com.itextpdf.io.codec.Jbig2SegmentReader Maven / Gradle / Ivy
/*
This file is part of the iText (R) project.
Copyright (c) 1998-2023 Apryse Group NV
Authors: Apryse Software.
This program is offered under a commercial and under the AGPL license.
For commercial licensing, contact us at https://itextpdf.com/sales. For AGPL licensing, see below.
AGPL licensing:
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see segments = new TreeMap<>();
private final Map pages = new TreeMap<>();
private final Set globals = new TreeSet<>();
private RandomAccessFileOrArray ra;
private boolean sequential;
private boolean number_of_pages_known;
private int number_of_pages = -1;
private boolean read = false;
/**
* Inner class that holds information about a JBIG2 segment.
*/
public static class Jbig2Segment implements Comparable {
public final int segmentNumber;
public long dataLength = -1;
public int page = -1;
public int[] referredToSegmentNumbers = null;
public boolean[] segmentRetentionFlags = null;
public int type = -1;
public boolean deferredNonRetain = false;
public int countOfReferredToSegments = -1;
public byte[] data = null;
public byte[] headerData = null;
public boolean page_association_size = false;
public int page_association_offset = -1;
public Jbig2Segment(int segment_number) {
this.segmentNumber = segment_number;
}
public int compareTo(Jbig2Segment s) {
return this.segmentNumber - s.segmentNumber;
}
}
/**
* Inner class that holds information about a JBIG2 page.
*/
public static class Jbig2Page {
public final int page;
private final Jbig2SegmentReader sr;
private final Map segs = new TreeMap<>();
public int pageBitmapWidth = -1;
public int pageBitmapHeight = -1;
public Jbig2Page(int page, Jbig2SegmentReader sr) {
this.page = page;
this.sr = sr;
}
/**
* return as a single byte array the header-data for each segment in segment number
* order, EMBEDDED organization, but I am putting the needed segments in SEQUENTIAL organization.
* if for_embedding, skip the segment types that are known to be not for acrobat.
*
* @param for_embedding True if the bytes represents embedded data, false otherwise
* @throws java.io.IOException if an I/O error occurs.
* @return a byte array
*/
public byte[] getData(boolean for_embedding) throws java.io.IOException {
ByteArrayOutputStream os = new ByteArrayOutputStream();
for (int sn : segs.keySet()) {
Jbig2Segment s = segs.get(sn);
// pdf reference 1.4, section 3.3.6 Jbig2Decode Filter
// D.3 Embedded organisation
if (for_embedding &&
(s.type == END_OF_FILE || s.type == END_OF_PAGE)) {
continue;
}
if (for_embedding) {
// change the page association to page 1
byte[] headerData_emb = copyByteArray(s.headerData);
if (s.page_association_size) {
headerData_emb[s.page_association_offset] = 0x0;
headerData_emb[s.page_association_offset + 1] = 0x0;
headerData_emb[s.page_association_offset + 2] = 0x0;
headerData_emb[s.page_association_offset + 3] = 0x1;
} else {
headerData_emb[s.page_association_offset] = 0x1;
}
os.write(headerData_emb);
} else {
os.write(s.headerData);
}
os.write(s.data);
}
os.close();
return os.toByteArray();
}
public void addSegment(Jbig2Segment s) {
segs.put(s.segmentNumber, s);
}
}
public Jbig2SegmentReader(RandomAccessFileOrArray ra) {
this.ra = ra;
}
public static byte[] copyByteArray(byte[] b) {
byte[] bc = new byte[b.length];
System.arraycopy(b, 0, bc, 0, b.length);
return bc;
}
public void read() throws java.io.IOException {
if (this.read) {
throw new IllegalStateException("already.attempted.a.read.on.this.jbig2.file");
}
this.read = true;
readFileHeader();
// Annex D
if (this.sequential) {
// D.1
do {
Jbig2Segment tmp = readHeader();
readSegment(tmp);
segments.put(tmp.segmentNumber, tmp);
} while (this.ra.getPosition() < this.ra.length());
} else {
// D.2
Jbig2Segment tmp;
do {
tmp = readHeader();
segments.put(tmp.segmentNumber, tmp);
} while (tmp.type != END_OF_FILE);
for (int integer : segments.keySet()) {
readSegment(segments.get(integer));
}
}
}
void readSegment(Jbig2Segment s) throws java.io.IOException {
int ptr = (int) ra.getPosition();
//TODO DEVSIX-6406 7.2.7 not supported
if (s.dataLength == 0xffffffffl) {
return;
}
byte[] data = new byte[(int) s.dataLength];
ra.read(data);
s.data = data;
if (s.type == PAGE_INFORMATION) {
int last = (int) ra.getPosition();
ra.seek(ptr);
int page_bitmap_width = ra.readInt();
int page_bitmap_height = ra.readInt();
ra.seek(last);
Jbig2Page p = pages.get(s.page);
if (p == null) {
throw new IOException("Referring to widht or height of a page we haven't seen yet: {0}").setMessageParams(s.page);
}
p.pageBitmapWidth = page_bitmap_width;
p.pageBitmapHeight = page_bitmap_height;
}
}
Jbig2Segment readHeader() throws java.io.IOException {
int ptr = (int) ra.getPosition();
// 7.2.1
int segment_number = ra.readInt();
Jbig2Segment s = new Jbig2Segment(segment_number);
// 7.2.3
int segment_header_flags = ra.read();
boolean deferred_non_retain = (segment_header_flags & 0x80) == 0x80;
