word2text.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

/*
 * word2text.c
 * Copyright (C) 1998-2005 A.J. van Os; Released under GNU GPL
 *
 * Description:
 * MS Word to "text" functions
 */

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <ctype.h>
#if defined(__riscos)
#include "DeskLib:Hourglass.h"
#include "drawfile.h"
#endif /* __riscos */
#include "antiword.h"


#define INITIAL_SIZE		40
#define EXTENTION_SIZE		20


/* Macros to make sure all such statements will be identical */
#define OUTPUT_LINE()		\
	do {\
		vAlign2Window(pDiag, pAnchor, lWidthMax, ucAlignment);\
		TRACE_MSG("after vAlign2Window");\
		pAnchor = pStartNewOutput(pAnchor, NULL);\
		pOutput = pAnchor;\
	} while(0)

#define RESET_LINE()		\
	do {\
		pAnchor = pStartNewOutput(pAnchor, NULL);\
		pOutput = pAnchor;\
	} while(0)

#if defined(__riscos)
/* Length of the document in characters */
static ULONG	ulDocumentLength;
/* Number of characters processed so far */
static ULONG	ulCharCounter;
static int	iCurrPct, iPrevPct;
#endif /* __riscos */
/* The document is in the format belonging to this version of Word */
static int	iWordVersion = -1;
/* Special treatment for files from Word 4/5/6 on an Apple Macintosh */
static BOOL	bOldMacFile = FALSE;
/* Section Information */
static const section_block_type	*pSection = NULL;
static const section_block_type	*pSectionNext = NULL;
/* All the (command line) options */
static options_type	tOptions;
/* Needed for reading a complete table row */
static const row_block_type	*pRowInfo = NULL;
static BOOL	bStartRow = FALSE;
static BOOL	bEndRowNorm = FALSE;
static BOOL	bEndRowFast = FALSE;
static BOOL	bIsTableRow = FALSE;
/* Index of the next style and font information */
static USHORT	usIstdNext = ISTD_NORMAL;
/* Needed for finding the start of a style */
static const style_block_type	*pStyleInfo = NULL;
static style_block_type		tStyleNext;
static BOOL	bStartStyle = FALSE;
static BOOL	bStartStyleNext = FALSE;
/* Needed for finding the start of a font */
static const font_block_type	*pFontInfo = NULL;
static font_block_type		tFontNext;
static BOOL	bStartFont = FALSE;
static BOOL	bStartFontNext = FALSE;
/* Needed for finding an image */
static ULONG	ulFileOffsetImage = FC_INVALID;


/*
 * vUpdateCounters - Update the counters for the hourglass
 */
static void
vUpdateCounters(void)
{
#if defined(__riscos)
	ulCharCounter++;
	iCurrPct = (int)((ulCharCounter * 100) / ulDocumentLength);
	if (iCurrPct != iPrevPct) {
		Hourglass_Percentage(iCurrPct);
		iPrevPct = iCurrPct;
	}
#endif /* __riscos */
} /* end of vUpdateCounters */

/*
 * bOutputContainsText - see if the output contains more than white space
 */
BOOL
bOutputContainsText(const output_type *pAnchor)
{
	const output_type	*pCurr;
	size_t	tIndex;

	fail(pAnchor == NULL);

	for (pCurr = pAnchor; pCurr != NULL; pCurr = pCurr->pNext) {
		fail(pCurr->lStringWidth < 0);
		for (tIndex = 0; tIndex < pCurr->tNextFree; tIndex++) {
			if (isspace((int)(UCHAR)pCurr->szStorage[tIndex])) {
				continue;
			}
#if defined(DEBUG)
			if (pCurr->szStorage[tIndex] == FILLER_CHAR) {
				continue;
			}
#endif /* DEBUG */
			return TRUE;
		}
	}
	return FALSE;
} /* end of bOutputContainsText */

/*
 * lTotalStringWidth - compute the total width of the output string
 */
static long
lTotalStringWidth(const output_type *pAnchor)
{
	const output_type	*pCurr;
	long		lTotal;

	lTotal = 0;
	for (pCurr = pAnchor; pCurr != NULL; pCurr = pCurr->pNext) {
		DBG_DEC_C(pCurr->lStringWidth < 0, pCurr->lStringWidth);
		fail(pCurr->lStringWidth < 0);
		lTotal += pCurr->lStringWidth;
	}
	return lTotal;
} /* end of lTotalStringWidth */

