/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/* ***** BEGIN LICENSE BLOCK *****
 * Version: NPL 1.1/GPL 2.0/LGPL 2.1
 *
 * The contents of this file are subject to the Netscape Public License
 * Version 1.1 (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.mozilla.org/NPL/
 *
 * Software distributed under the License is distributed on an "AS IS" basis,
 * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
 * for the specific language governing rights and limitations under the
 * License.
 *
 * The Original Code is mozilla.org code.
 *
 * The Initial Developer of the Original Code is 
 * Netscape Communications Corporation.
 * Portions created by the Initial Developer are Copyright (C) 1999
 * the Initial Developer. All Rights Reserved.
 *
 * Contributor(s):
 *
 * Alternatively, the contents of this file may be used under the terms of
 * either the GNU General Public License Version 2 or later (the "GPL"), or 
 * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
 * in which case the provisions of the GPL or the LGPL are applicable instead
 * of those above. If you wish to allow use of your version of this file only
 * under the terms of either the GPL or the LGPL, and not to allow others to
 * use your version of this file under the terms of the NPL, indicate your
 * decision by deleting the provisions above and replace them with the notice
 * and other provisions required by the GPL or the LGPL. If you do not delete
 * the provisions above, a recipient may use your version of this file under
 * the terms of any one of the NPL, the GPL or the LGPL.
 *
 * ***** END LICENSE BLOCK ***** */

#include "msgCore.h"  // for pre-compiled headers

#include "nsImapCore.h"
#include "nsImapProtocol.h"
#include "nsIMAPGenericParser.h"
#include "nsString.h"
#include "nsReadableUtils.h"

/*************************************************
   The following functions are used to implement
   a thread safe strtok
 *************************************************/
/*
 * Get next token from string *stringp, where tokens are (possibly empty)
 * strings separated by characters from delim.  Tokens are separated
 * by exactly one delimiter iff the skip parameter is false; otherwise
 * they are separated by runs of characters from delim, because we
 * skip over any initial `delim' characters.
 *
 * Writes NULs into the string at *stringp to end tokens.
 * delim will usually, but need not, remain CONSTant from call to call.
 * On return, *stringp points past the last NUL written (if there might
 * be further tokens), or is NULL (if there are definitely no more tokens).
 *
 * If *stringp is NULL, strtoken returns NULL.
 */
static 
char *strtoken_r(char ** stringp, const char *delim, int skip)
{
  char *s;
  const char *spanp;
  int c, sc;
  char *tok;
  
  if ((s = *stringp) == NULL)
    return (NULL);
  
  if (skip) {
		/*
                * Skip (span) leading delimiters (s += strspn(s, delim)).
    */
cont:
		c = *s;
                for (spanp = delim; (sc = *spanp++) != 0;) {
                  if (c == sc) {
                    s++;
                    goto cont;
                  }
                }
                if (c == 0) {		/* no token found */
                  *stringp = NULL;
                  return (NULL);
                }
  }
  
  /*
	 * Scan token (scan for delimiters: s += strcspn(s, delim), sort of).
         * Note that delim must have one NUL; we stop if we see that, too.
	 */
  for (tok = s;;) {
    c = *s++;
    spanp = delim;
    do {
      if ((sc = *spanp++) == c) {
        if (c == 0)
          s = NULL;
        else
          s[-1] = 0;
        *stringp = s;
        return( (char *) tok );
      }
    } while (sc != 0);
  }
  /* NOTREACHED */
  return (NULL);
}


/* static */ char *nsIMAPGenericParser::Imapstrtok_r(char *s1, const char *s2, char **lasts)
{
  if (s1)
    *lasts = s1;
  return (strtoken_r(lasts, s2, 1));
}


