[apple/icu.git] / icuSources / samples / ugrep / ugrep.cpp

/*************************************************************************
*
*   © 2016 and later: Unicode, Inc. and others.
*   License & terms of use: http://www.unicode.org/copyright.html#License
*
**************************************************************************
**************************************************************************
*
*   Copyright (C) 2002-2010, International Business Machines
*   Corporation and others.  All Rights Reserved.
*
***************************************************************************
*/

//
//   ugrep  - an ICU sample program illustrating the use of ICU Regular Expressions.
//
//            The use of the ICU Regex API all occurs within the main()
//            function.  The rest of the code deals with opening files,
//            encoding conversions, printing results, etc.
//
//            This is not a full-featured grep program.  The command line options
//            have been kept to a minimum to avoid complicating the sample code.
//


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

#include "unicode/utypes.h"
#include "unicode/ustring.h"
#include "unicode/regex.h"
#include "unicode/ucnv.h"
#include "unicode/uclean.h"

using namespace icu;

//
//  The following variables contain parameters that may be set from the command line.
//
const char *pattern = NULL;     // The regular expression
int        firstFileNum;        //  argv index of the first file name
UBool      displayFileName = FALSE;
UBool      displayLineNum  = FALSE;


//
//  Info regarding the file currently being processed
//
const char *fileName;      
int         fileLen;              // Length, in UTF-16 Code Units.  

UChar      *ucharBuf = 0;         // Buffer, holds converted file.  (Simple minded program, always reads
                                  //   the whole file at once.

char       *charBuf = 0;          // Buffer, for original, unconverted file data.


//
//  Info regarding the line currently being processed
//
int      lineStart;     // Index of first char of the current line in the file buffer
int      lineEnd;       // Index of char following the new line sequence for the current line
int      lineNum;

//
//  Converter, used on output to convert Unicode data back to char *
//             so that it will display in non-Unicode terminal windows.
//
UConverter  *outConverter = 0;

//
//  Function forward declarations
//
void processOptions(int argc, const char **argv);
void nextLine(int start);
void printMatch();
void printUsage();
void readFile(const char *name);


//------------------------------------------------------------------------------------------
//
//   main          for ugrep
//
//           Structurally, all use of the ICU Regular Expression API is in main(),
//           and all of the supporting stuff necessary to make a running program, but
//           not directly related to regular expressions, is factored out into these other
//           functions.
//
//------------------------------------------------------------------------------------------
int main(int argc, const char** argv) {
    UBool     matchFound = FALSE;

    //
    //  Process the command line options.
    //
    processOptions(argc, argv);

    //
    // Create a RegexPattern object from the user supplied pattern string.
    //
    UErrorCode status = U_ZERO_ERROR;   // All ICU operations report success or failure
                                        //   in a status variable.

    UParseError    parseErr;            // In the event of a syntax error in the regex pattern,
                                        //   this struct will contain the position of the
                                        //   error.

    RegexPattern  *rePat = RegexPattern::compile(pattern, parseErr, status);
                                        // Note that C++ is doing an automatic conversion
                                        //  of the (char *) pattern to a temporary
                                        //  UnicodeString object.
    if (U_FAILURE(status)) {
        fprintf(stderr, "ugrep:  error in pattern: \"%s\" at position %d\n",
            u_errorName(status), parseErr.offset);
        exit(-1);
    }

    //
    // Create a RegexMatcher from the newly created pattern.
    //
    UnicodeString empty;
    RegexMatcher *matcher = rePat->matcher(empty, status);
    if (U_FAILURE(status)) {
        fprintf(stderr, "ugrep:  error in creating RegexMatcher: \"%s\"\n",
            u_errorName(status));
        exit(-1);
    }

    //
    // Loop, processing each of the input files.
    //
    for (int fileNum=firstFileNum; fileNum < argc; fileNum++) {
        readFile(argv[fileNum]);

        //
        //  Loop through the lines of a file, trying to match the regex pattern on each.
        //
        for (nextLine(0); lineStart<fileLen; nextLine(lineEnd)) {
            UnicodeString s(FALSE, ucharBuf+lineStart, lineEnd-lineStart);
            matcher->reset(s);
            if (matcher->find()) {
                matchFound = TRUE;
                printMatch();
            }
        }
    }

