diff --git a/src/common/mod.rs b/src/common/mod.rs index 259a385..fd52779 100644 --- a/src/common/mod.rs +++ b/src/common/mod.rs @@ -2873,7 +2873,7 @@ impl ECIEncoderSet { pub fn new( stringToEncode: &str, priorityCharset: &'static dyn encoding::Encoding, - fnc1: u16, + fnc1: i16, ) -> Self { // List of encoders that potentially encode characters not in ISO-8859-1 in one byte. let mut ENCODERS = Vec::new(); @@ -3028,7 +3028,7 @@ impl ECIEncoderSet { return self.priorityEncoderIndex; } - pub fn canEncode(&self, c: u16, encoderIndex: usize) -> bool { + pub fn canEncode(&self, c: i16, encoderIndex: usize) -> bool { assert!(encoderIndex < self.len()); let encoder = self.encoders[encoderIndex]; let enc_data = encoder.encode(&c.to_string(), encoding::EncoderTrap::Strict); @@ -3085,7 +3085,7 @@ static COST_PER_ECI: usize = 3; */ pub struct MinimalECIInput { bytes: Vec, - fnc1: u16, + fnc1: i16, } impl ECIInput for MinimalECIInput { @@ -3244,7 +3244,7 @@ impl MinimalECIInput { * @param fnc1 denotes the character in the input that represents the FNC1 character or -1 if this is not GS1 * input. */ - pub fn new(stringToEncode: &str, priorityCharset: &'static dyn Encoding, fnc1: u16) -> Self { + pub fn new(stringToEncode: &str, priorityCharset: &'static dyn Encoding, fnc1: i16) -> Self { let encoderSet = ECIEncoderSet::new(stringToEncode, priorityCharset, fnc1); let bytes = if encoderSet.len() == 1 { //optimization for the case when all can be encoded without ECI in ISO-8859-1 @@ -3252,7 +3252,7 @@ impl MinimalECIInput { for i in 0..stringToEncode.len() { // for (int i = 0; i < bytes.length; i++) { let c = stringToEncode.chars().nth(i).unwrap(); - bytes_hld[i] = if c as u16 == fnc1 { 1000 } else { c as u16 }; + bytes_hld[i] = if c as i16 == fnc1 { 1000 } else { c as u16 }; } bytes_hld } else { @@ -3265,7 +3265,7 @@ impl MinimalECIInput { } } - pub fn getFNC1Character(&self) -> u16 { + pub fn getFNC1Character(&self) -> i16 { self.fnc1 } @@ -3305,14 +3305,14 @@ impl MinimalECIInput { edges: &mut Vec>>>, from: usize, previous: Option>, - fnc1: u16, + fnc1: i16, ) { - let ch = stringToEncode.chars().nth(from).unwrap() as u16; + let ch = stringToEncode.chars().nth(from).unwrap() as i16; let mut start = 0; let mut end = encoderSet.len(); if encoderSet.getPriorityEncoderIndex() >= 0 - && (ch as u16 == fnc1 || encoderSet.canEncode(ch, encoderSet.getPriorityEncoderIndex())) + && (ch as i16 == fnc1 || encoderSet.canEncode(ch, encoderSet.getPriorityEncoderIndex())) { start = encoderSet.getPriorityEncoderIndex(); end = start + 1; @@ -3320,7 +3320,7 @@ impl MinimalECIInput { for i in start..end { // for (int i = start; i < end; i++) { - if ch as u16 == fnc1 || encoderSet.canEncode(ch, i) { + if ch as i16 == fnc1 || encoderSet.canEncode(ch, i) { Self::addEdge( edges, from + 1, @@ -3333,7 +3333,7 @@ impl MinimalECIInput { pub fn encodeMinimally( stringToEncode: &str, encoderSet: &ECIEncoderSet, - fnc1: u16, + fnc1: i16, ) -> Vec { let inputLength = stringToEncode.len(); @@ -3413,18 +3413,18 @@ impl MinimalECIInput { } struct InputEdge { - c: u16, + c: i16, encoderIndex: usize, //the encoding of this edge previous: Option>, cachedTotalSize: usize, } impl InputEdge { pub fn new( - c: u16, + c: i16, encoderSet: &ECIEncoderSet, encoderIndex: usize, previous: Option>, - fnc1: u16, + fnc1: i16, ) -> Self { let mut size = if c == 1000 { 1 @@ -3440,7 +3440,7 @@ impl InputEdge { size += prev.cachedTotalSize; Self { - c: if c as u16 == fnc1 { 1000 } else { c as u16 }, + c: if c as i16 == fnc1 { 1000 } else { c as i16 }, encoderIndex, previous: Some(prev.clone()), cachedTotalSize: size, @@ -3452,7 +3452,7 @@ impl InputEdge { } Self { - c: if c as u16 == fnc1 { 1000 } else { c as u16 }, + c: if c as i16 == fnc1 { 1000 } else { c as i16 }, encoderIndex, previous: None, cachedTotalSize: size, diff --git a/src/lib.rs b/src/lib.rs index e14b82a..fa4f4cc 100644 --- a/src/lib.rs +++ b/src/lib.rs @@ -25,6 +25,12 @@ mod PlanarYUVLuminanceSourceTestCase; #[cfg(test)] mod RGBLuminanceSourceTestCase; + + +pub type EncodingHintDictionary = HashMap; +pub type DecodingHintDictionary= HashMap; +pub type MetadataDictionary = HashMap; + /* * Copyright 2007 ZXing authors * diff --git a/src/qrcode/decoder/mode.rs b/src/qrcode/decoder/mode.rs index 99d4b48..68fd4e2 100644 --- a/src/qrcode/decoder/mode.rs +++ b/src/qrcode/decoder/mode.rs @@ -24,7 +24,7 @@ use super::Version; * * @author Sean Owen */ -#[derive(Debug, PartialEq, Eq)] +#[derive(Debug, PartialEq, Eq, Copy, Clone)] pub enum Mode { TERMINATOR, //(new int[]{0, 0, 0}, 0x00), // Not really a mode... NUMERIC, //(new int[]{10, 12, 14}, 0x01), diff --git a/src/qrcode/decoder/version.rs b/src/qrcode/decoder/version.rs index c3665e5..aa67481 100755 --- a/src/qrcode/decoder/version.rs +++ b/src/qrcode/decoder/version.rs @@ -22,6 +22,8 @@ use super::{ErrorCorrectionLevel, FormatInformation}; use lazy_static::lazy_static; +pub type VersionRef = &'static Version; + lazy_static! { static ref VERSIONS: Vec = Version::buildVersions(); } diff --git a/src/qrcode/encoder/MinimalEncoder.java b/src/qrcode/encoder/MinimalEncoder.java deleted file mode 100644 index 3f97fc6..0000000 --- a/src/qrcode/encoder/MinimalEncoder.java +++ /dev/null @@ -1,667 +0,0 @@ -/* - * Copyright 2021 ZXing authors - * - * Licensed under the Apache License, Version 2.0 (the "License"); - * you may not use this file except in compliance with the License. - * You may obtain a copy of the License at - * - * http://www.apache.org/licenses/LICENSE-2.0 - * - * Unless required by applicable law or agreed to in writing, software - * distributed under the License is distributed on an "AS IS" BASIS, - * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. - * See the License for the specific language governing permissions and - * limitations under the License. - */ - -package com.google.zxing.qrcode.encoder; - -import com.google.zxing.qrcode.decoder.Mode; -import com.google.zxing.qrcode.decoder.Version; -import com.google.zxing.common.BitArray; -import com.google.zxing.common.ECIEncoderSet; -import com.google.zxing.WriterException; -import com.google.zxing.qrcode.decoder.ErrorCorrectionLevel; - -import java.nio.charset.Charset; -import java.util.ArrayList; -import java.util.List; - - -/** - * Encoder that encodes minimally - * - * Algorithm: - * - * The eleventh commandment was "Thou Shalt Compute" or "Thou Shalt Not Compute" - I forget which (Alan Perilis). - * - * This implementation computes. As an alternative, the QR-Code specification suggests heuristics like this one: - * - * If initial input data is in the exclusive subset of the Alphanumeric character set AND if there are less than - * [6,7,8] characters followed by data from the remainder of the 8-bit byte character set, THEN select the 8- - * bit byte mode ELSE select Alphanumeric mode; - * - * This is probably right for 99.99% of cases but there is at least this one counter example: The string "AAAAAAa" - * encodes 2 bits smaller as ALPHANUMERIC(AAAAAA), BYTE(a) than by encoding it as BYTE(AAAAAAa). - * Perhaps that is the only counter example but without having proof, it remains unclear. - * - * ECI switching: - * - * In multi language content the algorithm selects the most compact representation using ECI modes. - * For example the most compact representation of the string "\u0150\u015C" (O-double-acute, S-circumflex) is - * ECI(UTF-8), BYTE(\u0150\u015C) while prepending one or more times the same leading character as in - * "\u0150\u0150\u015C", the most compact representation uses two ECIs so that the string is encoded as - * ECI(ISO-8859-2), BYTE(\u0150\u0150), ECI(ISO-8859-3), BYTE(\u015C). - * - * @author Alex Geller - */ -final class MinimalEncoder { - - private enum VersionSize { - SMALL("version 1-9"), - MEDIUM("version 10-26"), - LARGE("version 27-40"); - - private final String description; - - VersionSize(String description) { - this.description = description; - } - - public String toString() { - return description; - } - } - - private final String stringToEncode; - private final boolean isGS1; - private final ECIEncoderSet encoders; - private final ErrorCorrectionLevel ecLevel; - - /** - * Creates a MinimalEncoder - * - * @param stringToEncode The string to encode - * @param priorityCharset The preferred {@link Charset}. When the value of the argument is null, the algorithm - * chooses charsets that leads to a minimal representation. Otherwise the algorithm will use the priority - * charset to encode any character in the input that can be encoded by it if the charset is among the - * supported charsets. - * @param isGS1 {@code true} if a FNC1 is to be prepended; {@code false} otherwise - * @param ecLevel The error correction level. - * @see RXingResultList#getVersion - */ - MinimalEncoder(String stringToEncode, Charset priorityCharset, boolean isGS1, ErrorCorrectionLevel ecLevel) { - this.stringToEncode = stringToEncode; - this.isGS1 = isGS1; - this.encoders = new ECIEncoderSet(stringToEncode, priorityCharset, -1); - this.ecLevel = ecLevel; - } - - /** - * Encodes the string minimally - * - * @param stringToEncode The string to encode - * @param version The preferred {@link Version}. A minimal version is computed (see - * {@link RXingResultList#getVersion method} when the value of the argument is null - * @param priorityCharset The preferred {@link Charset}. When the value of the argument is null, the algorithm - * chooses charsets that leads to a minimal representation. Otherwise the algorithm will use the priority - * charset to encode any character in the input that can be encoded by it if the charset is among the - * supported charsets. - * @param isGS1 {@code true} if a FNC1 is to be prepended; {@code false} otherwise - * @param ecLevel The error correction level. - * @return An instance of {@code RXingResultList} representing the minimal solution. - * @see RXingResultList#getBits - * @see RXingResultList#getVersion - * @see RXingResultList#getSize - */ - static RXingResultList encode(String stringToEncode, Version version, Charset priorityCharset, boolean isGS1, - ErrorCorrectionLevel ecLevel) throws WriterException { - return new MinimalEncoder(stringToEncode, priorityCharset, isGS1, ecLevel).encode(version); - } - - RXingResultList encode(Version version) throws WriterException { - if (version == null) { // compute minimal encoding trying the three version sizes. - Version[] versions = { getVersion(VersionSize.SMALL), - getVersion(VersionSize.MEDIUM), - getVersion(VersionSize.LARGE) }; - RXingResultList[] results = { encodeSpecificVersion(versions[0]), - encodeSpecificVersion(versions[1]), - encodeSpecificVersion(versions[2]) }; - int smallestSize = Integer.MAX_VALUE; - int smallestRXingResult = -1; - for (int i = 0; i < 3; i++) { - int size = results[i].getSize(); - if (Encoder.willFit(size, versions[i], ecLevel) && size < smallestSize) { - smallestSize = size; - smallestRXingResult = i; - } - } - if (smallestRXingResult < 0) { - throw new WriterException("Data too big for any version"); - } - return results[smallestRXingResult]; - } else { // compute minimal encoding for a given version - RXingResultList result = encodeSpecificVersion(version); - if (!Encoder.willFit(result.getSize(), getVersion(getVersionSize(result.getVersion())), ecLevel)) { - throw new WriterException("Data too big for version" + version); - } - return result; - } - } - - static VersionSize getVersionSize(Version version) { - return version.getVersionNumber() <= 9 ? VersionSize.SMALL : version.getVersionNumber() <= 26 ? - VersionSize.MEDIUM : VersionSize.LARGE; - } - - static Version getVersion(VersionSize versionSize) { - switch (versionSize) { - case SMALL: - return Version.getVersionForNumber(9); - case MEDIUM: - return Version.getVersionForNumber(26); - case LARGE: - default: - return Version.getVersionForNumber(40); - } - } - - static boolean isNumeric(char c) { - return c >= '0' && c <= '9'; - } - - static boolean isDoubleByteKanji(char c) { - return Encoder.isOnlyDoubleByteKanji(String.valueOf(c)); - } - - static boolean isAlphanumeric(char c) { - return Encoder.getAlphanumericCode(c) != -1; - } - - boolean canEncode(Mode mode, char c) { - switch (mode) { - case KANJI: return isDoubleByteKanji(c); - case ALPHANUMERIC: return isAlphanumeric(c); - case NUMERIC: return isNumeric(c); - case BYTE: return true; // any character can be encoded as byte(s). Up to the caller to manage splitting into - // multiple bytes when String.getBytes(Charset) return more than one byte. - default: - return false; - } - } - - static int getCompactedOrdinal(Mode mode) { - if (mode == null) { - return 0; - } - switch (mode) { - case KANJI: - return 0; - case ALPHANUMERIC: - return 1; - case NUMERIC: - return 2; - case BYTE: - return 3; - default: - throw new IllegalStateException("Illegal mode " + mode); - } - } - - void addEdge(Edge[][][] edges, int position, Edge edge) { - int vertexIndex = position + edge.characterLength; - Edge[] modeEdges = edges[vertexIndex][edge.charsetEncoderIndex]; - int modeOrdinal = getCompactedOrdinal(edge.mode); - if (modeEdges[modeOrdinal] == null || modeEdges[modeOrdinal].cachedTotalSize > edge.cachedTotalSize) { - modeEdges[modeOrdinal] = edge; - } - } - - void addEdges(Version version, Edge[][][] edges, int from, Edge previous) { - int start = 0; - int end = encoders.length(); - int priorityEncoderIndex = encoders.getPriorityEncoderIndex(); - if (priorityEncoderIndex >= 0 && encoders.canEncode(stringToEncode.charAt(from),priorityEncoderIndex)) { - start = priorityEncoderIndex; - end = priorityEncoderIndex + 1; - } - - for (int i = start; i < end; i++) { - if (encoders.canEncode(stringToEncode.charAt(from), i)) { - addEdge(edges, from, new Edge(Mode.BYTE, from, i, 1, previous, version)); - } - } - - if (canEncode(Mode.KANJI, stringToEncode.charAt(from))) { - addEdge(edges, from, new Edge(Mode.KANJI, from, 0, 1, previous, version)); - } - - int inputLength = stringToEncode.length(); - if (canEncode(Mode.ALPHANUMERIC, stringToEncode.charAt(from))) { - addEdge(edges, from, new Edge(Mode.ALPHANUMERIC, from, 0, from + 1 >= inputLength || - !canEncode(Mode.ALPHANUMERIC, stringToEncode.charAt(from + 1)) ? 1 : 2, previous, version)); - } - - if (canEncode(Mode.NUMERIC, stringToEncode.charAt(from))) { - addEdge(edges, from, new Edge(Mode.NUMERIC, from, 0, from + 1 >= inputLength || - !canEncode(Mode.NUMERIC, stringToEncode.charAt(from + 1)) ? 1 : from + 2 >= inputLength || - !canEncode(Mode.NUMERIC, stringToEncode.charAt(from + 2)) ? 2 : 3, previous, version)); - } - } - RXingResultList encodeSpecificVersion(Version version) throws WriterException { - - @SuppressWarnings("checkstyle:lineLength") - /* A vertex represents a tuple of a position in the input, a mode and a character encoding where position 0 - * denotes the position left of the first character, 1 the position left of the second character and so on. - * Likewise the end vertices are located after the last character at position stringToEncode.length(). - * - * An edge leading to such a vertex encodes one or more of the characters left of the position that the vertex - * represents and encodes it in the same encoding and mode as the vertex on which the edge ends. In other words, - * all edges leading to a particular vertex encode the same characters in the same mode with the same character - * encoding. They differ only by their source vertices who are all located at i+1 minus the number of encoded - * characters. - * - * The edges leading to a vertex are stored in such a way that there is a fast way to enumerate the edges ending - * on a particular vertex. - * - * The algorithm processes the vertices in order of their position thereby performing the following: - * - * For every vertex at position i the algorithm enumerates the edges ending on the vertex and removes all but the - * shortest from that list. - * Then it processes the vertices for the position i+1. If i+1 == stringToEncode.length() then the algorithm ends - * and chooses the the edge with the smallest size from any of the edges leading to vertices at this position. - * Otherwise the algorithm computes all possible outgoing edges for the vertices at the position i+1 - * - * Examples: - * The process is illustrated by showing the graph (edges) after each iteration from left to right over the input: - * An edge is drawn as follows "(" + fromVertex + ") -- " + encodingMode + "(" + encodedInput + ") (" + - * accumulatedSize + ") --> (" + toVertex + ")" - * - * Example 1 encoding the string "ABCDE": - * Note: This example assumes that alphanumeric encoding is only possible in multiples of two characters so that - * the example is both short and showing the principle. In reality this restriction does not exist. - * - * Initial situation - * (initial) -- BYTE(A) (20) --> (1_BYTE) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) - * - * Situation after adding edges to vertices at position 1 - * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) - * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) - * - * Situation after adding edges to vertices at position 2 - * (initial) -- BYTE(A) (20) --> (1_BYTE) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) - * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) - * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- BYTE(C) (44) --> (3_BYTE) - * (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) - * - * Situation after adding edges to vertices at position 3 - * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) - * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- BYTE(D) (64) --> (4_BYTE) - * (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) - * (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) - * - * Situation after adding edges to vertices at position 4 - * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) -- BYTE(D) (44) --> (4_BYTE) - * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) -- BYTE(E) (55) --> (5_BYTE) - * - * Situation after adding edges to vertices at position 5 - * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) -- BYTE(D) (44) --> (4_BYTE) -- BYTE(E) (52) --> (5_BYTE) - * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) - * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) - * - * Encoding as BYTE(ABCDE) has the smallest size of 52 and is hence chosen. The encodation ALPHANUMERIC(ABCD), - * BYTE(E) is longer with a size of 55. - * - * Example 2 encoding the string "XXYY" where X denotes a character unique to character set ISO-8859-2 and Y a - * character unique to ISO-8859-3. Both characters encode as double byte in UTF-8: - * - * Initial situation - * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) - * - * Situation after adding edges to vertices at position 1 - * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) - * (1_BYTE_ISO-8859-2) -- BYTE(X) (72) --> (2_BYTE_UTF-8) - * (1_BYTE_ISO-8859-2) -- BYTE(X) (72) --> (2_BYTE_UTF-16BE) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) - * - * Situation after adding edges to vertices at position 2 - * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) - * (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) - * (2_BYTE_ISO-8859-2) -- BYTE(Y) (80) --> (3_BYTE_UTF-8) - * (2_BYTE_ISO-8859-2) -- BYTE(Y) (80) --> (3_BYTE_UTF-16BE) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) - * - * Situation after adding edges to vertices at position 3 - * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) - * (3_BYTE_ISO-8859-3) -- BYTE(Y) (80) --> (4_BYTE_ISO-8859-3) - * (3_BYTE_ISO-8859-3) -- BYTE(Y) (112) --> (4_BYTE_UTF-8) - * (3_BYTE_ISO-8859-3) -- BYTE(Y) (112) --> (4_BYTE_UTF-16BE) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) -- BYTE(Y) (72) --> (3_BYTE_UTF-8) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) -- BYTE(Y) (72) --> (3_BYTE_UTF-16BE) - * - * Situation after adding edges to vertices at position 4 - * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) -- BYTE(Y) (80) --> (4_BYTE_ISO-8859-3) - * (3_BYTE_UTF-8) -- BYTE(Y) (88) --> (4_BYTE_UTF-8) - * (3_BYTE_UTF-16BE) -- BYTE(Y) (88) --> (4_BYTE_UTF-16BE) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) -- BYTE(Y) (72) --> (3_BYTE_UTF-8) - * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) -- BYTE(Y) (72) --> (3_BYTE_UTF-16BE) - * - * Encoding as ECI(ISO-8859-2),BYTE(XX),ECI(ISO-8859-3),BYTE(YY) has the smallest size of 80 and is hence chosen. - * The encodation ECI(UTF-8),BYTE(XXYY) is longer with a size of 88. - */ - - int inputLength = stringToEncode.length(); - - // Array that represents vertices. There is a vertex for every character, encoding and mode. The vertex contains - // a list of all edges that lead to it that have the same encoding and mode. - // The lists are created lazily - - // The last