s.deferredNonRetain = deferred_non_retain;
boolean page_association_size = (segment_header_flags & 0x40) == 0x40;
int segment_type = segment_header_flags & 0x3f;
s.type = segment_type;
//7.2.4
int referred_to_byte0 = ra.read();
int count_of_referred_to_segments = (referred_to_byte0 & 0xE0) >> 5;
int[] referred_to_segment_numbers = null;
boolean[] segment_retention_flags = null;
if (count_of_referred_to_segments == 7) {
// at least five bytes
ra.seek(ra.getPosition() - 1);
count_of_referred_to_segments = ra.readInt() & 0x1fffffff;
segment_retention_flags = new boolean[count_of_referred_to_segments + 1];
int i = 0;
int referred_to_current_byte = 0;
do {
int j = i % 8;
if (j == 0) {
referred_to_current_byte = ra.read();
}
segment_retention_flags[i] = (0x1 << j & referred_to_current_byte) >> j == 0x1;
i++;
} while (i <= count_of_referred_to_segments);
} else if (count_of_referred_to_segments <= 4) {
// only one byte
segment_retention_flags = new boolean[count_of_referred_to_segments + 1];
referred_to_byte0 &= 0x1f;
for (int i = 0; i <= count_of_referred_to_segments; i++) {
segment_retention_flags[i] = (0x1 << i & referred_to_byte0) >> i == 0x1;
}
} else if (count_of_referred_to_segments == 5 || count_of_referred_to_segments == 6) {
throw new IOException("Count of referred-to segments has forbidden value in the header for segment {0} starting at {1}")
.setMessageParams(segment_number, ptr);
}
s.segmentRetentionFlags = segment_retention_flags;
s.countOfReferredToSegments = count_of_referred_to_segments;
// 7.2.5
referred_to_segment_numbers = new int[count_of_referred_to_segments + 1];
for (int i = 1; i <= count_of_referred_to_segments; i++) {
if (segment_number <= 256) {
referred_to_segment_numbers[i] = ra.read();
} else if (segment_number <= 65536) {
referred_to_segment_numbers[i] = ra.readUnsignedShort();
} else {
referred_to_segment_numbers[i] = (int) ra.readUnsignedInt();
}
}
s.referredToSegmentNumbers = referred_to_segment_numbers;
// 7.2.6
int segment_page_association;
int page_association_offset = (int) ra.getPosition() - ptr;
if (page_association_size) {
segment_page_association = ra.readInt();
} else {
segment_page_association = ra.read();
}
if (segment_page_association < 0) {
throw new IOException("Page {0} is invalid for segment {1} starting at {2}")
.setMessageParams(segment_page_association, segment_number, ptr);
}
s.page = segment_page_association;
// so we can change the page association at embedding time.
s.page_association_size = page_association_size;
s.page_association_offset = page_association_offset;
if (segment_page_association > 0 && !pages.containsKey(segment_page_association)) {
pages.put(segment_page_association, new Jbig2Page(segment_page_association, this));
}
if (segment_page_association > 0) {
pages.get(segment_page_association).addSegment(s);
} else {
globals.add(s);
}
// 7.2.7
long segment_data_length = ra.readUnsignedInt();
//TODO DEVSIX-6406 the 0xffffffff value that might be here, and how to understand those afflicted segments
s.dataLength = segment_data_length;
int end_ptr = (int) ra.getPosition();
ra.seek(ptr);
byte[] header_data = new byte[end_ptr - ptr];
ra.read(header_data);
s.headerData = header_data;
return s;
}
void readFileHeader() throws java.io.IOException {
ra.seek(0);
byte[] idstring = new byte[8];
ra.read(idstring);
byte[] refidstring = {(byte) 0x97, 0x4A, 0x42, 0x32, 0x0D, 0x0A, 0x1A, 0x0A};
for (int i = 0; i < idstring.length; i++) {
if (idstring[i] != refidstring[i]) {
throw new IOException("File header idstring is not good at byte {0}").setMessageParams(i);
}
}
int fileheaderflags = ra.read();
this.sequential = (fileheaderflags & 0x1) == 0x1;
this.number_of_pages_known = (fileheaderflags & 0x2) == 0x0;
if ((fileheaderflags & 0xfc) != 0x0) {
throw new IOException("File header flags bits from 2 to 7 should be 0, some not");
}
if (this.number_of_pages_known) {
this.number_of_pages = ra.readInt();
}
}
public int numberOfPages() {
return pages.size();
}
public int getPageHeight(int i) {
return pages.get(i).pageBitmapHeight;
}
public int getPageWidth(int i) {
return pages.get(i).pageBitmapWidth;
}
public Jbig2Page getPage(int page) {
return pages.get(page);
}
public byte[] getGlobal(boolean for_embedding){
ByteArrayOutputStream os = new ByteArrayOutputStream();
byte[] streamBytes = null;
try {
for (Object element : globals) {
Jbig2Segment s = (Jbig2Segment) element;
if (for_embedding &&
(s.type == END_OF_FILE || s.type == END_OF_PAGE)) {
continue;
}
os.write(s.headerData);
os.write(s.data);
}
if (os.size() > 0) {
streamBytes = os.toByteArray();
}
os.close();
} catch (java.io.IOException e) {
Logger logger = LoggerFactory.getLogger(Jbig2SegmentReader.class);
logger.debug(e.getMessage());
}
return streamBytes;
}
@Override
public String toString() {
if (this.read) {
return "Jbig2SegmentReader: number of pages: " + this.numberOfPages();
} else {
return "Jbig2SegmentReader in indeterminate state.";
}
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy