/*# proc: binary_subimage_copy_8 - copies a binary subimage that is a multiple# proc:                          of 8 bits wide.# proc: binary_subimage_copy_gt - special condition when copying a# proc:                           binary subimage.# proc: binary_subimage_copy_lt - special condition when copying a# proc:                           binary subimage.# proc: binary_subimage_copy_eq - special condition when copying a# proc:                           binary subimage.# proc: binary_subimage_copy - copies a binary subimage to another image# proc:                        at a specified location.*//* LINTLIBRARY *//****************************************************************//*								*//*	Routines:	binary_subimage_<op>_8()		*//*			binary_subimage_<op>_gt()		*//*			binary_subimage_<op>_lt()		*//*			binary_subimage_<op>_eq()		*//*				where <op> is copy, and, or,	*//*				xor, invert, zero, or one	*//*								*//*	Author:		Stan Janet				*//*	Date:		11/16/90				*//*								*//* binary_subimage_copy_8() is a bit-level copy utility for	*//*	subimages that are a multiple of 8 bits wide.		*//* binary_subimage_copy_{gt,lt,eq}() are bit-level copy		*//*	utilities for subimages that are not a multiple of 8	*//*	bits wide and where the space remaining in the last	*//*	byte in every destination scan line will be (respect-	*//*	ively) greater than, less than, or equal to the		*//*	number of bits by which the subimage width exceeds 8.	*//* If binary_subimage_copy() is told to copy a subimage		*//*	that is a multiple of 8 bits wide, it calls		*//*	binary_subimage_copy_8(). Otherwise it calls one of	*//*	the other functions as appropriate.			*//* binary_subimage_or_8() is a bit-level logical or utility for	*//*	subimages that are a multiple of 8 bits wide.		*//* binary_subimage_{or,and,xor,invert}_{8,gt,lt,eq}() are	*//*	bit-level logical or, and, xor and invert utilities for *//*	subimages that are analogous to	the copy utilities	*//*	described above.					*//* binary_subimage_{zero,one}_{8,gt,lt,eq} are bit-level util-	*//*	ities for setting all bits in a subimage to zero and	*//*	one, respectively.					*//* All dimensions are taken to be in bit units.			*//****************************************************************/#include <stdio.h>#include <memory.h>#include <values.h>#include <sys/types.h>#include <masks.h>#include <bitmasks.h>#include <bits.h>#include <binops.h>void binary_subimage_copy_8 (		src, srcw, srch,		dst, dstw, dsth,		srcx, srcy,		cpw, cph,		dstx, dsty		)	register u_char *src, *dst;	int srcx,srcy,dstx,dsty, srcw,srch,dstw,dsth, cpw,cph;{int i, j, i_inv, j_inv, cpbw, srcbw, dstbw, dx, dy;u_char m0, m1, m2;register int bytes;register u_char c;if (src == (u_char *) NULL)	fatalerr("binary_subimage_copy_8","Null source image pointer",(char *)NULL);if (dst == (u_char *) NULL)	fatalerr("binary_subimage_copy_8","Null destination image pointer",(char *)NULL);if ((srcw < 0) || (srch < 0))	fatalerr("binary_subimage_copy_8","Negative source image dimension(s)",(char *)NULL);if ((dstw < 0) || (dsth < 0))	fatalerr("binary_subimage_copy_8","Negative destination image dimension(s)",(char *)NULL);if ((cpw < 0) || (cph < 0))	fatalerr("binary_subimage_copy_8","Negative subimage dimension(s)",(char *)NULL);if (srcw % BITSPERBYTE)	fatalerr("binary_subimage_copy_8",		"Source image width must be a multiple of 8",(char *)NULL);if (dstw % BITSPERBYTE)	fatalerr("binary_subimage_copy_8",		"Destination image width must be byte aligned",(char *)NULL);if (((srcx + cpw) > srcw) || ((srcy + cph) > srch))	fatalerr("binary_subimage_copy_8",		"Subimage exceeds