XML文件解析：如何开始

我需要一些解析xml文件的帮助。这是我第一次做这种工作，我希望能得到一些建议或帮助。我有一个像这样的大文件：

<Response success="true" start_row="0" num_rows="100" total_rows="100">
<ncbi-genes>
    <ncbi-gene>
        <acronym>Accn1</acronym>
        <alias-tags>BNC1 BNaC1 ACIC2 ASIC2 Mdeg BNaC1a</alias-tags>
        <data-sets>
            <data-set>
                <blue-channel nil="true"/>
                <delegate type="boolean">true</delegate>
                <specimen>
                    <chemotherapy nil="true"/>
                    <donor-id type="integer">9456</donor-id>
                    <donor>
                        <age-id type="integer">1</age-id>
                        <condition-description>TS26</condition-description>
                        <age>
                            <age-group-id type="integer">1</age-group-id>
                            <days type="float">18.5</days>
                        </age>
                    </donor>
                </specimen>
                <differential-expression-rankings type="array">
                    <differential-expression-ranking>
                        <structure>
                            <acronym>PPH</acronym>
                            <name>prepontine hindbrain</name>
                        </structure>
                    </differential-expression-ranking>
                    <differential-expression-ranking>
                        <structure>
                            <acronym>p3</acronym>
                            <name>prosomere 3</name>
                        </structure>
                    </differential-expression-ranking>
                </differential-expression-rankings>
            </data-set>
            <data-set>
              (...same fields as before...)
            </data-set>
        </data-sets>
   </ncbi-gene>
</ncbi-genes>

这个循环不起作用。。我想是因为ncbi基因的内部。我已将此字段更改为NCBIGENES，现在错误为：

Not a HASH reference at xml_parser.pl line 19.
HASH(0x29d7ca0)

调用哈希时出现问题。。。 正如我所说的，我对这种数据是新手，这是我第一次使用xml模块。因此，任何关于自我定位的建议都将不胜感激

---

提前感谢。

这里是一个使用解析的快速示例；LibXML使您能够轻松访问，这是一种XML查询语言，允许您根据节点的标记名、值、属性和/或与其他节点的关系来选择节点集。使用XPath可以很容易地挑出y节点下的所有x节点，或具有属性z且子节点id为w的所有x节点，或类似的复杂查询

use strict;
use warnings;
use feature qw(say);
use Data::Dumper;
use XML::LibXML;

my $tree = XML::LibXML->load_xml( IO => \*DATA );

## make sure that we have some genes!
die "Could not find any genes!" if ! $tree->exists('//ncbi-gene');

# for every 'ncbi-gene' node:
for my $gene ( $tree->findnodes('//ncbi-gene') ) {
    my %data;
    # is there an acronym as direct child of the node?
    $data{acronym} = $gene->findvalue('acronym') if $gene->exists('acronym');

    # find the donor age in days using the path specified
    # to get the value of each node, run to_literal on it
    $data{donor_age_days} = [ map { $_->to_literal }
        $gene->findnodes('data-sets/data-set/specimen/donor/age/days') ];

    # find all the 'name' nodes under a 'structure' node that is a descendant of $gene
    $data{structures} = [ map { $_->to_literal }
        $gene->findnodes( 'descendant::structure/name', $gene ) ];

    # this will find any 'name' node under a structure node anywhere in the tree
    $data{all_structures} = [ map { $_->to_literal } 
        $gene->findnodes('//structure/name') ];

    # an example of using findvalue on a query that returns an array: only the
    # first value is returned.
    $data{acronyms_str} = [ $gene->findvalue('//structure/acronym') ];

    say Dumper( \%data );
}

__DATA__
<Response success="true" start_row="0" num_rows="100" total_rows="100">
<ncbi-genes>
    <ncbi-gene>
        <acronym>Accn1</acronym>
        <alias-tags>BNC1 BNaC1 ACIC2 ASIC2 Mdeg BNaC1a</alias-tags>
        <data-sets>
            <data-set>
                <blue-channel nil="true"/>
                <delegate type="boolean">true</delegate>
                <specimen>
                    <chemotherapy nil="true"/>
                    <donor-id type="integer">9456</donor-id>
                    <donor>
                        <age-id type="integer">1</age-id>
                        <condition-description>TS26</condition-description>
                        <age>
                            <age-group-id type="integer">1</age-group-id>
                            <days type="float">18.5</days>
                        </age>
                    </donor>
                </specimen>
                <differential-expression-rankings type="array">
                    <differential-expression-ranking>
                        <structure>
                            <acronym>PPH</acronym>
                            <name>prepontine hindbrain</name>
                        </structure>
                    </differential-expression-ranking>
                    <differential-expression-ranking>
                        <structure>
                            <acronym>p3</acronym>
                            <name>prosomere 3</name>
                        </structure>
                    </differential-expression-ranking>
                </differential-expression-rankings>
            </data-set>
            <data-set>
              (...same fields as before...)
            </data-set>
        </data-sets>
   </ncbi-gene>
   <ncbi-favourite-places>
      <structure>
         <name>Eiffel Tower</name>
      </structure>
   </ncbi-favourite-places>
</ncbi-genes>
</Response>

