/**
* Decode a character according to the expected encoding.
*
* @param c first character of the possibly encoded character sequence
* @param in <code>InputStream</code> with possible extra encoded characters.
* @return decoded character
* @throws IOException if an i/o error occurs in the underlying input stream
*/
protected int decode(int c, InputStream in) throws IOException {
switch (encoding) {
case ENC_UTF8:
c = utf8.readUtf8(c, in);
bytesOut.write(utf8.chars_read);
bValidChar = utf8.bValidChar;
if (c != -1) {
if (!bValidChar) {
// Invalid UTF-8 char
bfErrors |= E_BIT_INVALID_UTF8_ENCODING;
}
}
break;
case ENC_US_ASCII:
bValidChar = (c <= 127);
if (!bValidChar) {
// Invalid US-ASCII char
bfErrors |= E_BIT_INVALID_US_ASCII_CHAR;
}
break;
case ENC_ISO8859_1:
// ISO-8859-1 utilizes all 8-bits and requires no decoding.
case ENC_RAW:
// Raw 8-bit character needs no decoding.
default:
bValidChar = true;
break;
}
return c;
}
/**
* Reads a header/line according to the configuration. After calling the readLine method
* additional information is available from public fields on the reader.
*
* @param in <code>InputStream</code> with characters
* @return result wrapped in a <code>HeaderLine</code> object
* @throws IOException if an i/o error occurs in the underlying input stream
*/
public HeaderLine readLine(PushbackInputStream in) throws IOException {
HeaderLine headerLine = new HeaderLine();
int state;
if (!bNameValue) {
state = S_LINE;
} else {
state = S_NAME;
}
lineSb.setLength(0);
nvSb.setLength(0);
bytesOut = new ByteArrayOutputStreamWithUnread();
bfErrors = 0;
int c;
bCr = false;
boolean bLoop = true;
while (bLoop) {
c = in.read();
if (c != -1) {
bytesOut.write(c);
}
switch (state) {
case S_LINE:
switch (c) {
case -1:
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
bLoop = false;
break;
case '\r':
bCr = true;
break;
case '\n':
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
// Check EOL.
check_eol();
bLoop = false;
break;
default:
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
// Decode character.
c = decode(c, in);
if (c == -1) {
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
bLoop = false;
} else {
if (bValidChar && encoding != ENC_RAW) {
if (c < 256 && ((charCharacteristicsTab[c] & CC_CONTROL) == CC_CONTROL)) {
bValidChar = false;
// Invalid control char
bfErrors |= E_BIT_INVALID_CONTROL_CHAR;
}
}
if (bValidChar) {
lineSb.append((char) c);
}
}
break;
}
break;
case S_NAME:
switch (c) {
case -1:
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
nvSb.setLength(0);
bLoop = false;
break;
case '\r':
bCr = true;
break;
case '\n':
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
nvSb.setLength(0);
// Check EOL.
check_eol();
bLoop = false;
break;
case ':':
headerLine.type = HeaderLine.HLT_HEADERLINE;
headerLine.name = nvSb.toString();
lineSb.setLength(0);
nvSb.setLength(0);
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
state = S_VALUE;
break;
default:
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
// Decode character.
c = decode(c, in);
if (c == -1) {
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.type = HeaderLine.HLT_LINE;
headerLine.line = lineSb.toString();
lineSb.setLength(0);
nvSb.setLength(0);
bLoop = false;
} else {
if (bValidChar && encoding != ENC_RAW) {
if (c < 256 && ((charCharacteristicsTab[c] & CC_CONTROL) == CC_CONTROL)) {
bValidChar = false;
// Invalid control char
bfErrors |= E_BIT_INVALID_CONTROL_CHAR;
}
}
if (bValidChar) {
lineSb.append((char) c);
if (c < 256
&& ((charCharacteristicsTab[c] & CC_SEPARATOR_WS) == CC_SEPARATOR_WS)) {
bValidChar = false;
// Invalid separator in name
bfErrors |= E_BIT_INVALID_SEPARATOR_CHAR;
}
}
if (bValidChar) {
nvSb.append((char) c);
}
}
break;
}
break;
case S_VALUE:
switch (c) {
case -1:
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
case '\r':
bCr = true;
break;
case '\n':
// Check EOL.
check_eol();
if (bLWS) {
state = S_LWS;
} else {
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
}
break;
default:
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
// Decode character.