////////////////// nsIMAPGenericParser /////////////////////////


nsIMAPGenericParser::nsIMAPGenericParser() :
fNextToken(nsnull),
fCurrentLine(nsnull),
fLineOfTokens(nsnull),
fStartOfLineOfTokens(nsnull),
fCurrentTokenPlaceHolder(nsnull),
fAtEndOfLine(PR_FALSE),
fTokenizerAdvanced(PR_FALSE),
fSyntaxErrorLine(nsnull),
fSyntaxError(PR_FALSE),
fDisconnected(PR_FALSE)
{
}

nsIMAPGenericParser::~nsIMAPGenericParser()
{
  PR_FREEIF( fCurrentLine );
  PR_FREEIF( fStartOfLineOfTokens);
  PR_FREEIF( fSyntaxErrorLine );
}

void nsIMAPGenericParser::HandleMemoryFailure()
{
  SetConnected(PR_FALSE);
}

void nsIMAPGenericParser::ResetLexAnalyzer()
{
  PR_FREEIF( fCurrentLine );
  PR_FREEIF( fStartOfLineOfTokens );
  fTokenizerAdvanced = PR_FALSE;
  
  fCurrentLine = fNextToken = fLineOfTokens = fStartOfLineOfTokens = fCurrentTokenPlaceHolder = nsnull;
  fAtEndOfLine = PR_FALSE;
}

PRBool nsIMAPGenericParser::LastCommandSuccessful()
{
	return Connected() && !SyntaxError();
}

void nsIMAPGenericParser::SetSyntaxError(PRBool error)
{
  fSyntaxError = error;
  PR_FREEIF( fSyntaxErrorLine );
  if (error)
  {
    NS_ASSERTION(PR_FALSE, "syntax error in generic parser");	
    fSyntaxErrorLine = PL_strdup(fCurrentLine);
  }
  else
    fSyntaxErrorLine = NULL;
}

char *nsIMAPGenericParser::CreateSyntaxErrorLine()
{
  return PL_strdup(fSyntaxErrorLine);
}


PRBool nsIMAPGenericParser::SyntaxError()
{
  return fSyntaxError;
}

void nsIMAPGenericParser::SetConnected(PRBool connected)
{
  fDisconnected = !connected;
}

PRBool nsIMAPGenericParser::Connected()
{
  return !fDisconnected;
}

PRBool nsIMAPGenericParser::ContinueParse()
{
  return !fSyntaxError && !fDisconnected;
}


PRBool nsIMAPGenericParser::at_end_of_line()
{
  return (fAtEndOfLine || (nsCRT::strcmp(fNextToken, CRLF) == 0));
}

void nsIMAPGenericParser::skip_to_CRLF()
{
  while (Connected() && !at_end_of_line())
    fNextToken = GetNextToken();
}

// fNextToken initially should point to
// a string after the initial open paren ("(")
// After this call, fNextToken points to the
// first character after the matching close
// paren.  Only call GetNextToken to get the NEXT
// token after the one returned in fNextToken.
void nsIMAPGenericParser::skip_to_close_paren()
{
  int numberOfCloseParensNeeded = 1;
  if (fNextToken && *fNextToken == ')')
  {
    numberOfCloseParensNeeded--;
    fNextToken++;
    if (!fNextToken || !*fNextToken)
      fNextToken = GetNextToken();
  }
  
  while (ContinueParse() && numberOfCloseParensNeeded > 0)
  {
    // go through fNextToken, count the number
    // of open and close parens, to account
    // for nested parens which might come in
    // the response
    char *loc = 0;
    for (loc = fNextToken; loc && *loc; loc++)
    {
      if (*loc == '(')
        numberOfCloseParensNeeded++;
      else if (*loc == ')')
        numberOfCloseParensNeeded--;
      if (numberOfCloseParensNeeded == 0)
      {
        fNextToken = loc + 1;
        if (!fNextToken || !*fNextToken)
          fNextToken = GetNextToken();
        break;	// exit the loop
      }
    }
    
    if (numberOfCloseParensNeeded > 0)
      fNextToken = GetNextToken();
  }
}

char *nsIMAPGenericParser::GetNextToken()
{
  if (!fCurrentLine || fAtEndOfLine)
    AdvanceToNextLine();
  else if (Connected())
  {
    if (fTokenizerAdvanced)
    {
      fNextToken = Imapstrtok_r(fLineOfTokens, WHITESPACE, &fCurrentTokenPlaceHolder);
      fTokenizerAdvanced = PR_FALSE;
    }
    else
    {
      fNextToken = Imapstrtok_r(nsnull, WHITESPACE, &fCurrentTokenPlaceHolder);
    }
    if (!fNextToken)
    {
      fAtEndOfLine = PR_TRUE;
      fNextToken = CRLF;
    }
  }
  
  return fNextToken;
}

void nsIMAPGenericParser::AdvanceToNextLine()
{
  PR_FREEIF( fCurrentLine );
  PR_FREEIF( fStartOfLineOfTokens);
  fTokenizerAdvanced = PR_FALSE;
  
