finddup

Borland C++ Compiler 5.5

finddup.cpp

#include <io.h>
#include <mbstring.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

#include <map>
#include <string>

#define BUFFER_SIZE	1000000
#define READ_FIRST	16384

// 型定義
typedef unsigned long	u_long;

typedef struct {			// ファイル情報構造体
	std::string	strPath;
	int		iDupFlag;	// 重複フラグ
} SFileInfo;

typedef std::multimap<u_long, SFileInfo>	MMapULongFI;

// グローバル変数
static MMapULongFI	g_mmap;

// プロトタイプ宣言
int files(const char *pcDir);
int filecmp(const char *pcFile1, const char *pcFile2);

int main(int argc, char **argv)
{
	MMapULongFI::iterator	pmmap;
	MMapULongFI::iterator	pmmapTarget;
	SFileInfo	*pfi;
	SFileInfo	*pfiTarget;
	char	acDir[_MAX_DIR];
	int	iDirLen;
	int	iCount;
	int	i;

if (argc < 2) {
		fprintf(stderr, "usage: finddup dir [...]\n");
		return 1;
	}

for (i = 1; i < argc; i++) {
		iDirLen = strlen(argv[i]);
		if (_MAX_DIR <= iDirLen) {
			fprintf(stderr, "error: dir too long\n");
			return 1;
		}
		strcpy(acDir, argv[i]);
		if (0 < iDirLen && _mbsrchr(acDir, '\\') == acDir + iDirLen - 1) {
			acDir[iDirLen - 1] = '\0';
		}
		if (files(acDir) != 0) {
			return 1;
		}
	}

for (pmmap = g_mmap.begin(); pmmap != g_mmap.end(); pmmap++) {
		pfi = &(pmmap->second);
		if (pfi->iDupFlag != 0) {
			continue;
		}
		iCount = 0;
		for (pmmapTarget = pmmap; ++pmmapTarget != g_mmap.end(); ) {
			if (pmmapTarget->first != pmmap->first) {
				break;
			}
			pfiTarget = &(pmmapTarget->second);
			if (pfiTarget->iDupFlag != 0) {
				continue;
			}
			if (filecmp(pfiTarget->strPath.c_str(), pfi->strPath.c_str()) == 0) {
				pfiTarget->iDupFlag = 1;
				if (iCount == 0) {
					printf("\n%u KB (%u B)\n",
						(pmmap->first + 1023) / 1024, pmmap->first);
					printf("%s\n", pfi->strPath.c_str());
				}
				printf("%s\n", pfiTarget->strPath.c_str());
				iCount++;
			}
		}
	}

return 0;
}

int files(const char *pcDir)
{
	struct _finddata_t	fd;
	SFileInfo	fi;
	char	acPath[_MAX_PATH];
	long	lHandle;
	int	iDirLen;

fi.iDupFlag = 0;

iDirLen = strlen(pcDir);
	sprintf(acPath, "%s\\*.*", pcDir);
	lHandle = _findfirst(acPath, &fd);
	if (lHandle == -1) {
		fprintf(stderr, "error: _findfirst(%s)\n", acPath);
		return -1;
	}

do {
		if (fd.attrib & _A_SUBDIR) {
			if (strcmp(fd.name, ".") && strcmp(fd.name, "..")) {
				if (_MAX_DIR <= iDirLen + 1 + strlen(fd.name)) {
					fprintf(stderr, "error: dir too long\n");
					return -1;
				}
				sprintf(acPath, "%s\\%s", pcDir, fd.name);
				files(acPath);
			}
		} else {
			if (_MAX_PATH <= iDirLen + 1 + strlen(fd.name)) {
				fprintf(stderr, "error: path too long\n");
				fprintf(stderr, "[%s]\n[%s]\n", pcDir, fd.name);
			} else {
				sprintf(acPath, "%s\\%s", pcDir, fd.name);
				fi.strPath = acPath;
				g_mmap.insert(MMapULongFI::value_type(fd.size, fi));
			}
		}
	} while (_findnext(lHandle, &fd) == 0);

_findclose(lHandle);
	return 0;
}

int filecmp(const char *pcFile1, const char *pcFile2)
{
	static char	acBuf1[BUFFER_SIZE];
	static char	acBuf2[BUFFER_SIZE];
	FILE	*pfile1;
	FILE	*pfile2;
	int	iReadCount;
	int	iRead1;
	int	iRead2;
	int	iRetVal;
	u_long	ulCount;

pfile1 = fopen(pcFile1, "rb");
	if (pfile1 == NULL) {
		fprintf(stderr, "error: fopen(%s)\n", pcFile1);
		return -1;
	}

pfile2 = fopen(pcFile2, "rb");
	if (pfile2 == NULL) {
		fprintf(stderr, "error: fopen(%s)\n", pcFile2);
		fclose(pfile1);
		return -1;
	}

ulCount = 0;
	for (iReadCount = READ_FIRST; ; iReadCount = BUFFER_SIZE) {
		iRead1 = fread(acBuf1, sizeof(char), iReadCount, pfile1);
		iRead2 = fread(acBuf2, sizeof(char), iReadCount, pfile2);
		if (iRead1 != iRead2) {
			iRetVal = 1;
			break;
		}
		if (memcmp(acBuf1, acBuf2, iRead1) != 0) {
			if (iReadCount != READ_FIRST) {
//				fprintf(stderr, "\n%u\n[%s]\n[%s]\n", ulCount, pcFile1, pcFile2);
			}
			iRetVal = 1;
			break;
		}
		if (iRead1 != iReadCount) {
			iRetVal = 0;
			break;
		}
		ulCount += iReadCount;
	}

fclose(pfile2);
	fclose(pfile1);
	return iRetVal;
}