<dependency>
      <groupId>org.apache.solr</groupId>
      <artifactId>solr-solrj</artifactId>
      <version>8.11.2</version>
  </dependency>

[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  1.192 s
[INFO] Finished at: 2021-05-16T22:33:55+09:00
[INFO] ------------------------------------------------------------------------

package org.yasundial.crawler.app;

import java.io.IOException;
import java.util.Map;

import org.apache.solr.client.solrj.SolrClient;
import org.apache.solr.client.solrj.SolrServerException;
import org.apache.solr.client.solrj.impl.HttpSolrClient;
import org.apache.solr.common.SolrInputDocument;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class MySolr {
    private static final MySolr INSTANCE = new MySolr();

    private SolrClient client;

    private MyConfig config = MyConfig.getInstance();
    private Logger logger = LoggerFactory.getLogger(getClass());

    private MySolr() {
        client = new HttpSolrClient.Builder(config.getString("MYC_SOLR_URL")).build();
    }

    public static MySolr getInstance() {
        return INSTANCE;
    }

    public void addDocument(Map<String, Object> document) {
        try {
            SolrInputDocument solrDocument = new SolrInputDocument();

            document.entrySet().forEach(entry -> solrDocument.addField(entry.getKey(), entry.getValue()));

            client.add(solrDocument);
        } catch (IOException | SolrServerException e) {
            logger.info("add Document error.", e);
        }
    }

    public void commit() {
        try {
            client.commit();
        } catch (IOException | SolrServerException e) {
            logger.info("commit error.", e);
        }
    }

    public void close() {
        try {
            client.close();
        } catch (IOException e) {
            logger.info("close error.", e);
        }
    }
}

        Document jsoupDoc = Jsoup.parse(html);
        System.out.println("title: " + jsoupDoc.title());
        System.out.println(" text: " + jsoupDoc.select("body").text());

        Map<String,Object> document = new LinkedHashMap<>();
        document.put("id", url);
        document.put("content", jsoupDoc.select("body").text());
        MySolr.getInstance().addDocument(document);
        MySolr.getInstance().commit();
    }
}

    private MySolr() {
        client = new HttpSolrClient.Builder(config.getString("MYC_SOLR_URL")).build();
    }

{
  "responseHeader":{
    "status":0,
    "QTime":233,
    "params":{
      "q":"content:*モデル*"}},
  "response":{"numFound":1,"start":0,"numFoundExact":true,"docs":[
      {
        "id":"https://www.u-aizu.ac.jp/~yasu-abe/ja/activity/",
        "content":["YasuhiroABE プロフィール 活動概要 学内のみ Home 活動概要  .... "],
        "_version_":1699964908869779456}]
}}

require 'parslet'

class QueryParser < Parslet::Parser
  rule(:space) { match('\s').repeat(1) }
  rule(:type1) { str('-').maybe >> match('[^"\s]').repeat(1) }
  rule(:type2) { str('-').maybe >> str('"') >> match('[^"]').repeat(1) >> str('"') }
  rule(:query) { (type1.as(:t1) | type2.as(:t1) | space).repeat(1) }
  root(:query)
end

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: solr
spec:
  replicas: 1
  selector:
    matchLabels:
      app: solr
  template:
    metadata:
      labels:
        app: solr
    spec:
      containers:
      - name: solr
        image: solr:8.11

---
apiVersion: v1
kind: Service
metadata:
  name: solr
  labels:
    app: solr
spec:
  type: ClusterIP
  selector:
    app: solr
  ports:
  - port: 8983
    targetPort: 8983

---
apiVersion: batch/v1
kind: CronJob
metadata:
  name: solr-crawler
spec:
  schedule: "10 9 * * *"
  concurrencyPolicy: Forbid
  startingDeadlineSeconds: 30
  successfulJobsHistoryLimit: 5
  failedJobsHistoryLimit: 5
  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: Never
          containers:
          - name: solr-crawler
            image: inovtst9.u-aizu.ac.jp/library/solr-crawler:1.1.0
            env:
            - name: LC_CTYPE
              value: ja_JP.UTF-8
            - name: TARGET_URL
              value: "https://u-aizu.ac.jp/~yasu-abe/ja/"
            - name: VISIT_URL_PATTERN
              value: "^https://u-aizu.ac.jp/~yasu-abe/.+$"
            - name: OK_FILTER
              value: ".*(\\.(text|txt|html|htm|yaml|yml|csv|json))$|.*/$"
            - name: MYC_SOLR_URL
              value: "http://solr:8983/solr/testcore"

## deploying for production

FROM docker.io/library/eclipse-temurin:17-jre-alpine

RUN apk update && \
    apk add --no-cache tzdata bash ca-certificates

RUN mkdir /jobs
WORKDIR /jobs
COPY target/crawlerdemo-1.0-SNAPSHOT.jar /jobs/
COPY run.sh /jobs/
RUN chmod +x /jobs/run.sh

ENV TARGET_URL="https://example.com/~user01/"
ENV VISIT_URL_PATTERN="^https://example.com/~user01/.+$"
ENV CRAWLER4J_STORAGE_DIR="data/crawl/strage"
ENV OK_FILTER=".*(\\.(text|txt|html|htm|yaml|yml|csv|json))$|.*/$"
ENV MYC_SOLR_URL="http://example.com:8983/solr/test"

RUN mkdir /jobs/data
RUN addgroup crawler
RUN adduser -S -G crawler crawler
RUN chown crawler /jobs/data
RUN chgrp -R crawler /jobs
RUN chmod -R g+rwX /jobs
USER crawler

ENTRYPOINT ["./run.sh"]

#!/bin/bash

exec java -jar crawlerdemo-1.0-SNAPSHOT.jar