  PRBool ok = GetNextLineForParser(&fCurrentLine);
  if (!ok)
  {
    SetConnected(PR_FALSE);
    fStartOfLineOfTokens = nsnull;
    fLineOfTokens = nsnull;
    fCurrentTokenPlaceHolder = nsnull;
    fNextToken = CRLF;
  }
  else if (fCurrentLine)	// might be NULL if we are would_block ?
  {
    fStartOfLineOfTokens = PL_strdup(fCurrentLine);
    if (fStartOfLineOfTokens)
    {
      fLineOfTokens = fStartOfLineOfTokens;
      fNextToken = Imapstrtok_r(fLineOfTokens, WHITESPACE, &fCurrentTokenPlaceHolder);
      if (!fNextToken)
      {
        fAtEndOfLine = PR_TRUE;
        fNextToken = CRLF;
      }
      else
        fAtEndOfLine = PR_FALSE;
    }
    else
      HandleMemoryFailure();
  }
  else
    HandleMemoryFailure();
}

void nsIMAPGenericParser::AdvanceTokenizerStartingPoint(int32 bytesToAdvance)
{
  PRInt32 startingDiff = fLineOfTokens - fStartOfLineOfTokens;
  PRInt32 nextTokenOffset;
  
  // save off offset into fStartOfLineOfTokens of fNextToken so we can set it appropriately
  // when we destroy the current line and create a new one. I'm pretty sure fNextToken must
  // point somewhere in the current line.
  nextTokenOffset = fNextToken - fStartOfLineOfTokens;
  
  PR_FREEIF(fStartOfLineOfTokens);
  if (fCurrentLine)
  {
    fStartOfLineOfTokens = PL_strdup(fCurrentLine);
    fNextToken = fStartOfLineOfTokens + nextTokenOffset;
    
    if (fStartOfLineOfTokens && ((int32) strlen(fStartOfLineOfTokens) >= bytesToAdvance))
    {
      fLineOfTokens = fStartOfLineOfTokens + bytesToAdvance  + startingDiff;
      fCurrentTokenPlaceHolder = fLineOfTokens;
      fTokenizerAdvanced = PR_TRUE;
    }
    else
      HandleMemoryFailure();
  }
  else
    HandleMemoryFailure();
}

// Lots of things in the IMAP protocol are defined as an "astring."
// An astring is either an atom or a string.
// An atom is just a series of one or more characters such as:  hello
// A string can either be quoted or literal.
// Quoted:  "Test Folder 1"
// Literal: {13}Test Folder 1
// This function leaves us off with fCurrentTokenPlaceHolder immediately after
// the end of the Astring.  Call GetNextToken() to get the token after it.
char *nsIMAPGenericParser::CreateAstring()
{
  if (*fNextToken == '{')
  {
    return CreateLiteral();		// literal
  }
  else if (*fNextToken == '"')
  {
    return CreateQuoted();		// quoted
  }
  else
  {
    return CreateAtom();		// atom
  }
}


// Create an atom
// This function does not advance the parser.
// Call GetNextToken() to get the next token after the atom.
char *nsIMAPGenericParser::CreateAtom()
{
  char *rv = PL_strdup(fNextToken);
  //fNextToken = GetNextToken();
  return (rv);
}