    //
    //  Clean up
    //
    delete matcher;
    delete rePat;
    free(ucharBuf);
    free(charBuf);
    ucnv_close(outConverter);
    
    u_cleanup();       // shut down ICU, release any cached data it owns.

    return matchFound? 0: 1;
}


//------------------------------------------------------------------------------------------
//
//   doOptions          Run through the command line options, and set
//                      the global variables accordingly.
//
//                      exit without returning if an error occurred and
//                      ugrep should not proceed further.
//
//------------------------------------------------------------------------------------------
void processOptions(int argc, const char **argv) {
    int            optInd;
    UBool          doUsage   = FALSE;
    UBool          doVersion = FALSE;
    const char    *arg;


    for(optInd = 1; optInd < argc; ++optInd) {
        arg = argv[optInd];
        
        /* version info */
        if(strcmp(arg, "-V") == 0 || strcmp(arg, "--version") == 0) {
            doVersion = TRUE;
        }
        /* usage info */
        else if(strcmp(arg, "--help") == 0) {
            doUsage = TRUE;
        }
        else if(strcmp(arg, "-n") == 0 || strcmp(arg, "--line-number") == 0) {
            displayLineNum = TRUE;
        }
        /* POSIX.1 says all arguments after -- are not options */
        else if(strcmp(arg, "--") == 0) {
            /* skip the -- */
            ++optInd;
            break;
        }
        /* unrecognized option */
        else if(strncmp(arg, "-", strlen("-")) == 0) {
            printf("ugrep: invalid option -- %s\n", arg+1);
            doUsage = TRUE;
        }
        /* done with options */
        else {
            break;
        }
    }

    if (doUsage) {
        printUsage();
        exit(0);
    }

    if (doVersion) {
        printf("ugrep version 0.01\n");
        if (optInd == argc) {
            exit(0);
        }
    }

    int  remainingArgs = argc-optInd;     // pattern file ...
    if (remainingArgs < 2) {
        fprintf(stderr, "ugrep:  files or pattern are missing.\n");
        printUsage();
        exit(1);
    }

    if (remainingArgs > 2) {
        // More than one file to be processed.   Display file names with match output.
        displayFileName = TRUE;
    }

    pattern      = argv[optInd];
    firstFileNum = optInd+1;
}

//------------------------------------------------------------------------------------------
//
//   printUsage
//
//------------------------------------------------------------------------------------------
void printUsage() {
    printf("ugrep [options] pattern file...\n"
        "     -V or --version     display version information\n"
        "     --help              display this help and exit\n"
        "     --                  stop further option processing\n"
        "-n,  --line-number       Prefix each line of output with the line number within its input file.\n"
        );
    exit(0);
}

//------------------------------------------------------------------------------------------
//
//    readFile          Read a file into memory, and convert it to Unicode.
//
//                      Since this is just a demo program, take the simple minded approach
//                      of always reading the whole file at once.  No intelligent buffering
//                      is done.
//
//------------------------------------------------------------------------------------------
void readFile(const char *name) {

    //
    //  Initialize global file variables
    //
    fileName = name;
    fileLen  = 0;      // zero length prevents processing in case of errors.


    //
    //  Open the file and determine its size.
    //
    FILE *file = fopen(name, "rb");
    if (file == 0 ) {
        fprintf(stderr, "ugrep: Could not open file \"%s\"\n", fileName);
        return;
    }
    fseek(file, 0, SEEK_END);
    int rawFileLen = ftell(file);
    fseek(file, 0, SEEK_SET);
    

    //
    //   Read in the file
    //
    charBuf    = (char *)realloc(charBuf, rawFileLen+1);   // Need error checking...
    int t = static_cast<int>(fread(charBuf, 1, rawFileLen, file));
    if (t != rawFileLen)  {
        fprintf(stderr, "Error reading file \"%s\"\n", fileName);
        fclose(file);
        return;
    }
    charBuf[rawFileLen]=0;
    fclose(file);

