{*******************************************************}
{                                                       }
{       Tocsg.Html                                      }
{                                                       }
{       Copyright (C) 2020 kku                          }
{                                                       }
{*******************************************************}

unit Tocsg.Html;

interface

uses
  Tocsg.Obj, System.SysUtils, System.Classes;

type
  TTgHtmlParser = class(TTgObject)
  private
    sHtml_: String;
    procedure SetHtmlStr(const sHtml: String);
    function GetTagData(sTagPath: String): String;
    function GetTagText(sTagPath: String): String;
  public
    Constructor Create;

    property Html: String read sHtml_ write SetHtmlStr;
    property Text[sTagpath: String]: String read GetTagText; default;
  end;

function ExtractElementToStrings(sTag, sHtml: String; EltList: TStrings): Integer;
function ExtractElement(sTag, sHtml: String): String;
function ExtractElementOnce(sTag, sHtml: String): String;
function ClearElement(sTag, sHtml: String): String; overload;
function ClearElement(sBTag, sETag, sHtml: String): String; overload;

function ExtractAttrToStrings(sHtml: String; AttrList: TStrings): Integer;

// 셋중에 하나로 정리 하자.... 19_1205 kku
function HtmlDecode(const AStr: String): String;
function StripHTMLTags(const strHTML: string): string;
function ConvHtmlEntities(sText: String): String;

implementation

uses
  EM.DomParser, Tocsg.Strings, Tocsg.Safe;

function ExtractElementToStrings(sTag, sHtml: String; EltList: TStrings): Integer;
var
  nPosB, nPosE, nPosE_Ck: Integer;
  nBEltC, nEEltC: Integer;
  sBTag, sETag, sCheckHtml: String;
begin
  EltList.Clear;

  sBTag := '<' + sTag;
  sETag := '</' + sTag + '>';

  Result := 0;
  nBEltC := Length(sBTag);
  while Length(sHtml) > 0 do
  begin
    nEEltC := Length(sETag);
    nPosB := Pos(sBTag, sHtml);
    if nPosB = 0 then
      break;

    Delete(sHtml, 1, nPosB - 1);
    if (sHtml[nBEltC+1] <> ' ') and (sHtml[nBEltC+1] <> '>') then
    begin
      Delete(sHtml, 1, nBEltC);
      continue;
    end;

//    nPosB := Pos(sBTag, PChar(@sHtml[nBEltC + 1]));
    sCheckHtml := Copy(sHtml, nBEltC + 1, Length(sHtml) - nBEltC + 1);
    nPosB := Pos(sBTag, sCheckHtml);
    nPosE := Pos(sETag, sHtml);
//    nPosE_Ck := Pos('<', sCheckHtml, nPosB + 1);

//    if nPosE > nPosE_Ck then
//    begin
//      nPosE := nPosE_Ck;
//      nEEltc := 0;
//    end else
    if (nPosE = 0) or ((nPosB > 0) and (nPosB < nPosE)) then
    begin
    // /> 이렇게 끝날수도 있다.
      nPosE_Ck := Pos('/>', sHtml);
      if nPosE > nPosE_Ck then
      begin
        nPosE := nPosE_Ck;
        nEEltC := 2;
      end;
    end;

    if nPosE <> 0 then
    begin
      EltList.Add(Copy(sHtml, 1, nPosE + nEEltC - 1));
      Delete(sHtml, 1, nPosE + nEEltC - 1);
    end else break;
  end;

  Result := EltList.Count;
end;

function ExtractElement(sTag, sHtml: String): String;
var
  nPosB, nPosE: Integer;
  nBEltC, nEEltC: Integer;
  sBTag, sETag: String;
begin
  Result := '';

  sBTag := '<' + sTag;
  sETag := '</' + sTag + '>';

  nBEltC := Length(sBTag);
  while Length(sHtml) > 0 do
  begin
    nEEltC := Length(sETag);
    nPosB := Pos(sBTag, sHtml);
    if nPosB = 0 then
      break;

    Delete(sHtml, 1, nPosB - 1);

