From 247cd995b8e49c7923c3832d08d89fdc7d56ebce Mon Sep 17 00:00:00 2001 From: Andy Stadler <> Date: Tue, 24 Mar 2009 19:53:01 -0700 Subject: [PATCH] Automated import from //branches/master/...@141239,141239 --- .../email/mail/internet/MimeMessage.java | 3 +- .../email/mail/internet/MimeUtility.java | 117 +++- .../james/mime4j/codec/EncoderUtil.java | 626 ++++++++++++++++++ .../james/mime4j/decoder/DecoderUtil.java | 234 ++++--- .../apache/james/mime4j/util/CharsetUtil.java | 67 ++ .../email/mail/internet/MimeMessageTest.java | 98 ++- .../email/mail/internet/MimeUtilityTest.java | 115 +++- 7 files changed, 1128 insertions(+), 132 deletions(-) create mode 100644 src/org/apache/james/mime4j/codec/EncoderUtil.java diff --git a/src/com/android/email/mail/internet/MimeMessage.java b/src/com/android/email/mail/internet/MimeMessage.java index 27d7aadb4..371c5b1de 100644 --- a/src/com/android/email/mail/internet/MimeMessage.java +++ b/src/com/android/email/mail/internet/MimeMessage.java @@ -247,7 +247,8 @@ public class MimeMessage extends Message { } public void setSubject(String subject) throws MessagingException { - setHeader("Subject", subject); + final int HEADER_NAME_LENGTH = 9; // "Subject: " + setHeader("Subject", MimeUtility.foldAndEncode2(subject, HEADER_NAME_LENGTH)); } public Address[] getFrom() throws MessagingException { diff --git a/src/com/android/email/mail/internet/MimeUtility.java b/src/com/android/email/mail/internet/MimeUtility.java index 7aa740221..008b60adf 100644 --- a/src/com/android/email/mail/internet/MimeUtility.java +++ b/src/com/android/email/mail/internet/MimeUtility.java @@ -25,6 +25,7 @@ import com.android.email.mail.Multipart; import com.android.email.mail.Part; import org.apache.commons.io.IOUtils; +import org.apache.james.mime4j.codec.EncoderUtil; import org.apache.james.mime4j.decoder.Base64InputStream; import org.apache.james.mime4j.decoder.DecoderUtil; import org.apache.james.mime4j.decoder.QuotedPrintableInputStream; @@ -37,14 +38,27 @@ import java.io.IOException; import java.io.InputStream; import java.io.OutputStream; import java.util.ArrayList; +import java.util.regex.Matcher; import java.util.regex.Pattern; public class MimeUtility { + + private final static Pattern PATTERN_CR_OR_LF = Pattern.compile("\r|\n"); + + /** + * Replace sequences of CRLF+WSP with WSP. Tries to preserve original string + * object whenever possible. + */ public static String unfold(String s) { if (s == null) { return null; } - return s.replaceAll("\r|\n", ""); + Matcher patternMatcher = PATTERN_CR_OR_LF.matcher(s); + if (patternMatcher.find()) { + patternMatcher.reset(); + s = patternMatcher.replaceAll(""); + } + return s; } public static String decode(String s) { @@ -59,9 +73,99 @@ public class MimeUtility { } // TODO implement proper foldAndEncode + // NOTE: When this really works, we *must* remove all calls to foldAndEncode2() to prevent + // duplication of encoding. public static String foldAndEncode(String s) { return s; } + + /** + * INTERIM version of foldAndEncode that will be used only by Subject: headers. + * This is safer than implementing foldAndEncode() (see above) and risking unknown damage + * to other headers. + * + * TODO: Copy this code to foldAndEncode(), get rid of this function, confirm all working OK. + * + * @param s original string to encode and fold + * @param usedCharacters number of characters already used up by header name + + * @return the String ready to be transmitted + */ + public static String foldAndEncode2(String s, int usedCharacters) { + // james.mime4j.codec.EncoderUtil.java + // encode: encodeIfNecessary(text, usage, numUsedInHeaderName) + // Usage.TEXT_TOKENlooks like the right thing for subjects + // use WORD_ENTITY for address/names + + String encoded = EncoderUtil.encodeIfNecessary(s, EncoderUtil.Usage.TEXT_TOKEN, + usedCharacters); + + return fold(encoded, usedCharacters); + } + + /** + * INTERIM: From newer version of org.apache.james (but we don't want to import + * the entire MimeUtil class). + * + * Splits the specified string into a multiple-line representation with + * lines no longer than 76 characters (because the line might contain + * encoded words; see RFC + * 2047 section 2). If the string contains non-whitespace sequences + * longer than 76 characters a line break is inserted at the whitespace + * character following the sequence resulting in a line longer than 76 + * characters. + * + * @param s + * string to split. + * @param usedCharacters + * number of characters already used up. Usually the number of + * characters for header field name plus colon and one space. + * @return a multiple-line representation of the given string. + */ + public static String fold(String s, int usedCharacters) { + final int maxCharacters = 76; + + final int length = s.length(); + if (usedCharacters + length <= maxCharacters) + return s; + + StringBuilder sb = new StringBuilder(); + + int lastLineBreak = -usedCharacters; + int wspIdx = indexOfWsp(s, 0); + while (true) { + if (wspIdx == length) { + sb.append(s.substring(Math.max(0, lastLineBreak))); + return sb.toString(); + } + + int nextWspIdx = indexOfWsp(s, wspIdx + 1); + + if (nextWspIdx - lastLineBreak > maxCharacters) { + sb.append(s.substring(Math.max(0, lastLineBreak), wspIdx)); + sb.append("\r\n"); + lastLineBreak = wspIdx; + } + + wspIdx = nextWspIdx; + } + } + + /** + * INTERIM: From newer version of org.apache.james (but we don't want to import + * the entire MimeUtil