rtfdecoder.h

此文件为RTF文件转换为TXT文件的解码程序。

//RtfDecoder.h
#ifndef CRTFDECODER
#define CRTFDECODER

// RTF parser declarations
#include <ctype.h>
#include <stddef.h>
#include "../include/decoderbase.h"


#define MIDBUFSIZE  (8*1024*1024)
#define OUTBUFSIZE  (8*1024*1024)
#define MAXCHARSET  (1000)


// RTF parser error codes
#define ecOK				0       // Everything's fine!
#define ecStackUnderflow    1       // Unmatched '}'
#define ecStackOverflow     2       // Too many '{' -- memory exhausted
#define ecUnmatchedBrace    3       // RTF ended during an open group.
#define ecInvalidHex        4       // invalid hex character found in data
#define ecBadTable          5       // RTF table (sym or prop) invalid
#define ecAssertion         6       // Assertion failure
#define ecEndOfFile         7       // End of file reached while reading RTF

#define fTrue 1
#define fFalse 0

typedef struct char_prop
{
    char fBold;
    char fUnderline;
    char fItalic;
} CHP;                  // CHaracter Properties

typedef enum {justL, justR, justC, justF } JUST;
typedef struct para_prop
{
    int xaLeft;                 // left indent in twips
    int xaRight;                // right indent in twips
    int xaFirst;                // first line indent in twips
    JUST just;                  // justification
} PAP;                  // PAragraph Properties

typedef enum {sbkNon, sbkCol, sbkEvn, sbkOdd, sbkPg} SBK;
typedef enum {pgDec, pgURom, pgLRom, pgULtr, pgLLtr} PGN;

typedef struct sect_prop
{
    int cCols;                  // number of columns
    SBK sbk;                    // section break type
    int xaPgn;                  // x position of page number in twips
    int yaPgn;                  // y position of page number in twips
    PGN pgnFormat;              // how the page number is formatted
} SEP;                  // SEction Properties

typedef struct doc_prop
{
    int xaPage;                 // page width in twips
    int yaPage;                 // page height in twips
    int xaLeft;                 // left margin in twips
    int yaTop;                  // top margin in twips
    int xaRight;                // right margin in twips
    int yaBottom;               // bottom margin in twips
    int pgnStart;               // starting page number in twips
    char fFacingp;              // facing pages enabled?
    char fLandscape;            // landscape or portrait??
} DOP;                  // DOcument Properties

typedef enum { rdsNorm, rdsSkip } RDS;              // Rtf Destination State
typedef enum { risNorm, risBin, risHex } RIS;       // Rtf Internal State

typedef struct save             // property save structure
{
    struct save *pNext;         // next save
    CHP chp;
    PAP pap;
    SEP sep;
    DOP dop;
    RDS rds;
    RIS ris;
} SAVE;

// What types of properties are there?
typedef enum {
	ipropBold, ipropItalic, ipropUnderline, ipropLeftInd,
	ipropRightInd, ipropFirstInd, ipropCols, ipropPgnX,
	ipropPgnY, ipropXaPage, ipropYaPage, ipropXaLeft,
	ipropXaRight, ipropYaTop, ipropYaBottom, ipropPgnStart,
	ipropSbk, ipropPgnFormat, ipropFacingp, ipropLandscape,
	ipropJust, ipropPard, ipropPlain, ipropSectd,
	ipropMax 
} IPROP;

typedef enum {actnSpec, actnByte, actnWord} ACTN;
typedef enum {propChp, propPap, propSep, propDop} PROPTYPE;

typedef struct propmod
{
    ACTN actn;              // size of value
    PROPTYPE prop;          // structure containing value
    int  offset;            // offset of value from base of structure
} PROP;

typedef enum {ipfnBin, ipfnHex, ipfnSkipDest } IPFN;
typedef enum {idestPict, idestSkip } IDEST;
typedef enum {kwdChar, kwdDest, kwdProp, kwdSpec} KWD;

typedef struct symbol
{
    char *szKeyword;        // RTF keyword
    int  dflt;              // default value to use
    bool fPassDflt;         // true to use default value from this table
    KWD  kwd;               // base action to take
    int  idx;               // index into property table if kwd == kwdProp
	// index into destination table if kwd == kwdDest
	// character to print if kwd == kwdChar
} SYM;


// RTF parser tables


//charset和code page转换表

//rtf解码类
class CRtfDecoder:public CDecoderBase
{
public:
	int  m_cGroup;
	int	 m_midbuflen;
	int  m_outbuflen;
	bool m_fSkipDestIfUnk;
	long m_cbBin;
	long m_lParam;
	long m_previousparam;			//前一个param值
	int	 m_curCharset;				//当前Charset值
	int  m_curCodePage;				//当前CodePage值
	int  m_previousCodeP;			//前一个CodePage值
	int  m_outCodePage;				//输出缓冲区CodePage值
	int  allCharset[MAXCHARSET];	//全部的Charset值表
	bool m_firstentry;				//第一次进入

	RDS rds;
	RIS ris;
	
	CHP chp;
	PAP pap;
	SEP sep;
	DOP dop;
	SAVE *psave;
	int isymMax;
	unsigned char *p_midbuf;
	unsigned char *p_outbuf;

public:
	CRtfDecoder(CPublicResource *ps=NULL);
	~CRtfDecoder();
	
public:
	int ecRtfParse(u_char *inbuf,int insize);
	int ecPushRtfState(void);
	int ecPopRtfState(void);
	int ecParseRtfKeyword(u_char *buf,int &pos,int size);
	int ecParseChar(int c);
	int ecTranslateKeyword(char *szKeyword, int param, bool fParam);
	int ecPrintChar(int ch);
	int ecEndGroupAction(RDS rds);
	int ecApplyPropChange(IPROP iprop, int val);
	int ecChangeDest(IDEST idest);
	int ecParseSpecialKeyword(IPFN ipfn);
	int ecParseSpecialProperty(IPROP iprop, int val);
	int ecParseHexByte(void);
	int ecMidBuftoOutBuf(void);
	int ecIntToStr(int ch,unsigned char *out);

public:
	bool reset();
	int is_me(u_char *buf,int len);
	bool decode(u_char *inbuf,int insize,u_char *outbuf,int outbufsize);
	u_char* get_result(int &retsize,int &type);
	bool continue_decode();
};

#endif