c = decode(c, in);
if (c == -1) {
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
} else {
if (bValidChar && encoding != ENC_RAW) {
if (c < 256 && ((charCharacteristicsTab[c] & CC_CONTROL) == CC_CONTROL)) {
bValidChar = false;
// Invalid control char
bfErrors |= E_BIT_INVALID_CONTROL_CHAR;
}
}
if (bValidChar) {
switch (c) {
case '\"':
nvSb.append((char) c);
if (bQuotedText) {
state = S_QUOTED_TEXT;
}
break;
case '=':
if (bEncodedWords) {
state = S_ENCODED_WORD_EQ;
} else {
nvSb.append((char) c);
}
break;
default:
nvSb.append((char) c);
break;
}
}
}
break;
}
break;
case S_LWS:
switch (c) {
case -1:
// EOF.
// bfErrors |= E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
case ' ':
case '\t':
nvSb.append(" ");
state = S_VALUE;
break;
default:
in.unread(c);
bytesOut.unread(c);
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
}
break;
case S_QUOTED_TEXT:
switch (c) {
case -1:
// EOF.
bfErrors |= E_BIT_MISSING_QUOTE | E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
case '\"':
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
nvSb.append((char) c);
state = S_VALUE;
break;
case '\\':
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
state = S_QUOTED_PAIR;
break;
case '\r':
bCr = true;
break;
case '\n':
// Check EOL.
check_eol();
if (bLWS) {
state = S_QUOTED_LWS;
} else {
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
}
break;
default:
if (bCr) {
// Misplaced CR.
bfErrors |= E_BIT_MISPLACED_CR;
bCr = false;
}
// Decode character.
c = decode(c, in);
if (c == -1) {
// EOF.
bfErrors |= E_BIT_MISSING_QUOTE | E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
} else {
if (bValidChar && encoding != ENC_RAW) {
if (c < 256 && ((charCharacteristicsTab[c] & CC_CONTROL) == CC_CONTROL)) {
bValidChar = false;
// Invalid control char
bfErrors |= E_BIT_INVALID_CONTROL_CHAR;
}
}
if (bValidChar) {
nvSb.append((char) c);
}
}
break;
}
break;
case S_QUOTED_PAIR:
switch (c) {
case -1:
nvSb.append('\\');
// EOF.
bfErrors |= E_BIT_MISSING_QUOTED_PAIR_CHAR | E_BIT_MISSING_QUOTE | E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
default:
// Decode character.
c = decode(c, in);
if (c == -1) {
// EOF.
bfErrors |= E_BIT_MISSING_QUOTED_PAIR_CHAR | E_BIT_MISSING_QUOTE | E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
} else {
nvSb.append('\\');
nvSb.append((char) c);
if (!bValidChar) {
bfErrors |= E_BIT_INVALID_QUOTED_PAIR_CHAR;
}
state = S_QUOTED_TEXT;
}
break;
}
break;
case S_QUOTED_LWS:
switch (c) {
case -1:
// EOF.
bfErrors |= E_BIT_MISSING_QUOTE;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
case ' ':
case '\t':
nvSb.append(" ");
state = S_QUOTED_TEXT;
break;
default:
in.unread(c);
bytesOut.unread(c);
bfErrors |= E_BIT_MISSING_QUOTE;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
}
break;
case S_ENCODED_WORD_EQ:
switch (c) {
case -1:
nvSb.append('=');
// EOF.
bfErrors |= E_BIT_EOF;
headerLine.value = trim(nvSb);
nvSb.setLength(0);
bLoop = false;
break;
case '?':
// Unread "=?", so it can be parsed as an EncodedWord which always starts with "=?"
in.unread('?');
in.unread('=');
bytesOut.unread('?');
bytesOut.unread('=');
EncodedWords ew = EncodedWords.parseEncodedWords(in, true);
/*
if (!ew.bIsValid) {
// TODO Decide whether to report encoded word errors or interpret as non encoded words.
}
*/
nvSb.append("=?");
in.unread(ew.line, 2, ew.line.length - 2);
bytesOut.write("=?".getBytes());
state = S_VALUE;
break;
default:
nvSb.append('=');
in.unread(c);
bytesOut.unread(c);
state = S_VALUE;
break;
}
break;
}
}
headerLine.raw = bytesOut.toByteArray();
headerLine.bfErrors = bfErrors;
bEof = (headerLine.raw.length == 0);
return headerLine;
}