class). + * + * Search for whitespace. + */ + private static int indexOfWsp(String s, int fromIndex) { + final int len = s.length(); + for (int index = fromIndex; index < len; index++) { + char c = s.charAt(index); + if (c == ' ' || c == '\t') + return index; + } + return len; + } /** * Returns the named parameter of a header field. If name is null the first @@ -69,6 +173,11 @@ public class MimeUtility { * field the entire field is returned. Otherwise the named parameter is * searched for in a case insensitive fashion and returned. If the parameter * cannot be found the method returns null. + * + * TODO: quite inefficient with the inner trimming & splitting. + * TODO: Also has a latent bug: uses "startsWith" to match the name, which can false-positive. + * TODO: The doc says that for a null name you get the first param, but you get the header. + * Should probably just fix the doc, but if other code assumes that behavior, fix the code. * * @param header * @param name @@ -78,13 +187,13 @@ public class MimeUtility { if (header == null) { return null; } - header = header.replaceAll("\r|\n", ""); - String[] parts = header.split(";"); + String[] parts = unfold(header).split(";"); if (name == null) { return parts[0]; } + String lowerCaseName = name.toLowerCase(); for (String part : parts) { - if (part.trim().toLowerCase().startsWith(name.toLowerCase())) { + if (part.trim().toLowerCase().startsWith(lowerCaseName)) { String parameter = part.split("=", 2)[1].trim(); if (parameter.startsWith("\"") && parameter.endsWith("\"")) { return parameter.substring(1, parameter.length() - 1); diff --git a/src/org/apache/james/mime4j/codec/EncoderUtil.java b/src/org/apache/james/mime4j/codec/EncoderUtil.java new file mode 100644 index 000000000..d6f3998d4 --- /dev/null +++ b/src/org/apache/james/mime4j/codec/EncoderUtil.java @@ -0,0 +1,626 @@ +/**************************************************************** + * Licensed to the Apache Software Foundation (ASF) under one * + * or more contributor license agreements. See the NOTICE file * + * distributed with this work for additional information * + * regarding copyright ownership. The ASF licenses this file * + * to you 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 org.apache.james.mime4j.codec; + +import java.nio.ByteBuffer; +import java.nio.charset.Charset; +import java.util.BitSet; +import java.util.Locale; + +import org.apache.james.mime4j.util.CharsetUtil; + +/** + * ANDROID: THIS CLASS IS COPIED FROM A NEWER VERSION OF MIME4J + */ + +/** + * Static methods for encoding header field values. This includes encoded-words + * as defined in RFC 2047 + * or display-names of an e-mail address, for example. + * + */ +public class EncoderUtil { + + // This array is a lookup table that translates 6-bit positive integer index + // values into their "Base64 Alphabet" equivalents as specified in Table 1 + // of RFC 2045. + // ANDROID: THIS TABLE IS COPIED FROM BASE64OUTPUTSTREAM + static final byte[] BASE64_TABLE = { 'A', 'B', 'C', 'D', 'E', 'F', + 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', + 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f', + 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', + 't', 'u', 'v', 'w', 'x', 'y', 'z', '0', '1', '2', '3', '4', '5', + '6', '7', '8', '9', '+', '/' }; + + // Byte used to pad output. + private static final byte BASE64_PAD = '='; + + private static final BitSet Q_REGULAR_CHARS = initChars("=_?"); + + private static final BitSet Q_RESTRICTED_CHARS = initChars("=_?\"#$%&'(),.:;<>@[\\]^`{|}~"); + + private static final int MAX_USED_CHARACTERS = 50; + + private static final String ENC_WORD_PREFIX = "=?"; + private static final String ENC_WORD_SUFFIX = "?="; + + private static final int ENCODED_WORD_MAX_LENGTH = 75; // RFC 2047 + + private static final BitSet TOKEN_CHARS = initChars("()<>@,;:\\\"/[]?="); + + private static final BitSet ATEXT_CHARS = initChars("()<>@.,;:\\\"[]"); + + private static BitSet initChars(String specials) { + BitSet bs = new BitSet(128); + for (char ch = 33; ch < 127; ch++) { + if (specials.indexOf(ch) == -1) { + bs.set(ch); + } + } + return bs; + } + + /** + * Selects one of the two encodings specified in RFC 2047. + */ + public enum Encoding { + /** The B encoding (identical to base64 defined in RFC 2045). */ + B, + /** The Q encoding (similar to quoted-printable defined in RFC 2045). */ + Q + } + + /** + * Indicates the intended usage of an encoded word. + */ + public enum Usage { + /** + * Encoded word is used to replace a 'text' token in any Subject or + * Comments header field. + */ + TEXT_TOKEN, + /** + * Encoded word is used to replace a 'word' entity within a 'phrase', + * for example, one that precedes an address in a From, To, or Cc + * header. + */ + WORD_ENTITY + } + + private EncoderUtil() { + } + + /** + * Encodes the display-name portion of an address. See RFC 5322 section 3.4 + * and RFC 2047 section + * 5.3. The specified string should not be folded. + * + * @param displayName + * display-name to encode. + * @return encoded display-name. + */ + public static String encodeAddressDisplayName(String displayName) { + // display-name = phrase + // phrase = 1*( encoded-word / word ) + // word = atom / quoted-string + // atom = [CFWS] 1*atext [CFWS] + // CFWS = comment or folding white space + + if (isAtomPhrase(displayName)) { + return displayName; + } else if (hasToBeEncoded(displayName, 0)) { + return encodeEncodedWord(displayName, Usage.WORD_ENTITY); + } else { + return quote(displayName); + } + } + + /** + * Encodes the local part of an address specification as described in RFC + * 5322 section 3.4.1. Leading and trailing CFWS should have been removed + * before calling this method. The specified string should not contain any + * illegal (control or non-ASCII) characters. + * + * @param localPart + * the local part to encode + * @return the encoded local part. + */ + public static String encodeAddressLocalPart(String localPart) { + // local-part = dot-atom / quoted-string + // dot-atom = [CFWS] dot-atom-text [CFWS] + // CFWS = comment or folding white space + + if (isDotAtomText(localPart)) { + return localPart; + } else { + return quote(localPart); + } + } + + /** + * Encodes the specified strings into a header parameter as described in RFC + * 2045 section 5.1 and RFC 2183 section 2. The specified strings should not + * contain any illegal (control or non-ASCII) characters. + * + * @param name + * parameter name. + * @param value + * parameter value. + * @return encoded result. + */ + public static String encodeHeaderParameter(String name, String value) { + name = name.toLowerCase(Locale.US); + + // value := token / quoted-string + if (isToken(value)) { + return name + "=" + value; + } else { + return name + "=" + quote(value); + } + } + + /** + * Shortcut method that encodes the specified text into an encoded-word if + * the text has to be encoded. + * + * @param text + * text to encode. + * @param usage + * whether the encoded-word is to be used to replace a text token + * or a word entity (see RFC 822). + * @param usedCharacters + * number of characters already used up (0 <= usedCharacters <= 50). + * @return the specified text if encoding is not necessary or an encoded + * word or a sequence of encoded words otherwise. + */ + public static String encodeIfNecessary(String text, Usage usage, + int usedCharacters) { + if (hasToBeEncoded(text, usedCharacters)) + return encodeEncodedWord(text, usage, usedCharacters); + else + return text; + } + + /** + * Determines if the specified string has to encoded into an encoded-word. + * Returns true if the text contains characters that don't + * fall into the printable ASCII character set or if the text contains a + * 'word' (sequence of non-whitespace characters) longer than 77 characters + * (including characters already used up in the line). + * + * @param text + * text to analyze. + * @param usedCharacters + * number of characters already used up (0 <= usedCharacters <= 50). + * @return true if the specified text has to be encoded into + * an encoded-word, false otherwise. + */ + public static boolean hasToBeEncoded(String text, int usedCharacters) { + if (text == null) + throw new IllegalArgumentException(); + if (usedCharacters < 0 || usedCharacters > MAX_USED_CHARACTERS) + throw new IllegalArgumentException(); + + int nonWhiteSpaceCount = usedCharacters; + + for (int idx = 0; idx < text.length(); idx++) { + char ch = text.charAt(idx); + if (ch == '\t' || ch == ' ') { + nonWhiteSpaceCount = 0; + } else { + nonWhiteSpaceCount++; + if (nonWhiteSpaceCount > 77) { + // Line cannot be folded into multiple lines with no more + // than 78 characters each. Encoding as encoded-words makes + // that possible. One character has to be reserved for + // folding white space; that leaves 77 characters. + return true; + } + + if (ch < 32 || ch >= 127) { + // non-printable ascii character has to be encoded + return true; + } + } + } + + return false; + } + + /** + * Encodes the specified text into an encoded word or a sequence of encoded + * words separated by space. The text is separated into a sequence of + * encoded words if it does not fit in a single one. + *

+ * The charset to encode the specified text into a byte array and the + * encoding to use for the encoded-word are detected automatically. + *

+ * This method assumes that zero characters have already been used up in the + * current line. + * + * @param text + * text to encode. + * @param usage + * whether the encoded-word is to be used to replace a text token + * or a word entity (see RFC 822). + * @return the encoded word (or sequence of encoded words if the given text + * does not fit in a single encoded word). + * @see #hasToBeEncoded(String, int) + */ + public static String encodeEncodedWord(String text, Usage usage) { + return encodeEncodedWord(text, usage, 0, null, null); + } + + /** + * Encodes the specified text into an encoded word or a sequence of encoded + * words separated by space. The text is separated into a sequence of + * encoded words if it does not fit in a single one. + *

+ * The charset to encode the specified text into a byte array and the + * encoding to use for the encoded-word are detected automatically. + * + * @param text + * text to encode. + * @param usage + * whether the encoded-word is to be used to replace a text token + * or a word entity (see RFC 822). + * @param usedCharacters + * number of characters already used up (0 <= usedCharacters <= 50). + * @return the encoded word (or sequence of encoded words if the given text + * does not fit in a single encoded word). + * @see #hasToBeEncoded(String, int) + */ + public static String encodeEncodedWord(String text, Usage usage, + int usedCharacters) { + return encodeEncodedWord(text, usage, usedCharacters, null, null); + } + + /** + * Encodes the specified text into an encoded word or a sequence of encoded + * words separated by space. The text