//    nPosB := Pos(sBTag, PChar(@sHtml[nBEltC + 1]));
    nPosB := Pos(sBTag, Copy(sHtml, nBEltC + 1, Length(sHtml) - nBEltC + 1));
    nPosE := Pos(sETag, sHtml);

    if (nPosE = 0) or ((nPosB > 0) and (nPosB < nPosE)) then
    begin
    // /> 이렇게 끝날수도 있다.
      nPosE := Pos('/>', sHtml);
      nEEltC := 2;
    end;

    if nPosE <> 0 then
    begin
      Result := Result + Copy(sHtml, 1, nPosE + nEEltC - 1);
      Delete(sHtml, 1, nPosE + nEEltC - 1);
    end else break;
  end;
end;

function ClearElement(sTag, sHtml: String): String;
var
  nPosB, nPosE: Integer;
  nBEltC, nEEltC: Integer;
  sBTag, sETag: String;
begin
  Result := '';

  sBTag := '<' + sTag;
  sETag := '</' + sTag + '>';

  nBEltC := Length(sBTag);
  while Length(sHtml) > 0 do
  begin
    nEEltC := Length(sETag);
    nPosB := Pos(sBTag, sHtml);
    if nPosB = 0 then
      break;

    Result := Result + Copy(sHtml, 1, nPosB - 1);
    Delete(sHtml, 1, nPosB + nBEltC - 1);

//    nPosB := Pos(sBTag, PChar(@sHtml[nBEltC + 1]));
    nPosB := Pos(sBTag, Copy(sHtml, nBEltC + 1, Length(sHtml) - nBEltC + 1));
    nPosE := Pos(sETag, sHtml);

    if (nPosE = 0) or ((nPosB > 0) and (nPosB < nPosE)) then
    begin
    // /> 이렇게 끝날수도 있다.
      nPosE := Pos('/>', sHtml);
      nEEltC := 2;
    end;

    if nPosE = 0 then
      break
    else
      Delete(sHtml, 1, nPosE + nEEltC - 1);
  end;
  Result := Result + sHtml;
end;

function ClearElement(sBTag, sETag, sHtml: String): String;
var
  nPosB, nPosE: Integer;
  nBEltC, nEEltC: Integer;
begin
  Result := '';

  nBEltC := Length(sBTag);
  while Length(sHtml) > 0 do
  begin
    nEEltC := Length(sETag);
    nPosB := Pos(sBTag, sHtml);
    if nPosB = 0 then
      break;

    Result := Result + Copy(sHtml, 1, nPosB - 1);
    Delete(sHtml, 1, nPosB + nBEltC - 1);

//    nPosB := Pos(sBTag, PChar(@sHtml[nBEltC + 1]));
    nPosB := Pos(sBTag, Copy(sHtml, nBEltC + 1, Length(sHtml) - nBEltC + 1));
    nPosE := Pos(sETag, sHtml);

    if (nPosE = 0) or ((nPosB > 0) and (nPosB < nPosE)) then
    begin
    // /> 이렇게 끝날수도 있다.
      nPosE := Pos('/>', sHtml);
      nEEltC := 2;
    end;

    if nPosE = 0 then
      break
    else
      Delete(sHtml, 1, nPosE + nEEltC - 1);
  end;
  Result := Result + sHtml;
end;

function ExtractElementOnce(sTag, sHtml: String): String;
var
  nPosB, nPosE: Integer;
  nBEltC, nEEltC: Integer;
  sBTag, sETag: String;
begin
  Result := '';

  sBTag := '<' + sTag;
  sETag := '</' + sTag + '>';

  nBEltC := Length(sBTag);
  while Length(sHtml) > 0 do
  begin
    nEEltC := Length(sETag);
    nPosB := Pos(sBTag, sHtml);
    if nPosB = 0 then
      break;

    Delete(sHtml, 1, nPosB - 1);

