Package gov.nysenate.openleg.util

Examples of gov.nysenate.openleg.util.TranscriptLine

29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
        StringBuffer fullText = new StringBuffer();
        StringBuffer fullTextProcessed = new StringBuffer();

        BufferedReader reader = new BufferedReader(new InputStreamReader(new FileInputStream(file), "latin1"));
        String lineText;
        TranscriptLine line;
        String date = null;
        String time = null;
        boolean firstPageParsed = false;
        boolean firstLineParsed = false;
        boolean skipFirstThreeLines = false;
        int numSkipped = 0;

        while ((lineText = reader.readLine()) != null) {
            line = new TranscriptLine(lineText);

            if (!firstPageParsed) {

                // Handle transcripts with 3 incorrect lines at start of transcript.
                if (!firstLineParsed) {
                    if (lineText.contains("SESSION")) {
                        skipFirstThreeLines = true;
                        numSkipped = 1;
                        continue;
                    }
                }
                if (skipFirstThreeLines == true && numSkipped <= 3) {
                    numSkipped++;
                    continue;
                }

                if (line.isLocation())
                    transcript.setLocation(line.removeLineNumber().trim());

                if (line.isDate())
                    date = line.getDateString();

                if (line.isTime())
                    time = line.getTimeString();

                if (line.isSession())
                    transcript.setType(line.removeLineNumber().trim());

                if (transcript.getLocation() != null && date != null && time != null && transcript.getType() != null)
                    firstPageParsed = true;
            }

            firstLineParsed = true;

            fullText.append(line.fullText()).append("\n");

            if (line.removeLineNumber().trim().length() > 0) {
                fullTextProcessed.append(line.removeLineNumber().trim()).append("\n");
            }
        }

        reader.close();
View Full Code Here
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
     */
    private TranscriptPage parseWithManualSpacing(List<String> pageLines) {
        TranscriptPage page = new TranscriptPage();
        int lineCount = 0;
        for (int i = 0; i < pageLines.size(); i++) {
            TranscriptLine line = new TranscriptLine(pageLines.get(i));

            if (line.isTranscriptNumber()) {
                page.setTranscriptNumber(line.removeInvalidCharacters());
                lineCount++;
            }
            else if (!line.isEmpty() && !line.isStenographer()) {
                page.addLine(line);
                lineCount++;

                if (line.fullText().trim().equals("NEW YORK STATE SENATE")) {
                    addBlankLines(page, 2);
                    lineCount += 2;
                }

                else if (line.fullText().trim().contains("STENOGRAPHIC RECORD")) {
                    addBlankLines(page, 2);
                    lineCount += 2;
                }

                else if (line.isTime()) {
                    addBlankLines(page, 2);
                    lineCount += 2;
                }

                else if (line.isSession()) {
                    addBlankLines(page, 3);
                    lineCount += 3;
                }
            }
        }
View Full Code Here
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    private TranscriptPage parseWithOriginalSpacing(List<String> pageLines) {
        TranscriptPage page = new TranscriptPage();
        int lineCount = 0;

        for (String pageLine : pageLines) {
            TranscriptLine line = new TranscriptLine(pageLine);

            if (line.isTranscriptNumber()) {
                page.setTranscriptNumber(line.removeInvalidCharacters());
                lineCount++;
            }
            else if (!line.isEmpty() && !line.isStenographer()) {
                page.addLine(line);
                lineCount++;
            }
        }
View Full Code Here
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
    private void fixErrorsOnFirstPage(List<List<String>> pages) {
        List<String> correctedFirstPage = new ArrayList<String>();
        List<String> firstPage = pages.get(0);

        for (int i = 0; i < firstPage.size(); i++) {
            TranscriptLine line = new TranscriptLine(firstPage.get(i));

            if (!line.isEmpty()) {
                if (line.fullText().endsWith(",") || line.fullText().endsWith(", Acting")) {
                    // Combine two lines into one; corrects formatting. i.e. 123096.v1
                    TranscriptLine nextLine = getNextLine(firstPage, i);
                    if (nextLine.fullText().trim().equals("President") || nextLine.fullText().trim().equals("Acting President")) {
                        line = new TranscriptLine(line.fullText() + " " + nextLine.fullText().trim());
                        // Skip next line since we combined it with the previous line.
                        i++;
                    }
                }
View Full Code Here
119
120
121
122
123
124
125
126
127
        pages.set(pages.indexOf(firstPage), correctedFirstPage);
    }

    private void addBlankLines(TranscriptPage page, int numLines) {
        for (int i = 0; i < numLines; i++) {
            TranscriptLine blankLine = new TranscriptLine(" ");
            page.addLine(blankLine);
        }
    }
View Full Code Here
126
127
128
129
130
131
132
133
134
135
136
137
        }
    }

    private boolean pageHasLineNumbers(List<String> pageLines) {
        for (String pageLine : pageLines) {
            TranscriptLine line = new TranscriptLine(pageLine);
            if (!line.isEmpty() && !line.isTranscriptNumber()) {
                return line.hasLineNumber();
            }
        }
        return false;
    }
View Full Code Here
141
142
143
144
145
146
147
148
149
        return rawPages.indexOf(pageLines) == 0;
    }

    private TranscriptLine getNextLine(List<String> pageLines, int i) {
        if (i + 1 < pageLines.size()) {
            return new TranscriptLine(pageLines.get(i + 1));
        }
        return null;
    }
View Full Code Here
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
    private List<List<String>> splitPages(String transcriptText) {
        List<String> page = new ArrayList<String>();
        List<List<String>> pages = new ArrayList<List<String>>();

        String[] line = transcriptText.split("\n");
        TranscriptLine nextLine;
        for (int i = 0; i < line.length; i++) {
            page.add(line[i]);

            // Ignore the first transcript number.
            if (i > 10) {
                if (i + 1 < line.length) {
                    nextLine = new TranscriptLine(line[i + 1]);

                    if (nextLine.isTranscriptNumber()) {
                        pages.add(page);
                        page = new ArrayList<String>();
                    }
                }
            }
View Full Code Here
TOP

Related Classes of gov.nysenate.openleg.util.TranscriptLine

  • gov.nysenate.openleg.converter.pdf.TranscriptPageParser
  • gov.nysenate.openleg.processors.TranscriptProcessor
  • java.text.SimpleDateFormat
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.