is separated into a sequence of + * encoded words if it does not fit in a single one. + * + * @param text + * text to encode. + * @param usage + * whether the encoded-word is to be used to replace a text token + * or a word entity (see RFC 822). + * @param usedCharacters + * number of characters already used up (0 <= usedCharacters <= 50). + * @param charset + * the Java charset that should be used to encode the specified + * string into a byte array. A suitable charset is detected + * automatically if this parameter is null. + * @param encoding + * the encoding to use for the encoded-word (either B or Q). A + * suitable encoding is automatically chosen if this parameter is + * null. + * @return the encoded word (or sequence of encoded words if the given text + * does not fit in a single encoded word). + * @see #hasToBeEncoded(String, int) + */ + public static String encodeEncodedWord(String text, Usage usage, + int usedCharacters, Charset charset, Encoding encoding) { + if (text == null) + throw new IllegalArgumentException(); + if (usedCharacters < 0 || usedCharacters > MAX_USED_CHARACTERS) + throw new IllegalArgumentException(); + + if (charset == null) + charset = determineCharset(text); + + String mimeCharset = CharsetUtil.toMimeCharset(charset.name()); + if (mimeCharset == null) { + // cannot happen if charset was originally null + throw new IllegalArgumentException("Unsupported charset"); + } + + byte[] bytes = encode(text, charset); + + if (encoding == null) + encoding = determineEncoding(bytes, usage); + + if (encoding == Encoding.B) { + String prefix = ENC_WORD_PREFIX + mimeCharset + "?B?"; + return encodeB(prefix, text, usedCharacters, charset, bytes); + } else { + String prefix = ENC_WORD_PREFIX + mimeCharset + "?Q?"; + return encodeQ(prefix, text, usage, usedCharacters, charset, bytes); + } + } + + /** + * Encodes the specified byte array using the B encoding defined in RFC + * 2047. + * + * @param bytes + * byte array to encode. + * @return encoded string. + */ + public static String encodeB(byte[] bytes) { + StringBuilder sb = new StringBuilder(); + + int idx = 0; + final int end = bytes.length; + for (; idx < end - 2; idx += 3) { + int data = (bytes[idx] & 0xff) << 16 | (bytes[idx + 1] & 0xff) << 8 + | bytes[idx + 2] & 0xff; + sb.append((char) BASE64_TABLE[data >> 18 & 0x3f]); + sb.append((char) BASE64_TABLE[data >> 12 & 0x3f]); + sb.append((char) BASE64_TABLE[data >> 6 & 0x3f]); + sb.append((char) BASE64_TABLE[data & 0x3f]); + } + + if (idx == end - 2) { + int data = (bytes[idx] & 0xff) << 16 | (bytes[idx + 1] & 0xff) << 8; + sb.append((char) BASE64_TABLE[data >> 18 & 0x3f]); + sb.append((char) BASE64_TABLE[data >> 12 & 0x3f]); + sb.append((char) BASE64_TABLE[data >> 6 & 0x3f]); + sb.append(BASE64_PAD); + + } else if (idx == end - 1) { + int data = (bytes[idx] & 0xff) << 16; + sb.append((char) BASE64_TABLE[data >> 18 & 0x3f]); + sb.append((char) BASE64_TABLE[data >> 12 & 0x3f]); + sb.append(BASE64_PAD); + sb.append(BASE64_PAD); + } + + return sb.toString(); + } + + /** + * Encodes the specified byte array using the Q encoding defined in RFC + * 2047. + * + * @param bytes + * byte array to encode. + * @param usage + * whether the encoded-word is to be used to replace a text token + * or a word entity (see RFC 822). + * @return encoded string. + */ + public static String encodeQ(byte[] bytes, Usage usage) { + BitSet qChars = usage == Usage.TEXT_TOKEN ? Q_REGULAR_CHARS + : Q_RESTRICTED_CHARS; + + StringBuilder sb = new StringBuilder(); + + final int end = bytes.length; + for (int idx = 0; idx < end; idx++) { + int v = bytes[idx] & 0xff; + if (v == 32) { + sb.append('_'); + } else if (!qChars.get(v)) { + sb.append('='); + sb.append(hexDigit(v >>> 4)); + sb.append(hexDigit(v & 0xf)); + } else { + sb.append((char) v); + } + } + + return sb.toString(); + } + + /** + * Tests whether the specified string is a token as defined in RFC 2045 + * section 5.1. + * + * @param str + * string to test. + * @return true if the specified string is a RFC 2045 token, + * false otherwise. + */ + public static boolean isToken(String str) { + // token := 1* + // tspecials := "(" / ")" / "<" / ">" / "@" / "," / ";" / ":" / "\" / + // <"> / "/" / "[" / "]" / "?" / "=" + // CTL := 0.- 31., 127. + + final int length = str.length(); + if (length == 0) + return false; + + for (int idx = 0; idx < length; idx++) { + char ch = str.charAt(idx); + if (!TOKEN_CHARS.get(ch)) + return false; + } + + return true; + } + + private static boolean isAtomPhrase(String str) { + // atom = [CFWS] 1*atext [CFWS] + + boolean containsAText = false; + + final int length = str.length(); + for (int idx = 0; idx < length; idx++) { + char ch = str.charAt(idx); + if (ATEXT_CHARS.get(ch)) { + containsAText = true; + } else if (!CharsetUtil.isWhitespace(ch)) { + return false; + } + } + + return containsAText; + } + + // RFC 5322 section 3.2.3 + private static boolean isDotAtomText(String str) { + // dot-atom-text = 1*atext *("." 