////////////////////////////////////////////////////////////////////////////
// Copyright KAZ 2021							  //
// 									  //
// contact (at) kaz.bzh							  //
// 									  //
// This software is a filter to shrink email by attachment extraction.	  //
// 									  //
// This software is governed by the CeCILL-B license under French law and //
// abiding by  the rules of distribution  of free software. You  can use, //
// modify  and/or  redistribute  the  software under  the  terms  of  the //
// CeCILL-B license as circulated by CEA, CNRS and INRIA at the following //
// URL "http://www.cecill.info".					  //
// 									  //
// As a counterpart to the access to  the source code and rights to copy, //
// modify and  redistribute granted  by the  license, users  are provided //
// only with a limited warranty and  the software's author, the holder of //
// the economic  rights, and the  successive licensors have  only limited //
// liability.								  //
// 									  //
// In this respect, the user's attention is drawn to the risks associated //
// with loading,  using, modifying  and/or developing or  reproducing the //
// software by the user in light of its specific status of free software, //
// that may  mean that  it is  complicated to  manipulate, and  that also //
// therefore means  that it  is reserved  for developers  and experienced //
// professionals having in-depth computer  knowledge. Users are therefore //
// encouraged  to load  and test  the software's  suitability as  regards //
// their  requirements  in  conditions  enabling the  security  of  their //
// systems and/or  data to  be ensured  and, more  generally, to  use and //
// operate it in the same conditions as regards security.		  //
// 									  //
// The fact that  you are presently reading this means  that you have had //
// knowledge of the CeCILL-B license and that you accept its terms.	  //
////////////////////////////////////////////////////////////////////////////

#include <iostream>
#include <vector>
#include <set>
#include <fstream>
#include <iomanip>
#include <math.h>
#include <algorithm>
#include <unistd.h>
#include <boost/algorithm/string.hpp>

#include "kazDebug.hpp"
#include "kazMisc.hpp"
#include "SizeArg.hpp"
#include "Attachment.hpp"

using namespace std;
using namespace kaz;

// ================================================================================
const string Attachment::contentTypeToken		("content-type");
const string Attachment::contentDispositionToken	("content-disposition");
const string Attachment::contentTransferEncodingToken	("content-transfer-encoding");
const string Attachment::base64Token			("base64");
const string Attachment::quotedPrintableToken		("quoted-printable");
const string Attachment::contentIDToken			("content-id");
const string Attachment::PLAIN				("plain");
const string Attachment::HTML				("html");
const string Attachment::RELATED			("related");
const string Attachment::ALTERNATIVE			("alternative");


const regex Attachment::nameCharsetRegEx		(".*name\\*=(.*)");
const regex Attachment::nameRegEx			(".*name=\"([^\"]*)\".*");
// boundary="----=_Part_796779_1154936629.1668080348646"
// boundary="------------040709000505010508040808"
// boundary="----------=_1668606031-941125-91"
// boundary="_004_PAVPR10MB6792713B313048E3A259B215B2079PAVPR10MB6792EURP_";
// boundary="_000_PAVPR10MB6792713B313048E3A259B215B2079PAVPR10MB6792EURP_"
// boundary=--boundary_1351_64006126-2b0e-4a3b-98ac-4797d1634188
// boundary=--boundary_1352_7e294c9a-cfab-44a0-bfb3-7310380ac7cb;
const regex Attachment::boundaryRegEx			(".*boundary=\"?([^\"; ]*)\"?;?.*");
const regex Attachment::cidDefRegEx			(".*<([^>]*)>.*");
const regex Attachment::textRegEx			(".*text/("+PLAIN+"|"+HTML+").*");
const regex Attachment::multiRegEx			("\\s*multipart/(mixed|"+RELATED+"|"+ALTERNATIVE+").*");

const string Attachment::IMG_BEGIN			("<IMG");
const string Attachment::IMG_END			(">");



static const string SRC_BEGIN ("SRC=\"");
static const string RFC822 ("message/rfc822");