// CreateNilString creates either NIL (reutrns NULL) or a string
// Call with fNextToken pointing to the thing which we think is the nilstring.
// This function leaves us off with fCurrentTokenPlaceHolder immediately after
// the end of the string, if it is a string, or at the NIL.
// Regardless of type, call GetNextToken() to get the token after it.
char *nsIMAPGenericParser::CreateNilString()
{
  if (!PL_strncasecmp(fNextToken, "NIL", 3))
  {
    if (strlen(fNextToken) != 3)
      fNextToken += 3;
    //fNextToken = GetNextToken();
    return NULL;
  }
  else
    return CreateString();
}


// Create a string, which can either be quoted or literal,
// but not an atom.
// This function leaves us off with fCurrentTokenPlaceHolder immediately after
// the end of the String.  Call GetNextToken() to get the token after it.
char *nsIMAPGenericParser::CreateString()
{
  if (*fNextToken == '{')
  {
    char *rv = CreateLiteral();		// literal
    return (rv);
  }
  else if (*fNextToken == '"')
  {
    char *rv = CreateQuoted();		// quoted
    //fNextToken = GetNextToken();
    return (rv);
  }
  else
  {
    SetSyntaxError(PR_TRUE);
    return NULL;
  }
}


// This function leaves us off with fCurrentTokenPlaceHolder immediately after
// the end of the closing quote.  Call GetNextToken() to get the token after it.
// Note that if the current line ends without the
// closed quote then we have to fetch another line from the server, until
// we find the close quote.
char *nsIMAPGenericParser::CreateQuoted(PRBool /*skipToEnd*/)
{
  char *currentChar = fCurrentLine + 
    (fNextToken - fStartOfLineOfTokens)
    + 1;	// one char past opening '"'
  
  int  charIndex = 0;
  int  escapeCharsCut = 0;
  PRBool closeQuoteFound = PR_FALSE;
  nsCString returnString(currentChar);
  
  while (!closeQuoteFound && ContinueParse())
  {
    if (!returnString.CharAt(charIndex))
    {
      AdvanceToNextLine();
      returnString += fCurrentLine;
      charIndex++;
    }
    else if (returnString.CharAt(charIndex) == '"')
    {
      // don't check to see if it was escaped, 
      // that was handled in the next clause
      closeQuoteFound = PR_TRUE;
    }
    else if (returnString.CharAt(charIndex) == '\\')
    {
      // eat the escape character
      returnString.Cut(charIndex, 1);
      // whatever the escaped character was, we want it
      charIndex++;
      
      // account for charIndex not reflecting the eat of the escape character
      escapeCharsCut++;
    }
    else
      charIndex++;
  }
  
  if (closeQuoteFound)
  {
    returnString.Truncate(charIndex);
    //if ((charIndex == 0) && skipToEnd)	// it's an empty string.  Why skip to end?
    //	skip_to_CRLF();
    //else if (charIndex == strlen(fCurrentLine))	// should we have this?
    //AdvanceToNextLine();
    //else 
    if (charIndex < (int) (strlen(fNextToken) - 2))	// -2 because of the start and end quotes
    {
      // the quoted string was fully contained within fNextToken,
      // and there is text after the quote in fNextToken that we
      // still need
      //			int charDiff = strlen(fNextToken) - charIndex - 1;
      //			fCurrentTokenPlaceHolder -= charDiff;
      //			if (!nsCRT::strcmp(fCurrentTokenPlaceHolder, CRLF))
      //				fAtEndOfLine = PR_TRUE;
      AdvanceTokenizerStartingPoint ((fNextToken - fLineOfTokens) + returnString.Length() + escapeCharsCut + 2);
      if (!nsCRT::strcmp(fLineOfTokens, CRLF))
        fAtEndOfLine = PR_TRUE;
    }
    else
    {
      fCurrentTokenPlaceHolder += escapeCharsCut + charIndex + 1 - strlen(fNextToken);
      if (!*fCurrentTokenPlaceHolder)
        *fCurrentTokenPlaceHolder = ' ';	// put the token delimiter back
                                                /*	if (!nsCRT::strcmp(fNextToken, CRLF))
                                                fAtEndOfLine = PR_TRUE;
      */
    }
  }
  else
    NS_ASSERTION(PR_FALSE, "didn't find close quote");
  