1*atext) + // atext = ALPHA / DIGIT / "!" / "#" / "$" / "%" / "&" / "'" / "*" / + // "+" / "-" / "/" / "=" / "?" / "^" / "_" / "`" / "{" / "|" / "}" / "~" + + char prev = '.'; + + final int length = str.length(); + if (length == 0) + return false; + + for (int idx = 0; idx < length; idx++) { + char ch = str.charAt(idx); + + if (ch == '.') { + if (prev == '.' || idx == length - 1) + return false; + } else { + if (!ATEXT_CHARS.get(ch)) + return false; + } + + prev = ch; + } + + return true; + } + + // RFC 5322 section 3.2.4 + private static String quote(String str) { + // quoted-string = [CFWS] DQUOTE *([FWS] qcontent) [FWS] DQUOTE [CFWS] + // qcontent = qtext / quoted-pair + // qtext = %d33 / %d35-91 / %d93-126 + // quoted-pair = ("\" (VCHAR / WSP)) + // VCHAR = %x21-7E + // DQUOTE = %x22 + + String escaped = str.replaceAll("[\\\\\"]", "\\\\$0"); + return "\"" + escaped + "\""; + } + + private static String encodeB(String prefix, String text, + int usedCharacters, Charset charset, byte[] bytes) { + int encodedLength = bEncodedLength(bytes); + + int totalLength = prefix.length() + encodedLength + + ENC_WORD_SUFFIX.length(); + if (totalLength <= ENCODED_WORD_MAX_LENGTH - usedCharacters) { + return prefix + encodeB(bytes) + ENC_WORD_SUFFIX; + } else { + String part1 = text.substring(0, text.length() / 2); + byte[] bytes1 = encode(part1, charset); + String word1 = encodeB(prefix, part1, usedCharacters, charset, + bytes1); + + String part2 = text.substring(text.length() / 2); + byte[] bytes2 = encode(part2, charset); + String word2 = encodeB(prefix, part2, 0, charset, bytes2); + + return word1 + " " + word2; + } + } + + private static int bEncodedLength(byte[] bytes) { + return (bytes.length + 2) / 3 * 4; + } + + private static String encodeQ(String prefix, String text, Usage usage, + int usedCharacters, Charset charset, byte[] bytes) { + int encodedLength = qEncodedLength(bytes, usage); + + int totalLength = prefix.length() + encodedLength + + ENC_WORD_SUFFIX.length(); + if (totalLength <= ENCODED_WORD_MAX_LENGTH - usedCharacters) { + return prefix + encodeQ(bytes, usage) + ENC_WORD_SUFFIX; + } else { + String part1 = text.substring(0, text.length() / 2); + byte[] bytes1 = encode(part1, charset); + String word1 = encodeQ(prefix, part1, usage, usedCharacters, + charset, bytes1); + + String part2 = text.substring(text.length() / 2); + byte[] bytes2 = encode(part2, charset); + String word2 = encodeQ(prefix, part2, usage, 0, charset, bytes2); + + return word1 + " " + word2; + } + } + + private static int qEncodedLength(byte[] bytes, Usage usage) { + BitSet qChars = usage == Usage.TEXT_TOKEN ? Q_REGULAR_CHARS + : Q_RESTRICTED_CHARS; + + int count = 0; + + for (int idx = 0; idx < bytes.length; idx++) { + int v = bytes[idx] & 0xff; + if (v == 32) { + count++; + } else if (!qChars.get(v)) { + count += 3; + } else { + count++; + } + } + + return count; + } + + private static byte[] encode(String text, Charset charset) { + ByteBuffer buffer = charset.encode(text); + byte[] bytes = new byte[buffer.limit()]; + buffer.get(bytes); + return bytes; + } + + private static Charset determineCharset(String text) { + // it is an important property of iso-8859-1 that it directly maps + // unicode code points 0000 to 00ff to byte values 00 to ff. + boolean ascii = true; + final int len = text.length(); + for (int index = 0; index < len; index++) { + char ch = text.charAt(index); + if (ch > 0xff) { + return CharsetUtil.UTF_8; + } + if (ch > 0x7f) { + ascii = false; + } + } + return ascii ? CharsetUtil.US_ASCII : CharsetUtil.ISO_8859_1; + } + + private static Encoding determineEncoding(byte[] bytes, Usage usage) { + if (bytes.length == 0) + return Encoding.Q; + + BitSet qChars = usage == Usage.TEXT_TOKEN ? Q_REGULAR_CHARS + : Q_RESTRICTED_CHARS; + + int qEncoded = 0; + for (int i = 0; i < bytes.length; i++) { + int v = bytes[i] & 0xff; + if (v != 32 && !qChars.get(v)) { + qEncoded++; + } + } + + int percentage = qEncoded * 100 / bytes.length; + return percentage > 30 ? Encoding.B : Encoding.Q; + } + + private static char hexDigit(int i) { + return i < 10 ? (char) (i + '0') : (char) (i - 10 + 'A'); + } +} diff --git a/src/org/apache/james/mime4j/decoder/DecoderUtil.java b/src/org/apache/james/mime4j/decoder/DecoderUtil.java index 9bd2c512d..f7b9fa8f8 100644 --- a/src/org/apache/james/mime4j/decoder/DecoderUtil.java +++ b/src/org/apache/james/mime4j/decoder/DecoderUtil.java @@ -19,15 +19,15 @@ package org.apache.james.mime4j.decoder; +import org.apache.commons.logging.Log; +import org.apache.commons.logging.LogFactory; +import org.apache.james.mime4j.util.CharsetUtil; + import java.io.ByteArrayInputStream; import java.io.ByteArrayOutputStream; import java.io.IOException; import java.io.UnsupportedEncodingException; -import org.apache.commons.logging.Log; -import org.apache.commons.logging.LogFactory; -import org.apache.james.mime4j.util.CharsetUtil; - /** * Static methods for decoding strings, byte arrays and encoded words. * @@ -146,131 +146,119 @@ public class DecoderUtil { * =?charset?enc?Encoded word?= where enc is either 'Q' or 'q' for * quoted-printable and 'B' or 'b' for Base64. * + * ANDROID: COPIED FROM A NEWER VERSION OF MIME4J + * * @param body the string to decode. * @return the decoded string. */ public static String decodeEncodedWords(String body) { - StringBuffer sb = new StringBuffer(); - int p1 = 0; - int p2 = 0; - - try { - - /* - * Encoded words in headers have the form - * =?charset?enc?Encoded word?= where enc is either 'Q' or 'q' for - * quoted printable and 'B' and 'b' for Base64 - */ - - while (p2 < body.length()) { - /* - * Find beginning of first encoded word - */ - p1 = body.indexOf("=?", p2); - if (p1 == -1) { - /* - * None found. Emit the rest of the header and exit. - */ - sb.append(body.substring(p2)); - break; - } - - /* - * p2 points to the previously found end marker or the start - * of the entire header text. Append the text between that - * marker and the one pointed to by p1. - */ - if (p1 - p2 > 0) { - sb.append(body.substring(p2, p1)); - } - - /* - * Find the first and second '?':s after the marker pointed to - * by p1. - */ - int t1 = body.indexOf('?', p1 + 2); - int t2 = t1 != -1 ? body.indexOf('?', t1 + 1) : -1; - - /* - * Find this words end marker. - */ - p2 = t2 != -1 ? body.indexOf("?