// ================================================================================
string
Attachment::getUnknown (const string &contentType) {
  DEF_LOG ("Attachment::getUnknown", "contentType: " << contentType);
  static time_t now (time (NULL));
  static int count (0);

  tm *ltm = localtime (&now);
  ostringstream nameStream;
  nameStream << "U-"
	     << std::setfill ('0') << std::setw (2) << (ltm->tm_year-100)
	     << std::setfill ('0') << std::setw (2) << (1 + ltm->tm_mon)
	     << std::setfill ('0') << std::setw (2) << ltm->tm_mday
	     << std::setfill ('0') << std::setw (2) << ltm->tm_hour
	     << std::setfill ('0') << std::setw (2) << ltm->tm_min
	     << std::setfill ('0') << std::setw (2) << ltm->tm_sec
	     << "-" << count;
  const string::size_type subTypePos (contentType.find ("/"));
  if (subTypePos != string::npos)
    nameStream << "." << contentType.substr (subTypePos+1);
  ++count;
  LOG ("name: " << nameStream.str ());
  return nameStream.str ();
}

// ================================================================================
void
Attachment::removeSection (string &content, const string &beginTag, const string &endTag) {
  DEF_LOG ("Attachment::removeSection", "beginTag: " << beginTag <<  " endTag: " << endTag);
  for (string::size_type startPos (0);
       (startPos = caseInsensitiveFind (content, beginTag, startPos)) != string::npos;
       ) {
    string::size_type stopPos = caseInsensitiveFind (content, endTag, startPos);

    LOG_BUG (stopPos == startPos, content.erase (startPos, endTag.length ()); continue, "eMailShrinker: bug A1: removeSection: no " << beginTag);
    LOG_BUG (stopPos == string::npos, content.erase (startPos, beginTag.length ()); break, "eMailShrinker: bug A2: removeSection: no " << endTag);
    LOG ("KAZ start: " << startPos << " stop: " << stopPos);

    content.erase (startPos, stopPos+endTag.length ()-startPos);
  }
}

// ================================================================================
string
Attachment::getSection (const string &content, const string &beginTag, const string &endTag) {
  DEF_LOG ("Attachment::getSection", "beginTag: " << beginTag << " endTag: " << endTag <<  " content: " << content);
  vector<string> list;
  getSection (content, beginTag, endTag, list);
  size_t sum (0);
  for (const string &s : list)
    sum += s.length ();
  string result;
  result.reserve (sum);
  for (const string &s : list)
    result += s;
  LOG ("result: " << result);
 return result;
}

// ================================================================================
void
Attachment::getSection (const string &content, const string &beginTag, const string &endTag, vector<string> &result) {
  DEF_LOG ("Attachment::getSection", "beginTag: " << beginTag << " endTag: " << endTag <<  " content: " << content);
  for (string::size_type startPos (0);
       (startPos = caseInsensitiveFind (content, beginTag, startPos)) != string::npos;
       ) {
    LOG (beginTag << ": " << startPos);
    string::size_type stopPos = caseInsensitiveFind (content, endTag, startPos);

    LOG_BUG (stopPos == string::npos, break, "eMailShrinker: bug A3: " << endTag << " not found! at: " << startPos);
    LOG ("start: " << startPos << " stop: " << stopPos);

    LOG_BUG (startPos == stopPos, /**/, "eMailShrinker: bug A4: " << endTag << " without " << beginTag << " at: " << startPos);
    if (startPos != stopPos) {
      startPos += beginTag.length ();
      result.push_back (content.substr (startPos, stopPos-startPos));
    }
    startPos = stopPos+endTag.length ();
  }
}

// ================================================================================
const string
Attachment::getContentType () const {
  map<string, string>::const_iterator it (env.find (contentTypeToken));
  if (it == env.end ())
    return "";
  const string &contentTypeVal (it->second);
  const string::size_type semicolonPos = contentTypeVal.find (';');
  if (semicolonPos == string::npos)
    return contentTypeVal;
  return contentTypeVal.substr (0, semicolonPos);
}

