// Licensed to Elasticsearch B.V. under one or more contributor // license agreements. See the NOTICE file distributed with // this work for additional information regarding copyright // ownership. Elasticsearch B.V. licenses this file to you under // the Apache License, Version 2.0 (the "License"); you may // not use this file except in compliance with the License. // You may obtain a copy of the License at // // http://www.apache.org/licenses/LICENSE-2.0 // // Unless required by applicable law or agreed to in writing, // software distributed under the License is distributed on an // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY // KIND, either express or implied. See the License for the // specific language governing permissions and limitations // under the License. // // Code generated from specification version 7.17.10: DO NOT EDIT package esapi import ( "context" "io" "net/http" "strconv" "strings" ) func newMLFlushJobFunc(t Transport) MLFlushJob { return func(job_id string, o ...func(*MLFlushJobRequest)) (*Response, error) { var r = MLFlushJobRequest{JobID: job_id} for _, f := range o { f(&r) } return r.Do(r.ctx, t) } } // ----- API Definition ------------------------------------------------------- // MLFlushJob - Forces any buffered data to be processed by the job. // // See full documentation at https://www.elastic.co/guide/en/elasticsearch/reference/current/ml-flush-job.html. type MLFlushJob func(job_id string, o ...func(*MLFlushJobRequest)) (*Response, error) // MLFlushJobRequest configures the ML Flush Job API request. type MLFlushJobRequest struct { Body io.Reader JobID string AdvanceTime string CalcInterim *bool End string SkipTime string Start string Pretty bool Human bool ErrorTrace bool FilterPath []string Header http.Header ctx context.Context } // Do executes the request and returns response or error. func (r MLFlushJobRequest) Do(ctx context.Context, transport Transport) (*Response, error) { var ( method string path strings.Builder params map[string]string ) method = "POST" path.Grow(1 + len("_ml") + 1 + len("anomaly_detectors") + 1 + len(r.JobID) + 1 + len("_flush")) path.WriteString("/") path.WriteString("_ml") path.WriteString("/") path.WriteString("anomaly_detectors") path.WriteString("/") path.WriteString(r.JobID) path.WriteString("/") path.WriteString("_flush") params = make(map[string]string) if r.AdvanceTime != "" { params["advance_time"] = r.AdvanceTime } if r.CalcInterim != nil { params["calc_interim"] = strconv.FormatBool(*r.CalcInterim) } if r.End != "" { params["end"] = r.End } if r.SkipTime != "" { params["skip_time"] = r.SkipTime } if r.Start != "" { params["start"] = r.Start } if r.Pretty { params["pretty"] = "true" } if r.Human { params["human"] = "true" } if r.ErrorTrace { params["error_trace"] = "true" } if len(r.FilterPath) > 0 { params["filter_path"] = strings.Join(r.FilterPath, ",") } req, err := newRequest(method, path.String(), r.Body) if err != nil { return nil, err } if len(params) > 0 { q := req.URL.Query() for k, v := range params { q.Set(k, v) } req.URL.RawQuery = q.Encode() } if len(r.Header) > 0 { if len(req.Header) == 0 { req.Header = r.Header } else { for k, vv := range r.Header { for _, v := range vv { req.Header.Add(k, v) } } } } if r.Body != nil && req.Header.Get(headerContentType) == "" { req.Header[headerContentType] = headerContentTypeJSON } if ctx != nil { req = req.WithContext(ctx) } res, err := transport.Perform(req) if err != nil { return nil, err } response := Response{ StatusCode: res.StatusCode, Body: res.Body, Header: res.Header, } return &response, nil } // WithContext sets the request context. func (f MLFlushJob) WithContext(v context.Context) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.ctx = v } } // WithBody - Flush parameters. func (f MLFlushJob) WithBody(v io.Reader) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.Body = v } } // WithAdvanceTime - advances time to the given value generating results and updating the model for the advanced interval. func (f MLFlushJob) WithAdvanceTime(v string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.AdvanceTime = v } } // WithCalcInterim - calculates interim results for the most recent bucket or all buckets within the latency period. func (f MLFlushJob) WithCalcInterim(v bool) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.CalcInterim = &v } } // WithEnd - when used in conjunction with calc_interim, specifies the range of buckets on which to calculate interim results. func (f MLFlushJob) WithEnd(v string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.End = v } } // WithSkipTime - skips time to the given value without generating results or updating the model for the skipped interval. func (f MLFlushJob) WithSkipTime(v string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.SkipTime = v } } // WithStart - when used in conjunction with calc_interim, specifies the range of buckets on which to calculate interim results. func (f MLFlushJob) WithStart(v string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.Start = v } } // WithPretty makes the response body pretty-printed. func (f MLFlushJob) WithPretty() func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.Pretty = true } } // WithHuman makes statistical values human-readable. func (f MLFlushJob) WithHuman() func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.Human = true } } // WithErrorTrace includes the stack trace for errors in the response body. func (f MLFlushJob) WithErrorTrace() func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.ErrorTrace = true } } // WithFilterPath filters the properties of the response body. func (f MLFlushJob) WithFilterPath(v ...string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { r.FilterPath = v } } // WithHeader adds the headers to the HTTP request. func (f MLFlushJob) WithHeader(h map[string]string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { if r.Header == nil { r.Header = make(http.Header) } for k, v := range h { r.Header.Add(k, v) } } } // WithOpaqueID adds the X-Opaque-Id header to the HTTP request. func (f MLFlushJob) WithOpaqueID(s string) func(*MLFlushJobRequest) { return func(r *MLFlushJobRequest) { if r.Header == nil { r.Header = make(http.Header) } r.Header.Set("X-Opaque-Id", s) } }