source: other-projects/hathitrust/solr-extracted-features/trunk/src/main/java/org/hathitrust/extractedfeatures/ProcessForSolrIngest.java@ 31001

package org.hathitrust.extractedfeatures;

import java.io.Serializable;
import org.apache.commons.cli.*;

import org.apache.spark.api.java.*;
import org.apache.spark.util.DoubleAccumulator;
import org.hathitrust.extractedfeatures.PagedJSON;
import org.json.JSONObject;
import org.apache.spark.SparkConf;

public class ProcessForSolrIngest implements Serializable
{
    private static final long serialVersionUID = 1L;

    // Following details on number of partitions to use given in 
    //  "Parallelized collections" section of:
    //   https://spark.apache.org/docs/2.0.1/programming-guide.html
    //
    // For a more detailed discussion see:
    //   http://blog.cloudera.com/blog/2015/03/how-to-tune-your-apache-spark-jobs-part-2/
    
    public static final int NUM_CORES = 6;
    public static final int NUM_PARTITIONS = 2*NUM_CORES; // default would appear to be 2
    
    protected String _input_dir;
    protected String _json_list_filename;
    protected String _solr_url;
    protected String _output_dir;
    
    protected int    _verbosity;

    public ProcessForSolrIngest(String input_dir, String json_list_filename, 
                                String solr_url, String output_dir, int verbosity)
    {
        _input_dir = input_dir;
        _json_list_filename = (json_list_filename != null) ? json_list_filename : input_dir;

        _solr_url   = solr_url;
        _output_dir = output_dir;
        _verbosity  = verbosity;
    }

    public void execPerVolume()
    {   
        String spark_app_name = "[Per Volume] Extract Features: Process for Solr Ingest";
        spark_app_name += " [" + _json_list_filename + "]";

        SparkConf conf = new SparkConf().setAppName(spark_app_name);
        JavaSparkContext jsc = new JavaSparkContext(conf);
        
        if (_verbosity >= 2) {
            System.out.println("Default Minimum Partions: " + jsc.defaultMinPartitions());
            System.out.println("Default Parallelism: " + jsc.defaultParallelism());
        }
                
        JavaRDD<String> json_list_data = jsc.textFile(_json_list_filename,NUM_PARTITIONS).cache();

        long num_volumes = json_list_data.count();
        double per_vol = 100.0/(double)num_volumes;
        
        DoubleAccumulator progress_accum = jsc.sc().doubleAccumulator("Progress Percent");
        
        PerVolumeJSON per_vol_json = new PerVolumeJSON(_input_dir,_solr_url,_output_dir,_verbosity, progress_accum,per_vol);

        json_list_data.foreach(per_vol_json);
        
        long num_ids = num_volumes;
        
        System.out.println("");
        System.out.println("############");
        System.out.println("# Number of volume ids: " + num_ids);
        System.out.println("############");
        System.out.println("");

        jsc.close();
    }

    public void execPerPage()
    {   
        String spark_app_name = "[Per Page] Extract Features: Process for Solr Ingest";
        spark_app_name += " [" + _json_list_filename + "]";

        SparkConf conf = new SparkConf().setAppName(spark_app_name);
        JavaSparkContext jsc = new JavaSparkContext(conf);
        
        if (_verbosity >= 2) {
            System.out.println("Default Minimum Partions: " + jsc.defaultMinPartitions());
            System.out.println("Default Parallelism: " + jsc.defaultParallelism());
        }
                
        JavaRDD<String> json_list_data = jsc.textFile(_json_list_filename,NUM_PARTITIONS).cache();

        long num_volumes = json_list_data.count();
        double per_vol = 100.0/(double)num_volumes;
        
        DoubleAccumulator progress_accum = jsc.sc().doubleAccumulator("Progress Percent");
        
        PagedJSON paged_json = new PagedJSON(_input_dir,_solr_url,_output_dir,_verbosity, progress_accum,per_vol);
        JavaRDD<JSONObject> json_per_page_ids = json_list_data.flatMap(paged_json).cache();
        
        json_per_page_ids.foreach(paged_json);

/*
        System.out.println("");
        System.out.println("############");
        System.out.println("# Progress Accumulator: " + progress_accum.value());
        System.out.println("############");
        System.out.println("");
*/
        
        long num_page_ids = json_per_page_ids.count();
        
        System.out.println("");
        System.out.println("############");
        System.out.println("# Number of page ids: " + num_page_ids);
        System.out.println("############");
        System.out.println("");

        /*
        if (_output_dir != null) {
            String rdd_save_file = "rdd-solr-json-page-files";
            json_ids.saveAsTextFile(rdd_save_file);
            System.out.println("############");
            System.out.println("# Saved RDD of Solr JSON page files, top-level, as:");
            System.out.println("#  " + rdd_save_file);
            System.out.println("############");
            System.out.println("");
        }
        */
        
        jsc.close();
    }

    
    

    public static void print_usage(HelpFormatter formatter, Options options)
    {
        formatter.printHelp("RUN.bash [options] input-dir json-filelist.txt", options);
    }
    
    public static void main(String[] args) {
        Options options = new Options();

        Option verbosity_opt = new Option("v", "verbosity", true, 
                "Set to control the level of debugging output [0=none, 1=some, 2=lots]");
        verbosity_opt.setRequired(false);
        options.addOption(verbosity_opt);
        
        Option output_dir_opt = new Option("o", "output-dir", true, 
                "If specified, save BZipped Solr JSON files to this directory");
        output_dir_opt.setRequired(false);
        options.addOption(output_dir_opt);
        
        Option solr_url_opt = new Option("u", "solr-url", true, 
                "If specified, the URL to post the Solr JSON data to");
        solr_url_opt.setRequired(false);
        options.addOption(solr_url_opt);
        
        Option read_only_opt = new Option("r", "read-only", false, 
                "Used to initiate a run where the files are all read in, but nothing is ingested/saved");
        read_only_opt.setRequired(false);
        options.addOption(read_only_opt);
        
        // Need to work with CLI v1.2 as this is the JAR that is bundled with Hadoop/Spark 
        CommandLineParser parser = new GnuParser(); 
        //CommandLineParser parser = new DefaultParser(); // if working with CLI v1.3 and above
        
        HelpFormatter formatter = new HelpFormatter();
        CommandLine cmd = null;

        try {
            cmd = parser.parse(options, args);
        }
        catch (ParseException e) {
            System.err.println(e.getMessage());
            print_usage(formatter,options);
            System.exit(1);
        }

        
        String verbosity_str = cmd.getOptionValue("verbosity","0");
        int verbosity = Integer.parseInt(verbosity_str);

        String output_dir = cmd.getOptionValue("output-dir",null);
        String solr_url   = cmd.getOptionValue("solr-url",null);
        boolean read_only   = cmd.hasOption("read-only");
        
        String[] filtered_args = cmd.getArgs();

        if (filtered_args.length != 2) {
            print_usage(formatter,options);
            System.exit(1);
        }
        
        if (!read_only && ((output_dir == null) && (solr_url==null))) {
            System.err.println("Need to specify either --solr-url or --output-dir otherwise generated files are not ingested/saved");
            print_usage(formatter,options);
            System.exit(1);
        }
        if (read_only) {
            // For this case, need to ensure solr-url and output-dir are null
            output_dir = null;
            solr_url = null;
        }
        
        String input_dir  = filtered_args[0];
        String json_list_filename = filtered_args[1];
        
        ProcessForSolrIngest prep_for_ingest 
            = new ProcessForSolrIngest(input_dir,json_list_filename,solr_url,output_dir,verbosity);
        prep_for_ingest.execPerVolume();
    }
}