dimension in the array below encodes the 4 modes KANJI, ALPHANUMERIC, NUMERIC and BYTE via the - // function getCompactedOrdinal(Mode) - Edge[][][] edges = new Edge[inputLength + 1][encoders.length()][4]; - addEdges(version, edges, 0, null); - - for (int i = 1; i <= inputLength; i++) { - for (int j = 0; j < encoders.length(); j++) { - for (int k = 0; k < 4; k++) { - if (edges[i][j][k] != null && i < inputLength) { - addEdges(version, edges, i, edges[i][j][k]); - } - } - } - - } - int minimalJ = -1; - int minimalK = -1; - int minimalSize = Integer.MAX_VALUE; - for (int j = 0; j < encoders.length(); j++) { - for (int k = 0; k < 4; k++) { - if (edges[inputLength][j][k] != null) { - Edge edge = edges[inputLength][j][k]; - if (edge.cachedTotalSize < minimalSize) { - minimalSize = edge.cachedTotalSize; - minimalJ = j; - minimalK = k; - } - } - } - } - if (minimalJ < 0) { - throw new WriterException("Internal error: failed to encode \"" + stringToEncode + "\""); - } - return new RXingResultList(version, edges[inputLength][minimalJ][minimalK]); - } - - private final class Edge { - private final Mode mode; - private final int fromPosition; - private final int charsetEncoderIndex; - private final int characterLength; - private final Edge previous; - private final int cachedTotalSize; - - private Edge(Mode mode, int fromPosition, int charsetEncoderIndex, int characterLength, Edge previous, - Version version) { - this.mode = mode; - this.fromPosition = fromPosition; - this.charsetEncoderIndex = mode == Mode.BYTE || previous == null ? charsetEncoderIndex : - previous.charsetEncoderIndex; // inherit the encoding if not of type BYTE - this.characterLength = characterLength; - this.previous = previous; - - int size = previous != null ? previous.cachedTotalSize : 0; - - boolean needECI = mode == Mode.BYTE && - (previous == null && this.charsetEncoderIndex != 0) || // at the beginning and charset is not ISO-8859-1 - (previous != null && this.charsetEncoderIndex != previous.charsetEncoderIndex); - - if (previous == null || mode != previous.mode || needECI) { - size += 4 + mode.getCharacterCountBits(version); - } - switch (mode) { - case KANJI: - size += 13; - break; - case ALPHANUMERIC: - size += characterLength == 1 ? 6 : 11; - break; - case NUMERIC: - size += characterLength == 1 ? 4 : characterLength == 2 ? 7 : 10; - break; - case BYTE: - size += 8 * encoders.encode(stringToEncode.substring(fromPosition, fromPosition + characterLength), - charsetEncoderIndex).length; - if (needECI) { - size += 4 + 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long - } - break; - } - cachedTotalSize = size; - } - } - - final class RXingResultList { - - private final List list = new ArrayList<>(); - private final Version version; - - RXingResultList(Version version, Edge solution) { - int length = 0; - Edge current = solution; - boolean containsECI = false; - - while (current != null) { - length += current.characterLength; - Edge previous = current.previous; - - boolean needECI = current.mode == Mode.BYTE && - (previous == null && current.charsetEncoderIndex != 0) || // at the beginning and charset is not ISO-8859-1 - (previous != null && current.charsetEncoderIndex != previous.charsetEncoderIndex); - - if (needECI) { - containsECI = true; - } - - if (previous == null || previous.mode != current.mode || needECI) { - list.add(0, new RXingResultNode(current.mode, current.fromPosition, current.charsetEncoderIndex, length)); - length = 0; - } - - if (needECI) { - list.add(0, new RXingResultNode(Mode.ECI, current.fromPosition, current.charsetEncoderIndex, 0)); - } - current = previous; - } - - // prepend FNC1 if needed. If the bits contain an ECI then the FNC1 must be preceeded by an ECI. - // If there is no ECI at the beginning then we put an ECI to the default charset (ISO-8859-1) - if (isGS1) { - RXingResultNode first = list.get(0); - if (first != null && first.mode != Mode.ECI && containsECI) { - // prepend a default character set ECI - list.add(0, new RXingResultNode(Mode.ECI, 0, 0, 0)); - } - first = list.get(0); - // prepend or insert a FNC1_FIRST_POSITION after the ECI (if any) - list.add(first.mode != Mode.ECI ? 0 : 1, new RXingResultNode(Mode.FNC1_FIRST_POSITION, 0, 0, 0)); - } - - // set version to smallest version into which the bits fit. - int versionNumber = version.getVersionNumber(); - int lowerLimit; - int upperLimit; - switch (getVersionSize(version)) { - case SMALL: - lowerLimit = 1; - upperLimit = 9; - break; - case MEDIUM: - lowerLimit = 10; - upperLimit = 26; - break; - case LARGE: - default: - lowerLimit = 27; - upperLimit = 40; - break; - } - int size = getSize(version); - // increase version if needed - while (versionNumber < upperLimit && !Encoder.willFit(size, Version.getVersionForNumber(versionNumber), - ecLevel)) { - versionNumber++; - } - // shrink version if possible - while (versionNumber > lowerLimit && Encoder.willFit(size, Version.getVersionForNumber(versionNumber - 1), - ecLevel)) { - versionNumber--; - } - this.version = Version.getVersionForNumber(versionNumber); - } - - /** - * returns the size in bits - */ - int getSize() { - return getSize(version); - } - - private int getSize(Version version) { - int result = 0; - for (RXingResultNode resultNode : list) { - result += resultNode.getSize(version); - } - return result; - } - - /** - * appends the bits - */ - void getBits(BitArray bits) throws WriterException { - for (RXingResultNode resultNode : list) { - resultNode.getBits(bits); - } - } - - Version getVersion() { - return version; - } - - public String toString() { - StringBuilder result = new StringBuilder(); - RXingResultNode previous = null; - for (RXingResultNode current : list) { - if (previous != null) { - result.append(","); - } - result.append(current.toString()); - previous = current; - } - return result.toString(); - } - - final class RXingResultNode { - - private final Mode mode; - private final int fromPosition; - private final int charsetEncoderIndex; - private final int characterLength; - - RXingResultNode(Mode mode, int fromPosition, int charsetEncoderIndex, int characterLength) { - this.mode = mode; - this.fromPosition = fromPosition; - this.charsetEncoderIndex = charsetEncoderIndex; - this.characterLength = characterLength; - } - - /** - * returns the size in bits - */ - private int getSize(Version version) { - int size = 4 + mode.getCharacterCountBits(version); - switch (mode) { - case KANJI: - size += 13 * characterLength; - break; - case ALPHANUMERIC: - size += (characterLength / 2) * 11; - size += (characterLength % 2) == 1 ? 6 : 0; - break; - case NUMERIC: - size += (characterLength / 3) * 10; - int rest = characterLength % 3; - size += rest == 1 ? 4 : rest == 2 ? 7 : 0; - break; - case BYTE: - size += 8 * getCharacterCountIndicator(); - break; - case ECI: - size += 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long - } - return size; - } - - /** - * returns the length in characters according to the specification (differs from getCharacterLength() in BYTE mode - * for multi byte encoded characters) - */ - private int getCharacterCountIndicator() { - return mode == Mode.BYTE ? - encoders.encode(stringToEncode.substring(fromPosition, fromPosition + characterLength), - charsetEncoderIndex).length : characterLength; - } - - /** - * appends the bits - */ - private void getBits(BitArray bits) throws WriterException { - bits.appendBits(mode.getBits(), 4); - if (characterLength > 0) { - int length = getCharacterCountIndicator(); - bits.appendBits(length, mode.getCharacterCountBits(version)); - } - if (mode == Mode.ECI) { - bits.appendBits(encoders.getECIValue(charsetEncoderIndex), 8); - } else if (characterLength > 0) { - // append data - Encoder.appendBytes(stringToEncode.substring(fromPosition, fromPosition + characterLength), mode, bits, - encoders.getCharset(charsetEncoderIndex)); - } - } - - public String toString() { - StringBuilder result = new StringBuilder(); - result.append(mode).append('('); - if (mode == Mode.ECI) { - result.append(encoders.getCharset(charsetEncoderIndex).displayName()); - } else { - result.append(makePrintable(stringToEncode.substring(fromPosition, fromPosition + characterLength))); - } - result.append(')'); - return result.toString(); - } - - private String makePrintable(String s) { - StringBuilder result = new StringBuilder(); - for (int i = 0; i < s.length(); i++) { - if (s.charAt(i) < 32 || s.charAt(i) > 126) { - result.append('.'); - } else { - result.append(s.charAt(i)); - } - } - return result.toString(); - } - } - } -} diff --git a/src/qrcode/encoder/Encoder.java b/src/qrcode/encoder/encoder.rs similarity index 57% rename from src/qrcode/encoder/Encoder.java rename to src/qrcode/encoder/encoder.rs index c0f9f58..9e89758 100644 --- a/src/qrcode/encoder/Encoder.java +++ b/src/qrcode/encoder/encoder.rs @@ -14,53 +14,58 @@ * limitations under the License. */ -package com.google.zxing.qrcode.encoder; +// package com.google.zxing.qrcode.encoder; -import com.google.zxing.EncodeHintType; -import com.google.zxing.WriterException; -import com.google.zxing.common.BitArray; -import com.google.zxing.common.StringUtils; -import com.google.zxing.common.CharacterSetECI; -import com.google.zxing.common.reedsolomon.GenericGF; -import com.google.zxing.common.reedsolomon.ReedSolomonEncoder; -import com.google.zxing.qrcode.decoder.ErrorCorrectionLevel; -import com.google.zxing.qrcode.decoder.Mode; -import com.google.zxing.qrcode.decoder.Version; +// import com.google.zxing.EncodeHintType; +// import com.google.zxing.WriterException; +// import com.google.zxing.common.BitArray; +// import com.google.zxing.common.StringUtils; +// import com.google.zxing.common.CharacterSetECI; +// import com.google.zxing.common.reedsolomon.GenericGF; +// import com.google.zxing.common.reedsolomon.ReedSolomonEncoder; +// import com.google.zxing.qrcode.decoder.ErrorCorrectionLevel; +// import com.google.zxing.qrcode.decoder.Mode; +// import com.google.zxing.qrcode.decoder.Version; -import java.nio.charset.Charset; -import java.nio.charset.StandardCharsets; -import java.util.ArrayList; -import java.util.Collection; -import java.util.Map; +// import java.nio.charset.Charset; +// import java.nio.charset.StandardCharsets; +// import java.util.ArrayList; +// import java.util.Collection; +// import java.util.Map; + +use std::collections::HashMap; + +use encoding::EncodingRef; + +use