const string
Attachment::getAttachName () const {
  DEF_LOG ("Attachment::getAttachName", "");
  string result = getProp (contentTypeToken, nameRegEx);
  if (result.length ()) {
    LOG ("name=: " << result);
    encodedWord (result);
    return result;
  }
  result = getProp (contentTypeToken, nameCharsetRegEx);
  if (result.length ()) {
    LOG ("name*=: " << result);
    charsetValue (result);
    return result;
  }
  // XXX il faut composer s'il y a plusieurs ligne filename*x=
  result = getProp (contentDispositionToken, nameRegEx);
  if (result.length ()) {
    LOG ("filename=: " << result);
    encodedWord (result);
    return result;
  }
  // XXX il faut composer s'il y a plusieurs ligne filename*x*=
  result = getProp (contentDispositionToken, nameRegEx);
  if (result.length ()) {
    LOG ("filename*=: " << result);
    charsetValue (result);
    return result;
  }
  return getUnknown (getContentType ());
}

const string &
Attachment::getBoundary () const {
  return boundary;
}

const streamoff
Attachment::getSize () const {
  return endPos-beginPos;
}

const string
Attachment::getProp (const string &token, const regex &regEx) const {
  DEF_LOG ("Attachment::getProp", "token: " << token);
  map<string, string>::const_iterator it (env.find (token));
  if (it == env.end ()) {
    LOG ("no token");
    return "";
  }
  const string &val (it->second);
  LOG ("val: " << val);
  if (!regex_match (val.begin (), val.end (), regEx)) {
    LOG ("no prop");
    return "";
  }
  return regex_replace (val, regEx, "$1");
}

const bool
Attachment::isBase64Encoding () const {
  return isDefProp (contentTransferEncodingToken, base64Token);
}

const bool
Attachment::isQuotedPrintableEnconding () const {
  return isDefProp (contentTransferEncodingToken, quotedPrintableToken);
}

const bool
Attachment::isTextBase64 () const {
  return !getProp (contentTypeToken, textRegEx).empty () && isBase64Encoding ();
}

const bool
Attachment::isDefProp (const string &token, const string &val) const {
  DEF_LOG ("Attachment::getProp", "getProp token: " << token << " val: " << val);
  map<string, string>::const_iterator it (env.find (token));
  if (it == env.end ())
    return false;
  // XXX case insensitive ??
  return it->second.find (val) != string::npos;
}

// ================================================================================
Attachment::Attachment (ifstream &mbox, const int &level, const streamoff beginInParent, streamoff &curPos)
  : level (level),
    beginInParent (beginInParent),
    beginPos (curPos),
    contentPos (0),
    endPos (0),
    toExtract (false),
    toUpdate (false),
    toDisclaim (false),
    boundaryMiddleSize (0) {
  DEF_LOG ("Attachment::Attachment", "curPos: " << curPos << " level: " << level);
  readMime (mbox, curPos);
  readBoundaries (mbox, curPos);
}

// ================================================================================
void
Attachment::readMime (ifstream &mbox, streamoff &curPos) {
  DEF_LOG ("Attachment::readMime", "curPos: " << curPos);
  string lastVar;
  string line;
  for (; getline (mbox, line); ) {
    LOG ("pos: " << curPos << " line: " << line);
    curPos += line.length () + 1;
    if (line.empty ())
      break;
    if (line[0] == ' ' || line[0] == '\t') {
      if (lastVar.empty ()) {

	LOG_BUG (true, /**/, "eMailShrinker: bug A5: not compliant MIME. pos: " << (curPos - (line.length () + 1)) << " line: " << line);
      } else {
	LOG ("add line to var: " << line);
	env.find (lastVar)->second += line;
	LOG ("new val: " << env.find (lastVar)->second);
      }
      continue;
    }
    string::size_type colonPos = line.find (':');
    if (colonPos != string::npos) {
      lastVar = line.substr (0, colonPos);
      toLower (lastVar);
      LOG ("find var: " << lastVar);
      string val (line.length () >= colonPos+2 ? line.substr (colonPos+2) : ""); // XXX check RFC " " after ": "
      LOG ("new var: " << lastVar << " <=> " << val);
      env [lastVar] = val;
    }
  }
  LOG ("end of mime");

  contentPos = curPos;
  cid = getProp (contentIDToken, cidDefRegEx);
  boundary = getProp (contentTypeToken, boundaryRegEx);
  LOG ("boundary: " << boundary);
  if (boundary.length ()) {
    boundary = "--"+boundary+"--";
    boundaryMiddleSize = boundary.length () - 2;
  }
  LOG ("readMime contentPos: " << contentPos << " cid: " << cid << " boundary: " << boundary);
}