/*
 * vStoreByte - store one byte
 */
static void
vStoreByte(UCHAR ucChar, output_type *pOutput)
{
	fail(pOutput == NULL);

	if (ucChar == 0) {
		pOutput->szStorage[pOutput->tNextFree] = '\0';
		return;
	}

	while (pOutput->tNextFree + 2 > pOutput->tStorageSize) {
		pOutput->tStorageSize += EXTENTION_SIZE;
		pOutput->szStorage = xrealloc(pOutput->szStorage,
					pOutput->tStorageSize);
	}
	pOutput->szStorage[pOutput->tNextFree] = (char)ucChar;
	pOutput->szStorage[pOutput->tNextFree + 1] = '\0';
	pOutput->tNextFree++;
} /* end of vStoreByte */

/*
 * vStoreChar - store a character as one or more bytes
 */
static void
vStoreChar(ULONG ulChar, BOOL bChangeAllowed, output_type *pOutput)
{
	char	szResult[4];
	size_t	tIndex, tLen;

	fail(pOutput == NULL);

	if (tOptions.eEncoding == encoding_utf_8 && bChangeAllowed) {
		DBG_HEX_C(ulChar > 0xffff, ulChar);
		fail(ulChar > 0xffff);
		tLen = tUcs2Utf8(ulChar, szResult, sizeof(szResult));
		for (tIndex = 0; tIndex < tLen; tIndex++) {
			vStoreByte((UCHAR)szResult[tIndex], pOutput);
		}
	} else {
		DBG_HEX_C(ulChar > 0xff, ulChar);
		fail(ulChar > 0xff);
		vStoreByte((UCHAR)ulChar, pOutput);
		tLen = 1;
	}
	pOutput->lStringWidth += lComputeStringWidth(
				pOutput->szStorage + pOutput->tNextFree - tLen,
				tLen,
				pOutput->tFontRef,
				pOutput->usFontSize);
} /* end of vStoreChar */

/*
 * vStoreCharacter - store one character
 */
static void
vStoreCharacter(ULONG ulChar, output_type *pOutput)
{
	vStoreChar(ulChar, TRUE, pOutput);
} /* end of vStoreCharacter */

/*
 * vStoreString - store a string
 */
static void
vStoreString(const char *szString, size_t tStringLength, output_type *pOutput)
{
	size_t	tIndex;

	fail(szString == NULL || pOutput == NULL);

	for (tIndex = 0; tIndex < tStringLength; tIndex++) {
		vStoreCharacter((ULONG)(UCHAR)szString[tIndex], pOutput);
	}
} /* end of vStoreString */

/*
 * vStoreNumberAsDecimal - store a number as a decimal number
 */
static void
vStoreNumberAsDecimal(UINT uiNumber, output_type *pOutput)
{
	size_t	tLen;
	char	szString[3 * sizeof(UINT) + 1];

	fail(uiNumber == 0);
	fail(pOutput == NULL);

	tLen = (size_t)sprintf(szString, "%u", uiNumber);
	vStoreString(szString, tLen, pOutput);
} /* end of vStoreNumberAsDecimal */

/*
 * vStoreNumberAsRoman - store a number as a roman numerical
 */
static void
vStoreNumberAsRoman(UINT uiNumber, output_type *pOutput)
{
	size_t	tLen;
	char	szString[15];

	fail(uiNumber == 0);
	fail(pOutput == NULL);

	tLen = tNumber2Roman(uiNumber, FALSE, szString);
	vStoreString(szString, tLen, pOutput);
} /* end of vStoreNumberAsRoman */

/*
 * vStoreStyle - store a style
 */
static void
vStoreStyle(diagram_type *pDiag, output_type *pOutput,
	const style_block_type *pStyle)
{
	size_t	tLen;
	char	szString[120];

	fail(pDiag == NULL);
	fail(pOutput == NULL);
	fail(pStyle == NULL);

	if (tOptions.eConversionType == conversion_xml) {
		vSetHeaders(pDiag, pStyle->usIstd);
	} else {
		tLen = tStyle2Window(szString, sizeof(szString),
					pStyle, pSection);
		vStoreString(szString, tLen, pOutput);
	}
} /* end of vStoreStyle */

