Changeset 31271 for other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/PerVolumePOSStreamFlatmap.java
- Timestamp:
- 2016-12-28T14:04:19+13:00 (7 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/PerVolumePOSStreamFlatmap.java
r31258 r31271 41 41 JSONObject extracted_feature_record = JSONClusterFileIO.readJSONFile(full_json_file_in); 42 42 43 ArrayList<String> all_ word_list = new ArrayList<String>();43 ArrayList<String> all_pos_list = new ArrayList<String>(); 44 44 45 45 if (extracted_feature_record != null) { … … 62 62 } 63 63 64 if (_verbosity >= 2) {64 if (_verbosity >= 3) { 65 65 System.out.print(" Pages: "); 66 66 } … … 70 70 String page_id = volume_id + "." + formatted_i; 71 71 72 if (_verbosity >= 2) {72 if (_verbosity >= 3) { 73 73 if (i>0) { 74 74 System.out.print(", "); … … 78 78 79 79 if (i==(ef_page_count-1)) { 80 if (_verbosity >= 2) {80 if (_verbosity >= 3) { 81 81 System.out.println(); 82 82 } … … 87 87 if (ef_page != null) { 88 88 89 ArrayList<String> page_ word_list = SolrDocJSON.generateTokenPosCountPOSLabels(volume_id, page_id, ef_page);90 all_ word_list.addAll(page_word_list);89 ArrayList<String> page_pos_list = SolrDocJSON.generateTokenPosCountPOSLabels(volume_id, page_id, ef_page); 90 all_pos_list.addAll(page_pos_list); 91 91 } 92 92 else { … … 109 109 _progress_accum.add(_progress_step); 110 110 111 return all_ word_list.iterator();111 return all_pos_list.iterator(); 112 112 } 113 113
Note:
See TracChangeset
for help on using the changeset viewer.