  return ToNewCString(returnString);
}


// This function leaves us off with fCurrentTokenPlaceHolder immediately after
// the end of the literal string.  Call GetNextToken() to get the token after it
// the literal string.
char *nsIMAPGenericParser::CreateLiteral()
{
  int32 numberOfCharsInMessage = atoi(fNextToken + 1);
  int32 charsReadSoFar = 0, currentLineLength = 0;
  int32 bytesToCopy = 0;
  
  uint32 numBytes = numberOfCharsInMessage + 1;
  NS_ASSERTION(numBytes, "overflow!");
  if (!numBytes)
    return nsnull;
  
  char *returnString = (char *) PR_Malloc(numBytes);
  
  if (returnString)
  {
    *(returnString + numberOfCharsInMessage) = 0; // Null terminate it first
    
    PRBool terminatedLine = PR_FALSE;
    while (ContinueParse() && (charsReadSoFar < numberOfCharsInMessage))
    {
      if (!terminatedLine)
      {
        if (fCurrentTokenPlaceHolder &&
          *fCurrentTokenPlaceHolder == nsCRT::LF &&
          *(fCurrentTokenPlaceHolder+1))
        {
          // This is a static buffer, with a CRLF between the literal size ({91}) and
          // the string itself
          fCurrentTokenPlaceHolder++;
        }
        else
        {
          // We have to read the next line from AdvanceToNextLine().
          terminatedLine = PR_TRUE;
          AdvanceToNextLine();
        }
      }
      else
        AdvanceToNextLine();
      if (ContinueParse())
      {
        currentLineLength = strlen(terminatedLine ? fCurrentLine : fCurrentTokenPlaceHolder);
        bytesToCopy = (currentLineLength > numberOfCharsInMessage - charsReadSoFar ?
          numberOfCharsInMessage - charsReadSoFar : currentLineLength);
        NS_ASSERTION (bytesToCopy, "0 length literal?");
        
        memcpy(returnString + charsReadSoFar, terminatedLine ? fCurrentLine : fCurrentTokenPlaceHolder, bytesToCopy); 
        charsReadSoFar += bytesToCopy;
      }
    }
    
    if (ContinueParse())
    {
      if (bytesToCopy == 0)
      {
        skip_to_CRLF();
        fAtEndOfLine = PR_TRUE;
      }
      else if (currentLineLength == bytesToCopy)
      {
        fAtEndOfLine = PR_TRUE;
      }
      else
      {
        // Move fCurrentTokenPlaceHolder
        if (terminatedLine)
          AdvanceTokenizerStartingPoint (bytesToCopy);
        else
          AdvanceTokenizerStartingPoint (	bytesToCopy + 
          strlen(fNextToken) + 
          2 /* CRLF */ +
          (fNextToken - fLineOfTokens)
          );
        if (!*fCurrentTokenPlaceHolder)	// landed on a token boundary
          fCurrentTokenPlaceHolder++;
        if (!nsCRT::strcmp(fCurrentTokenPlaceHolder, CRLF))
          fAtEndOfLine = PR_TRUE;
      }	
    }
  }
  
  return returnString;
}


// Call this to create a buffer containing all characters within
// a given set of parentheses.
// Call this with fNextToken[0]=='(', that is, the open paren
// of the group.
// It will allocate and return all characters up to and including the corresponding
// closing paren, and leave the parser in the right place afterwards.
char *nsIMAPGenericParser::CreateParenGroup()
{
#ifdef DEBUG_bienvenu
  NS_ASSERTION(fNextToken[0] == '(', "we don't have a paren group!");
#endif
  
  int numOpenParens = 1;
  
  // build up a buffer with the paren group.
  // start with an initial chunk, expand later if necessary
  nsCString buf;
  nsCString returnString;
  int bytesUsed = 0;
  
  // count the number of parens in the current token
  int count, tokenLen = strlen(fNextToken);
  for (count = 1; (count < tokenLen) && (numOpenParens > 0); count++)
  {
    if (fNextToken[count] == '(')
      numOpenParens++;
    else if (fNextToken[count] == ')')
      numOpenParens--;
  }
  
  if ((numOpenParens > 0) && ContinueParse())
  {
    // Copy that first token from before
    returnString =fNextToken;
    returnString.Append(" ");	// space that got stripped off the token
    
