <word lang="sp" id="0">azul</word>
</table>

中间表：

<table id="translation id">
  <en_sp id="0"> <!-- en_sp means English-to-Spanish -->
    <en>0</en>
    <sp>0</sp>
  </en_sp>
  <en_sp>
    <en>1</en>
    <sp>0</sp>
  </en_sp>
</table>

Answer 1

--更新

假设像这样的 xml：

<catalog>
  <cd>
    <entry id="1">
      <word>blue</word>
      <word>beryl</word>
      <word lang="SP">azul</word>
    </entry>
  </cd>
  <cd>
    <entry id="2">
  ...

试试这个：

$super = array();
$url = "original.xml";
if ($xml = @simplexml_load_file($url, 'SimpleXMLElement', LIBXML_NOCDATA)) {
  foreach($xml->cd as $cd) {
     foreach ($cd->entry as $entry) {
      $id = (string)$entry['id'];
        foreach($entry->word as $word) {
            $lang = isset($word['lang']) ? (string)$word['lang'] : 'EN';
            $super[$id][$lang][] = (string)$word;
        }
     }
  }
}

显示使用：

print "<pre>";
print_r($super);
print "</pre>";

注意： 这是另一种方法，基本上是您需要理解的方法使用 xml 对象以及更一般地使用数组，您可以通过创建基于父级 -> 的结构化层次结构来存储数据。孩子;在这种情况下，我创建了一个像这样的数组 $super[$id][$lang][] = (string)$word; 其中 $id 是父级$lang 是 $word 的父级，而 $word 分别是两者的子级；这将产生一个像这样的数组：

Array
(
    [1] => Array
        (
            [EN] => Array
                (
                    [0] => blue
                    [1] => beryl
                )

            [SP] => Array
                (
                    [0] => azul
                )

        )
      ...

其他需要考虑的事情是，

1. 如何获取匹配标签的属性，例如id或lang，在我的示例中我使用了$entry['id'] 但 $cd->entry['id'] 也是有效的。

2. 如何将 xml-dom-object 转换为有效字符串，以便您可以将其重用为数组索引或值，如 (string)$word

---

从我从您的示例中看到的：

<catalog>
  <cd>
    <ent_seq>1</ent_seq>
    <title>For Your Love</title>
    <artist>
    ...

尝试

$super = array();
$url = "original.xml";
if ($xml = @simplexml_load_file($url, 'SimpleXMLElement', LIBXML_NOCDATA)) {
  $xml_array = @json_decode(@json_encode($xml), 1);
  foreach ($xml_array['cd'] as $val) {
  $key = $val['ent_seq'];
    if (is_array($val)) {
      foreach ($val as $k1 => $v1) {
        if (is_array($v1)) {
          switch ($k1) {
            case 'artist':
              foreach ($v1 as $k2 => $v2) {
                if (is_array($v2)) {
                  foreach ($v2 as $v3) {
                    $super[$k2][$key] = $v3;
                  }
                }
                else {
                  $super[$k2][$key] = $v2;
                }
              }
              break;
          }
        }
        else {
          switch ($k1) {
            case 'title':
              $super[$k1][$key] = $v1;
              break;
          }
        }
      }
    }
  }
}

显示迭代的结果像这样的数组：

foreach( $super as $key => $val) {
  echo "<table id='{$key}'>\n";
   foreach($val as $key2 => $val2) {
    echo "<$key id='$key2'> " . $val2." </$key>\n";
    }
    echo "</table>\n";                

}

为了更好地查看数组结构你可以像这样打印它：

print "<pre>";
print_r($super);
print "</pre>";

这将显示一个像这样的数组：

Array
(
    [title] => Array
        (
            [1] => For Your Love
            [2] => Splish Splash
            [3] => How Great Thuo Art
            [4] => Big Willie style
            [5] => Empire Burlesque
        )

    [name] => Array
        (
            [1] => Eric Clapton
            [2] => Roberto Carlos
            [3] => Elvis Presley
            [4] => Will Smith
            [5] => Boby Rockhammer
        )

    [nickname] => Array
        (
            [1] => Slowhand
            [2] => The King
            [3] => The King of Rock 'n Roll
        )

    [band] => Array
        (
            [4] => Will Smith
            [5] => Bob Dylan and Boby Rockhammer
        )

)

注意：正如你所看到的，我使用了switch-case，因为你的xml标签不是始终具有相同的一致性，并且在某些情况下它们具有相似的名称，例如 和 ；您可以创建自己的案例。

然而，就像现在一样，它可以很好地处理您想要获取的字段，如 da 示例中所示。

Answer 2

澄清一下，您是否正在尝试获取输入 XML 文档，使用 XSL/T 将其转换为另一个（不同格式的）XML 文档，然后获取生成的 XML 并将其存储在 MySQL 数据库中？

我是堆栈溢出的新手，所以我不确定如何向原始帖子添加评论。