    //
    // Look for a Unicode Signature (BOM) in the data
    //
    int32_t        signatureLength;
    const char *   charDataStart = charBuf;
    UErrorCode     status        = U_ZERO_ERROR;
    const char*    encoding      = ucnv_detectUnicodeSignature(
                           charDataStart, rawFileLen, &signatureLength, &status);
    if (U_FAILURE(status)) {
        fprintf(stderr, "ugrep: ICU Error \"%s\" from ucnv_detectUnicodeSignature()\n",
            u_errorName(status));
        return;
    }
    if(encoding!=NULL ){
        charDataStart  += signatureLength;
        rawFileLen     -= signatureLength;
    }

    //
    // Open a converter to take the file to UTF-16
    //
    UConverter* conv;
    conv = ucnv_open(encoding, &status);
    if (U_FAILURE(status)) {
        fprintf(stderr, "ugrep: ICU Error \"%s\" from ucnv_open()\n", u_errorName(status));
        return;
    }

    //
    // Convert the file data to UChar.
    //  Preflight first to determine required buffer size.
    //
    uint32_t destCap = ucnv_toUChars(conv,
                       NULL,           //  dest,
                       0,              //  destCapacity,
                       charDataStart,
                       rawFileLen,
                       &status);
    if (status != U_BUFFER_OVERFLOW_ERROR) {
        fprintf(stderr, "ugrep: ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status));
        return;
    };
    
    status = U_ZERO_ERROR;
    ucharBuf = (UChar *)realloc(ucharBuf, (destCap+1) * sizeof(UChar));
    ucnv_toUChars(conv,
        ucharBuf,           //  dest,
        destCap+1,
        charDataStart,
        rawFileLen,
        &status);
    if (U_FAILURE(status)) {
        fprintf(stderr, "ugrep: ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status));
        return;
    };
    ucnv_close(conv);
    
    //
    //  Successful conversion.  Set the global size variables so that
    //     the rest of the processing will proceed for this file.
    //
    fileLen = destCap;
}
    
    
//------------------------------------------------------------------------------------------
//
//   nextLine           Advance the line index variables, starting at the
//                      specified position in the input file buffer, by
//                      scanning forward until the next end-of-line.
//
//                      Need to take into account all of the possible Unicode
//                      line ending sequences.
//
//------------------------------------------------------------------------------------------
void nextLine(int  startPos) {
    if (startPos == 0) {
        lineNum = 0;
    } else {
        lineNum++;
    }
    lineStart = lineEnd = startPos;

    for (;;) {
        if (lineEnd >= fileLen) {
            return;
        }
        UChar c = ucharBuf[lineEnd];
        lineEnd++;
        if (c == 0x0a   ||       // Line Feed
            c == 0x0c   ||       // Form Feed
            c == 0x0d   ||       // Carriage Return
            c == 0x85   ||       // Next Line
            c == 0x2028 ||       // Line Separator
            c == 0x2029)         // Paragraph separator
        { 
            break;
        }
    }

    // Check for CR/LF sequence, and advance over the LF if we're in the middle of one.
    if (lineEnd < fileLen           &&
        ucharBuf[lineEnd-1] == 0x0d &&
        ucharBuf[lineEnd]   == 0x0a) 
    {
        lineEnd++;
    }
}


//------------------------------------------------------------------------------------------
//
//   printMatch         Called when a matching line has been located.
//                      Print out the line from the file with the match, after
//                         converting it back to the default code page.
//
//------------------------------------------------------------------------------------------
void printMatch() {
    char                buf[2000];
    UErrorCode         status       = U_ZERO_ERROR;

    // If we haven't already created a converter for output, do it now.
    if (outConverter == 0) {
        outConverter = ucnv_open(NULL, &status);
        if (U_FAILURE(status)) {
            fprintf(stderr, "ugrep:  Error opening default converter: \"%s\"\n",
                u_errorName(status));
            exit(-1);
        }
    };

    // Convert the line to be printed back to the default 8 bit code page.
    //   If the line is too long for our buffer, just truncate it.
    ucnv_fromUChars(outConverter,
                    buf,                   // destination buffer for conversion
                    sizeof(buf),           // capacity of destination buffer
                    &ucharBuf[lineStart],   // Input to conversion
                    lineEnd-lineStart,     // number of UChars to convert
                    &status);
    buf[sizeof(buf)-1] = 0;                // Add null for use in case of too long lines.
                                           // The converter null-terminates its output unless
                                           //   the buffer completely fills.
   