// ================================================================================
void
Attachment::readBoundaries (ifstream &mbox, streamoff &curPos) {
  DEF_LOG ("Attachment::readBoundaries", "curPos: " << curPos);

  if (caseInsensitiveFind (getContentType (), RFC822) != string::npos) {
    subAttachements.push_back (Attachment (mbox, level+1, curPos, curPos));
    subAttachements.back ().endPos = curPos;
    return;
  }
  if (boundary.empty ())
    return;
  for (; nextBondary (mbox, curPos); )
    ;
}

bool
Attachment::nextBondary (ifstream &mbox, streamoff &curPos) {
  DEF_LOG ("Attachment::nextBondary", "curPos: " << curPos << " boundary: " << boundary);
  bool isTextBase64 (subAttachements.size () && subAttachements.back ().isTextBase64 ());
  LOG ("isTextBase64: " << isTextBase64 << " attach: " << *this);
  for (string prev, line; getline (mbox, line); ) {
    LOG ("curPos: " << curPos << " line: " << line);
    streamoff lastPos = curPos;
    curPos += line.length () + 1;

    string::size_type bpos = line.find (boundary.c_str (), 0, boundaryMiddleSize);
    if (bpos == string::npos) {
      string clearLine (line);
      if (isTextBase64)
	base64Decode (clearLine);
      string couple (prev+clearLine);
      for (vector <string>::iterator it = stringsToUpdate.begin ();
	   it != stringsToUpdate.end ();
	   ++it)
	if (couple.find (*it) != string::npos) {
	  LOG ("find: "+ *it);
	  LOG ("size: " << subAttachements.size ());
	  if (subAttachements.size ())
	    subAttachements.back ().toUpdate = true;
	  else
	    LOG_BUG (true, continue, "eMailShrinker: bug A10: boundary format ? " << *this);
	}
      prev = clearLine;
      continue;
    }
    LOG ("find: " << boundary);
    LOG ("lastPos: " << lastPos << " bpos: " << bpos << " boundaryMiddleSize: " << boundaryMiddleSize);
    if (subAttachements.size ())
      subAttachements.back ().endPos = lastPos;
    LOG ("line: " << line << "bpos+boundaryMiddleSize: " << (bpos+boundaryMiddleSize) << " find: " << line.find ("--", bpos+boundaryMiddleSize));
    bpos += boundaryMiddleSize;
    if (line.find ("--", bpos) == bpos) {
      LOG ("end");
      return false;
    }
    subAttachements.push_back (Attachment (mbox, level+1, lastPos, curPos));
    return true;
  }
  endPos = curPos;
  return false;
}

// ================================================================================
void
Attachment::markDisclaim (bool &plainMarked, bool &htmlMarked) {
  if (plainMarked && htmlMarked)
    return;
  string multiProp = getProp (contentTypeToken, multiRegEx);
  // LOG_BUG (multiProp == ALTERNATIVE && subAttachements.size () != 2, continue, "eMailShrinker: bug A6: alternative give not 1 case (" << subAttachements.size () << ").");
  if (multiProp.length ())
    for (Attachment &subAttach : subAttachements)
      subAttach.markDisclaim (plainMarked, htmlMarked);
  string textProp = getProp (contentTypeToken, textRegEx);
  if (textProp.empty ())
    return;
  if (!plainMarked && textProp == PLAIN)
    plainMarked = toUpdate = toDisclaim = true;
  if (!htmlMarked && textProp == HTML)
    htmlMarked = toUpdate = toDisclaim = true;
}

// ================================================================================
bool
Attachment::markSignificant (const string &parentMultiProp, const streamoff &minAttachSize, ifstream &mbox, vector<Attachment *> &allMarkedPtrs) {
  DEF_LOG ("Attachment::markSignificant", "parentMultiProp: " << parentMultiProp << " minAttachSize: " << minAttachSize);
  string textProp = getProp (contentTypeToken, textRegEx);
  bool cantBeExtract ((parentMultiProp == ALTERNATIVE && (textProp == PLAIN || textProp == HTML)) ||
		      (parentMultiProp == RELATED && textProp == HTML));
  string multiProp = getProp (contentTypeToken, multiRegEx);
  for (Attachment &sub : subAttachements)
    cantBeExtract |= sub.markSignificant (multiProp, minAttachSize, mbox, allMarkedPtrs);
  if (getProp (contentTypeToken, textRegEx) == HTML) {
    string content = getContent (mbox);
    vector<string> imgs;
    getSection (content, IMG_BEGIN, IMG_END, imgs);
    EmbeddedData::fillEmbeddedData (imgs, minAttachSize, embeddedData);
    if (embeddedData.size ())
      toUpdate = true;
  }
  cantBeExtract |= toUpdate;
  if (boundary.empty () && getSize () >= minAttachSize && !cantBeExtract)
    cantBeExtract = toExtract = true; // XXX cantBeExtract ?
  if (toExtract || toUpdate || toDisclaim)
    allMarkedPtrs.push_back (this);
  return cantBeExtract;
}