crate::{EncodingHintDictionary, common::{BitArray, CharacterSetECI, reedsolomon::{ReedSolomonEncoder, get_predefined_genericgf, PredefinedGenericGF}, StringUtils}, Exceptions, qrcode::decoder::{ErrorCorrectionLevel, Mode, VersionRef, Version}}; + +use super::{mask_util, ByteMatrix, QRCode, matrix_util}; /** * @author satorux@google.com (Satoru Takabayashi) - creator * @author dswitkin@google.com (Daniel Switkin) - ported from C++ */ -public final class Encoder { // The original table is defined in the table 5 of JISX0510:2004 (p.19). - private static final int[] ALPHANUMERIC_TABLE = { + const ALPHANUMERIC_TABLE : [i8;96]= [ -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 0x00-0x0f -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, // 0x10-0x1f 36, -1, -1, -1, 37, 38, -1, -1, -1, -1, 39, 40, -1, 41, 42, 43, // 0x20-0x2f 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 44, -1, -1, -1, -1, -1, // 0x30-0x3f -1, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, // 0x40-0x4f 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, -1, -1, -1, -1, -1, // 0x50-0x5f - }; + ]; - static final Charset DEFAULT_BYTE_MODE_ENCODING = StandardCharsets.ISO_8859_1; + const DEFAULT_BYTE_MODE_ENCODING : EncodingRef = encoding::all::ISO_8859_1; - private Encoder() { - } // The mask penalty calculation is complicated. See Table 21 of JISX0510:2004 (p.45) for details. // Basically it applies four rules and summate all penalties. - private static int calculateMaskPenalty(ByteMatrix matrix) { - return MaskUtil.applyMaskPenaltyRule1(matrix) - + MaskUtil.applyMaskPenaltyRule2(matrix) - + MaskUtil.applyMaskPenaltyRule3(matrix) - + MaskUtil.applyMaskPenaltyRule4(matrix); + pub fn calculateMaskPenalty( matrix:&ByteMatrix) -> u32{ + return mask_util::applyMaskPenaltyRule1(matrix) + + mask_util::applyMaskPenaltyRule2(matrix) + + mask_util::applyMaskPenaltyRule3(matrix) + + mask_util::applyMaskPenaltyRule4(matrix); } /** @@ -70,26 +75,26 @@ public final class Encoder { * @throws WriterException if encoding can't succeed, because of for example invalid content * or configuration */ - public static QRCode encode(String content, ErrorCorrectionLevel ecLevel) throws WriterException { - return encode(content, ecLevel, null); + pub fn encode( content:&str, ecLevel:&ErrorCorrectionLevel) -> Result { + return encode_with_hints(content, ecLevel, HashMap::new()); } - public static QRCode encode(String content, - ErrorCorrectionLevel ecLevel, - Map hints) throws WriterException { + pub fn encode_with_hints( content:&str, + ecLevel:&ErrorCorrectionLevel, + hints:EncodingHintDictionary) -> Result { - Version version; - BitArray headerAndDataBits; - Mode mode; + let version; + let headerAndDataBits; + let mode; - boolean hasGS1FormatHint = hints != null && hints.containsKey(EncodeHintType.GS1_FORMAT) && - Boolean.parseBoolean(hints.get(EncodeHintType.GS1_FORMAT).toString()); - boolean hasCompactionHint = hints != null && hints.containsKey(EncodeHintType.QR_COMPACT) && - Boolean.parseBoolean(hints.get(EncodeHintType.QR_COMPACT).toString()); + let hasGS1FormatHint = hints != null && hints.containsKey(EncodeHintType::GS1_FORMAT) && + Boolean.parseBoolean(hints.get(EncodeHintType::GS1_FORMAT).toString()); + let hasCompactionHint = hints != null && hints.containsKey(EncodeHintType::QR_COMPACT) && + Boolean.parseBoolean(hints.get(EncodeHintType::QR_COMPACT).toString()); // Determine what character encoding has been specified by the caller, if any - Charset encoding = DEFAULT_BYTE_MODE_ENCODING; - boolean hasEncodingHint = hints != null && hints.containsKey(EncodeHintType.CHARACTER_SET); + let encoding = DEFAULT_BYTE_MODE_ENCODING; + let hasEncodingHint = hints != null && hints.containsKey(EncodeHintType::CHARACTER_SET); if (hasEncodingHint) { encoding = Charset.forName(hints.get(EncodeHintType.CHARACTER_SET).toString()); } @@ -191,7 +196,7 @@ public final class Encoder { qrCode.setMaskPattern(maskPattern); // Build the matrix and set it to "qrCode". - MatrixUtil.buildMatrix(finalBits, ecLevel, version, maskPattern, matrix); + matrix_util::buildMatrix(finalBits, ecLevel, version, maskPattern, matrix); qrCode.setMatrix(matrix); return qrCode; @@ -202,25 +207,25 @@ public final class Encoder { * * @throws WriterException if the data cannot fit in any version */ - private static Version recommendVersion(ErrorCorrectionLevel ecLevel, - Mode mode, - BitArray headerBits, - BitArray dataBits) throws WriterException { + fn recommendVersion( ecLevel:&ErrorCorrectionLevel, + mode:Mode, + headerBits:&BitArray, + dataBits:&BitArray) -> Result { // Hard part: need to know version to know how many bits length takes. But need to know how many // bits it takes to know version. First we take a guess at version by assuming version will be // the minimum, 1: - int provisionalBitsNeeded = calculateBitsNeeded(mode, headerBits, dataBits, Version.getVersionForNumber(1)); - Version provisionalVersion = chooseVersion(provisionalBitsNeeded, ecLevel); + let provisionalBitsNeeded = calculateBitsNeeded(mode, headerBits, dataBits, Version::getVersionForNumber(1)); + let provisionalVersion = chooseVersion(provisionalBitsNeeded, ecLevel); // Use that guess to calculate the right version. I am still not sure this works in 100% of cases. - int bitsNeeded = calculateBitsNeeded(mode, headerBits, dataBits, provisionalVersion); + let bitsNeeded = calculateBitsNeeded(mode, headerBits, dataBits, provisionalVersion); return chooseVersion(bitsNeeded, ecLevel); } - private static int calculateBitsNeeded(Mode mode, - BitArray headerBits, - BitArray dataBits, - Version version) { + fn calculateBitsNeeded( mode:Mode, + headerBits:&BitArray, + dataBits:&BitArray, + version:VersionRef) -> u32 { return headerBits.getSize() + mode.getCharacterCountBits(version) + dataBits.getSize(); } @@ -228,22 +233,22 @@ public final class Encoder { * @return the code point of the table used in alphanumeric mode or * -1 if there is no corresponding code in the table. */ - static int getAlphanumericCode(int code) { - if (code < ALPHANUMERIC_TABLE.length) { + pub fn getAlphanumericCode( code:u32) -> u32{ + if code < ALPHANUMERIC_TABLE.len() { return ALPHANUMERIC_TABLE[code]; } return -1; } - public static Mode chooseMode(String content) { - return chooseMode(content, null); + pub fn chooseMode( content:&str) -> Mode{ + return chooseModeWithEncoding(content, None); } /** * Choose the best mode by examining the content. Note that 'encoding' is used as a hint; * if it is Shift_JIS, and the input is only double-byte Kanji, then we return {@link Mode#KANJI}. */ - private static Mode chooseMode(String content, Charset encoding) { + fn chooseModeWithEncoding( content:&str, encoding:Option) -> Mode{ if (StringUtils.SHIFT_JIS_CHARSET.equals(encoding) && isOnlyDoubleByteKanji(content)) { // Choose Kanji mode if all input are double-byte characters return Mode.KANJI; @@ -269,32 +274,36 @@ public final class Encoder { return Mode.BYTE; } - static boolean isOnlyDoubleByteKanji(String content) { - byte[] bytes = content.getBytes(StringUtils.SHIFT_JIS_CHARSET); - int length = bytes.length; - if (length % 2 != 0) { + pub fn isOnlyDoubleByteKanji( content:&str) -> bool{ + let bytes = content.getBytes(StringUtils::SHIFT_JIS_CHARSET); + let length = bytes.len(); + if length % 2 != 0 { return false; } - for (int i = 0; i < length; i += 2) { - int byte1 = bytes[i] & 0xFF; - if ((byte1 < 0x81 || byte1 > 0x9F) && (byte1 < 0xE0 || byte1 > 0xEB)) { + let mut i = 0; + while i < length { + // for (int i = 0; i < length; i += 2) { + let byte1 = bytes[i] & 0xFF; + if (byte1 < 0x81 || byte1 > 0x9F) && (byte1 < 0xE0 || byte1 > 0xEB) { return false; } + i+=2; } return true; } - private static int chooseMaskPattern(BitArray bits, - ErrorCorrectionLevel ecLevel, - Version version, - ByteMatrix matrix) throws WriterException { + fn chooseMaskPattern( bits:&BitArray, + ecLevel:&ErrorCorrectionLevel, + version:VersionRef, + matrix:&ByteMatrix) -> Result { - int minPenalty = Integer.MAX_VALUE; // Lower penalty is better. - int bestMaskPattern = -1; + let minPenalty = u32::MAX; // Lower penalty is better. + let bestMaskPattern = -1; // We try all mask patterns to choose the best one. - for (int maskPattern = 0; maskPattern < QRCode.NUM_MASK_PATTERNS; maskPattern++) { - MatrixUtil.buildMatrix(bits, ecLevel, version, maskPattern, matrix); - int penalty = calculateMaskPenalty(matrix); + for maskPattern in 0..QRCode::NUM_MASK_PATTERNS { + // for (int maskPattern = 0; maskPattern < QRCode.NUM_MASK_PATTERNS; maskPattern++) { + matrix_util::buildMatrix(bits, ecLevel, version, maskPattern, matrix); + let penalty = calculateMaskPenalty(matrix); if (penalty < minPenalty) { minPenalty = penalty; bestMaskPattern = maskPattern; @@ -303,38 +312,39 @@ public final class Encoder { return bestMaskPattern; } - private static Version chooseVersion(int numInputBits, ErrorCorrectionLevel ecLevel) throws WriterException { - for (int versionNum = 1; versionNum <= 40; versionNum++) { - Version version = Version.getVersionForNumber(versionNum); - if (willFit(numInputBits, version, ecLevel)) { + fn chooseVersion( numInputBits:u32, ecLevel:&ErrorCorrectionLevel) -> Result { + for versionNum in 1..