    if (displayFileName) {
        printf("%s:", fileName);
    }
    if (displayLineNum) {
        printf("%d:", lineNum);
    }
    printf("%s", buf);
}
Commit	Line	Data
f3c0d7a5 A	1	/*************************************************************************
	2	*
	3	* © 2016 and later: Unicode, Inc. and others.
	4	* License & terms of use: http://www.unicode.org/copyright.html#License
	5	*
	6	**************************************************************************
	7	**************************************************************************
b75a7d8f	8	*
729e4ab9	9	* Copyright (C) 2002-2010, International Business Machines
b75a7d8f A	10	* Corporation and others. All Rights Reserved.
	11	*
	12	***************************************************************************
	13	*/
	14
	15	//
	16	// ugrep - an ICU sample program illustrating the use of ICU Regular Expressions.
	17	//
	18	// The use of the ICU Regex API all occurs within the main()
0f5d89e8	19	// function. The rest of the code deals with opening files,
b75a7d8f A	20	// encoding conversions, printing results, etc.
	21	//
	22	// This is not a full-featured grep program. The command line options
	23	// have been kept to a minimum to avoid complicating the sample code.
	24	//
	25
	26
	27
	28	#include <stdio.h>
	29	#include <stdlib.h>
	30	#include <string.h>
	31
	32	#include "unicode/utypes.h"
	33	#include "unicode/ustring.h"
	34	#include "unicode/regex.h"
	35	#include "unicode/ucnv.h"
	36	#include "unicode/uclean.h"
	37
0f5d89e8	38	using namespace icu;
b75a7d8f A	39
b75a7d8f A	40	//
0f5d89e8	41	// The following variables contain parameters that may be set from the command line.
b75a7d8f A	42	//
	43	const char *pattern = NULL; // The regular expression
	44	int firstFileNum; // argv index of the first file name
	45	UBool displayFileName = FALSE;
	46	UBool displayLineNum = FALSE;
	47
	48
	49	//
	50	// Info regarding the file currently being processed
	51	//
	52	const char *fileName;
	53	int fileLen; // Length, in UTF-16 Code Units.
	54
	55	UChar *ucharBuf = 0; // Buffer, holds converted file. (Simple minded program, always reads
	56	// the whole file at once.
	57
	58	char *charBuf = 0; // Buffer, for original, unconverted file data.
	59
	60
	61	//
	62	// Info regarding the line currently being processed
	63	//
	64	int lineStart; // Index of first char of the current line in the file buffer
	65	int lineEnd; // Index of char following the new line sequence for the current line
	66	int lineNum;
	67
	68	//
	69	// Converter, used on output to convert Unicode data back to char *
	70	// so that it will display in non-Unicode terminal windows.
	71	//
	72	UConverter *outConverter = 0;
	73
	74	//
	75	// Function forward declarations
	76	//
	77	void processOptions(int argc, const char **argv);
	78	void nextLine(int start);
	79	void printMatch();
	80	void printUsage();
	81	void readFile(const char *name);
	82
	83
	84
	85	//------------------------------------------------------------------------------------------
	86	//
	87	// main for ugrep
	88	//
	89	// Structurally, all use of the ICU Regular Expression API is in main(),
	90	// and all of the supporting stuff necessary to make a running program, but
	91	// not directly related to regular expressions, is factored out into these other
	92	// functions.
	93	//
	94	//------------------------------------------------------------------------------------------
	95	int main(int argc, const char** argv) {
	96	UBool matchFound = FALSE;
	97
	98	//
0f5d89e8	99	// Process the command line options.
b75a7d8f A	100	//
	101	processOptions(argc, argv);
	102
	103	//
	104	// Create a RegexPattern object from the user supplied pattern string.
	105	//
	106	UErrorCode status = U_ZERO_ERROR; // All ICU operations report success or failure
	107	// in a status variable.
	108
