Section:

Extract Text, Read, Parse PDF - TextExtract - Java Sample Code

Sample code for using Apryse SDK to read a PDF (parse and extract text), provided in Python, C++, C#, Java, Node.js (JavaScript), PHP, Ruby, Go and VB. If you'd like to search text on PDF pages, see our code sample for text search. Learn more about our Server SDK and PDF Data Extraction SDK Capabilities.

1//---------------------------------------------------------------------------------------
2// Copyright (c) 2001-2024 by Apryse Software Inc. All Rights Reserved.
3// Consult legal.txt regarding legal and license information.
4//---------------------------------------------------------------------------------------
5
6import java.awt.Color;
7import com.pdftron.common.PDFNetException;
8import com.pdftron.pdf.*;
9import java.text.DecimalFormat; 
10
11
12// This sample illustrates the basic text extraction capabilities of PDFNet.
13public class TextExtractTest {
14
15    public static void main(String[] args) {
16        PDFNet.initialize(PDFTronLicense.Key());
17
18        // Relative path to the folder containing test files.
19        String input_path = "../../TestFiles/";
20        // string output_path = "../../TestFiles/Output/";
21        boolean example1_basic = false;
22        boolean example2_xml = false;
23        boolean example3_wordlist = false;
24        boolean example4_advanced = true;
25        boolean example5_low_level = false;
26
27        // Sample code showing how to use high-level text extraction APIs.
28        try (PDFDoc doc = new PDFDoc(input_path + "newsletter.pdf")) {
29            doc.initSecurityHandler();
30
31            Page page = doc.getPage(1);
32            if (page == null) {
33                System.out.println("Page not found.");
34            }
35
36            TextExtractor txt = new TextExtractor();
37            txt.begin(page);  // Read the page.
38            // Other options you may want to consider...
39            // txt.Begin(page, 0, TextExtractor.ProcessingFlags.e_no_dup_remove);
40            // txt.Begin(page, 0, TextExtractor.ProcessingFlags.e_remove_hidden_text);
41            // ...
42
43            // Example 1. Get all text on the page in a single string.
44            // Words will be separated with space or new line characters.
45            if (example1_basic) {
46                // Get the word count.
47                System.out.println("Word Count: " + txt.getWordCount());
48
49                System.out.println("\n\n- GetAsText --------------------------\n" + txt.getAsText());
50                System.out.println("-----------------------------------------------------------");
51            }
52
53            // Example 2. Get XML logical structure for the page.
54            if (example2_xml) {
55                String text = txt.getAsXML(TextExtractor.e_words_as_elements | TextExtractor.e_output_bbox | TextExtractor.e_output_style_info);
56                System.out.println("\n\n- GetAsXML  --------------------------\n" + text);
57                System.out.println("-----------------------------------------------------------");
58            }
59
60            // Example 3. Extract words one by one.
61            if (example3_wordlist) {
62                TextExtractor.Word word;
63                for (TextExtractor.Line line = txt.getFirstLine(); line.isValid(); line = line.getNextLine()) {
64                    for (word = line.getFirstWord(); word.isValid(); word = word.getNextWord()) {
65                        System.out.println(word.getString());
66                    }
67                }
68                System.out.println("-----------------------------------------------------------");
69            }
70
71            // Example 4. A more advanced text extraction example.
72            // The output is XML structure containing paragraphs, lines, words,
73            // as well as style and positioning information.
74            if (example4_advanced) {
75                Rect bbox;
76                int cur_flow_id = -1, cur_para_id = -1;
77
78                TextExtractor.Line line;
79                TextExtractor.Word word;
80                TextExtractor.Style s, line_style;
81
82                System.out.println("<PDFText>");
83                // For each line on the page...
84                for (line = txt.getFirstLine(); line.isValid(); line = line.getNextLine()) {
85                    if (line.getNumWords() == 0)
86                        continue;
87                    if (cur_flow_id != line.getFlowID()) {
88                        if (cur_flow_id != -1) {
89                            if (cur_para_id != -1) {
90                                cur_para_id = -1;
91                                System.out.println("</Para>");
92                            }
93                            System.out.println("</Flow>");
94                        }
95                        cur_flow_id = line.getFlowID();
96                        System.out.println("<Flow id=\"" + cur_flow_id + "\">");
97                    }
98
99                    if (cur_para_id != line.getParagraphID()) {
100                        if (cur_para_id != -1)
101                            System.out.println("</Para>");
102                        cur_para_id = line.getParagraphID();
103                        System.out.println("<Para id=\"" + cur_para_id + "\">");
104                    }
105
106                    bbox = line.getBBox();
107                    line_style = line.getStyle();
108                    System.out.print("<Line box=\"" +  String.format("%.2f, %.2f, %.2f, %.2f", bbox.getX1(), bbox.getY1(), bbox.getX2(), bbox.getY2()) + "\"");
109                    printStyle(line_style);
110                    System.out.println(" cur_num=\"" + line.getCurrentNum() + "\">");
111                 
112
113                    // For each word in the line...
114                    for (word = line.getFirstWord(); word.isValid(); word = word.getNextWord()) {
115                        // Output the bounding box for the word.
116                        bbox = word.getBBox();
117                        System.out.print("<Word box=\"" +  String.format("%.2f, %.2f, %.2f, %.2f", bbox.getX1(), bbox.getY1(), bbox.getX2(), bbox.getY2()) + "\"");
118                        System.out.print(" cur_num=\"" + word.getCurrentNum() + "\"");
119                        int sz = word.getStringLen();
120                        if (sz == 0) continue;
121
122                        // If the word style is different from the parent style, output the new style.
123                        s = word.getStyle();
124                        if (!s.equals(line_style)) {
125                            printStyle(s);
126                        }
127
128                        System.out.print(">" + word.getString());
129                        System.out.println("</Word>");
130                    }
131                    System.out.println("</Line>");
132                }
133
134                if (cur_flow_id != -1) {
135                    if (cur_para_id != -1) {
136                        cur_para_id = -1;
137                        System.out.println("</Para>");
138                    }
139                    System.out.println("</Flow>");
140                }
141            }
142            txt.destroy();
143            System.out.println("</PDFText>");
144        } catch (PDFNetException e) {
145            System.out.println(e);
146        }
147
148        // Sample code showing how to use low-level text extraction APIs.
149        if (example5_low_level) {
150            try (PDFDoc doc = new PDFDoc((input_path + "newsletter.pdf"))) {
151                doc.initSecurityHandler();
152
153                // Example 1. Extract all text content from the document
154
155                ElementReader reader = new ElementReader();
156                //  Read every page
157                for (PageIterator itr = doc.getPageIterator(); itr.hasNext(); ) {
158                    reader.begin(itr.next());
159                    DumpAllText(reader);
160                    reader.end();
161                }
162
163                // Example 2. Extract text content based on the
164                // selection rectangle.
165                System.out.print("\n----------------------------------------------------");
166                System.out.print("\nExtract text based on the selection rectangle.");
167                System.out.println("\n----------------------------------------------------");
168
169                Page first_page = doc.getPageIterator().next();
170                String s1 = ReadTextFromRect(first_page, new Rect(27, 392, 563, 534), reader);
171                System.out.print("\nField 1: " + s1);
172
173                s1 = ReadTextFromRect(first_page, new Rect(28, 551, 106, 623), reader);
174                System.out.print("\nField 2: " + s1);
175
176                s1 = ReadTextFromRect(first_page, new Rect(208, 550, 387, 621), reader);
177                System.out.print("\nField 3: " + s1);
178
179                // ...
180                System.out.println("Done.");
181            } catch (Exception e) {
182                e.printStackTrace();
183            }
184        }
185
186        PDFNet.terminate();
187    }
188
189
190    static void printStyle(TextExtractor.Style s) {
191        Color rgb = s.getColor();
192        String rgb_hex =  String.format("%02X%02X%02X;", rgb.getRed(), rgb.getGreen(), rgb.getBlue() );
193        DecimalFormat df = new DecimalFormat("#.#");
194        System.out.print(" style=\"font-family:" + s.getFontName() + "; "
195                + "font-size:" + df.format(s.getFontSize()) + ";"
196                + (s.isSerif() ? " sans-serif; " : " ")
197                + "color:#" + rgb_hex + "\"");
198    }
199
200    // A utility method used to dump all text content in the console window.
201    static void DumpAllText(ElementReader reader) throws PDFNetException {
202        Element element;
203        while ((element = reader.next()) != null) {
204            switch (element.getType()) {
205                case Element.e_text_begin:
206                    System.out.println("\n--> Text Block Begin");
207                    break;
208                case Element.e_text_end:
209                    System.out.println("\n--> Text Block End");
210                    break;
211                case Element.e_text: {
212                    Rect bbox = element.getBBox();
213                    if (bbox == null) continue;
214                    System.out.println("\n--> BBox: " + bbox.getX1() + ", "
215                            + bbox.getY1() + ", "
216                            + bbox.getX2() + ", "
217                            + bbox.getY2());
218
219                    String arr = element.getTextString();
220                    System.out.println(arr);
221                }
222                break;
223                case Element.e_text_new_line:
224                    System.out.println("\n--> New Line");
225                    break;
226                case Element.e_form:                // Process form XObjects
227                    reader.formBegin();
228                    DumpAllText(reader);
229                    reader.end();
230                    break;
231            }
232        }
233    }
234
235    // A helper method for ReadTextFromRect
236    static String RectTextSearch(ElementReader reader, Rect pos) throws PDFNetException {
237        Element element;
238        String srch_str = new String();
239        while ((element = reader.next()) != null) {
240            switch (element.getType()) {
241                case Element.e_text: {
242                    Rect bbox = element.getBBox();
243                    if (bbox == null) continue;
244                    if (bbox.intersectRect(bbox, pos)) {
245                        String arr = element.getTextString();
246                        srch_str += arr;
247                        srch_str += "\n"; // add a new line?
248                    }
249                    break;
250                }
251                case Element.e_text_new_line: {
252                    break;
253                }
254                case Element.e_form: // Process form XObjects
255                {
256                    reader.formBegin();
257                    srch_str += RectTextSearch(reader, pos);
258                    reader.end();
259                    break;
260                }
261            }
262        }
263        return srch_str;
264    }
265
266    // A utility method used to extract all text content from
267    // a given selection rectangle. The rectangle coordinates are
268    // expressed in PDF user/page coordinate system.
269    static String ReadTextFromRect(Page page, Rect pos, ElementReader reader) throws PDFNetException {
270        reader.begin(page);
271        String srch_str = RectTextSearch(reader, pos);
272        reader.end();
273        return srch_str;
274    }
275}

Did you find this helpful?

Trial setup questions?

Ask experts on Discord

Need other help?

Contact Support

Pricing or product questions?

Contact Sales