=40 { + // for (int versionNum = 1; versionNum <= 40; versionNum++) { + let version = Version::getVersionForNumber(versionNum); + if willFit(numInputBits, version, ecLevel) { return version; } } - throw new WriterException("Data too big"); + Err(Exceptions::WriterException("Data too big".to_owned())); } /** * @return true if the number of input bits will fit in a code with the specified version and * error correction level. */ - static boolean willFit(int numInputBits, Version version, ErrorCorrectionLevel ecLevel) { + pub fn willFit( numInputBits:u32, version:VersionRef, ecLevel:&ErrorCorrectionLevel) -> bool { // In the following comments, we use numbers of Version 7-H. // numBytes = 196 - int numBytes = version.getTotalCodewords(); + let numBytes = version.getTotalCodewords(); // getNumECBytes = 130 - Version.ECBlocks ecBlocks = version.getECBlocksForLevel(ecLevel); - int numEcBytes = ecBlocks.getTotalECCodewords(); + let ecBlocks = version.getECBlocksForLevel(ecLevel); + let numEcBytes = ecBlocks.getTotalECCodewords(); // getNumDataBytes = 196 - 130 = 66 - int numDataBytes = numBytes - numEcBytes; - int totalInputBytes = (numInputBits + 7) / 8; + let numDataBytes = numBytes - numEcBytes; + let totalInputBytes = (numInputBits + 7) / 8; return numDataBytes >= totalInputBytes; } /** * Terminate bits as described in 8.4.8 and 8.4.9 of JISX0510:2004 (p.24). */ - static void terminateBits(int numDataBytes, BitArray bits) throws WriterException { - int capacity = numDataBytes * 8; + pub fn terminateBits( numDataBytes:u32, bits:&BitArray) -> Result<(),Exceptions> { + let capacity = numDataBytes * 8; if (bits.getSize() > capacity) { throw new WriterException("data bits cannot fit in the QR Code" + bits.getSize() + " > " + capacity); @@ -366,13 +376,13 @@ public final class Encoder { * the result in "numDataBytesInBlock", and "numECBytesInBlock". See table 12 in 8.5.1 of * JISX0510:2004 (p.30) */ - static void getNumDataBytesAndNumECBytesForBlockID(int numTotalBytes, - int numDataBytes, - int numRSBlocks, - int blockID, - int[] numDataBytesInBlock, - int[] numECBytesInBlock) throws WriterException { - if (blockID >= numRSBlocks) { + pub fn getNumDataBytesAndNumECBytesForBlockID( numTotalBytes:u32, + numDataBytes:u32, + numRSBlocks:u32, + blockID:u32, + numDataBytesInBlock:&[u32], + numECBytesInBlock:&[u32]) -> Result<(),Exceptions> { + if blockID >= numRSBlocks { throw new WriterException("Block ID too large"); } // numRsBlocksInGroup2 = 196 % 5 = 1 @@ -422,36 +432,37 @@ public final class Encoder { * Interleave "bits" with corresponding error correction bytes. On success, store the result in * "result". The interleave rule is complicated. See 8.6 of JISX0510:2004 (p.37) for details. */ - static BitArray interleaveWithECBytes(BitArray bits, - int numTotalBytes, - int numDataBytes, - int numRSBlocks) throws WriterException { + pub fn interleaveWithECBytes( bits:&BitArray, + numTotalBytes:u32, + numDataBytes:u32, + numRSBlocks:u32) -> Result { // "bits" must have "getNumDataBytes" bytes of data. - if (bits.getSizeInBytes() != numDataBytes) { - throw new WriterException("Number of bits and data bytes does not match"); + if bits.getSizeInBytes() != numDataBytes { + return Err(Exceptions::WriterException("Number of bits and data bytes does not match".to_owned())) } // Step 1. Divide data bytes into blocks and generate error correction bytes for them. We'll // store the divided data bytes blocks and error correction bytes blocks into "blocks". - int dataBytesOffset = 0; - int maxNumDataBytes = 0; - int maxNumEcBytes = 0; + let dataBytesOffset = 0; + let maxNumDataBytes = 0; + let maxNumEcBytes = 0; // Since, we know the number of reedsolmon blocks, we can initialize the vector with the number. - Collection blocks = new ArrayList<>(numRSBlocks); + let blocks = Vec::new(); - for (int i = 0; i < numRSBlocks; ++i) { - int[] numDataBytesInBlock = new int[1]; - int[] numEcBytesInBlock = new int[1]; + for i in 0..numRSBlocks { + // for (int i = 0; i < numRSBlocks; ++i) { + let numDataBytesInBlock = new int[1]; + let numEcBytesInBlock = new int[1]; getNumDataBytesAndNumECBytesForBlockID( numTotalBytes, numDataBytes, numRSBlocks, i, numDataBytesInBlock, numEcBytesInBlock); - int size = numDataBytesInBlock[0]; - byte[] dataBytes = new byte[size]; + let size = numDataBytesInBlock[0]; + let dataBytes = new byte[size]; bits.toBytes(8 * dataBytesOffset, dataBytes, 0, size); - byte[] ecBytes = generateECBytes(dataBytes, numEcBytesInBlock[0]); + let ecBytes = generateECBytes(dataBytes, numEcBytesInBlock[0]); blocks.add(new BlockPair(dataBytes, ecBytes)); maxNumDataBytes = Math.max(maxNumDataBytes, size); @@ -459,10 +470,10 @@ public final class Encoder { dataBytesOffset += numDataBytesInBlock[0]; } if (numDataBytes != dataBytesOffset) { - throw new WriterException("Data bytes does not match offset"); + return Err(Exceptions::WriterException("Data bytes does not match offset".to_owned())) } - BitArray result = new BitArray(); + let result = BitArray::new(); // First, place data blocks. for (int i = 0; i < maxNumDataBytes; ++i) { @@ -490,17 +501,20 @@ public final class Encoder { return result; } - static byte[] generateECBytes(byte[] dataBytes, int numEcBytesInBlock) { - int numDataBytes = dataBytes.length; - int[] toEncode = new int[numDataBytes + numEcBytesInBlock]; - for (int i = 0; i < numDataBytes; i++) { + pub fn generateECBytes( dataBytes:&[u8], numEcBytesInBlock:u32) -> Vec { + let numDataBytes = dataBytes.length; + let toEncode = vec![0;numDataBytes + numEcBytesInBlock]; + for i in 0..numDataBytes { + // for (int i = 0; i < numDataBytes; i++) { toEncode[i] = dataBytes[i] & 0xFF; } - new ReedSolomonEncoder(GenericGF.QR_CODE_FIELD_256).encode(toEncode, numEcBytesInBlock); + + ReedSolomonEncoder::new(get_predefined_genericgf(PredefinedGenericGF::QrCodeField256)).encode(&mut toEncode, numEcBytesInBlock); - byte[] ecBytes = new byte[numEcBytesInBlock]; - for (int i = 0; i < numEcBytesInBlock; i++) { - ecBytes[i] = (byte) toEncode[numDataBytes + i]; + let ecBytes = vec![0u8;numEcBytesInBlock]; + for i in 0..numEcBytesInBlock { + // for (int i = 0; i < numEcBytesInBlock; i++) { + ecBytes[i] = toEncode[numDataBytes + i]; } return ecBytes; } @@ -508,7 +522,7 @@ public final class Encoder { /** * Append mode info. On success, store the result in "bits". */ - static void appendModeInfo(Mode mode, BitArray bits) { + pub fn appendModeInfo( mode:Mode, bits:&BitArray) { bits.appendBits(mode.getBits(), 4); } @@ -516,75 +530,84 @@ public final class Encoder { /** * Append length info. On success, store the result in "bits". */ - static void appendLengthInfo(int numLetters, Version version, Mode mode, BitArray bits) throws WriterException { - int numBits = mode.getCharacterCountBits(version); - if (numLetters >= (1 << numBits)) { - throw new WriterException(numLetters + " is bigger than " + ((1 << numBits) - 1)); + pub fn appendLengthInfo( numLetters:u32, version:VersionRef, mode:Mode, bits:&BitArray) -> Result<(),Exceptions> { + let numBits = mode.getCharacterCountBits(version); + if numLetters >= (1 << numBits) { + return Err(Exceptions::WriterExceptin(format!("{} is bigger than {}" ,numLetters , ((1 << numBits) - 1)))) } bits.appendBits(numLetters, numBits); + Ok(()) } /** * Append "bytes" in "mode" mode (encoding) into "bits". On success, store the result in "bits". */ - static void appendBytes(String content, - Mode mode, - BitArray bits, - Charset encoding) throws WriterException { - switch (mode) { - case NUMERIC: - appendNumericBytes(content, bits); - break; - case ALPHANUMERIC: - appendAlphanumericBytes(content, bits); - break; - case BYTE: - append8BitBytes(content, bits, encoding); - break; - case KANJI: - appendKanjiBytes(content, bits); - break; - default: - throw new WriterException("Invalid mode: " + mode); - } + pub fn appendBytes( content:&str, + mode:Mode, + bits:&BitArray, + encoding:EncodingRef) -> Result<(),Exceptions>{ + match mode { + Mode::NUMERIC => Ok(appendNumericBytes(content, bits)), + Mode::ALPHANUMERIC => Ok(appendAlphanumericBytes(content, bits)), + Mode::BYTE => Ok(append8BitBytes(content, bits, encoding)), + Mode::KANJI => Ok(appendKanjiBytes(content, bits)), + _=> Err(Exceptions::WriterException(format!("Invalid mode: {}" , mode))) + } + // switch (mode) { + // case NUMERIC: + // appendNumericBytes(content, bits); + // break; + // case ALPHANUMERIC: + // appendAlphanumericBytes(content, bits); + // break; + // case BYTE: + // append8BitBytes(content, bits, encoding); + // break; + // case KANJI: + // appendKanjiBytes(content, bits); + // break; + // default: + // throw new WriterException("Invalid mode: " + mode); + // } } - static void appendNumericBytes(CharSequence content, BitArray bits) { - int length = content.length(); - int i = 0; - while (i < length) { - int num1 = content.charAt(i) - '0'; - if (i + 2 < length) { + pub fn appendNumericBytes( content:&str, bits:&BitArray) { + let length = content.length(); + let i = 0; + while i < length { + let num1 = content.charAt(i) - '0'; + if i + 2 < length { // Encode three numeric letters in ten bits. - int num2 = content.charAt(i + 1) - '0'; - int num3 = content.charAt(i + 2) - '0'; + let num2 = content.charAt(i + 1) - '0'; + let num3 = content.charAt(i + 2) - '0'; bits.appendBits(num1 * 100 + num2 * 10 + num3, 10); i += 3; - } else if (i + 1 < length) { + } else if i + 1 < length { // Encode two numeric letters in seven bits. - int num2 = content.charAt(i + 1) - '0'; + let num2 = content.charAt(i + 1) - '0'; bits.appendBits(num1 * 10 + num2, 7); i += 2; } else { // Encode one numeric letter in four bits. bits.appendBits(num1, 4); - i++; + i+=1; } } } - static void appendAlphanumericBytes(CharSequence content, BitArray bits) throws WriterException { - int length = content.length(); - int i = 0; - while (i < length) { - int code1 = getAlphanumericCode(content.charAt(i)); - if (code1 == -1) { - throw new WriterException(); + pub fn appendAlphanumericBytes( content:&str, bits:&BitArray) -> Result<(),Exceptions> { + let length = content.len(); + let i = 0; + while i < length { + let code1 = getAlphanumericCode(content.charAt(i)); + if code1 == -1 { + return Err(Exceptions::WriterException("".to_owned())); } - if (i + 1 < length) { - int code2 = getAlphanumericCode(content.charAt(i + 1)); + if i + 1 < length { + let code2 = getAlphanumericCode(content.charAt(i + 1)); if (code2 == -1) { - throw new WriterException(); + return Err(Exceptions::WriterException("".to_owned())); + } // Encode two alphanumeric letters in 11 bits. bits.appendBits(code1 * 45 + code2, 11); @@ -592,46 +615,52 @@ public final class Encoder { } else { // Encode one alphanumeric letter in six bits. bits.appendBits(code1, 6); - i++; + i+=1; } } + Ok(()) } - static void append8BitBytes(String content, BitArray bits, Charset encoding) { - byte[] bytes = content.getBytes(encoding); - for (byte b : bytes) { + fn append8BitBytes( content:&str, bits:&BitArray, encoding:EncodingRef) { + let bytes = content.getBytes(encoding); + for b in bytes { + // for (byte b : bytes) { bits.appendBits(b, 8); } } - static void appendKanjiBytes(String content, BitArray bits) throws WriterException { - byte[] bytes = content.getBytes(StringUtils.SHIFT_JIS_CHARSET); - if (bytes.length % 2 != 0) { - throw new WriterException("Kanji byte size not even"); + fn appendKanjiBytes( content:&str, bits:&BitArray) -> Result<(),Exceptions> { + let bytes = content.getBytes(StringUtils::SHIFT_JIS_CHARSET); + if bytes.length % 2 != 0 { + return Err(Exceptions::WriterException("Kanji byte size not