/*
 * vPutIndentation - output the specified amount of indentation
 */
static void
vPutIndentation(diagram_type *pDiag, output_type *pOutput,
	BOOL bNoMarks, BOOL bFirstLine,
	UINT uiListNumber, UCHAR ucNFC, const char *szListChar,
	long lLeftIndentation, long lLeftIndentation1)
{
	long	lWidth;
	size_t	tIndex, tNextFree;
	char	szLine[30];

	fail(pDiag == NULL);
	fail(pOutput == NULL);
	fail(szListChar == NULL);
	fail(lLeftIndentation < 0);

	if (tOptions.eConversionType == conversion_xml) {
		/* XML does its own indentation at rendering time */
		return;
	}

	if (bNoMarks) {
		if (bFirstLine) {
			lLeftIndentation += lLeftIndentation1;
		}
		if (lLeftIndentation < 0) {
			lLeftIndentation = 0;
		}
		vSetLeftIndentation(pDiag, lLeftIndentation);
		return;
	}
	if (lLeftIndentation <= 0) {
		DBG_HEX_C(ucNFC != 0x00, ucNFC);
		vSetLeftIndentation(pDiag, 0);
		return;
	}

#if defined(DEBUG)
	if (tOptions.eEncoding == encoding_utf_8) {
		fail(strlen(szListChar) > 3);
	} else {
		DBG_HEX_C(iscntrl((int)szListChar[0]), szListChar[0]);
		fail(iscntrl((int)szListChar[0]));
		fail(szListChar[1] != '\0');
	}
#endif /* DEBUG */

	switch (ucNFC) {
	case LIST_ARABIC_NUM:
	case LIST_NUMBER_TXT:
		tNextFree = (size_t)sprintf(szLine, "%u", uiListNumber);
		break;
	case LIST_UPPER_ROMAN:
	case LIST_LOWER_ROMAN:
		tNextFree = tNumber2Roman(uiListNumber,
				ucNFC == LIST_UPPER_ROMAN, szLine);
		break;
	case LIST_UPPER_ALPHA:
	case LIST_LOWER_ALPHA:
		tNextFree = tNumber2Alpha(uiListNumber,
				ucNFC == LIST_UPPER_ALPHA, szLine);
		break;
	case LIST_ORDINAL_NUM:
	case LIST_ORDINAL_TXT:
		if (uiListNumber % 10 == 1 && uiListNumber != 11) {
			tNextFree =
				(size_t)sprintf(szLine, "%ust", uiListNumber);
		} else if (uiListNumber % 10 == 2 && uiListNumber != 12) {
			tNextFree =
				(size_t)sprintf(szLine, "%und", uiListNumber);
		} else if (uiListNumber % 10 == 3 && uiListNumber != 13) {
			tNextFree =
				(size_t)sprintf(szLine, "%urd", uiListNumber);
		} else {
			tNextFree =
				(size_t)sprintf(szLine, "%uth", uiListNumber);
		}
		break;
	case LIST_OUTLINE_NUM:
		tNextFree = (size_t)sprintf(szLine, "%02u", uiListNumber);
		break;
	case LIST_SPECIAL:
	case LIST_SPECIAL2:
	case LIST_BULLETS:
		tNextFree = 0;
		break;
	default:
		DBG_HEX(ucNFC);
		DBG_FIXME();
		tNextFree = (size_t)sprintf(szLine, "%u", uiListNumber);
		break;
	}
	tNextFree += (size_t)sprintf(szLine + tNextFree, "%.3s", szListChar);
	szLine[tNextFree++] = ' ';
	szLine[tNextFree] = '\0';
	lWidth = lComputeStringWidth(szLine, tNextFree,
				pOutput->tFontRef, pOutput->usFontSize);
	lLeftIndentation -= lWidth;
	if (lLeftIndentation < 0) {
		lLeftIndentation = 0;
	}
	vSetLeftIndentation(pDiag, lLeftIndentation);
	for (tIndex = 0; tIndex < tNextFree; tIndex++) {
		vStoreChar((ULONG)(UCHAR)szLine[tIndex], FALSE, pOutput);
	}
} /* end of vPutIndentation */

