summary.c

这是一个同样来自贝尔实验室的和UNIX有着渊源的操作系统, 其简洁的设计和实现易于我们学习和理解

/*
 * summary.c
 * Copyright (C) 2002-2005 A.J. van Os; Released under GNU GPL
 *
 * Description:
 * Read the summary information of a Word document
 */

#include <time.h>
#include <string.h>
#include "antiword.h"

#define P_HEADER_SZ		28
#define P_SECTIONLIST_SZ	20
#define P_LENGTH_SZ		 4
#define P_SECTION_MAX_SZ	(2 * P_SECTIONLIST_SZ + P_LENGTH_SZ)
#define P_SECTION_SZ(x)		((x) * P_SECTIONLIST_SZ + P_LENGTH_SZ)

#define PID_TITLE		 2
#define PID_SUBJECT		 3
#define PID_AUTHOR		 4
#define PID_CREATE_DTM		12
#define PID_LASTSAVE_DTM	13
#define PID_APPNAME		18

#define PIDD_MANAGER		14
#define PIDD_COMPANY		15

#define VT_LPSTR		30
#define VT_FILETIME		64

#define TIME_OFFSET_HI		0x019db1de
#define TIME_OFFSET_LO		0xd53e8000

static char	*szTitle = NULL;
static char	*szSubject = NULL;
static char	*szAuthor = NULL;
static time_t	tCreateDtm = (time_t)-1;
static time_t	tLastSaveDtm= (time_t)-1;
static char	*szAppName = NULL;
static char	*szManager = NULL;
static char	*szCompany = NULL;
static USHORT	usLid = (USHORT)-1;


/*
 * vDestroySummaryInfo - destroy the summary information
 */
void
vDestroySummaryInfo(void)
{
	TRACE_MSG("vDestroySummaryInfo");

	szTitle = xfree(szTitle);
	szSubject = xfree(szSubject);
	szAuthor = xfree(szAuthor);
	tCreateDtm = (time_t)-1;
	tLastSaveDtm = (time_t)-1;
	szAppName = xfree(szAppName);
	szManager = xfree(szManager);
	szCompany = xfree(szCompany);
	usLid = (USHORT)-1;
} /* end of vDestroySummaryInfo */

/*
 * tConvertDosDate - convert DOS date format
 *
 * returns Unix time_t or -1
 */
static time_t
tConvertDosDate(const char *szDosDate)
{
	struct tm	tTime;
	const char	*pcTmp;
	time_t		tResult;

	memset(&tTime, 0, sizeof(tTime));
	pcTmp = szDosDate;
	/* Get the month */
	if (!isdigit(*pcTmp)) {
		return (time_t)-1;
	}
	tTime.tm_mon = (int)(*pcTmp - '0');
	pcTmp++;
	if (isdigit(*pcTmp)) {
		tTime.tm_mon *= 10;
		tTime.tm_mon += (int)(*pcTmp - '0');
		pcTmp++;
	}
	/* Get the first separater */
	if (isalnum(*pcTmp)) {
		return (time_t)-1;
	}
	pcTmp++;
	/* Get the day */
	if (!isdigit(*pcTmp)) {
		return (time_t)-1;
	}
	tTime.tm_mday = (int)(*pcTmp - '0');
	pcTmp++;
	if (isdigit(*pcTmp)) {
		tTime.tm_mday *= 10;
		tTime.tm_mday += (int)(*pcTmp - '0');
		pcTmp++;
	}
	/* Get the second separater */
	if (isalnum(*pcTmp)) {
		return (time_t)-1;
	}
	pcTmp++;
	/* Get the year */
	if (!isdigit(*pcTmp)) {
		return (time_t)-1;
	}
	tTime.tm_year = (int)(*pcTmp - '0');
	pcTmp++;
	if (isdigit(*pcTmp)) {
		tTime.tm_year *= 10;
		tTime.tm_year += (int)(*pcTmp - '0');
		pcTmp++;
	}
	/* Check the values */
	if (tTime.tm_mon == 0 || tTime.tm_mday == 0 || tTime.tm_mday > 31) {
		return (time_t)-1;
	}
	/* Correct the values */
	tTime.tm_mon--;		/* From 01-12 to 00-11 */
	if (tTime.tm_year < 80) {
		tTime.tm_year += 100;	/* 00 means 2000 is 100 */
	}
	tTime.tm_isdst = -1;
	tResult = mktime(&tTime);
	NO_DBG_MSG(ctime(&tResult));
	return tResult;
} /* end of tConvertDosDate */