有一些很好的XPath教程，在浏览XML文档时应该很方便。请注意，基于XML:：libxml的libxml只实现了XPath1.0

下面是为每个数据集节点收集数据的快速示例：

---

此脚本没有26行，哪一行是错误？您阅读了中此模块部分的状态了吗？旁注：使用引号，$genelist->{'ncbi-gene'}，以避免重命名元素。@cucurbit对用户非常友好-您可能会发现更好地开始使用XML:：LibXML.maging是没有意义的。它可以表示1Mb到100Gb之间的任何内容。这有很大的区别，它可能会影响你如何去做。效果非常好，非常感谢，现在，我将试着理解所有的行：有什么方法可以把和年龄相关的结构分组吗？我的意思是，我想知道每个年龄段都包括哪些结构。。也许将年龄保存为键并构造该键的数组？是的-您可能希望依次遍历每个数据集，以确保将年龄与结构相关联。我将在答案中添加一个示例。

use strict;
use warnings;
use feature qw(say);
use Data::Dumper;
use XML::LibXML;

my $tree = XML::LibXML->load_xml( IO => \*DATA );

## make sure that we have some genes!
die "Could not find any genes!" if ! $tree->exists('//ncbi-gene');

# for every 'ncbi-gene' node:
for my $gene ( $tree->findnodes('//ncbi-gene') ) {
    my %data;
    # is there an acronym as direct child of the node?
    $data{acronym} = $gene->findvalue('acronym') if $gene->exists('acronym');

    # find the donor age in days using the path specified
    # to get the value of each node, run to_literal on it
    $data{donor_age_days} = [ map { $_->to_literal }
        $gene->findnodes('data-sets/data-set/specimen/donor/age/days') ];

    # find all the 'name' nodes under a 'structure' node that is a descendant of $gene
    $data{structures} = [ map { $_->to_literal }
        $gene->findnodes( 'descendant::structure/name', $gene ) ];

    # this will find any 'name' node under a structure node anywhere in the tree
    $data{all_structures} = [ map { $_->to_literal } 
        $gene->findnodes('//structure/name') ];

    # an example of using findvalue on a query that returns an array: only the
    # first value is returned.
    $data{acronyms_str} = [ $gene->findvalue('//structure/acronym') ];

    say Dumper( \%data );
}

__DATA__
<Response success="true" start_row="0" num_rows="100" total_rows="100">
<ncbi-genes>
    <ncbi-gene>
        <acronym>Accn1</acronym>
        <alias-tags>BNC1 BNaC1 ACIC2 ASIC2 Mdeg BNaC1a</alias-tags>
        <data-sets>
            <data-set>
                <blue-channel nil="true"/>
                <delegate type="boolean">true</delegate>
                <specimen>
                    <chemotherapy nil="true"/>
                    <donor-id type="integer">9456</donor-id>
                    <donor>
                        <age-id type="integer">1</age-id>
                        <condition-description>TS26</condition-description>
                        <age>
                            <age-group-id type="integer">1</age-group-id>
                            <days type="float">18.5</days>
                        </age>
                    </donor>
                </specimen>
                <differential-expression-rankings type="array">
                    <differential-expression-ranking>
                        <structure>
                            <acronym>PPH</acronym>
                            <name>prepontine hindbrain</name>
                        </structure>
                    </differential-expression-ranking>
                    <differential-expression-ranking>
                        <structure>
                            <acronym>p3</acronym>
                            <name>prosomere 3</name>
                        </structure>
                    </differential-expression-ranking>
                </differential-expression-rankings>
            </data-set>
            <data-set>
              (...same fields as before...)
            </data-set>
        </data-sets>
   </ncbi-gene>
   <ncbi-favourite-places>
      <structure>
         <name>Eiffel Tower</name>
      </structure>
   </ncbi-favourite-places>
</ncbi-genes>
</Response>

$VAR1 = {
  'acronym' => 'Accn1',
  'donor_age_days' => [
    '18.5'
  ],
  'structures' => [
    'prepontine hindbrain',
    'prosomere 3'
  ],
  'acronyms_str' => [
    'PPHp3'
  ],
  'all_structures' => [
    'prepontine hindbrain',
    'prosomere 3',
    'Eiffel Tower'
  ]
};

for my $gene ( $tree->findnodes('//ncbi-gene') ) {
    my $data;
    for my $ds ( $gene->findnodes('data-sets/data-set')) {

        # get the age in days -- assumes there is only one age per <data-set>
        my $age = $ds->findvalue('specimen/donor/age/days');

        # get the structures associated with that age
        my @structures = map { $_->to_literal } 
                     $ds->findnodes('descendant::structure/name');
        # you can now save them however you like--e.g.
        push @{$data->{$age}}, @structures;
    }
}