=", t2 + 1) : -1; - if (p2 == -1) { - if (t2 != -1 && (body.length() - 1 == t2 || body.charAt(t2 + 1) == '=')) { - /* - * Treat "=?charset?enc?" and "=?charset?enc?=" as - * empty strings. - */ - p2 = t2; - } else { - /* - * No end marker was found. Append the rest of the - * header and exit. - */ - sb.append(body.substring(p1)); - break; - } - } - - /* - * [p1+2, t1] -> charset - * [t1+1, t2] -> encoding - * [t2+1, p2] -> encoded word - */ - - String decodedWord = null; - if (t2 == p2) { - /* - * The text is empty - */ - decodedWord = ""; - } else { - - String mimeCharset = body.substring(p1 + 2, t1); - String enc = body.substring(t1 + 1, t2); - String encodedWord = body.substring(t2 + 1, p2); - - /* - * Convert the MIME charset to a corresponding Java one. - */ - String charset = CharsetUtil.toJavaCharset(mimeCharset); - if (charset == null) { - decodedWord = body.substring(p1, p2 + 2); - if (log.isWarnEnabled()) { - log.warn("MIME charset '" + mimeCharset - + "' in header field doesn't have a " - +"corresponding Java charset"); - } - } else if (!CharsetUtil.isDecodingSupported(charset)) { - decodedWord = body.substring(p1, p2 + 2); - if (log.isWarnEnabled()) { - log.warn("Current JDK doesn't support decoding " - + "of charset '" + charset - + "' (MIME charset '" - + mimeCharset + "')"); - } - } else { - if (enc.equalsIgnoreCase("Q")) { - decodedWord = DecoderUtil.decodeQ(encodedWord, charset); - } else if (enc.equalsIgnoreCase("B")) { - decodedWord = DecoderUtil.decodeB(encodedWord, charset); - } else { - decodedWord = encodedWord; - if (log.isWarnEnabled()) { - log.warn("Warning: Unknown encoding in " - + "header field '" + enc + "'"); - } - } - } - } - p2 += 2; - sb.append(decodedWord); - } - } catch (Throwable t) { - log.error("Decoding header field body '" + body + "'", t); + // ANDROID: Most strings will not include "=?" so a quick test can prevent unneeded + // object creation. This could also be handled via lazy creation of the StringBuilder. + if (body.indexOf("=?") == -1) { + return body; } - return sb.toString(); + int previousEnd = 0; + boolean previousWasEncoded = false; + + StringBuilder sb = new StringBuilder(); + + while (true) { + int begin = body.indexOf("=?", previousEnd); + int end = begin == -1 ? -1 : body.indexOf("?=", begin + 2); + if (end == -1) { + if (previousEnd == 0) + return body; + + sb.append(body.substring(previousEnd)); + return sb.toString(); + } + end += 2; + + String sep = body.substring(previousEnd, begin); + + String decoded = decodeEncodedWord(body, begin, end); + if (decoded == null) { + sb.append(sep); + sb.append(body.substring(begin, end)); + } else { + if (!previousWasEncoded || !CharsetUtil.isWhitespace(sep)) { + sb.append(sep); + } + sb.append(decoded); + } + + previousEnd = end; + previousWasEncoded = decoded != null; + } + } + + // return null on error + private static String decodeEncodedWord(String body, int begin, int end) { + int qm1 = body.indexOf('?', begin + 2); + if (qm1 == end - 2) + return null; + + int qm2 = body.indexOf('?', qm1 + 1); + if (qm2 == end - 2) + return null; + + String mimeCharset = body.substring(begin + 2, qm1); + String encoding = body.substring(qm1 + 1, qm2); + String encodedText = body.substring(qm2 + 1, end - 2); + + String charset = CharsetUtil.toJavaCharset(mimeCharset); + if (charset == null) { + if (log.isWarnEnabled()) { + log.warn("MIME charset '" + mimeCharset + "' in encoded word '" + + body.substring(begin, end) + "' doesn't have a " + + "corresponding Java charset"); + } + return null; + } else if (!CharsetUtil.isDecodingSupported(charset)) { + if (log.isWarnEnabled()) { + log.warn("Current JDK doesn't support decoding of charset '" + + charset + "' (MIME charset '" + mimeCharset + + "' in encoded word '" + body.substring(begin, end) + + "')"); + } + return null; + } + + if (encodedText.length() == 0) { + if (log.isWarnEnabled()) { + log.warn("Missing encoded text in encoded word: '" + + body.substring(begin, end) + "'"); + } + return null; + } + + try { + if (encoding.equalsIgnoreCase("Q")) { + return DecoderUtil.decodeQ(encodedText, charset); + } else if (encoding.equalsIgnoreCase("B")) { + return DecoderUtil.decodeB(encodedText, charset); + } else { + if (log.isWarnEnabled()) { + log.warn("Warning: Unknown encoding in encoded word '" + + body.substring(begin, end) + "'"); + } + return null; + } + } catch (UnsupportedEncodingException e) { + // should not happen because of isDecodingSupported check above + if (log.isWarnEnabled()) { + log.warn("Unsupported encoding in encoded word '" + + body.substring(begin, end) + "'", e); + } + return null; + } catch (RuntimeException e) { + if (log.isWarnEnabled()) { + log.warn("Could not decode encoded word '" + + body.substring(begin, end) + "'", e); + } + return null; + } } } diff --git a/src/org/apache/james/mime4j/util/CharsetUtil.java b/src/org/apache/james/mime4j/util/CharsetUtil.java index f5026b1f8..8bb044039 100644 --- a/src/org/apache/james/mime4j/util/CharsetUtil.java +++ b/src/org/apache/james/mime4j/util/CharsetUtil.java @@ -1056,6 +1056,73 @@ public class CharsetUtil { } } + /** + * ANDROID: THE