/*
 * szLpstr - get a zero terminate string property
 */
static char *
szLpstr(ULONG ulOffset, const UCHAR *aucBuffer)
{
	char	*szStart, *szResult, *szTmp;
	size_t	tSize;

	tSize = (size_t)ulGetLong(ulOffset + 4, aucBuffer);
	NO_DBG_DEC(tSize);
	if (tSize == 0) {
		return NULL;
	}
	/* Remove white space from the start of the string */
	szStart = (char *)aucBuffer + ulOffset + 8;
	NO_DBG_MSG(szStart);
	fail(strlen(szStart) >= tSize);
	while (isspace(*szStart)) {
		szStart++;
	}
	if (szStart[0] == '\0') {
		return NULL;
	}
	szResult = xstrdup(szStart);
	/* Remove white space from the end of the string */
	szTmp = szResult + strlen(szResult) - 1;
	while (isspace(*szTmp)) {
		*szTmp = '\0';
		szTmp--;
	}
	NO_DBG_MSG(szResult);
	return szResult;
} /* end of szLpstr */

/*
 * tFiletime - get a filetime property
 */
static time_t
tFiletime(ULONG ulOffset, const UCHAR *aucBuffer)
{
	double	dHi, dLo, dTmp;
	ULONG	ulHi, ulLo;
	time_t	tResult;

	ulLo = ulGetLong(ulOffset + 4, aucBuffer);
	ulHi = ulGetLong(ulOffset + 8, aucBuffer);
	NO_DBG_HEX(ulHi);
	NO_DBG_HEX(ulLo);

	/* Move the starting point from 01 Jan 1601 to 01 Jan 1970 */
	dHi = (double)ulHi - (double)TIME_OFFSET_HI;
	dLo = (double)ulLo - (double)TIME_OFFSET_LO;
	NO_DBG_FLT(dHi);
	NO_DBG_FLT(dLo);

	/* Combine the values and divide by 10^7 to get seconds */
	dTmp  = dLo / 10000000.0;	/* 10^7 */
	dTmp += dHi * 429.4967926;	/* 2^32 / 10^7 */
	NO_DBG_FLT(dTmp);

	/* Make a time_t */
	if (dTmp - 0.5 < TIME_T_MIN || dTmp + 0.5 > TIME_T_MAX) {
		return (time_t)-1;
	}
	tResult = dTmp < 0.0 ? (time_t)(dTmp - 0.5) : (time_t)(dTmp + 0.5);
	NO_DBG_MSG(ctime(&tResult));
	return tResult;
} /* end of tFiletime */

/*
 * vAnalyseSummaryInfo - analyse the summary information
 */
static void
vAnalyseSummaryInfo(const UCHAR *aucBuffer)
{
	ULONG	ulOffset;
	size_t	tIndex, tCount, tPropID, tPropType;

	tCount = (size_t)ulGetLong(4, aucBuffer);
	DBG_DEC(tCount);
	for (tIndex = 0; tIndex < tCount; tIndex++) {
		tPropID = (size_t)ulGetLong(8 + tIndex * 8, aucBuffer);
		ulOffset = ulGetLong(12 + tIndex * 8, aucBuffer);
		NO_DBG_DEC(tPropID);
		NO_DBG_HEX(ulOffset);
		tPropType = (size_t)ulGetLong(ulOffset, aucBuffer);
		NO_DBG_DEC(tPropType);
		switch (tPropID) {
		case PID_TITLE:
			if (tPropType == VT_LPSTR && szTitle == NULL) {
				szTitle = szLpstr(ulOffset, aucBuffer);
			}
			break;
		case PID_SUBJECT:
			if (tPropType == VT_LPSTR && szSubject == NULL) {
				szSubject = szLpstr(ulOffset, aucBuffer);
			}
			break;
		case PID_AUTHOR:
			if (tPropType == VT_LPSTR && szAuthor == NULL) {
				szAuthor = szLpstr(ulOffset, aucBuffer);
			}
			break;
		case PID_CREATE_DTM:
			if (tPropType == VT_FILETIME &&
			    tCreateDtm == (time_t)-1) {
				tCreateDtm = tFiletime(ulOffset, aucBuffer);
			}
			break;
		case PID_LASTSAVE_DTM:
			if (tPropType == VT_FILETIME &&
			    tLastSaveDtm == (time_t)-1) {
				tLastSaveDtm = tFiletime(ulOffset, aucBuffer);
			}
			break;
		case PID_APPNAME:
			if (tPropType == VT_LPSTR && szAppName == NULL) {
				szAppName = szLpstr(ulOffset, aucBuffer);
			}
			break;
		default:
			break;
		}
	}
} /* end of vAnalyseSummaryInfo */

