TextSearch

1//---------------------------------------------------------------------------------------
2// Copyright (c) 2001-2019 by PDFTron Systems Inc. All Rights Reserved.
3// Consult legal.txt regarding legal and license information.
4//---------------------------------------------------------------------------------------
5
6package com.pdftron.android.pdfnetsdksamples.samples;
7
8import com.pdftron.android.pdfnetsdksamples.OutputListener;
9import com.pdftron.android.pdfnetsdksamples.PDFNetSample;
10import com.pdftron.android.pdfnetsdksamples.R;
11import com.pdftron.android.pdfnetsdksamples.util.Utils;
12import com.pdftron.common.PDFNetException;
13import com.pdftron.pdf.Action;
14import com.pdftron.pdf.Highlights;
15import com.pdftron.pdf.PDFDoc;
16import com.pdftron.pdf.Page;
17import com.pdftron.pdf.Rect;
18import com.pdftron.pdf.TextSearch;
19import com.pdftron.pdf.TextSearchResult;
20import com.pdftron.sdf.SDFDoc;
21
22import java.util.ArrayList;
23
24public class TextSearchTest extends PDFNetSample {
25
26	private static OutputListener mOutputListener;
27
28	private static ArrayList<String> mFileList = new ArrayList<>();
29
30    public TextSearchTest() {
31        setTitle(R.string.sample_textsearch_title);
32        setDescription(R.string.sample_textsearch_description);
33    }
34
35	@Override
36	public void run(OutputListener outputListener) {
37		super.run(outputListener);
38		mOutputListener = outputListener;
39		mFileList.clear();
40		printHeader(outputListener);
41
42        try (PDFDoc doc = new PDFDoc(Utils.getAssetTempFile(INPUT_PATH + "credit card numbers.pdf").getAbsolutePath())) {
43            doc.initSecurityHandler();
44
45            TextSearch txt_search = new TextSearch();
46            int mode = TextSearch.e_whole_word | TextSearch.e_page_stop;
47
48            String pattern = "joHn sMiTh";
49
50            //PDFDoc doesn't allow simultaneous access from different threads. If this
51            //document could be used from other threads (e.g., the rendering thread inside
52            //PDFView/PDFViewCtrl, if used), it is good practice to lock it.
53            //Notice: don't forget to call doc.Unlock() to avoid deadlock.
54            doc.lock();
55
56            //call Begin() method to initialize the text search.
57            txt_search.begin(doc, pattern, mode, -1, -1);
58
59            int step = 0;
60
61            //call Run() method iteratively to find all matching instances.
62            while (true) {
63                TextSearchResult result = txt_search.run();
64
65                if (result.getCode() == TextSearchResult.e_found) {
66                    if (step == 0) {
67                        //step 0: found "John Smith"
68                        //note that, here, 'ambient_string' and 'hlts' are not written to,
69                        //as 'e_ambient_string' and 'e_highlight' are not set.
70                        mOutputListener.println(result.getResultStr() + "'s credit card number is: ");
71
72                        //now switch to using regular expressions to find John's credit card number
73                        mode = txt_search.getMode();
74                        mode |= TextSearch.e_reg_expression | TextSearch.e_highlight;
75                        txt_search.setMode(mode);
76                        String new_pattern = "\\d{4}-\\d{4}-\\d{4}-\\d{4}"; //or "(\\d{4}-){3}\\d{4}"
77                        txt_search.setPattern(new_pattern);
78
79                        step = step + 1;
80                    } else if (step == 1) {
81                        //step 1: found John's credit card number
82                        mOutputListener.println("  " + result.getResultStr());
83
84                        //note that, here, 'hlts' is written to, as 'e_highlight' has been set.
85                        //output the highlight info of the credit card number
86                        Highlights hlts = result.getHighlights();
87                        hlts.begin(doc);
88                        while (hlts.hasNext()) {
89                            mOutputListener.println("The current highlight is from page: " + hlts.getCurrentPageNumber());
90                            hlts.next();
91                        }
92
93                        //see if there is an AMEX card number
94                        String new_pattern = "\\d{4}-\\d{6}-\\d{5}";
95                        txt_search.setPattern(new_pattern);
96
97                        step = step + 1;
98                    } else if (step == 2) {
99                        //found an AMEX card number
100                        mOutputListener.println("\nThere is an AMEX card number:");
101                        mOutputListener.println("  " + result.getResultStr());
102
103                        //change mode to find the owner of the credit card; supposedly, the owner's
104                        //name proceeds the number
105                        mode = txt_search.getMode();
106                        mode |= TextSearch.e_search_up;
107                        txt_search.setMode(mode);
108                        String new_pattern = "[A-z]++ [A-z]++";
109                        txt_search.setPattern(new_pattern);
110
111                        step = step + 1;
112                    } else if (step == 3) {
113                        //found the owner's name of the AMEX card
114                        mOutputListener.println("Is the owner's name:");
115                        mOutputListener.println("  " + result.getResultStr() + "?");
116
117                        //add a link annotation based on the location of the found instance
118                        Highlights hlts = result.getHighlights();
119                        hlts.begin(doc);
120                        while (hlts.hasNext()) {
121                            Page cur_page = doc.getPage(hlts.getCurrentPageNumber());
122                            double[] q = hlts.getCurrentQuads();
123                            int quad_count = q.length / 8;
124                            for (int i = 0; i < quad_count; ++i) {
125                                //assume each quad is an axis-aligned rectangle
126                                int offset = 8 * i;
127                                double x1 = Math.min(Math.min(Math.min(q[offset + 0], q[offset + 2]), q[offset + 4]), q[offset + 6]);
128                                double x2 = Math.max(Math.max(Math.max(q[offset + 0], q[offset + 2]), q[offset + 4]), q[offset + 6]);
129                                double y1 = Math.min(Math.min(Math.min(q[offset + 1], q[offset + 3]), q[offset + 5]), q[offset + 7]);
130                                double y2 = Math.max(Math.max(Math.max(q[offset + 1], q[offset + 3]), q[offset + 5]), q[offset + 7]);
131                                com.pdftron.pdf.annots.Link hyper_link = com.pdftron.pdf.annots.Link.create(doc, new Rect(x1, y1, x2, y2), Action.createURI(doc, "http://www.pdftron.com"));
132                                cur_page.annotPushBack(hyper_link);
133                            }
134                            hlts.next();
135                        }
136                        doc.save(Utils.createExternalFile("credit card numbers_linked.pdf", mFileList).getAbsolutePath(), SDFDoc.SaveMode.LINEARIZED, null);
137                        break;
138                    }
139                } else if (result.getCode() == TextSearchResult.e_page) {
140                    //you can update your UI here, if needed
141                } else {
142                    break;
143                }
144            }
145
146            doc.unlock();
147        } catch (PDFNetException e) {
148            mOutputListener.printError(e.getStackTrace());
149        }
150
151		for (String file : mFileList) {
152			addToFileList(file);
153		}
154		printFooter(outputListener);
155	}
156
157}