source image dimension(s)",(char *)NULL);if (((dstx + cpw) > dstw) || ((dsty + cph) > dsth))	fatalerr("binary_subimage_copy_8",		"Subimage exceeds destination image dimension(s)",(char *)NULL);if (!cpw || !cph)	return;if (!srcw || !srch)	return;if (!dstw || !dsth)	return;	i = srcx % BITSPERBYTE;	i_inv = BITSPERBYTE - i;j = dstx % BITSPERBYTE;j_inv = BITSPERBYTE - j;	if (cpw % BITSPERBYTE)		fatalerr("binary_subimage_copy_8",			"Copy width not a multiple of eight",(char *)NULL);	srcbw = srcw / BITSPERBYTE;dstbw = dstw / BITSPERBYTE;cpbw  = cpw  / BITSPERBYTE;	/* Increment src to the byte in the upper left corner of the */	/* subimage to be copied. Increment dst to the upper left corner */	/* of the byte in the destination that will be modified. */	src += srcy * srcbw + srcx / BITSPERBYTE;dst += dsty * dstbw + dstx / BITSPERBYTE;	/* Set dx (dy) to be the increment needed to get src (dst) to the */	/* first byte in the next scan line to be copied (or modified) from */	/* the last byte in the current scan line. */	dx = srcbw - cpbw;dy = dstbw - cpbw;if (!i && !j) {			/* Both src & dst are byte-aligned. */	while (cph--) {		bytes = cpbw;		while (bytes--)			*dst++ = *src++;		src += dx;		dst += dy;	}	return;}if (!j) {			/* Dst is byte-aligned. */	while (cph--) {		bytes = cpbw;		while (bytes--) {			c = (*src << i) | (*(src+1) >> i_inv);			*dst++ = c;			src++;		}		src += dx;		dst += dy;	}	return;}if (!i) {					/* Src is byte-aligned. */	m1 = mask_end_0[j_inv];				/* eg. 11110000 */	m2 = mask_begin_0[j];				/* eg. 00001111 */	while (cph--) {		bytes = cpbw;		while (bytes--) {			c = *src++;			*dst = (*dst & m1) | (c >> j);			dst++;			*dst = (*dst & m2) | (c << j_inv);		}		;		src += dx;		dst += dy;	}	return;}if (i == j) {	m0 = mask_begin_1[i];	m1 = mask_end_0[i_inv];	m2 = mask_end_1[i_inv];	/* Neither src nor dst is byte-aligned, but they */	/* are offset by the same amount, so we can copy */	/* (cpbw - 1) bytes directly from src to dst. The */	/* bits in the other byte are on the ends of the */	/* scan line and are treated specially. */	if (cpbw == 1) {		while (cph--) {			*dst = (*dst & m0) | (*src++ & m2);			dst++;			*dst = (*dst & m2) | (*src & m0);			;			src += dx;			dst += dy;		}		return;	}	while (cph--) {		*dst = (*dst & m0) | (*src++ & m2);		dst++;		bytes = cpbw - 1;		while (bytes--)			*dst++ = *src++;		*dst = (*dst & m2) | (*src & m0);		;		src += dx;		dst += dy;	}	return;}m1 = mask_end_0[j_inv];		/* eg. 11111000 */m2 = mask_begin_0[j];		/* eg. 00000111 */while (cph--) {				/* The absolute worst case. Neither */	bytes = cpbw;			/* src nor dst is byte-aligned, and */	while (bytes--) {		/* they are offset by diff. amounts. */		c = (*src << i) | (*(src+1) >> i_inv);		*dst = (*dst & m1) | (c >> j);		dst++;		*dst = (*dst & m2) | (c << j_inv);		src++;	}	;	src += dx;	dst += dy;}}/* LINTLIBRARY *//****************************************************************//*								*//*	Routines:	binary_subimage_<op>_8()		*//*			binary_subimage_<op>_gt()		*//*			binary_subimage_<op>_lt()		*//*			binary_subimage_<op>_eq()		*//*				where <op> is copy, and, or,	*//*				xor, invert, zero, or one	*//*								*//*	Author:		Stan Janet				*//*	Date:		11/16/90				*//*								*//* binary_subimage_copy_8() is a bit-level copy utility for	*//*	subimages that are a multiple of 8 bits wide.		*//* binary_subimage_copy_{gt,lt,eq}() are bit-level copy		*//*	utilities for subimages that are not a multiple of 8	*//*	bits wide and where the space remaining in the last	*//*	byte in every destination scan line will be (respect-	*//*	ively) greater than, less than, or equal to the		*//*	number of bits by which the subimage width exceeds 8.	