1 | package org.hathitrust.extractedfeatures;
|
---|
2 |
|
---|
3 | import java.io.BufferedReader;
|
---|
4 | import java.io.FileReader;
|
---|
5 | import java.io.IOException;
|
---|
6 | import java.io.PrintWriter;
|
---|
7 | import java.util.ArrayList;
|
---|
8 | import java.util.HashMap;
|
---|
9 |
|
---|
10 | import javax.servlet.ServletConfig;
|
---|
11 | import javax.servlet.ServletException;
|
---|
12 | import javax.servlet.annotation.WebServlet;
|
---|
13 | import javax.servlet.http.HttpServlet;
|
---|
14 | import javax.servlet.http.HttpServletRequest;
|
---|
15 | import javax.servlet.http.HttpServletResponse;
|
---|
16 |
|
---|
17 | /**
|
---|
18 | * Servlet implementation class VolumeCheck
|
---|
19 | */
|
---|
20 | @WebServlet("/VolumeCheck")
|
---|
21 | public class VolumeCheck extends HttpServlet {
|
---|
22 | private static final long serialVersionUID = 1L;
|
---|
23 |
|
---|
24 | protected static int HASHMAP_INIT_SIZE = 13800000;
|
---|
25 | protected static String DATA_DIR="D:/cygwin64/home/davidb/research/code-managed/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest";
|
---|
26 | protected static HashMap<String,Boolean> id_check_ = null;
|
---|
27 |
|
---|
28 |
|
---|
29 | public VolumeCheck() {
|
---|
30 | if (id_check_ == null) {
|
---|
31 | id_check_ = new HashMap<String,Boolean>(HASHMAP_INIT_SIZE);
|
---|
32 | storeIDs(DATA_DIR + "/full-listing.txt");
|
---|
33 | }
|
---|
34 | }
|
---|
35 |
|
---|
36 | protected void storeIDs(String filename)
|
---|
37 | {
|
---|
38 | try {
|
---|
39 | FileReader fr = new FileReader(filename);
|
---|
40 | BufferedReader br = new BufferedReader(fr);
|
---|
41 |
|
---|
42 | long line_num = 1;
|
---|
43 | String line;
|
---|
44 | System.err.println("Loading in volume IDS: " + filename);
|
---|
45 |
|
---|
46 | String opt_file_ext = ".json.bz2";
|
---|
47 |
|
---|
48 | while ((line = br.readLine()) != null) {
|
---|
49 |
|
---|
50 | String file_id = line.substring(line.lastIndexOf("/")+1);
|
---|
51 | String id = null;
|
---|
52 | if (file_id.endsWith(opt_file_ext)) {
|
---|
53 | id = file_id.substring(0,file_id.lastIndexOf(opt_file_ext));
|
---|
54 | }
|
---|
55 | else {
|
---|
56 | id = file_id;
|
---|
57 | }
|
---|
58 |
|
---|
59 | id = id.replaceAll("\\+", ":").replaceAll("=", "/");
|
---|
60 |
|
---|
61 | id_check_.put(id, true);
|
---|
62 |
|
---|
63 | if ((line_num % 100000) == 0) {
|
---|
64 | System.err.println("sample id = " + id);
|
---|
65 | System.err.println("Passed line: " + line_num);
|
---|
66 | }
|
---|
67 | line_num++;
|
---|
68 |
|
---|
69 | }
|
---|
70 |
|
---|
71 | br.close();
|
---|
72 | }
|
---|
73 | catch (Exception e) {
|
---|
74 | e.printStackTrace();
|
---|
75 | }
|
---|
76 |
|
---|
77 | }
|
---|
78 | /**
|
---|
79 | * @see Servlet#init(ServletConfig)
|
---|
80 | */
|
---|
81 | public void init(ServletConfig config) throws ServletException {
|
---|
82 | // currently everything we need to happen at start up can be done in the constructor
|
---|
83 | }
|
---|
84 |
|
---|
85 | /**
|
---|
86 | * @see HttpServlet#doGet(HttpServletRequest request, HttpServletResponse response)
|
---|
87 | */
|
---|
88 | protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
|
---|
89 | PrintWriter pw = response.getWriter();
|
---|
90 |
|
---|
91 | String cgi_ids = request.getParameter("ids");
|
---|
92 | String cgi_id = request.getParameter("id");
|
---|
93 |
|
---|
94 | if (cgi_ids != null) {
|
---|
95 | response.setContentType("application/json");
|
---|
96 |
|
---|
97 | String[] ids = cgi_ids.split(",");
|
---|
98 | int ids_len = ids.length;
|
---|
99 |
|
---|
100 | pw.append("{");
|
---|
101 |
|
---|
102 | for (int i=0; i<ids_len; i++) {
|
---|
103 | String id = ids[i];
|
---|
104 |
|
---|
105 | Boolean exists = id_check_.get(id);
|
---|
106 | boolean status = (exists != null) ? exists : false;
|
---|
107 |
|
---|
108 | if (i>0) {
|
---|
109 | pw.append(",");
|
---|
110 | }
|
---|
111 | pw.append("\"" + id + "\":" + status );
|
---|
112 | }
|
---|
113 | pw.append("}");
|
---|
114 |
|
---|
115 | }
|
---|
116 | else if (cgi_id != null) {
|
---|
117 | response.setContentType("application/json");
|
---|
118 |
|
---|
119 | String id = cgi_id;
|
---|
120 | Boolean exists = id_check_.get(id);
|
---|
121 | boolean status = (exists != null) ? exists : false;
|
---|
122 | pw.append("{'" + id + "':" + status + "}");
|
---|
123 | }
|
---|
124 | else {
|
---|
125 |
|
---|
126 | pw.append("General Info: Number of HTRC Volumes in check-list = " + id_check_.size());
|
---|
127 |
|
---|
128 | }
|
---|
129 | //pw.close();
|
---|
130 |
|
---|
131 | }
|
---|
132 |
|
---|
133 | /**
|
---|
134 | * @see HttpServlet#doPost(HttpServletRequest request, HttpServletResponse response)
|
---|
135 | */
|
---|
136 | protected void doPost(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
|
---|
137 | doGet(request, response);
|
---|
138 | }
|
---|
139 |
|
---|
140 | }
|
---|