/*
 * vPutSeparatorLine - output a separator line
 *
 * A separator line is a horizontal line two inches long.
 * Two inches equals 144000 millipoints.
 */
static void
vPutSeparatorLine(output_type *pOutput)
{
	long	lCharWidth;
	int	iCounter, iChars;
	char	szOne[2];

	fail(pOutput == NULL);

	szOne[0] = OUR_EM_DASH;
	szOne[1] = '\0';
	lCharWidth = lComputeStringWidth(szOne, 1,
				pOutput->tFontRef, pOutput->usFontSize);
	NO_DBG_DEC(lCharWidth);
	iChars = (int)((144000 + lCharWidth / 2) / lCharWidth);
	NO_DBG_DEC(iChars);
	for (iCounter = 0; iCounter < iChars; iCounter++) {
		vStoreCharacter((ULONG)(UCHAR)OUR_EM_DASH, pOutput);
	}
} /* end of vPutSeparatorLine */

/*
 * pStartNextOutput - start the next output record
 *
 * returns a pointer to the next record
 */
static output_type *
pStartNextOutput(output_type *pCurrent)
{
	output_type	*pNew;

	TRACE_MSG("pStartNextOutput");

	if (pCurrent->tNextFree == 0) {
		/* The current record is empty, re-use */
		fail(pCurrent->szStorage[0] != '\0');
		fail(pCurrent->lStringWidth != 0);
		return pCurrent;
	}
	/* The current record is in use, make a new one */
	pNew = xmalloc(sizeof(*pNew));
	pCurrent->pNext = pNew;
	pNew->tStorageSize = INITIAL_SIZE;
	pNew->szStorage = xmalloc(pNew->tStorageSize);
	pNew->szStorage[0] = '\0';
	pNew->tNextFree = 0;
	pNew->lStringWidth = 0;
	pNew->ucFontColor = FONT_COLOR_DEFAULT;
	pNew->usFontStyle = FONT_REGULAR;
	pNew->tFontRef = (drawfile_fontref)0;
	pNew->usFontSize = DEFAULT_FONT_SIZE;
	pNew->pPrev = pCurrent;
	pNew->pNext = NULL;
	return pNew;
} /* end of pStartNextOutput */

/*
 * pStartNewOutput
 */
static output_type *
pStartNewOutput(output_type *pAnchor, output_type *pLeftOver)
{
	output_type	*pCurr, *pNext;
	USHORT		usFontStyle, usFontSize;
	drawfile_fontref	tFontRef;
	UCHAR		ucFontColor;

	TRACE_MSG("pStartNewOutput");

	ucFontColor = FONT_COLOR_DEFAULT;
	usFontStyle = FONT_REGULAR;
	tFontRef = (drawfile_fontref)0;
	usFontSize = DEFAULT_FONT_SIZE;
	/* Free the old output space */
	pCurr = pAnchor;
	while (pCurr != NULL) {
		TRACE_MSG("Free the old output space");
		pNext = pCurr->pNext;
		pCurr->szStorage = xfree(pCurr->szStorage);
		if (pCurr->pNext == NULL) {
			ucFontColor = pCurr->ucFontColor;
			usFontStyle = pCurr->usFontStyle;
			tFontRef = pCurr->tFontRef;
			usFontSize = pCurr->usFontSize;
		}
		pCurr = xfree(pCurr);
		pCurr = pNext;
	}
	if (pLeftOver == NULL) {
		/* Create new output space */
		TRACE_MSG("Create new output space");
		pLeftOver = xmalloc(sizeof(*pLeftOver));
		pLeftOver->tStorageSize = INITIAL_SIZE;
		NO_DBG_DEC(pLeftOver->tStorageSize);
		TRACE_MSG("before 2nd xmalloc");
		pLeftOver->szStorage = xmalloc(pLeftOver->tStorageSize);
		TRACE_MSG("after 2nd xmalloc");
		pLeftOver->szStorage[0] = '\0';
		pLeftOver->tNextFree = 0;
		pLeftOver->lStringWidth = 0;
		pLeftOver->ucFontColor = ucFontColor;
		pLeftOver->usFontStyle = usFontStyle;
		pLeftOver->tFontRef = tFontRef;
		pLeftOver->usFontSize = usFontSize;
		pLeftOver->pPrev = NULL;
		pLeftOver->pNext = NULL;
	}
	fail(!bCheckDoubleLinkedList(pLeftOver));
	return pLeftOver;
} /* end of pStartNewOutput */

/*
 * ulGetChar - get the next character from the specified list
 *
 * returns the next character of EOF
 */