fieldTypeAdd.py

#-*- coding: utf-8 -*-
from __future__ import print_function
from __future__ import unicode_literals
from functools import partial
import codecs
import click
from xml.dom.minidom import parse

@click.command()
@click.argument('core')
def main(core):
    solr_path = '/var/searchengine/solr-7.4.0/server/solr/{0}/conf/schema.xml'
    tree = parse(solr_path.format(core))
    
    def line_break(tree, space_width, new_line_count=1):
        break_str = "\n" * new_line_count
        break_str = "{0}{1}".format(break_str, " " * space_width)
        return tree.createTextNode(break_str)
    
    def createElement(tree, field_name, **options):
        node = tree.createElement(field_name)
        for attr_name, attr_value in options.items():
            if attr_name.endswith("_"):
                attr_name = attr_name[:-1]
            node.setAttribute(attr_name, attr_value)
        return node
    
    break_node = partial(line_break, tree)
    make_element = partial(createElement, tree)
    
    # 마지막으로 정의된 fieldType 가져오기
    last_field_type = tree.getElementsByTagName("fieldType")[-1]
    
    # 줄바꿈 태그 만들기
    
    # 새로운 필드 타입 생성
    alpha_num_tag = make_element('fieldType', name='alphaNumeric',
                                 sortMissingLast='false', omitNorms='true',
                                 class_='solr.TextField')
    
    last_field_type.parentNode.insertBefore(alpha_num_tag, last_field_type.nextSibling)
    last_field_type.parentNode.insertBefore(break_node(4, 2), last_field_type.nextSibling)
    
    alpha_num_tag.appendChild(break_node(8))
    analyzer = tree.createElement('analyzer')
    alpha_num_tag.appendChild(analyzer)
    alpha_num_tag.appendChild(break_node(4))
    
    analyzer.appendChild(break_node(12))
    tokenizer = make_element('tokenizer', class_='solr.KeywordTokenizerFactory')
    analyzer.appendChild(tokenizer)
    
    analyzer.appendChild(break_node(12))
    filter_lower = make_element('filter', class_='solr.LowerCaseFilterFactory')
    analyzer.appendChild(filter_lower)
    
    analyzer.appendChild(break_node(12))
    filter_trim = make_element('filter', class_='solr.TrimFilterFactory')
    analyzer.appendChild(filter_trim)
    
    analyzer.appendChild(break_node(12))
    filter_pattern = make_element('filter', class_='solr.PatternReplaceFilterFactory', pattern="(\d+)", replacement="00000$1", replace='all')
    analyzer.appendChild(filter_pattern)
    
    analyzer.appendChild(break_node(12))
    filter_pattern = make_element('filter', class_='solr.PatternReplaceFilterFactory', pattern="0*([0-9]{6,})", replacement="$1", replace='all')
    analyzer.appendChild(filter_pattern)
    
    analyzer.appendChild(break_node(12))
    filter_pattern = make_element('filter', class_='solr.PatternReplaceFilterFactory', pattern="([^a-z0-9])", replacement="", replace='all')
    analyzer.appendChild(filter_pattern)
    
    analyzer.appendChild(break_node(8))
    
    # reg_no는 새로 추가한 fieldType으로 변경해야 함
    # <field indexed="true" name="reg_no" stored="true" type="string"/>
    field_elements = tree.getElementsByTagName('field')
    for element in field_elements:
        if element.getAttribute('name') == 'reg_no':
            element.setAttribute('type', 'alphaNumeric')
    
    tree.writexml(codecs.open(solr_path.format(core), 'w', 'utf-8'))


if __name__ == '__main__':
    main()