JSON 到文本，特殊序列

Question

我想从以下 JSON 文档创建一个文本文档：

[
    {
        "id": 12345,
        "url": "https://www.w3schools.com",
        "person": {
            "firstname": "John",
            "lastname": "Doe"
        },
        "department": "IT"
    },
    {
        "id": 12346,
        "url": "https://www.w3schools.com",
        "person": {
            "firstname": "Anna",
            "lastname": "Jackson"
        },
        "department": "LOG"
    }
]

我的 JSON 架构如下所示：

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "id": { "type": "integer" },
            "url": { "type": "string", "format": "uri" },
            "person": {
                "type": "object",
                "properties": {
                    "firstname": { "type": "string" },
                    "lastname": { "type": "string" }
                }
            },
            "department": { "enum": [ "IT", "LOG"] }
        }
    }
}

文本文档的结构应如下：

pid:        12345
dep-abb:    IT
surname:    Doe
name:       John

pid:        12346
dep-abb:    LOG
surname:    Jackson
name:       Anna

我是 Perl 和 JSON 新手，正在寻找可以通过扩展模式（例如 txt_seq_no 和 txt_label）来处理这种方法的 Perl 库。文本文件中的标签应按 txt_seq_no ASC 排序并按 txt_label 重命名。有没有可能这么简单地解决这个问题？然后架构可能如下所示：

{
    "type": "array",
    "items": {
        "type": "object",
        "properties": {
            "id": { "type": "integer", "txt_seq_no"=10, "txt_label"="pid" },
            "url": { "type": "string", "format": "uri" },
            "person": {
                "type": "object",
                "properties": {
                    "firstname": { "type": "string", "txt_seq_no"=40, "txt_label"="name" },
                    "lastname": { "type": "string", "txt_seq_no"=30, "txt_label"="surname" }
                }
            },
            "department": { "enum": [ "IT", "LOG", "PROD" ], "txt_seq_no"=20, "txt_label"="dep-abb" }
        }
    }
}

Answer 1

基本方法是使用 JSON 模块将 JSON 解码为 Perl 数据结构，然后遍历该结构以构建您想要的输出。

#!/usr/bin/perl

use strict;
use warnings;
use feature 'say';

use JSON;

my $json = JSON->new;

my $json_str = '[
    {
        "id": 12345,
        "url": "https://www.w3schools.com",
        "person": {
            "firstname": "John",
            "lastname": "Doe"
        },
        "department": "IT"
    },
    {
        "id": 12346,
        "url": "https://www.w3schools.com",
        "person": {
            "firstname": "Anna",
            "lastname": "Jackson"
        },
        "department": "LOG"
    }
]';

my $data = $json->decode($json_str);

for (@$data) {
  say "pid:     $_->{id}";
  say "dep-abb: $_->{department}";
  say "surname: $_->{person}{lastname}";
  say "name:    $_->{person}{firstname}\n";
}

Answer 2

感谢大家的提示和支持！我是 Perl 的初学者，欢迎提供反馈！（我省略了函数参数检查以保存 space）

使用的包：

use Data::Traverse;
use Data::Path;

子变换

# key identifier[0]: mark JSON key that write into text file
# key identifier[1]: optional in JSON schema: new label in text file
my @KEY_IDENTIFIER = ("txt_seq_no", "txt_label");
# keys in JSON SCHEMA, can not be found in JSON document
my @DELETE_FROM_PATH = ( "items", "properties" );

# transform a JSON document into text file
# use identifier in JSON schema to define keys to transform to text
sub transform {
my ( $doc, $schema, $key_identifier) = @_;
if( not defined $key_identifier ) { $key_identifier = \@KEY_IDENTIFIER; }

my $transformator = get_transformator4schema( $schema, $key_identifier, $log );

my $data = get_data4transformator( $doc, $transformator, $log );

print_text( $data, $transformator, $key_identifier, $log );

}

sub get_transformator4schema

sub get_transformator4schema {
my ( $schema, $key_identifier) = @_;
my $t= Data::Traverse->new( $schema);

my %transformator;

$t->traverse(
    sub {
        if( $t->item_key eq @{$key_identifier}[0] or $t->item_key eq @{$key_identifier}[1]) {
            my $path = $t->path;
            my $key;

            # delete path that only exists in schema, not in doc
            foreach my $del_path (@DELETE_FROM_PATH) {
                $path =~ s/\/$del_path//g;
            }

            # delete last element from path
            if( $t->item_key eq @{$key_identifier}[0]) {
                $key = @{$key_identifier}[0];
                $path =~ s/\/@{$key_identifier}[0]$//g;
            }
            if( $t->item_key eq @{$key_identifier}[1]) {
                $key = @{$key_identifier}[1];
                $path =~ s/\/@{$key_identifier}[1]$//g;
            }

            # add key => { key => value }
            if( not exists $transformator{$path} ) {
                my %key_val = ( $key => $_ );
                $transformator{$path} = \%key_val;
            } else {
                my $nested_hash = $transformator{$path};
                $nested_hash->{$key} = $_;
            }
        }
    }
);
return \%transformator;
}

sub get_data4transformator

# fetch needed data from document
sub get_data4transformator {
my ( $document, $transformator, $log ) = @_;

my @template;

foreach my $doc (@{$document}) {
    my $doc_path= Data::Path->new( $doc );

    my %th = ();

    # load values for key = path
    foreach my $path (keys %{$transformator}) {
        my $val = $doc_path->get($path);
        # add value from doc to hash
        $th{$path} = $val;
    }
    push @template, \%th;
}
return \@template;
}

sub print_text:

sub print_text {

my ( $data, $transformator, $key_ids, $log ) = @_;

# sort by 1st item of $key_ids
my $sort_by = @{$key_ids}[0];
my $opt_name = @{$key_ids}[1];
print "\nsort data by '$sort_by' ASC\n";
my @sorted_keys = sort {
    $transformator->{$a}->{$sort_by}
    <=>
    $transformator->{$b}->{$sort_by}
} keys %{$transformator};

foreach my $tdoc ( @{$data} ) {
    print "\nread document item:\n";
    foreach my $key ( @sorted_keys ) {
        my $name = $key;
        $name =~ s/^.*\/(\w+)$//g;
        my $alias_name = "";
        if( defined $transformator->{$key}->{$opt_name} ) {
            $alias_name = $transformator->{$key}->{$opt_name};
        }

        print   " $name => $tdoc->{$key}";
        if( not $alias_name eq "" ) {
            print " alias '$alias_name'\n";
        } else {
            print "\n";
        }
    }
}
}

此功能允许控制哪个键进入文本文件及其位置。至少有一个参数 txt_seq_no。第二个参数 txt_label 是可选的，将替代标签存储在文本文件中。

输出：

 id => 12345 alias 'pid'
 department => IT alias 'dep-abb'
 lastname => Doe
 firstname => John alias 'name'

 id => 12346 alias 'pid'
 department => LOG alias 'dep-abb'
 lastname => Jackson
 firstname => Anna alias 'name'

 id => 12347 alias 'pid'
 department => PROD alias 'dep-abb'
 lastname => Smith
 firstname => Peter alias 'name'

JSON 到文本，特殊序列

JSON to text, special sequence

json

perl

text

string-conversion