据我所知,不幸的是,Huma 不支持这样的数组查询过滤器:filters[]=filter1&filters[]=filter2(也不保留括号,例如filter=filter1&filter=filter2)。我遇到了这个 Github 问题,它给出了一个用逗号分隔过滤器的示例 https://github.com/danielgtaylor/huma/issues/325,所以这就是我们最终要做的:filters=postcode:eq:RM7(EX,创建时间:gt:2024-01-01
与主体参数不同,主体参数可以简单地指定为结构,然后在文档中对其进行验证和生成,过滤器的文档和验证必须单独完成。
文档可以简单地添加到 Huma.Param 对象的描述属性下(在操作下):
Parameters: []*huma.Param{{ Name: "filters", In: "query", Description: "Filter properties by various fields. Separate filters by comma.\n\n" + "Format: field:operator:value\n\n" + "Supported fields:\n" + "- postcode (operator: eq)\n" + "- created (operators: gt, lt, gte, lte)\n", Schema: &huma.Schema{ Type: "string", Items: &huma.Schema{ Type: "string", Pattern: "^[a-zA-Z_]+:(eq|neq|gt|lt|gte|lte):[a-zA-Z0-9-:.]+$", }, Examples: []any{ "postcode:eq:RM7 8EX", "created:gt:2024-01-01", }, }, Required: false, }},
我们现在可以定义 PropertyFilterParams 结构进行验证:
type FilterParam struct { Field string Operator string Value interface{} } type PropertyFilterParams struct { Items []FilterParam } func (s *PropertyFilterParams) UnmarshalText(text []byte) error { equalityFields := []string{"postcode"} greaterSmallerFields := []string{} dateFields := []string{"created"} for _, item := range strings.Split(string(text), ",") { filterParam, err := parseAndValidateFilterItem(item, equalityFields, greaterSmallerFields, dateFields) if err != nil { return err } s.Items = append(s.Items, filterParam) } return nil } func (s *PropertyFilterParams) Schema(registry huma.Registry) *huma.Schema { return &huma.Schema{ Type: huma.TypeString, } } func parseAndValidateFilterItem(item string, equalityFields []string, greaterSmallerFields []string, dateFields []string) (FilterParam, error) { parts := strings.SplitN(item, ":", 3) field := parts[0] operator := parts[1] value := parts[2] if contains(equalityFields, field) { if operator != "eq" && operator != "neq" { return FilterParam{}, fmt.Errorf("Unsupported operator %s for field %s. Only 'eq' and 'neq' are supported.", operator, field) } } else if contains(greaterSmallerFields, field) { if !validation.IsValidCompareGreaterSmallerOperator(operator) { return FilterParam{}, fmt.Errorf("Unsupported operator %s for field %s. Supported operators: eq, neq, gt, lt, gte, lte.", operator, field) } } else if contains(dateFields, field) { if !validation.IsValidCompareGreaterSmallerOperator(operator) { return FilterParam{}, fmt.Errorf("Unsupported operator %s for field %s. Supported operators: eq, neq, gt, lt, gte, lte.", operator, field) } if !validation.IsValidDate(value) { return FilterParam{}, fmt.Errorf("Invalid date format: %s. Expected: YYYY-MM-DD", value) } } else { return FilterParam{}, fmt.Errorf("Unsupported filter field: %s", field) } return FilterParam{Field: field, Operator: operator, Value: value}, nil }
我将 PropertyFilterParams 添加到 PropertyQueryParams 结构中:
type PropertyQueryParams struct { PaginationParams Filter PropertyFilterParams `query:"filters" doc:"Filter properties by various fields"` Sort PropertySortParams `query:"sorts" doc:"Sort properties by various fields"` }
这就是将 PropertyQueryParams 添加到路由的样子(请注意,操作代码本身,包括过滤器描述,位于 getAllPropertyOperation 下 - 我没有粘贴完整的代码,但希望您能理解它的要点) 。如果验证失败,它将抛出 422 响应。我还添加了如何循环遍历通过的过滤器值:
huma.Register(api, getAllPropertyOperation(schema, "get-properties", "/properties", []string{"Properties"}), func(ctx context.Context, input *struct { models.Headers models.PropertyQueryParams }) (*models.MultiplePropertyOutput, error) { for _, filter := range input.Filter.Items { fmt.Println(filter) } return mockMultiplePropertyResponse(), err }) }
我希望这对某人有帮助。如果您找到更好的解决方案,请在评论中告诉我。
以上是在 Go Huma 中添加过滤查询参数的详细内容。更多信息请关注PHP中文网其他相关文章!