FOLLOWING SET OF STATIC STRINGS ARE COPIED FROM A NEWER VERSION OF MIME4J + */ + + /** carriage return - line feed sequence */ + public static final String CRLF = "\r\n"; + + /** US-ASCII CR, carriage return (13) */ + public static final int CR = '\r'; + + /** US-ASCII LF, line feed (10) */ + public static final int LF = '\n'; + + /** US-ASCII SP, space (32) */ + public static final int SP = ' '; + + /** US-ASCII HT, horizontal-tab (9)*/ + public static final int HT = '\t'; + + public static final java.nio.charset.Charset US_ASCII = java.nio.charset.Charset + .forName("US-ASCII"); + + public static final java.nio.charset.Charset ISO_8859_1 = java.nio.charset.Charset + .forName("ISO-8859-1"); + + public static final java.nio.charset.Charset UTF_8 = java.nio.charset.Charset + .forName("UTF-8"); + + /** + * Returns true if the specified character is a whitespace + * character (CR, LF, SP or HT). + * + * ANDROID: COPIED FROM A NEWER VERSION OF MIME4J + * + * @param ch + * character to test. + * @return true if the specified character is a whitespace + * character, false otherwise. + */ + public static boolean isWhitespace(char ch) { + return ch == SP || ch == HT || ch == CR || ch == LF; + } + + /** + * Returns true if the specified string consists entirely of + * whitespace characters. + * + * ANDROID: COPIED FROM A NEWER VERSION OF MIME4J + * + * @param s + * string to test. + * @return true if the specified string consists entirely of + * whitespace characters, false otherwise. + */ + public static boolean isWhitespace(final String s) { + if (s == null) { + throw new IllegalArgumentException("String may not be null"); + } + final int len = s.length(); + for (int i = 0; i < len; i++) { + if (!isWhitespace(s.charAt(i))) { + return false; + } + } + return true; + } + /** * Determines if the VM supports encoding (chars to bytes) the * specified character set. NOTE: the given character set name may diff --git a/tests/src/com/android/email/mail/internet/MimeMessageTest.java b/tests/src/com/android/email/mail/internet/MimeMessageTest.java index d1fde4a6e..45f223e4c 100644 --- a/tests/src/com/android/email/mail/internet/MimeMessageTest.java +++ b/tests/src/com/android/email/mail/internet/MimeMessageTest.java @@ -35,6 +35,26 @@ import junit.framework.TestCase; */ @SmallTest public class MimeMessageTest extends TestCase { + + /** up arrow, down arrow, left arrow, right arrow */ + private final String SHORT_UNICODE = "\u2191\u2193\u2190\u2192"; + private final String SHORT_UNICODE_ENCODED = "=?UTF-8?B?4oaR4oaT4oaQ4oaS?="; + + /** a string without any unicode */ + private final String SHORT_PLAIN = "abcd"; + + /** longer unicode strings */ + private final String LONG_UNICODE_16 = SHORT_UNICODE + SHORT_UNICODE + + SHORT_UNICODE + SHORT_UNICODE; + private final String LONG_UNICODE_64 = LONG_UNICODE_16 + LONG_UNICODE_16 + + LONG_UNICODE_16 + LONG_UNICODE_16; + + /** longer plain strings (with fold points) */ + private final String LONG_PLAIN_16 = "abcdefgh ijklmno"; + private final String LONG_PLAIN_64 = + LONG_PLAIN_16 + LONG_PLAIN_16 + LONG_PLAIN_16 + LONG_PLAIN_16; + private final String LONG_PLAIN_256 = + LONG_PLAIN_64 + LONG_PLAIN_64 + LONG_PLAIN_64 + LONG_PLAIN_64; // TODO: more tests. @@ -98,7 +118,7 @@ public class MimeMessageTest extends TestCase { assertEquals("set and get Message-ID", testId2, message2.getMessageId()); } - /* + /** * Confirm getContentID() correctly works. */ public void testGetContentId() throws MessagingException { @@ -116,4 +136,80 @@ public class MimeMessageTest extends TestCase { message.setHeader(MimeHeader.HEADER_CONTENT_ID, "<" + cid1 + ">"); assertEquals(cid1, message.getContentId()); } + + /** + * Confirm that setSubject() works with plain strings + */ + public void testSetSubjectPlain() throws MessagingException { + MimeMessage message = new MimeMessage(); + + message.setSubject(SHORT_PLAIN); + + // test 1: readback + assertEquals("plain subjects", SHORT_PLAIN, message.getSubject()); + + // test 2: raw readback is not escaped + String rawHeader = message.getFirstHeader("Subject"); + assertEquals("plain subject not encoded", -1, rawHeader.indexOf("=?")); + + // test 3: long subject (shouldn't fold) + message.setSubject(LONG_PLAIN_64); + rawHeader = message.getFirstHeader("Subject"); + String[] split = rawHeader.split("\r\n"); + assertEquals("64 shouldn't fold", 1, split.length); + + // test 4: very long subject (should fold) + message.setSubject(LONG_PLAIN_256); + rawHeader = message.getFirstHeader("Subject"); + split = rawHeader.split("\r\n"); + assertTrue("long subject should fold", split.length > 1); + for (String s : split) { + assertTrue("split lines max length 78", s.length() <= 76); // 76+\r\n = 78 + String trimmed = s.trim(); + assertFalse("split lines are not encoded", trimmed.startsWith("=?")); + } + } + + /** + * Confirm that setSubject() works with unicode strings + */ + public void testSetSubject() throws MessagingException { + MimeMessage message = new MimeMessage(); + + message.setSubject(SHORT_UNICODE); + + // test 1: readback in unicode + assertEquals("unicode readback", SHORT_UNICODE, message.getSubject()); + + // test 2: raw readback is escaped + String rawHeader = message.getFirstHeader("Subject"); + assertEquals("raw readback", SHORT_UNICODE_ENCODED, rawHeader); + } + + /** + * Confirm folding operations on unicode subjects + */ + public void testSetLongSubject() throws MessagingException { + MimeMessage message = new MimeMessage(); + + // test 1: long unicode - readback in unicode + message.setSubject(LONG_UNICODE_16); + assertEquals("unicode readback 16", LONG_UNICODE_16, message.getSubject()); + + // test 2: longer unicode (will fold) + message.setSubject(LONG_UNICODE_64); + assertEquals("unicode readback 64", LONG_UNICODE_64, message.getSubject()); + + // test 3: check folding & encoding + String rawHeader = message.getFirstHeader("Subject"); + String[] split = rawHeader.split("\r\n"); + assertTrue("long subject should fold", split.length > 1); + for (String s : split) { + assertTrue("split lines max length 78", s.length() <= 76); // 76+\r\n = 78 + String trimmed = s.trim(); + assertTrue("split lines are encoded", + trimmed.startsWith("=?") && trimmed.endsWith("?