	109	UParseError parseErr; // In the event of a syntax error in the regex pattern,
	110	// this struct will contain the position of the
	111	// error.
	112
	113	RegexPattern *rePat = RegexPattern::compile(pattern, parseErr, status);
	114	// Note that C++ is doing an automatic conversion
	115	// of the (char *) pattern to a temporary
	116	// UnicodeString object.
	117	if (U_FAILURE(status)) {
	118	fprintf(stderr, "ugrep: error in pattern: \"%s\" at position %d\n",
	119	u_errorName(status), parseErr.offset);
	120	exit(-1);
	121	}
	122
	123	//
	124	// Create a RegexMatcher from the newly created pattern.
	125	//
	126	UnicodeString empty;
	127	RegexMatcher *matcher = rePat->matcher(empty, status);
	128	if (U_FAILURE(status)) {
	129	fprintf(stderr, "ugrep: error in creating RegexMatcher: \"%s\"\n",
	130	u_errorName(status));
	131	exit(-1);
	132	}
	133
	134	//
	135	// Loop, processing each of the input files.
	136	//
	137	for (int fileNum=firstFileNum; fileNum < argc; fileNum++) {
	138	readFile(argv[fileNum]);
	139
	140	//
	141	// Loop through the lines of a file, trying to match the regex pattern on each.
	142	//
	143	for (nextLine(0); lineStart<fileLen; nextLine(lineEnd)) {
	144	UnicodeString s(FALSE, ucharBuf+lineStart, lineEnd-lineStart);
	145	matcher->reset(s);
	146	if (matcher->find()) {
	147	matchFound = TRUE;
	148	printMatch();
	149	}
	150	}
	151	}
	152
	153	//
	154	// Clean up
	155	//
	156	delete matcher;
	157	delete rePat;
	158	free(ucharBuf);
	159	free(charBuf);
	160	ucnv_close(outConverter);
	161
	162	u_cleanup(); // shut down ICU, release any cached data it owns.
	163
164	return matchFound? 0: 1;
165	}
166
167
168
169	//------------------------------------------------------------------------------------------
170	//
171	// doOptions Run through the command line options, and set
172	// the global variables accordingly.
173	//
0f5d89e8	174	// exit without returning if an error occurred and
b75a7d8f A	175	// ugrep should not proceed further.
	176	//
	177	//------------------------------------------------------------------------------------------
	178	void processOptions(int argc, const char **argv) {
	179	int optInd;
	180	UBool doUsage = FALSE;
	181	UBool doVersion = FALSE;
	182	const char *arg;
	183
	184
	185	for(optInd = 1; optInd < argc; ++optInd) {
	186	arg = argv[optInd];
	187
	188	/* version info */
	189	if(strcmp(arg, "-V") == 0 \|\| strcmp(arg, "--version") == 0) {
	190	doVersion = TRUE;
	191	}
	192	/* usage info */
	193	else if(strcmp(arg, "--help") == 0) {
	194	doUsage = TRUE;
	195	}
	196	else if(strcmp(arg, "-n") == 0 \|\| strcmp(arg, "--line-number") == 0) {
	197	displayLineNum = TRUE;
	198	}
	199	/* POSIX.1 says all arguments after -- are not options */
	200	else if(strcmp(arg, "--") == 0) {
	201	/* skip the -- */
	202	++optInd;
	203	break;
	204	}
	205	/* unrecognized option */
	206	else if(strncmp(arg, "-", strlen("-")) == 0) {
	207	printf("ugrep: invalid option -- %s\n", arg+1);
	208	doUsage = TRUE;
	209	}
	210	/* done with options */
	211	else {
	212	break;
	213	}
	214	}
	215
	216	if (doUsage) {
	217	printUsage();
	218	exit(0);
	219	}
	220
	221	if (doVersion) {
	222	printf("ugrep version 0.01\n");
	223	if (optInd == argc) {
	224	exit(0);
	225	}
	226	}
	227
	228	int remainingArgs = argc-optInd; // pattern file ...
	229	if (remainingArgs < 2) {
	230	fprintf(stderr, "ugrep: files or pattern are missing.\n");
	231	printUsage();
	232	exit(1);
	233	}
	234
	235	if (remainingArgs > 2) {
	236	// More than one file to be processed. Display file names with match output.
	237	displayFileName = TRUE;
	238	}
239
240	pattern = argv[optInd];
241	firstFileNum = optInd+1;
242	}
243
244	//------------------------------------------------------------------------------------------
245	//
246	// printUsage
247	//
248	//------------------------------------------------------------------------------------------