/*
 * vAnalyseDocumentSummaryInfo - analyse the document summary information
 */
static void
vAnalyseDocumentSummaryInfo(const UCHAR *aucBuffer)
{
	ULONG	ulOffset;
	size_t	tIndex, tCount, tPropID, tPropType;

	tCount = (size_t)ulGetLong(4, aucBuffer);
	DBG_DEC(tCount);
	for (tIndex = 0; tIndex < tCount; tIndex++) {
		tPropID = (size_t)ulGetLong(8 + tIndex * 8, aucBuffer);
		ulOffset = ulGetLong(12 + tIndex * 8, aucBuffer);
		NO_DBG_DEC(tPropID);
		NO_DBG_HEX(ulOffset);
		tPropType = (size_t)ulGetLong(ulOffset, aucBuffer);
		NO_DBG_DEC(tPropType);
		switch (tPropID) {
		case PIDD_MANAGER:
			if (tPropType == VT_LPSTR && szManager == NULL) {
				szManager = szLpstr(ulOffset, aucBuffer);
			}
			break;
		case PIDD_COMPANY:
			if (tPropType == VT_LPSTR && szCompany == NULL) {
				szCompany = szLpstr(ulOffset, aucBuffer);
			}
			break;
		default:
			break;
		}
	}
} /* end of vAnalyseDocumentSummaryInfo */

/*
 * pucAnalyseSummaryInfoHeader-
 */
static UCHAR *
pucAnalyseSummaryInfoHeader(FILE *pFile,
	ULONG ulStartBlock, ULONG ulSize,
	const ULONG *aulBBD, size_t tBBDLen,
	const ULONG *aulSBD, size_t tSBDLen)
{
	const ULONG	*aulBlockDepot;
	UCHAR	*aucBuffer;
	size_t	tBlockDepotLen, tBlockSize, tSectionCount, tLength;
	ULONG	ulTmp, ulOffset;
	USHORT	usLittleEndian, usEmpty, usOS, usVersion;
	UCHAR	aucHdr[P_HEADER_SZ], aucSecLst[P_SECTION_MAX_SZ];

	if (ulSize < MIN_SIZE_FOR_BBD_USE) {
		/* Use the Small Block Depot */
		aulBlockDepot = aulSBD;
		tBlockDepotLen = tSBDLen;
		tBlockSize = SMALL_BLOCK_SIZE;
	} else {
		/* Use the Big Block Depot */
		aulBlockDepot = aulBBD;
		tBlockDepotLen = tBBDLen;
		tBlockSize = BIG_BLOCK_SIZE;
	}

	if (tBlockDepotLen == 0) {
		DBG_MSG("The Block Depot length is zero");
		return NULL;
	}

	/* Read the Summery Information header */
	if (!bReadBuffer(pFile, ulStartBlock,
			aulBlockDepot, tBlockDepotLen, tBlockSize,
			aucHdr, 0, P_HEADER_SZ)) {
		return NULL;
	}
	NO_DBG_PRINT_BLOCK(aucHdr, P_HEADER_SZ);

