不同类型Mapper的Django弹性搜索DSL TransportError

Question

我正在使用Python3.6、Django(2.1)、ElasticSearch(5.1.1)和Elasticsearch-dsl(5.4.0)进行一个项目，其中我需要实现搜索功能。

以下是我尝试过的：

：models.py:

class searchdatamodel(models.Model):
    id = models.IntegerField(null=False, primary_key=True)
    company_name = models.TextField(blank=True, null=True)
    city = models.TextField(blank=True, null=True)
    state = models.TextField(blank=True, null=True)
    zip_codes = models.TextField(blank=True, null=True)
    street_address = models.TextField(blank=True, null=True)
    street_address_zip = models.TextField(blank=True, null=True)
    county = models.TextField(blank=True, null=True)
    phone_number = models.DecimalField(max_digits=65535, decimal_places=65535, blank=True, null=True)
    fax_number = models.DecimalField(max_digits=65535, decimal_places=65535, blank=True, null=True)
    web_address = models.TextField(blank=True, null=True)
    last_name = models.TextField(blank=True, null=True)
    first_name = models.TextField(blank=True, null=True)
    contact_title = models.TextField(blank=True, null=True)
    contact_gender = models.TextField(blank=True, null=True)
    actual_employee_size = models.IntegerField(blank=True, null=True)
    employee_size_range = models.TextField(blank=True, null=True)
    actual_sales_volume = models.IntegerField(blank=True, null=True)
    sales_volume_range = models.TextField(blank=True, null=True)
    primary_sic = models.IntegerField(blank=True, null=True)
    primary_sic_description = models.TextField(blank=True, null=True)
    secondary_sic_1 = models.IntegerField(blank=True, null=True)
    secondary_sic_description_1 = models.TextField(blank=True, null=True)
    secondary_sic_2 = models.IntegerField(blank=True, null=True)
    secondary_sic_description_2 = models.TextField(blank=True, null=True)
    credit_alpha_score = models.TextField(blank=True, null=True)
    credit_numeric_score = models.IntegerField(blank=True, null=True)
    headquarters_branch = models.TextField(blank=True, null=True)
    square_footage = models.TextField(blank=True, null=True)
    registry_date = models.IntegerField(blank=True, null=True)

    class Meta:
        managed = False
        db_table = 'searchdatamodel'

    # Implement indexing for SearchDataModel model
    def indexing(self):
        SearchDataIndex.init()
        obj = SearchDataIndex(
            id=self.id,
            company_name=self.company_name,
            city=self.city,
            state=self.state,
            zip_code=self.zip_codes,
            street_address=self.street_address,
            street_address_zip=self.street_address_zip,
            county=self.county,
            phone_number=self.phone_number,
            fax_number=self.fax_number,
            web_address=self.web_address,
            last_name=self.last_name,
            first_name=self.first_name,
            contact_title=self.contact_title,
            contact_gender=self.contact_gender,
            actual_employee_size=self.actual_employee_size,
            actual_sales_volume=self.actual_sales_volume,
            primary_sic=self.primary_sic,
            primary_sic_description=self.primary_sic_description,
            registry_date=self.registry_date
        )
        obj.save()
        return obj.to_dict(include_meta=True)

：serach.py:

class SearchDataIndex(DocType):
    id = Integer()
    company_name = Text()
    city = Text()
    state = Text()
    zip_codes = Text()
    street_address = Text()
    street_address_zip = Integer()
    county = Text()
    phone_number = Text()
    fax_number = Text()
    web_address = Text()
    last_name = Text()
    first_name = Text()
    contact_title = Text()
    contact_gender = Text()
    actual_employee_size = Integer()
    actual_sales_volume = Text()
    primary_sic = Text()
    primary_sic_description = Text()
    registry_date = Date()

    class Meta:
        index = 'data-search'

    # A method for bulk indexing
    def bulk_indexing():
        SearchDataIndex.init()
        es = Elasticsearch()
        bulk(client=es, actions=(b.indexing() for b in
                        models.searchdatamodel.objects.all().iterator()))

当我尝试运行bulk_indexing函数时，它返回一个错误，如下所示：

文件"/Users/abdul/PycharmProjects/Dmitry/DVirEnv/lib/python3.6/site-packages/elasticsearch/connection/base.py"，第125行，在_raise_error raise HTTP_EXCEPTIONS.get(status_code，TransportError)(status_code，error_message，additional_info)中 elasticsearch.exceptions.RequestError: TransportError(400，“非法_参数_异常”，“不同类型的映射器zip_codes，current_type整数，merged_type文本”)

我尝试过在模型中将zip_code类型更改为IntegerField，但没有成功。

这里有什么问题吗？

提前感谢！

Answer 1

不幸的是，你不能就这样改变类型。您必须重新映射索引，它只是简单地删除索引，然后再次运行映射函数。

若要删除索引，请运行

es = Elasticsearch()
es.indices.delete(index='data-search')

然后通过再次运行大容量索引器来映射索引。

注意:任何现有的数据都将维护旧的数据类型，所以如果您需要保存数据但更改数据类型，那么这是一个完全不同的蠕虫，包括创建一个新版本的索引，并将旧数据读入其中。