even".to_owned())) } - int maxI = bytes.length - 1; // bytes.length must be even - for (int i = 0; i < maxI; i += 2) { - int byte1 = bytes[i] & 0xFF; - int byte2 = bytes[i + 1] & 0xFF; - int code = (byte1 << 8) | byte2; - int subtracted = -1; - if (code >= 0x8140 && code <= 0x9ffc) { + let maxI = bytes.length - 1; // bytes.length must be even + let mut i = 0; + while i < maxI { + // for (int i = 0; i < maxI; i += 2) { + let byte1 = bytes[i] & 0xFF; + let byte2 = bytes[i + 1] & 0xFF; + let code = (byte1 << 8) | byte2; + let subtracted = -1; + if code >= 0x8140 && code <= 0x9ffc { subtracted = code - 0x8140; } else if (code >= 0xe040 && code <= 0xebbf) { subtracted = code - 0xc140; } - if (subtracted == -1) { - throw new WriterException("Invalid byte sequence"); + if subtracted == -1 { + return Err(Exceptions::WriterException("Invalid byte sequence".to_owned())) } - int encoded = ((subtracted >> 8) * 0xc0) + (subtracted & 0xff); + let encoded = ((subtracted >> 8) * 0xc0) + (subtracted & 0xff); bits.appendBits(encoded, 13); + + i+=2; } + Ok(()) } - private static void appendECI(CharacterSetECI eci, BitArray bits) { - bits.appendBits(Mode.ECI.getBits(), 4); + fn appendECI( eci:&CharacterSetECI, bits:&BitArray) { + bits.appendBits(Mode::ECI.getBits(), 4); // This is correct for values up to 127, which is all we need now. bits.appendBits(eci.getValue(), 8); } -} diff --git a/src/qrcode/encoder/minimal_encoder.rs b/src/qrcode/encoder/minimal_encoder.rs new file mode 100644 index 0000000..5f71626 --- /dev/null +++ b/src/qrcode/encoder/minimal_encoder.rs @@ -0,0 +1,818 @@ +/* + * Copyright 2021 ZXing authors + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +use std::{fmt, rc::Rc}; + +use encoding::EncodingRef; + +use crate::{common::{ECIEncoderSet, BitArray}, qrcode::decoder::{ErrorCorrectionLevel, Version, Mode, VersionRef}, Exceptions}; + +use super::encoder; + + +enum VersionSize { + SMALL,//("version 1-9"), + MEDIUM,//("version 10-26"), + LARGE,//("version 27-40"); + + // private final String description; + + // VersionSize(String description) { + // this.description = description; + // } + + // public String toString() { + // return description; + // } +} + +impl fmt::Display for VersionSize { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + write!( f, "{}", match self { + VersionSize::SMALL => "version 1-9", + VersionSize::MEDIUM => "version 10-26", + VersionSize::LARGE => "version 27-40", + }) + } +} + + +/** + * Encoder that encodes minimally + * + * Algorithm: + * + * The eleventh commandment was "Thou Shalt Compute" or "Thou Shalt Not Compute" - I forget which (Alan Perilis). + * + * This implementation computes. As an alternative, the QR-Code specification suggests heuristics like this one: + * + * If initial input data is in the exclusive subset of the Alphanumeric character set AND if there are less than + * [6,7,8] characters followed by data from the remainder of the 8-bit byte character set, THEN select the 8- + * bit byte mode ELSE select Alphanumeric mode; + * + * This is probably right for 99.99% of cases but there is at least this one counter example: The string "AAAAAAa" + * encodes 2 bits smaller as ALPHANUMERIC(AAAAAA), BYTE(a) than by encoding it as BYTE(AAAAAAa). + * Perhaps that is the only counter example but without having proof, it remains unclear. + * + * ECI switching: + * + * In multi language content the algorithm selects the most compact representation using ECI modes. + * For example the most compact representation of the string "\u0150\u015C" (O-double-acute, S-circumflex) is + * ECI(UTF-8), BYTE(\u0150\u015C) while prepending one or more times the same leading character as in + * "\u0150\u0150\u015C", the most compact representation uses two ECIs so that the string is encoded as + * ECI(ISO-8859-2), BYTE(\u0150\u0150), ECI(ISO-8859-3), BYTE(\u015C). + * + * @author Alex Geller + */ +pub struct MinimalEncoder { + + stringToEncode: String, + isGS1: bool, + encoders: ECIEncoderSet, + ecLevel:ErrorCorrectionLevel, +} + +impl MinimalEncoder { + + /** + * Creates a MinimalEncoder + * + * @param stringToEncode The string to encode + * @param priorityCharset The preferred {@link Charset}. When the value of the argument is null, the algorithm + * chooses charsets that leads to a minimal representation. Otherwise the algorithm will use the priority + * charset to encode any character in the input that can be encoded by it if the charset is among the + * supported charsets. + * @param isGS1 {@code true} if a FNC1 is to be prepended; {@code false} otherwise + * @param ecLevel The error correction level. + * @see RXingResultList#getVersion + */ + pub fn new( stringToEncode:String, priorityCharset: EncodingRef, isGS1:bool, ecLevel:ErrorCorrectionLevel) -> Self { + Self { + stringToEncode, + isGS1, + encoders: ECIEncoderSet::new(&stringToEncode, priorityCharset, -1), + ecLevel, + } + + // let encoders = ECIEncoderSet::new(&stringToEncode, priorityCharset, -1); + + } + + /** + * Encodes the string minimally + * + * @param stringToEncode The string to encode + * @param version The preferred {@link Version}. A minimal version is computed (see + * {@link RXingResultList#getVersion method} when the value of the argument is null + * @param priorityCharset The preferred {@link Charset}. When the value of the argument is null, the algorithm + * chooses charsets that leads to a minimal representation. Otherwise the algorithm will use the priority + * charset to encode any character in the input that can be encoded by it if the charset is among the + * supported charsets. + * @param isGS1 {@code true} if a FNC1 is to be prepended; {@code false} otherwise + * @param ecLevel The error correction level. + * @return An instance of {@code RXingResultList} representing the minimal solution. + * @see RXingResultList#getBits + * @see RXingResultList#getVersion + * @see RXingResultList#getSize + */ + pub fn encode_with_details<'a>( stringToEncode:String, version:VersionRef, priorityCharset:EncodingRef, isGS1:bool, + ecLevel:ErrorCorrectionLevel) -> Result,Exceptions> { + MinimalEncoder::new(stringToEncode, priorityCharset, isGS1, ecLevel).encode(Some(version)) + } + + pub fn encode(&self, version:Option) -> Result { + if version.is_none() { // compute minimal encoding trying the three version sizes. + let versions = [ Self::getVersion(VersionSize::SMALL), + Self::getVersion(VersionSize::MEDIUM), + Self::getVersion(VersionSize::LARGE) ]; + let results = [ self.encodeSpecificVersion(&versions[0]), + self.encodeSpecificVersion(&versions[1]), + self.encodeSpecificVersion(&versions[2]) ]; + let smallestSize = u32::MAX; + let smallestRXingResult = -1; + for i in 0..3 { + // for (int i = 0; i < 3; i++) { + let size = results[i].getSize(); + if encoder::willFit(size, versions[i], self.ecLevel) && size < smallestSize { + smallestSize = size; + smallestRXingResult = i; + } + } + if smallestRXingResult < 0 { + return Err(Exceptions::WriterException("Data too big for any version".to_owned())); + } + return results[smallestRXingResult]; + } else { // compute minimal encoding for a given version + let result = self.encodeSpecificVersion(version); + if !encoder::willFit(result.getSize(), Self::getVersion(Self::getVersionSize(result.getVersion())), self.ecLevel) { + return Err(Exceptions::WriterException(format!("Data too big for version {}" , version))); + } + return result; + } + } + + pub fn getVersionSize( version:&Version) -> VersionSize { + return if version.getVersionNumber() <= 9 { VersionSize::SMALL} else {if version.getVersionNumber() <= 26 + {VersionSize::MEDIUM} else {VersionSize::LARGE}}; + } + + pub fn getVersion( versionSize:VersionSize)->VersionRef { + match versionSize { + VersionSize::SMALL => Version::getVersionForNumber(9).expect("should always exist"), + VersionSize::MEDIUM => Version::getVersionForNumber(26).expect("should always exist"), + VersionSize::LARGE => Version::getVersionForNumber(40).expect("should always exist"), + } + // switch (versionSize) { + // case SMALL: + // return Version.getVersionForNumber(9); + // case MEDIUM: + // return Version.getVersionForNumber(26); + // case LARGE: + // default: + // return Version.getVersionForNumber(40); + // } + } + + pub fn isNumeric( c:char) -> bool{ + return c >= '0' && c <= '9'; + } + + pub fn isDoubleByteKanji( c:char) -> bool{ + return encoder::isOnlyDoubleByteKanji(String.valueOf(c)); + } + + pub fn isAlphanumeric( c: char) -> bool{ + return encoder::getAlphanumericCode(c) != -1; + } + + pub fn canEncode(&self, mode:&Mode, c:char) -> bool { + match mode { + Mode::NUMERIC => Self::isNumeric(c), + Mode::ALPHANUMERIC => Self::isAlphanumeric(c), + Mode::STRUCTURED_APPEND => todo!(), + Mode::BYTE => true, + Mode::KANJI => Self::isDoubleByteKanji(c), + _=> false,// any character can be encoded as byte(s). Up to the caller to manage splitting into + // multiple bytes when String.getBytes(Charset) return more than one byte. + } + } + + pub fn getCompactedOrdinal( mode:Option) -> Result{ + if mode.is_none() { + return Ok(0); + } + match &mode.unwrap() { + Mode::NUMERIC => Ok(2), + Mode::ALPHANUMERIC => Ok(1), + Mode::BYTE => Ok(3), + Mode::KANJI => Ok(0), + _=> Err(Exceptions::IllegalArgumentException(format!("Illegal mode {:?}", mode))), + } + // switch (mode) { + // case KANJI: + // return 0; + // case ALPHANUMERIC: + // return 1; + // case NUMERIC: + // return 2; + // case BYTE: + // return 3; + // default: + // throw new IllegalStateException("Illegal mode " + mode); + // } + } + + pub fn addEdge(&self, edges:&Vec>>, position:u32, edge:Option<&Edge>) { + let vertexIndex = position + edge.as_ref().unwrap().characterLength; + let modeEdges = edges[vertexIndex as usize][edge.as_ref().unwrap().charsetEncoderIndex as usize]; + let modeOrdinal = Self::getCompactedOrdinal(Some(edge.as_ref().unwrap().mode)).expect("value") as usize; + if /*modeEdges[modeOrdinal] == null ||*/ modeEdges[modeOrdinal].cachedTotalSize > (&edge).unwrap().cachedTotalSize { + modeEdges[modeOrdinal] = edge.unwrap(); + } + } + + pub fn addEdges( &self, version:&Edge, edges:&Vec>>, from:u32, previous:&Edge) { + let start = 0; + let end = self.encoders.len(); + let priorityEncoderIndex = self.encoders.getPriorityEncoderIndex(); + if priorityEncoderIndex >= 0 && self.encoders.canEncode(self.stringToEncode.chars().nth(from).unwrap(),priorityEncoderIndex) { + start = priorityEncoderIndex; + end = priorityEncoderIndex + 1; + } + + for i in start..end { + // for (int i = start; i < end; i++) { + if self.encoders.canEncode(self.stringToEncode.chars().nth(from).unwrap(), i) { + self.addEdge(edges, from, Edge::new(Mode::BYTE, from, i, 1, previous, version)); + } + } + + if self.canEncode(Mode::KANJI, self.stringToEncode.chars().nth(from).unwrap()) { + self.addEdge(edges, from, Edge::(Mode::KANJI, from, 0, 1, previous, version)); + } + + let inputLength = self.stringToEncode.len(); + if self.canEncode(Mode::ALPHANUMERIC, self.stringToEncode.charAt(from)) { + self.addEdge(edges, from, Edge::new(Mode::ALPHANUMERIC, from, 0, from + 1 >= inputLength || + !self.canEncode(Mode::ALPHANUMERIC, self.stringToEncode.charAt(from + 1)) ? 