[31007] | 1 | package org.hathitrust.extractedfeatures;
|
---|
| 2 |
|
---|
[31045] | 3 | import java.io.IOException;
|
---|
[31007] | 4 | import java.util.ArrayList;
|
---|
| 5 | import java.util.Iterator;
|
---|
| 6 | import org.apache.spark.api.java.function.FlatMapFunction;
|
---|
| 7 | import org.apache.spark.util.DoubleAccumulator;
|
---|
| 8 | import org.json.JSONArray;
|
---|
| 9 | import org.json.JSONObject;
|
---|
| 10 |
|
---|
| 11 | /*
|
---|
| 12 | class PagedJSON implements Function<String, Boolean> {
|
---|
| 13 |
|
---|
| 14 | private static final long serialVersionUID = 1L;
|
---|
| 15 |
|
---|
| 16 | public Boolean call(String s) { return s.contains("a"); }
|
---|
| 17 | }
|
---|
| 18 | */
|
---|
| 19 |
|
---|
| 20 |
|
---|
| 21 | class PerPageJSONFlatmap implements FlatMapFunction<String, JSONObject>
|
---|
| 22 | //public class PagedJSON implements VoidFunction<String>
|
---|
| 23 | {
|
---|
| 24 | private static final long serialVersionUID = 1L;
|
---|
| 25 |
|
---|
| 26 | protected String _input_dir;
|
---|
[31225] | 27 | protected String _whitelist_filename;
|
---|
[31220] | 28 |
|
---|
[31007] | 29 | protected String _solr_url;
|
---|
| 30 | protected String _output_dir;
|
---|
[31225] | 31 |
|
---|
[31007] | 32 | protected int _verbosity;
|
---|
| 33 |
|
---|
[31225] | 34 | protected WhitelistBloomFilter _whitelist_bloomfilter;
|
---|
[31375] | 35 | protected UniversalPOSLangMap _universal_langmap = null;
|
---|
[31225] | 36 |
|
---|
[31007] | 37 | protected DoubleAccumulator _progress_accum;
|
---|
| 38 | protected double _progress_step;
|
---|
| 39 |
|
---|
[31252] | 40 | boolean _icu_tokenize;
|
---|
[31045] | 41 | boolean _strict_file_io;
|
---|
| 42 |
|
---|
[31220] | 43 | public PerPageJSONFlatmap(String input_dir, String whitelist_filename,
|
---|
| 44 | String solr_url, String output_dir, int verbosity,
|
---|
[31045] | 45 | DoubleAccumulator progress_accum, double progress_step,
|
---|
[31252] | 46 | boolean icu_tokenize, boolean strict_file_io)
|
---|
[31007] | 47 | {
|
---|
| 48 | _input_dir = input_dir;
|
---|
[31225] | 49 | _whitelist_filename = whitelist_filename;
|
---|
[31220] | 50 |
|
---|
[31007] | 51 | _solr_url = solr_url;
|
---|
| 52 | _output_dir = output_dir;
|
---|
| 53 | _verbosity = verbosity;
|
---|
| 54 |
|
---|
| 55 | _progress_accum = progress_accum;
|
---|
| 56 | _progress_step = progress_step;
|
---|
[31045] | 57 |
|
---|
[31252] | 58 | _icu_tokenize = icu_tokenize;
|
---|
[31045] | 59 | _strict_file_io = strict_file_io;
|
---|
[31225] | 60 |
|
---|
| 61 | _whitelist_bloomfilter = null;
|
---|
[31007] | 62 | }
|
---|
| 63 |
|
---|
[31090] | 64 |
|
---|
[31045] | 65 | public Iterator<JSONObject> call(String json_file_in) throws IOException
|
---|
[31007] | 66 | {
|
---|
[31226] | 67 | if ((_whitelist_filename != null) && (_whitelist_bloomfilter == null)) {
|
---|
[31225] | 68 | _whitelist_bloomfilter = new WhitelistBloomFilter(_whitelist_filename,true);
|
---|
| 69 | }
|
---|
| 70 |
|
---|
[31220] | 71 | String full_json_file_in = _input_dir + "/" + json_file_in;
|
---|
[31045] | 72 | JSONObject extracted_feature_record = JSONClusterFileIO.readJSONFile(full_json_file_in);
|
---|
[31220] | 73 |
|
---|
[31045] | 74 | ArrayList<JSONObject> json_pages = new ArrayList<JSONObject>();
|
---|
[31007] | 75 |
|
---|
[31045] | 76 | if (extracted_feature_record != null) {
|
---|
| 77 | String volume_id = extracted_feature_record.getString("id");
|
---|
| 78 |
|
---|
| 79 | //JSONObject ef_metadata = extracted_feature_record.getJSONObject("metadata");
|
---|
| 80 | //String title= ef_metadata.getString("title");
|
---|
| 81 |
|
---|
| 82 | JSONObject ef_features = extracted_feature_record.getJSONObject("features");