//    nPosB := Pos(sBTag, PChar(@sHtml[nBEltC + 1]));
    nPosB := Pos(sBTag, Copy(sHtml, nBEltC + 1, Length(sHtml) - nBEltC + 1));
    nPosE := Pos(sETag, sHtml);

    if (nPosE = 0) or ((nPosB > 0) and (nPosB < nPosE)) then
    begin
    // /> 이렇게 끝날수도 있다.
      nPosE := Pos('/>', sHtml);
      nEEltC := 2;
    end;

    if nPosE <> 0 then
    begin
      Result := Result + Copy(sHtml, 1, nPosE + nEEltC - 1);
      exit;
//      Delete(sHtml, 1, nPosE + nEEltC - 1);
    end else break;
  end;
end;

function ExtractAttrToStrings(sHtml: String; AttrList: TStrings): Integer;
var
  sBTag: String;
  i, nPos, nLen: Integer;
begin
  Result := 0;
  AttrList.Clear;

  nPos := Pos('<', sHtml);
  if nPos > 0 then
  begin
    Delete(sHtml, 1, nPos);
    nPos := Pos('>', sHtml);
    if nPos > 0 then
    begin
      SetLength(sHtml, nPos - 1);
//      Delete(sHtml, nPos, Length(sHtml) - nPos);
      nLen := Length(sHtml);
      if sHtml[nLen] = '/' then
        SetLength(sHtml, nLen - 1);

      SplitString(sHtml, ' ', AttrList);

      for i := AttrList.Count - 1 downto 0 do
        if Pos('=', AttrList[i]) = 0 then
          AttrList.Delete(i);

      AttrList.Text := StringReplace(AttrList.Text, '"', '', [rfReplaceAll]);
      AttrList.Text := StringReplace(AttrList.Text, '''', '', [rfReplaceAll]);
      Result := AttrList.Count;
    end;
  end;
end;

{ TTgHtmlParser }

Constructor TTgHtmlParser.Create;
begin
  Inherited Create;
  sHtml_ := '';
end;

procedure TTgHtmlParser.SetHtmlStr(const sHtml: String);
begin
  if sHtml_ <> sHtml then
    sHtml_ := sHtml;
end;

function TTgHtmlParser.GetTagData(sTagPath: String): String;
var
  TagPaths: TStringList;
  i, b, e: Integer;
  sHtml: String;
begin
  Result := '';
  sHtml := sHtml_;
  Guard(TagPaths, TStringList.Create);
  SplitString(sTagpath, '/', TagPaths);
  for i := 0 to TagPaths.Count - 1 do
    sHtml := ExtractElementOnce(TagPaths[i], sHtml);

  Result := sHtml;
end;

function TTgHtmlParser.GetTagText(sTagPath: String): String;
var
  b, e: Integer;
  sAttr: String;
  AttrList: TStringList;
  i: Integer;
begin
  Result := '';
  try
    b := LastDelimiter(':', sTagPath);
    if b > 0 then
    begin
      e := Length(sTagPath);
      sAttr := Copy(sTagPath, b + 1, e - b);
      Delete(sTagPath, b, e - b + 1);
    end;

    Result := GetTagData(sTagpath);
    if Result <> '' then
    begin
      b := Pos('>', Result);
      e := LastDelimiter('<', Result);
      if b < e then
      begin
        if sAttr <> '' then
        begin
          Guard(AttrList, TStringList.Create);
          if ExtractAttrToStrings(Result, AttrList) > 0 then
          begin
            Result := '';
            for i := 0 to AttrList.Count - 1 do
              if Pos(sAttr + '=', AttrList[i]) = 1 then
              begin
                Result := Copy(AttrList[i], Length(sAttr) + 2, Length(AttrList[i]) - Length(sAttr) + 1);
                exit;
              end;
          end else Result := '';
        end else begin
          Result := Copy(Result, b + 1, e - b - 1);
          Result := ClearElement('<', '>', Result)
        end;

      end else Result := '';
    end;
  finally
    if Result <> '' then
    begin
      Result := StringReplace(Result, #13, ' ', [rfReplaceAll]);
      Result := StringReplace(Result, #10, '', [rfReplaceAll]);
      Result := Trim(Result);
    end;
  end;
end;