*//* If binary_subimage_copy() is told to copy a subimage		*//*	that is a multiple of 8 bits wide, it calls		*//*	binary_subimage_copy_8(). Otherwise it calls one of	*//*	the other functions as appropriate.			*//* binary_subimage_or_8() is a bit-level logical or utility for	*//*	subimages that are a multiple of 8 bits wide.		*//* binary_subimage_{or,and,xor,invert}_{8,gt,lt,eq}() are	*//*	bit-level logical or, and, xor and invert utilities for *//*	subimages that are analogous to	the copy utilities	*//*	described above.					*//* binary_subimage_{zero,one}_{8,gt,lt,eq} are bit-level util-	*//*	ities for setting all bits in a subimage to zero and	*//*	one, respectively.					*//* All dimensions are taken to be in bit units.			*//****************************************************************/#include <stdio.h>#include <memory.h>#include <values.h>#include <sys/types.h>#include <masks.h>#include <bitmasks.h>#include <bits.h>#include <binops.h>void binary_subimage_copy_gt (		src, srcw, srch,		dst, dstw, dsth,		srcx, srcy,		cpw, cph,		dstx, dsty		)	register u_char *src, *dst;	int srcx,srcy,dstx,dsty, srcw,srch,dstw,dsth, cpw,cph;{int i, j, i_inv, j_inv, cpbw, srcbw, dstbw, dx, dy;u_char m0, m1, m2;register int bytes;register u_char c;	u_char m3, m4;	int b, b_inv, diff;if (src == (u_char *) NULL)	fatalerr("binary_subimage_copy_gt","Null source image pointer",(char *)NULL);if (dst == (u_char *) NULL)	fatalerr("binary_subimage_copy_gt","Null destination image pointer",(char *)NULL);if ((srcw < 0) || (srch < 0))	fatalerr("binary_subimage_copy_gt","Negative source image dimension(s)",(char *)NULL);if ((dstw < 0) || (dsth < 0))	fatalerr("binary_subimage_copy_gt","Negative destination image dimension(s)",(char *)NULL);if ((cpw < 0) || (cph < 0))	fatalerr("binary_subimage_copy_gt","Negative subimage dimension(s)",(char *)NULL);if (srcw % BITSPERBYTE)	fatalerr("binary_subimage_copy_gt",		"Source image width must be a multiple of 8",(char *)NULL);if (dstw % BITSPERBYTE)	fatalerr("binary_subimage_copy_gt",		"Destination image width must be byte aligned",(char *)NULL);if (((srcx + cpw) > srcw) || ((srcy + cph) > srch))	fatalerr("binary_subimage_copy_gt",		"Subimage exceeds source image dimension(s)",(char *)NULL);if (((dstx + cpw) > dstw) || ((dsty + cph) > dsth))	fatalerr("binary_subimage_copy_gt",		"Subimage exceeds destination image dimension(s)",(char *)NULL);if (!cpw || !cph)	return;if (!srcw || !srch)	return;if (!dstw || !dsth)	return;	i = srcx % BITSPERBYTE;	i_inv = BITSPERBYTE - i;j = dstx % BITSPERBYTE;j_inv = BITSPERBYTE - j;		b = cpw % BITSPERBYTE;		b_inv = BITSPERBYTE - b;		diff = j_inv - b;		if (diff <= 0)		fatalerr("binary_subimage_copy_gt","bad diff",(char *)NULL);	srcbw = srcw / BITSPERBYTE;dstbw = dstw / BITSPERBYTE;cpbw  = cpw  / BITSPERBYTE;	/* Increment src to the byte in the upper left corner of the */	/* subimage to be copied. Increment dst to the upper left corner */	/* of the byte in the destination that will be modified. */	src += srcy * srcbw + srcx / BITSPERBYTE;dst += dsty * dstbw + dstx / BITSPERBYTE;	/* Set dx (dy) to be the increment needed to get src (dst) to the */	/* first byte in the next scan line to be copied (or modified) from */	/* the last byte in the current scan line. */	dx = srcbw - cpbw;dy = dstbw - cpbw;if (!i && !j) {			/* Both src & dst are byte-aligned. */	m3 = mask_end_1[b_inv];	/* eg. 00000111 */	m4 = mask_begin_1[b];	/* eg. 11111000 */	while (cph--) {		bytes = cpbw;		while (bytes--)			*dst++ = *src++;		*dst = (*dst & m3) | (*src & m4);		src += dx;