    PRBool extractReset = PR_TRUE;
    while (extractReset && ContinueParse())
    {
      extractReset = PR_FALSE;
      // Go through the current line and look for the last close paren.
      // We're not trying to parse it just yet, just separate it out.
      int len = strlen(fCurrentTokenPlaceHolder);
      for (count = 0; (count < len) && (numOpenParens > 0) && !extractReset; count++)
      {
        if (*fCurrentTokenPlaceHolder == '{')
        {
          fNextToken = GetNextToken();
          NS_ASSERTION(fNextToken, "out of memory?or invalid syntax");
          if (fNextToken)
          {
            tokenLen = strlen(fNextToken);
            if (fNextToken[tokenLen-1] == '}')
            {
              // ok, we're looking at a literal string here
              
              // first, flush buf
              if (bytesUsed > 0)
              {
                buf.Truncate(bytesUsed);
                returnString.Append(buf);
                buf.Truncate();
                bytesUsed = 0;
              }
              
              returnString.Append(fNextToken);	// append the {xx} to the buffer
              returnString.Append(CRLF);			// append a CRLF to the buffer
              char *lit = CreateLiteral();
              fTokenizerAdvanced = PR_FALSE;	// force it to use fCurrentTokenPlaceHolder
              NS_ASSERTION(lit, "syntax error or out of memory");
              if (lit)
              {
                returnString.Append(lit);
                //fCurrentTokenPlaceHolder += nsCRT::strlen(lit);
                //AdvanceTokenizerStartingPoint(nsCRT::strlen(lit));
                //fNextToken = GetNextToken();
                extractReset = PR_TRUE;
                PR_Free(lit);
              }
            }
            else
            {
#ifdef DEBUG_bienvenu
              NS_ASSERTION(PR_FALSE, "syntax error creating paren group");	// maybe not an error, but definitely a rare condition
#endif
            }
          }
        }
        else if (*fCurrentTokenPlaceHolder == '"')
        {
          // We're looking at a quoted string here.
          // Ignore the characters within it.
          
          // first, flush buf
          if (bytesUsed > 0)
          {
            buf.Truncate(bytesUsed);
            returnString.Append(buf);
            buf.Truncate();
            bytesUsed = 0;
          }
          
          fNextToken = GetNextToken();
          NS_ASSERTION(fNextToken, "syntax error or out of memory creating paren group");
          if (fNextToken)
          {
            char *q = CreateQuoted();
            fTokenizerAdvanced = PR_FALSE;	// force it to use fCurrentTokenPlaceHolder
            NS_ASSERTION(q, "syntax error or out of memory creating paren group");
            if (q)
            {
              returnString.Append("\"");
              returnString.Append(q);
              returnString.Append("\"");
              extractReset = PR_TRUE;
              PR_Free(q);
            }
          }
        }
        else if (*fCurrentTokenPlaceHolder == '(')
          numOpenParens++;
        else if (*fCurrentTokenPlaceHolder == ')')
          numOpenParens--;
        
        
        if (!extractReset)
        {
          // append this character to the buffer
          buf += *fCurrentTokenPlaceHolder;
          
          //.SetCharAt(*fCurrentTokenPlaceHolder, bytesUsed);
          bytesUsed++;
          fCurrentTokenPlaceHolder++;
        }
      }
    }
  }
  else if ((numOpenParens == 0) && ContinueParse())
  {
    // the whole paren group response was finished in a single token
    buf.Append(fNextToken);
  }
  
  
  if (numOpenParens != 0 || !ContinueParse())
  {
    SetSyntaxError(PR_TRUE);
    returnString.SetLength(0);
  }
  else
  {
    // flush buf the final time
    if (bytesUsed > 0)
    {
      buf.Truncate(bytesUsed);
      returnString.Append(buf);
      buf.Truncate();
    }
    fNextToken = GetNextToken();
  }
  
  return ToNewCString(returnString);
}