1 : 2, previous, version)); + } + + if self.canEncode(Mode::NUMERIC, self.stringToEncode.chars().nth(from).unwrap()) { + self.addEdge(edges, from, Edge::new(Mode::NUMERIC, from, 0, from + 1 >= inputLength || + !self.canEncode(Mode::NUMERIC, self.stringToEncode.charAt(from + 1)) ? 1 : from + 2 >= inputLength || + !self.canEncode(Mode::NUMERIC, self.stringToEncode.charAt(from + 2)) ? 2 : 3, previous, version)); + } + } + pub fn encodeSpecificVersion(&self, version:VersionRef) ->Result< RXingResultList, Exceptions >{ + + // @SuppressWarnings("checkstyle:lineLength") + /* A vertex represents a tuple of a position in the input, a mode and a character encoding where position 0 + * denotes the position left of the first character, 1 the position left of the second character and so on. + * Likewise the end vertices are located after the last character at position stringToEncode.length(). + * + * An edge leading to such a vertex encodes one or more of the characters left of the position that the vertex + * represents and encodes it in the same encoding and mode as the vertex on which the edge ends. In other words, + * all edges leading to a particular vertex encode the same characters in the same mode with the same character + * encoding. They differ only by their source vertices who are all located at i+1 minus the number of encoded + * characters. + * + * The edges leading to a vertex are stored in such a way that there is a fast way to enumerate the edges ending + * on a particular vertex. + * + * The algorithm processes the vertices in order of their position thereby performing the following: + * + * For every vertex at position i the algorithm enumerates the edges ending on the vertex and removes all but the + * shortest from that list. + * Then it processes the vertices for the position i+1. If i+1 == stringToEncode.length() then the algorithm ends + * and chooses the the edge with the smallest size from any of the edges leading to vertices at this position. + * Otherwise the algorithm computes all possible outgoing edges for the vertices at the position i+1 + * + * Examples: + * The process is illustrated by showing the graph (edges) after each iteration from left to right over the input: + * An edge is drawn as follows "(" + fromVertex + ") -- " + encodingMode + "(" + encodedInput + ") (" + + * accumulatedSize + ") --> (" + toVertex + ")" + * + * Example 1 encoding the string "ABCDE": + * Note: This example assumes that alphanumeric encoding is only possible in multiples of two characters so that + * the example is both short and showing the principle. In reality this restriction does not exist. + * + * Initial situation + * (initial) -- BYTE(A) (20) --> (1_BYTE) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) + * + * Situation after adding edges to vertices at position 1 + * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) + * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) + * + * Situation after adding edges to vertices at position 2 + * (initial) -- BYTE(A) (20) --> (1_BYTE) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) + * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) + * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- BYTE(C) (44) --> (3_BYTE) + * (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) + * + * Situation after adding edges to vertices at position 3 + * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) + * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- BYTE(D) (64) --> (4_BYTE) + * (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) + * (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) + * + * Situation after adding edges to vertices at position 4 + * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) -- BYTE(D) (44) --> (4_BYTE) + * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) -- BYTE(E) (55) --> (5_BYTE) + * + * Situation after adding edges to vertices at position 5 + * (initial) -- BYTE(A) (20) --> (1_BYTE) -- BYTE(B) (28) --> (2_BYTE) -- BYTE(C) (36) --> (3_BYTE) -- BYTE(D) (44) --> (4_BYTE) -- BYTE(E) (52) --> (5_BYTE) + * (1_BYTE) -- ALPHANUMERIC(BC) (44) --> (3_ALPHANUMERIC) -- ALPHANUMERIC(DE) (55) --> (5_ALPHANUMERIC) + * (initial) -- ALPHANUMERIC(AB) (24) --> (2_ALPHANUMERIC) -- ALPHANUMERIC(CD) (35) --> (4_ALPHANUMERIC) + * + * Encoding as BYTE(ABCDE) has the smallest size of 52 and is hence chosen. The encodation ALPHANUMERIC(ABCD), + * BYTE(E) is longer with a size of 55. + * + * Example 2 encoding the string "XXYY" where X denotes a character unique to character set ISO-8859-2 and Y a + * character unique to ISO-8859-3. Both characters encode as double byte in UTF-8: + * + * Initial situation + * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) + * + * Situation after adding edges to vertices at position 1 + * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) + * (1_BYTE_ISO-8859-2) -- BYTE(X) (72) --> (2_BYTE_UTF-8) + * (1_BYTE_ISO-8859-2) -- BYTE(X) (72) --> (2_BYTE_UTF-16BE) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) + * + * Situation after adding edges to vertices at position 2 + * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) + * (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) + * (2_BYTE_ISO-8859-2) -- BYTE(Y) (80) --> (3_BYTE_UTF-8) + * (2_BYTE_ISO-8859-2) -- BYTE(Y) (80) --> (3_BYTE_UTF-16BE) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) + * + * Situation after adding edges to vertices at position 3 + * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) + * (3_BYTE_ISO-8859-3) -- BYTE(Y) (80) --> (4_BYTE_ISO-8859-3) + * (3_BYTE_ISO-8859-3) -- BYTE(Y) (112) --> (4_BYTE_UTF-8) + * (3_BYTE_ISO-8859-3) -- BYTE(Y) (112) --> (4_BYTE_UTF-16BE) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) -- BYTE(Y) (72) --> (3_BYTE_UTF-8) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) -- BYTE(Y) (72) --> (3_BYTE_UTF-16BE) + * + * Situation after adding edges to vertices at position 4 + * (initial) -- BYTE(X) (32) --> (1_BYTE_ISO-8859-2) -- BYTE(X) (40) --> (2_BYTE_ISO-8859-2) -- BYTE(Y) (72) --> (3_BYTE_ISO-8859-3) -- BYTE(Y) (80) --> (4_BYTE_ISO-8859-3) + * (3_BYTE_UTF-8) -- BYTE(Y) (88) --> (4_BYTE_UTF-8) + * (3_BYTE_UTF-16BE) -- BYTE(Y) (88) --> (4_BYTE_UTF-16BE) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-8) -- BYTE(X) (56) --> (2_BYTE_UTF-8) -- BYTE(Y) (72) --> (3_BYTE_UTF-8) + * (initial) -- BYTE(X) (40) --> (1_BYTE_UTF-16BE) -- BYTE(X) (56) --> (2_BYTE_UTF-16BE) -- BYTE(Y) (72) --> (3_BYTE_UTF-16BE) + * + * Encoding as ECI(ISO-8859-2),BYTE(XX),ECI(ISO-8859-3),BYTE(YY) has the smallest size of 80 and is hence chosen. + * The encodation ECI(UTF-8),BYTE(XXYY) is longer with a size of 88. + */ + + let inputLength = self.stringToEncode.len(); + + // Array that represents vertices. There is a vertex for every character, encoding and mode. The vertex contains + // a list of all edges that lead to it that have the same encoding and mode. + // The lists are created lazily + + // The last dimension in the array below encodes the 4 modes KANJI, ALPHANUMERIC, NUMERIC and BYTE via the + // function getCompactedOrdinal(Mode) + let edges = vec![vec![vec![]]];//new Edge[inputLength + 1][encoders.length()][4]; + self. addEdges(version, &edges, 0, None); + + for i in 1..=inputLength { + // for (int i = 1; i <= inputLength; i++) { + for j in 0..self.encoders.len() { + // for (int j = 0; j < encoders.length(); j++) { + for k in 0..4 { + // for (int k = 0; k < 4; k++) { + if edges[i][j][k] != null && i < inputLength { + self.addEdges(version, edges, i, edges[i][j][k]); + } + } + } + + } + let minimalJ = -1; + let minimalK = -1; + let minimalSize = u32::MAX; + for j in 0..self.encoders.len() { + // for (int j = 0; j < encoders.length(); j++) { + for k in 0..4 { + // for (int k = 0; k < 4; k++) { + if edges[inputLength][j][k] != null { + let edge = edges[inputLength][j][k]; + if edge.cachedTotalSize < minimalSize { + minimalSize = edge.cachedTotalSize; + minimalJ = j; + minimalK = k; + } + } + } + } + if minimalJ < 0 { + return Err(Exceptions::WriterException(format!(r#"Internal error: failed to encode "{}"#,self.stringToEncode))); + } + Ok(RXingResultList::new(version, edges[inputLength][minimalJ][minimalK])) + } +} + + +struct Edge<'a> { + mode:Mode, + fromPosition:u32, + charsetEncoderIndex:u32, + characterLength:u32, + previous:Option<&'a Edge<'a>>, + cachedTotalSize:u32, + encoders:&'a ECIEncoderSet, + stringToEncode: &'a str, +} +impl Edge<'_> { + + pub fn new( mode:Mode, fromPosition:u32, charsetEncoderIndex:u32, characterLength:u32, previous:Option<&'_ Edge>, + version:&Version, encoders: &'_ ECIEncoderSet, stringToEncode: &'_ str) -> Self { + let nci = if mode == Mode::BYTE || previous.is_none() {charsetEncoderIndex} else + {previous.as_ref().unwrap().charsetEncoderIndex}; + Self { + mode, + fromPosition, + charsetEncoderIndex: nci, + characterLength, + previous, + stringToEncode, + cachedTotalSize: { + let size = if previous.is_some() {previous.as_ref().unwrap().cachedTotalSize} else {0}; + + let needECI = mode == Mode::BYTE && + (previous.is_none() && nci != 0) || // at the beginning and charset is not ISO-8859-1 + (previous.is_some() && nci != previous.as_ref().unwrap().charsetEncoderIndex); + + if previous.is_none()|| mode != previous.as_ref().unwrap().mode || needECI { + size += 4 + mode.getCharacterCountBits(&version) as u32; + } + match mode { + Mode::NUMERIC => size += if characterLength == 1 {4} else {if characterLength == 2 {7} else {10}}, + Mode::ALPHANUMERIC => size += if characterLength == 1 {6} else {11}, + Mode::BYTE =>{ + size += 8 * encoders.encode_string(&stringToEncode[fromPosition as usize..