// ================================================================================
string
Attachment::getContent (ifstream &mbox) const {
  DEF_LOG ("Attachment::getContent", "contentPos: " << contentPos);
  string content;
  content.resize (endPos-contentPos);
  mbox.seekg (contentPos, ios::beg);
  mbox.read (&content[0], endPos-contentPos);
  if (isBase64Encoding ())
    base64Decode (content);
  if (isQuotedPrintableEnconding ())
    quotedDecode (content);
  return content;
}

// ================================================================================
void
Attachment::println (ofstream &outbox, string content) const {
  DEF_LOG ("Attachment::println", "content: " << content);
  if (isBase64Encoding ())
    base64Encode (content);
  if (isQuotedPrintableEnconding ())
    quotedEncode (content);
  outbox << content;
  if (content.length () && content.back () != '\n')
    outbox << endl;
}

// ================================================================================
void
Attachment::replaceEmbedded (string &content) const {
  DEF_LOG ("Attachment::replaceEmbedded", "content.length: " << content.length ());
  if (!embeddedData.size ())
    return;
  int imgIdx (-1);
  string::size_type startPos (0);
  for (const EmbeddedData &embedded : embeddedData) {
    LOG ("embedded: " << embedded);
    for ( ; ; ) {
      startPos = caseInsensitiveFind (content, IMG_BEGIN, startPos);
      LOG_BUG (startPos == string::npos, return, "eMailShrinker: bug A7: can't find " << IMG_BEGIN);
      ++imgIdx;
      if (embedded.imgIdx >= imgIdx)
	break;
      startPos += IMG_BEGIN.length ();
    }
    startPos = caseInsensitiveFind (content, SRC_BEGIN, startPos);

    LOG_BUG (startPos == string::npos, return, "eMailShrinker: bug A8: can't find " << SRC_BEGIN );
    startPos += SRC_BEGIN.length ();
    const string::size_type endPos (content.find ("\"", startPos));

    LOG_BUG (endPos == string::npos, return, "eMailShrinker: bug A9: can't find end of " << SRC_BEGIN );
    content.replace (startPos, endPos-startPos, embedded.downloadUrl);
  }
}

// ================================================================================
ostream&
kaz::operator << (ostream& os, const Attachment& attachment) {
  string prop, sep;
  if (attachment.toExtract) { prop = "to extract"; sep = ", "; }
  if (attachment.toUpdate) { prop += sep+"need update"; sep = ", "; }
  if (attachment.toDisclaim) { prop += sep+"need diclaim"; sep = ", "; }
  if (attachment.embeddedData.size ()) { prop += sep+"embeddedData"; }
  if (prop.length ())
    prop = " ["+prop+"]";
  
  os << ("****************************************"+40-(attachment.level % 20)*2) << setw (10) << SizeArg (attachment.getSize ()) << "          " << attachment.getContentType ()
     << prop << (attachment.cid.length () ? " id: "+attachment.cid : "")
     << (attachment.boundary.length () ? " boundary: "+attachment.boundary : "")
     << " (" << attachment.beginPos << " / " << attachment.contentPos << " / " << attachment.endPos << ") " << endl;
  for (const EmbeddedData &embedded : attachment.embeddedData)
    os << setw (((attachment.level+1) % 20)*2) << "" << setw (10) << SizeArg (embedded.dataLength) << "          embedded [to extract] " << embedded;
  for (const Attachment &sub : attachment.subAttachements) {
    os << sub;
  }
  return os;
}

// ================================================================================