// By - https://stackoverflow.com/questions/1657105/delphi-html-decode
function HtmlDecode(const AStr: String): String;
var
  Sp, Rp, Cp, Tp: PChar;
  S: String;
  I, Code: Integer;
begin
  SetLength(Result, Length(AStr));
  Sp := PChar(AStr);
  Rp := PChar(Result);
  Cp := Sp;
  try
    while Sp^ <> #0 do
    begin
      case Sp^ of
        '&': begin
               Cp := Sp;
               Inc(Sp);
               case Sp^ of
                 'a': if AnsiStrPos(Sp, 'amp;') = Sp then  { do not localize }
                      begin
                        Inc(Sp, 3);
                        Rp^ := '&';
                      end;
                 'l',
                 'g': if (AnsiStrPos(Sp, 'lt;') = Sp) or (AnsiStrPos(Sp, 'gt;') = Sp) then { do not localize }
                      begin
                        Cp := Sp;
                        Inc(Sp, 2);
                        while (Sp^ <> ';') and (Sp^ <> #0) do
                          Inc(Sp);
                        if Cp^ = 'l' then
                          Rp^ := '<'
                        else
                          Rp^ := '>';
                      end;
                 'n': if AnsiStrPos(Sp, 'nbsp;') = Sp then  { do not localize }
                      begin
                        Inc(Sp, 4);
                        Rp^ := ' ';
                      end;
                 'q': if AnsiStrPos(Sp, 'quot;') = Sp then  { do not localize }
                      begin
                        Inc(Sp,4);
                        Rp^ := '"';
                      end;
                 '#': begin
                        Tp := Sp;
                        Inc(Tp);
                        while (Sp^ <> ';') and (Sp^ <> #0) do
                          Inc(Sp);
                        SetString(S, Tp, Sp - Tp);
                        Val(S, I, Code);
                        Rp^ := Chr((I));
                      end;
                 else
                   Exit;
               end;
           end
      else
        Rp^ := Sp^;
      end;
      Inc(Rp);
      Inc(Sp);
    end;
  except
  end;
  SetLength(Result, Rp - PChar(Result));
end;

// By - https://stackoverflow.com/questions/3001443/get-the-rendered-text-from-html-delphi
function StripHTMLTags(const strHTML: string): string;
var
  P: PChar;
  InTag: Boolean;
  i, intResultLength: Integer;
begin
  P := PChar(strHTML);
  Result := '';

  InTag := False;
  repeat
    case P^ of
      '<': InTag := True;
      '>': InTag := False;
      #13, #10: ; {do nothing}
      else
        if not InTag then
        begin
          if (P^ in [#9, #32]) and ((P+1)^ in [#10, #13, #32, #9, '<']) then
          else
            Result := Result + P^;
        end;
    end;
    Inc(P);
  until (P^ = #0);

  {convert system characters}
  Result := StringReplace(Result, '&quot;', '"',  [rfReplaceAll]);
  Result := StringReplace(Result, '&apos;', '''', [rfReplaceAll]);
  Result := StringReplace(Result, '&gt;',   '>',  [rfReplaceAll]);
  Result := StringReplace(Result, '&lt;',   '<',  [rfReplaceAll]);
  Result := StringReplace(Result, '&amp;',  '&',  [rfReplaceAll]);
  Result := StringReplace(Result, '&nbsp;',  ' ',  [rfReplaceAll]);
  {here you may add another symbols from RFC if you need}
end;

function ConvHtmlEntities(sText: String): String;
begin
  Result := StringReplace(sText, #9, '', [rfReplaceAll]);
  Result := StringReplace(Result, '&lt;', '<', [rfReplaceAll]);
  Result := StringReplace(Result, '&gt;', '>', [rfReplaceAll]);
  Result := StringReplace(Result, '&amp;', '&', [rfReplaceAll]);
  Result := StringReplace(Result, '&quot;', '"', [rfReplaceAll]);
  Result := StringReplace(Result, '&nbsp;', ' ', [rfReplaceAll]);
  Result := StringReplace(Result, '&#039;', '''', [rfReplaceAll]);
  Result := StringReplace(Result, '&middot;', '·', [rfReplaceAll]);
end;

end.