(fromPosition + characterLength)as usize], + charsetEncoderIndex as usize).len() as u32; + if needECI { + size += 4 + 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long + } + }, + Mode::KANJI => size += 13, + _=> {}, + } + // switch (mode) { + // case KANJI: + // size += 13; + // break; + // case ALPHANUMERIC: + // size += characterLength == 1 ? 6 : 11; + // break; + // case NUMERIC: + // size += characterLength == 1 ? 4 : characterLength == 2 ? 7 : 10; + // break; + // case BYTE: + // size += 8 * encoders.encode(stringToEncode.substring(fromPosition, fromPosition + characterLength), + // charsetEncoderIndex).length; + // if (needECI) { + // size += 4 + 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long + // } + // break; + // } + size + }, + encoders + } + // this.mode = mode; + // this.fromPosition = fromPosition; + // this.charsetEncoderIndex = mode == Mode.BYTE || previous == null ? charsetEncoderIndex : + // previous.charsetEncoderIndex; // inherit the encoding if not of type BYTE + // this.characterLength = characterLength; + // this.previous = previous; + + // int size = previous != null ? previous.cachedTotalSize : 0; + + // boolean needECI = mode == Mode.BYTE && + // (previous == null && this.charsetEncoderIndex != 0) || // at the beginning and charset is not ISO-8859-1 + // (previous != null && this.charsetEncoderIndex != previous.charsetEncoderIndex); + + // if (previous == null || mode != previous.mode || needECI) { + // size += 4 + mode.getCharacterCountBits(version); + // } + // switch (mode) { + // case KANJI: + // size += 13; + // break; + // case ALPHANUMERIC: + // size += characterLength == 1 ? 6 : 11; + // break; + // case NUMERIC: + // size += characterLength == 1 ? 4 : characterLength == 2 ? 7 : 10; + // break; + // case BYTE: + // size += 8 * encoders.encode(stringToEncode.substring(fromPosition, fromPosition + characterLength), + // charsetEncoderIndex).length; + // if (needECI) { + // size += 4 + 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long + // } + // break; + // } + // cachedTotalSize = size; + } +} + +struct RXingResultList<'a> { + list: Vec>, + version: VersionRef, +} +impl RXingResultList<'_> { + + pub fn new( version:VersionRef, solution:&'_ Edge, isGS1:bool, ecLevel: &ErrorCorrectionLevel) -> Self { + let length = 0; + let current = Some(solution); + let containsECI = false; + let list = Vec::new(); + + while current.is_some() { + length += current.as_ref().unwrap().characterLength; + let previous = current.as_ref().unwrap().previous; + + let needECI = current.as_ref().unwrap().mode == Mode::BYTE && + (previous.is_none() && current.as_ref().unwrap().charsetEncoderIndex != 0) || // at the beginning and charset is not ISO-8859-1 + (previous.is_some() && current.as_ref().unwrap().charsetEncoderIndex != previous.as_ref().unwrap().charsetEncoderIndex); + + if needECI { + containsECI = true; + } + + if previous.is_none() || previous.as_ref().unwrap().mode != current.as_ref().unwrap().mode || needECI { + list.push( RXingResultNode::new(current.mode, current.fromPosition, current.charsetEncoderIndex, length)); + length = 0; + } + + if needECI { + list.push( RXingResultNode::new(Mode::ECI, current.fromPosition, current.charsetEncoderIndex, 0)); + } + current = previous; + } + + // prepend FNC1 if needed. If the bits contain an ECI then the FNC1 must be preceeded by an ECI. + // If there is no ECI at the beginning then we put an ECI to the default charset (ISO-8859-1) + if isGS1 { + if let Some(first) = list.get(0){ + // if first != null && first.mode != Mode.ECI && containsECI { + if first.mode != Mode::ECI && containsECI { + // prepend a default character set ECI + list.push(0, RXingResultNode::new(Mode::ECI, 0, 0, 0)); + }} + let first = list.get(0).unwrap(); + // prepend or insert a FNC1_FIRST_POSITION after the ECI (if any) + // if first != null && first.mode != Mode.ECI && containsECI { + list.push( RXingResultNode::new(Mode::FNC1_FIRST_POSITION, 0, 0, 0)); + } + + // set version to smallest version into which the bits fit. + let versionNumber = version.getVersionNumber(); + let (lowerLimit,upperLimit) = + match MinimalEncoder::getVersionSize(&version) { + VersionSize::SMALL => + (1,9), + VersionSize::MEDIUM=>(10,26), + _=>(27,40), + }; + // let lowerLimit; + // let upperLimit; + // switch (getVersionSize(version)) { + // case SMALL: + // lowerLimit = 1; + // upperLimit = 9; + // break; + // case MEDIUM: + // lowerLimit = 10; + // upperLimit = 26; + // break; + // case LARGE: + // default: + // lowerLimit = 27; + // upperLimit = 40; + // break; + // } + let size = Self::internal_static_get_size(version,&list); + // increase version if needed + while versionNumber < upperLimit && !encoder::willFit(size, Version::getVersionForNumber(versionNumber), + ecLevel) { + versionNumber+=1; + } + // shrink version if possible + while versionNumber > lowerLimit && encoder::willFit(size, Version::getVersionForNumber(versionNumber - 1), + ecLevel) { + versionNumber-=1; + } + let version = Version::getVersionForNumber(versionNumber).unwrap(); + Self { + list, + version, + } + } + + /** + * returns the size in bits + */ + pub fn getSize(&self) -> u32{ + self. getSizeLocal(self.version) + } + + fn getSizeLocal(&self, version:VersionRef) -> u32{ + let result = 0; + for resultNode in &self.list { + result += resultNode.getSize(version); + } + return result; + } + + fn internal_static_get_size(version:VersionRef, list: &Vec) { + let result = 0; + for resultNode in list { + result += resultNode.getSize(version); + } + return result; + } + + /** + * appends the bits + */ + pub fn getBits(&self, bits:&BitArray) -> Result<(),Exceptions> { + for resultNode in &self.list { + resultNode.getBits(bits); + } + Ok(()) + } + + pub fn getVersion(&self) -> &Version{ + &self. version + } +} + +impl fmt::Display for RXingResultList<'_> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let mut result = String::new(); + let previous = None; + for current in &self.list { + // for (RXingResultNode current : list) { + if previous.is_some() { + result.push_str(","); + } + result.push_str(¤t.to_string()); + previous = Some(current); + } + write!(f,"{}", result) + } +} + +struct RXingResultNode<'a> { + + mode:Mode, + fromPosition:u32, + charsetEncoderIndex:u32, + characterLength:u32, + encoders:&'a ECIEncoderSet, + version: VersionRef, + stringToEncode: &'a str, +} + +impl RXingResultNode<'_> { + + pub fn new( mode:Mode, fromPosition:u32, charsetEncoderIndex:u32, characterLength:u32, encoders:&'_ ECIEncoderSet, stringToEncode: &'_ str, version: &'_ Version) -> Self { + Self { + mode, + fromPosition, + charsetEncoderIndex, + characterLength, + encoders, + stringToEncode, + version, + } + } + + /** + * returns the size in bits + */ + fn getSize(&self, version:&Version) -> u32{ + let size = 4 + self.mode.getCharacterCountBits(version) as u32; + match self.mode { + Mode::NUMERIC => { + size += (self.characterLength / 3) * 10; + let rest = self.characterLength % 3; + size += if rest == 1 {4} else { if rest == 2 {7} else {0}}; + }, + Mode::ALPHANUMERIC => { + size += (self.characterLength / 2) * 11; + size += if (self.characterLength % 2) == 1 {6} else {0}; + }, + Mode::BYTE => size += 8 * self.getCharacterCountIndicator(), + Mode::ECI => size += 8, + Mode::KANJI => size += 13 * self.characterLength, + _ => {}, + } + // switch (mode) { + // case KANJI: + // size += 13 * characterLength; + // break; + // case ALPHANUMERIC: + // size += (characterLength / 2) * 11; + // size += (characterLength % 2) == 1 ? 6 : 0; + // break; + // case NUMERIC: + // size += (characterLength / 3) * 10; + // int rest = characterLength % 3; + // size += rest == 1 ? 4 : rest == 2 ? 7 : 0; + // break; + // case BYTE: + // size += 8 * getCharacterCountIndicator(); + // break; + // case ECI: + // size += 8; // the ECI assignment numbers for ISO-8859-x, UTF-8 and UTF-16 are all 8 bit long + // } + size + } + + /** + * returns the length in characters according to the specification (differs from getCharacterLength() in BYTE mode + * for multi byte encoded characters) + */ + fn getCharacterCountIndicator(&self) -> u32{ + if self.mode == Mode::BYTE + {self.encoders.encode_string(&self.stringToEncode[self.fromPosition as usize..(self.fromPosition + self.characterLength) as usize], + self.charsetEncoderIndex as usize).len() as u32} else {self.characterLength} + } + + /** + * appends the bits + */ + fn getBits(&self, bits:&BitArray) -> Result<(),Exceptions> { + bits.appendBits(self.mode.getBits() as u32, 4); + if self.characterLength > 0 { + let length = self.getCharacterCountIndicator(); + bits.appendBits(length, self.mode.getCharacterCountBits(self.version.as_ref()) as usize); + } + if self.mode == Mode::ECI { + bits.appendBits(self.encoders.getECIValue(self.charsetEncoderIndex as usize), 8); + } else if self.characterLength > 0 { + // append data + encoder::appendBytes(self.stringToEncode[self.fromPosition as usize..(self.fromPosition + self.characterLength) as usize], self.mode, bits, + self.encoders.getCharset(self.charsetEncoderIndex as usize)); + } + Ok(()) + } + + fn makePrintable( s:&str) -> String { + let mut result = String::new(); + for i in 0..s.chars().count() { + // for (int i = 0; i < s.length(); i++) { + if (s.chars().nth(i).unwrap() as u8) < 32 || (s.chars().nth(i).unwrap() as u8) > 126 { + result.push('.'); + } else { + result.push(s.chars().nth(i).unwrap()); + } + } + result + } + } + + impl fmt::Display for RXingResultNode<'_> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let result = String::new(); + result.push_str(&format!("{:?}",self.mode)); + result.push('('); + if self.mode == Mode::ECI { + result.push_str(self.encoders.getCharset(self.charsetEncoderIndex as usize).name()); + } else { + result.push_str(&Self::makePrintable(&self.stringToEncode[self.fromPosition as usize..(self.fromPosition + self.characterLength) as usize])); + } + result.push(')'); + + write!(f,"{}", result) + } +} \ No newline at end of file diff --git a/src/qrcode/encoder/mod.rs b/src/qrcode/encoder/mod.rs index 7625dc5..5aca201 100644 --- a/src/qrcode/encoder/mod.rs +++ b/src/qrcode/encoder/mod.rs @@ -3,10 +3,13 @@ mod byte_matrix; mod block_pair; pub mod mask_util; pub mod matrix_util; +mod minimal_encoder; +pub mod encoder; pub use qr_code::*; pub use byte_matrix::*; pub use block_pair::*; +pub use minimal_encoder::*; #[cfg(test)] mod QRCodeTestCase;