=")); + } + } + } \ No newline at end of file diff --git a/tests/src/com/android/email/mail/internet/MimeUtilityTest.java b/tests/src/com/android/email/mail/internet/MimeUtilityTest.java index e894161be..b39e4cf0d 100644 --- a/tests/src/com/android/email/mail/internet/MimeUtilityTest.java +++ b/tests/src/com/android/email/mail/internet/MimeUtilityTest.java @@ -35,11 +35,120 @@ import junit.framework.TestCase; @SmallTest public class MimeUtilityTest extends TestCase { - // TODO: tests for unfold(String s) + /** up arrow, down arrow, left arrow, right arrow */ + private final String SHORT_UNICODE = "\u2191\u2193\u2190\u2192"; + private final String SHORT_UNICODE_ENCODED = "=?UTF-8?B?4oaR4oaT4oaQ4oaS?="; + + /** a string without any unicode */ + private final String SHORT_PLAIN = "abcd"; + + /** a typical no-param header */ + private final String HEADER_NO_PARAMETER = + "header"; + /** a typical multi-param header */ + private final String HEADER_MULTI_PARAMETER = + "header; Param1Name=Param1Value; Param2Name=Param2Value"; + + /** + * Test that decode/unfold is efficient when it can be + */ + public void testEfficientUnfoldAndDecode() { + String result1 = MimeUtility.unfold(SHORT_PLAIN); + String result2 = MimeUtility.decode(SHORT_PLAIN); + String result3 = MimeUtility.unfoldAndDecode(SHORT_PLAIN); + + assertSame(SHORT_PLAIN, result1); + assertSame(SHORT_PLAIN, result2); + assertSame(SHORT_PLAIN, result3); + } + + // TODO: more tests for unfold(String s) + + /** + * Test that decode is working for simple strings + */ + public void testDecodeSimple() { + String result1 = MimeUtility.decode(SHORT_UNICODE_ENCODED); + assertEquals(SHORT_UNICODE, result1); + } + // TODO: tests for decode(String s) + + /** + * Test that unfoldAndDecode is working for simple strings + */ + public void testUnfoldAndDecodeSimple() { + String result1 = MimeUtility.unfoldAndDecode(SHORT_UNICODE_ENCODED); + assertEquals(SHORT_UNICODE, result1); + } + // TODO: tests for unfoldAndDecode(String s) - // TODO: tests for foldAndEncode(String s) - // TODO: tests for getHeaderParameter(String header, String name) + + /** + * Test that fold/encode is efficient when it can be + */ + public void testEfficientFoldAndEncode() { + String result1 = MimeUtility.foldAndEncode(SHORT_PLAIN); + String result2 = MimeUtility.foldAndEncode2(SHORT_PLAIN, 10); + String result3 = MimeUtility.fold(SHORT_PLAIN, 10); + + assertSame(SHORT_PLAIN, result1); + assertSame(SHORT_PLAIN, result2); + assertSame(SHORT_PLAIN, result3); + } + + // TODO: more tests for foldAndEncode(String s) + + /** + * Test that foldAndEncode2 is working for simple strings + */ + public void testFoldAndEncode2() { + String result1 = MimeUtility.foldAndEncode2(SHORT_UNICODE, 10); + assertEquals(SHORT_UNICODE_ENCODED, result1); + } + + // TODO: more tests for foldAndEncode2(String s) + // TODO: more tests for fold(String s, int usedCharacters) + + /** + * Basic tests of getHeaderParameter() + * + * Typical header value: multipart/mixed; boundary="----E5UGTXUQQJV80DR8SJ88F79BRA4S8K" + * + * Function spec says: + * if header is null: return null + * if name is null: if params, return first param. else return full field + * else: if param is found (case insensitive) return it + * else return null + */ + public void testGetHeaderParameter() { + // if header is null, return null + assertNull("null header check", MimeUtility.getHeaderParameter(null, "name")); + + // if name is null, return first param or full header + // NOTE: The docs are wrong - it returns the header (no params) in that case +// assertEquals("null name first param per docs", "Param1Value", +// MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, null)); + assertEquals("null name first param per code", "header", + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, null)); + assertEquals("null name full header", HEADER_NO_PARAMETER, + MimeUtility.getHeaderParameter(HEADER_NO_PARAMETER, null)); + + // find name + assertEquals("get 1st param", "Param1Value", + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, "Param1Name")); + assertEquals("get 2nd param", "Param2Value", + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, "Param2Name")); + assertEquals("get missing param", null, + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, "Param3Name")); + + // case insensitivity + assertEquals("get 2nd param all LC", "Param2Value", + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, "param2name")); + assertEquals("get 2nd param all UC", "Param2Value", + MimeUtility.getHeaderParameter(HEADER_MULTI_PARAMETER, "PARAM2NAME")); + } + // TODO: tests for findFirstPartByMimeType(Part part, String mimeType) /** Tests for findPartByContentId(Part part, String contentId) */