|
---|
| 83 |
|
---|
| 84 |
|
---|
| 85 | int ef_page_count = ef_features.getInt("pageCount");
|
---|
| 86 |
|
---|
| 87 | if (_verbosity >= 1) {
|
---|
| 88 | System.out.println("Processing: " + json_file_in);
|
---|
| 89 | System.out.println(" pageCount = " + ef_page_count);
|
---|
[31030] | 90 | }
|
---|
[31045] | 91 |
|
---|
| 92 | JSONArray ef_pages = ef_features.getJSONArray("pages");
|
---|
| 93 | int ef_num_pages = ef_pages.length();
|
---|
| 94 | if (ef_num_pages != ef_page_count) {
|
---|
| 95 | System.err.println("Warning: number of page elements in JSON (" + ef_num_pages + ")"
|
---|
| 96 | +" does not match 'pageCount' metadata (" + ef_page_count + ")");
|
---|
| 97 | }
|
---|
| 98 |
|
---|
| 99 | // Make directory for page-level JSON output
|
---|
| 100 | String json_dir = ClusterFileIO.removeSuffix(json_file_in,".json.bz2");
|
---|
| 101 | String page_json_dir = json_dir + "/pages";
|
---|
| 102 |
|
---|
[31096] | 103 | if (_output_dir != null) {
|
---|
| 104 | // Only need to do this once per volume, so easier to here than in the per-page Map
|
---|
| 105 | ClusterFileIO.createDirectoryAll(_output_dir + "/" + page_json_dir);
|
---|
| 106 | }
|
---|
[31266] | 107 | if (_verbosity >= 3) {
|
---|
[31045] | 108 | System.out.print(" Pages: ");
|
---|
[31007] | 109 | }
|
---|
[31045] | 110 |
|
---|
| 111 | for (int i = 0; i < ef_page_count; i++) {
|
---|
| 112 | String formatted_i = String.format("page-%06d", i);
|
---|
| 113 | String page_id = volume_id + "." + formatted_i;
|
---|
| 114 |
|
---|
[31266] | 115 | if (_verbosity >= 3) {
|
---|
[31045] | 116 | if (i>0) {
|
---|
| 117 | System.out.print(", ");
|
---|
| 118 | }
|
---|
| 119 | System.out.print(page_id);
|
---|
[31030] | 120 | }
|
---|
[31045] | 121 |
|
---|
| 122 | String output_json_bz2 = page_json_dir +"/" + formatted_i + ".json.bz2";
|
---|
[31007] | 123 |
|
---|
[31045] | 124 | if (i==(ef_page_count-1)) {
|
---|
[31266] | 125 | if (_verbosity >= 3) {
|
---|
[31045] | 126 | System.out.println();
|
---|
| 127 | }
|
---|
[31266] | 128 | if (_verbosity >= 2) {
|
---|
| 129 | System.out.println("Sample output JSON page file: " + output_json_bz2);
|
---|
| 130 | }
|
---|
[31045] | 131 | }
|
---|
| 132 |
|
---|
| 133 | JSONObject ef_page = ef_pages.getJSONObject(i);
|
---|
| 134 |
|
---|
| 135 | if (ef_page != null) {
|
---|
| 136 | // Convert to Solr add form
|
---|
[31221] | 137 | JSONObject solr_add_doc_json
|
---|
[31375] | 138 | = SolrDocJSON.generateSolrDocJSON(volume_id, page_id, ef_page,
|
---|
| 139 | _whitelist_bloomfilter, _universal_langmap,
|
---|
| 140 | _icu_tokenize);
|
---|
[31045] | 141 | solr_add_doc_json.put("filename_json_bz2", output_json_bz2);
|
---|
| 142 |
|
---|
| 143 | json_pages.add(solr_add_doc_json);
|
---|
| 144 |
|
---|
| 145 |
|
---|
| 146 | }
|
---|
| 147 | else {
|
---|
| 148 | System.err.println("Skipping: " + page_id);
|
---|
| 149 | }
|
---|
| 150 |
|
---|
[31007] | 151 | }
|
---|
[31045] | 152 | }
|
---|
| 153 | else {
|
---|
| 154 | // File did not exist, or could not be parsed
|
---|
| 155 | String mess = "Failed to read in bzipped JSON file '" + full_json_file_in + "'";
|
---|
| 156 | if (_strict_file_io) {
|
---|
| 157 | throw new IOException(mess);
|
---|
| 158 | }
|
---|
[31007] | 159 | else {
|
---|
[31045] | 160 | System.err.println("Warning: " + mess);
|
---|
| 161 | System.out.println("Warning: " + mess);
|
---|
[31007] | 162 | }
|
---|
| 163 | }
|
---|
[31045] | 164 |
|
---|
[31007] | 165 | _progress_accum.add(_progress_step);
|
---|
| 166 |
|
---|
| 167 | return json_pages.iterator();
|
---|
| 168 | }
|
---|
| 169 |
|
---|
| 170 |
|
---|
| 171 | }
|
---|
| 172 |
|
---|