	/* Analyse the Summery Information header */
	usLittleEndian =  usGetWord(0, aucHdr);
	if (usLittleEndian != 0xfffe) {
		DBG_HEX(usLittleEndian);
		DBG_MSG_C(usLittleEndian == 0xfeff, "Big endian");
		return NULL;
	}
	usEmpty =  usGetWord(2, aucHdr);
	if (usEmpty != 0x0000) {
		DBG_DEC(usEmpty);
		return NULL;
	}
	ulTmp = ulGetLong(4, aucHdr);
	DBG_HEX(ulTmp);
	usOS = (USHORT)(ulTmp >> 16);
	usVersion = (USHORT)(ulTmp & 0xffff);
	switch (usOS) {
	case 0:
		DBG_MSG("Win16");
		DBG_HEX(usVersion);
		break;
	case 1:
		DBG_MSG("MacOS");
		DBG_HEX(usVersion);
		break;
	case 2:
		DBG_MSG("Win32");
		DBG_HEX(usVersion);
		break;
	default:
		DBG_DEC(usOS);
		DBG_HEX(usVersion);
		break;
	}
	tSectionCount = (size_t)ulGetLong(24, aucHdr);
	DBG_DEC_C(tSectionCount != 1 && tSectionCount != 2, tSectionCount);
	if (tSectionCount != 1 && tSectionCount != 2) {
		return NULL;
	}

	/* Read the Summery Information Section Lists */
	if (!bReadBuffer(pFile, ulStartBlock,
			aulBlockDepot, tBlockDepotLen, tBlockSize,
			aucSecLst, P_HEADER_SZ, P_SECTION_SZ(tSectionCount))) {
		return NULL;
	}
	NO_DBG_PRINT_BLOCK(aucSecLst, P_SECTION_SZ(tSectionCount));

	ulTmp = ulGetLong(0, aucSecLst);
	DBG_HEX(ulTmp);
	ulTmp = ulGetLong(4, aucSecLst);
	DBG_HEX(ulTmp);
	ulTmp = ulGetLong(8, aucSecLst);
	DBG_HEX(ulTmp);
	ulTmp = ulGetLong(12, aucSecLst);
	DBG_HEX(ulTmp);
	ulOffset = ulGetLong(16, aucSecLst);
	DBG_DEC_C(ulOffset != P_HEADER_SZ + P_SECTIONLIST_SZ &&
		ulOffset != P_HEADER_SZ + 2 * P_SECTIONLIST_SZ,
		ulOffset);
	fail(ulOffset != P_HEADER_SZ + P_SECTIONLIST_SZ &&
		ulOffset != P_HEADER_SZ + 2 * P_SECTIONLIST_SZ);
	tLength =
		(size_t)ulGetLong(tSectionCount * P_SECTIONLIST_SZ, aucSecLst);
	NO_DBG_HEX(tLength);
	fail(ulOffset + tLength > ulSize);

	/* Read the Summery Information */
	aucBuffer = xmalloc(tLength);
	if (!bReadBuffer(pFile, ulStartBlock,
			aulBlockDepot, tBlockDepotLen, tBlockSize,
			aucBuffer, ulOffset, tLength)) {
		aucBuffer = xfree(aucBuffer);
		return NULL;
	}
	NO_DBG_PRINT_BLOCK(aucBuffer, tLength);
	return aucBuffer;
} /* end of pucAnalyseSummaryInfoHeader */

/*
 * vSet0SummaryInfo - set summary information from a Word for DOS file
 */
void
vSet0SummaryInfo(FILE *pFile, const UCHAR *aucHeader)
{
	UCHAR	*aucBuffer;
	ULONG	ulBeginSumdInfo, ulBeginNextBlock;
	size_t	tLen;
	USHORT	usCodepage, usOffset;

	TRACE_MSG("vSet0SummaryInfo");

	fail(pFile == NULL || aucHeader == NULL);

	/* First check the header */
	usCodepage = usGetWord(0x7e, aucHeader);
	DBG_DEC(usCodepage);
	switch (usCodepage) {
	case 850: usLid = 0x0809; break; /* Latin1 -> British English */
	case 862: usLid = 0x040d; break; /* Hebrew */
	case 866: usLid = 0x0419; break; /* Russian */
	case 0:
	case 437:
	default: usLid = 0x0409; break; /* ASCII -> American English */
	}