249	void printUsage() {
250	printf("ugrep [options] pattern file...\n"
251	" -V or --version display version information\n"
252	" --help display this help and exit\n"
253	" -- stop further option processing\n"
254	"-n, --line-number Prefix each line of output with the line number within its input file.\n"
255	);
256	exit(0);
257	}
258
259	//------------------------------------------------------------------------------------------
260	//
261	// readFile Read a file into memory, and convert it to Unicode.
262	//
263	// Since this is just a demo program, take the simple minded approach
264	// of always reading the whole file at once. No intelligent buffering
265	// is done.
266	//
267	//------------------------------------------------------------------------------------------
268	void readFile(const char *name) {
269
270	//
271	// Initialize global file variables
272	//
273	fileName = name;
274	fileLen = 0; // zero length prevents processing in case of errors.
275
276
277	//
278	// Open the file and determine its size.
279	//
280	FILE *file = fopen(name, "rb");
281	if (file == 0 ) {
282	fprintf(stderr, "ugrep: Could not open file \"%s\"\n", fileName);
283	return;
284	}
285	fseek(file, 0, SEEK_END);
286	int rawFileLen = ftell(file);
287	fseek(file, 0, SEEK_SET);
288
289
290	//
291	// Read in the file
292	//
293	charBuf = (char *)realloc(charBuf, rawFileLen+1); // Need error checking...
0f5d89e8	294	int t = static_cast<int>(fread(charBuf, 1, rawFileLen, file));
b75a7d8f A	295	if (t != rawFileLen) {
b75a7d8f A	296	fprintf(stderr, "Error reading file \"%s\"\n", fileName);
729e4ab9	297	fclose(file);
b75a7d8f A	298	return;
	299	}
	300	charBuf[rawFileLen]=0;
	301	fclose(file);
	302
	303	//
	304	// Look for a Unicode Signature (BOM) in the data
	305	//
	306	int32_t signatureLength;
	307	const char * charDataStart = charBuf;
	308	UErrorCode status = U_ZERO_ERROR;
	309	const char* encoding = ucnv_detectUnicodeSignature(
	310	charDataStart, rawFileLen, &signatureLength, &status);
	311	if (U_FAILURE(status)) {
	312	fprintf(stderr, "ugrep: ICU Error \"%s\" from ucnv_detectUnicodeSignature()\n",
	313	u_errorName(status));
	314	return;
	315	}
	316	if(encoding!=NULL ){
	317	charDataStart += signatureLength;
	318	rawFileLen -= signatureLength;
	319	}
	320
	321	//
	322	// Open a converter to take the file to UTF-16
	323	//
	324	UConverter* conv;
	325	conv = ucnv_open(encoding, &status);
	326	if (U_FAILURE(status)) {
	327	fprintf(stderr, "ugrep: ICU Error \"%s\" from ucnv_open()\n", u_errorName(status));
	328	return;
	329	}
	330
	331	//
	332	// Convert the file data to UChar.
	333	// Preflight first to determine required buffer size.
	334	//
	335	uint32_t destCap = ucnv_toUChars(conv,
	336	NULL, // dest,
	337	0, // destCapacity,
	338	charDataStart,
	339	rawFileLen,
	340	&status);
	341	if (status != U_BUFFER_OVERFLOW_ERROR) {
	342	fprintf(stderr, "ugrep: ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status));
	343	return;
	344	};
	345
	346	status = U_ZERO_ERROR;
	347	ucharBuf = (UChar )realloc(ucharBuf, (destCap+1) sizeof(UChar));
	348	ucnv_toUChars(conv,
	349	ucharBuf, // dest,
	350	destCap+1,
	351	charDataStart,
	352	rawFileLen,
	353	&status);
	354	if (U_FAILURE(status)) {
	355	fprintf(stderr, "ugrep: ucnv_toUChars: ICU Error \"%s\"\n", u_errorName(status));
	356	return;
	357	};
	358	ucnv_close(conv);
	359
	360	//
	361	// Successful conversion. Set the global size variables so that
362	// the rest of the processing will proceed for this file.
363	//
364	fileLen = destCap;
365	}
366
367
368
369
370
371	//------------------------------------------------------------------------------------------
372	//
373	// nextLine Advance the line index variables, starting at the
374	// specified position in the input file buffer, by
0f5d89e8	375	// scanning forward until the next end-of-line.
b75a7d8f A	376	//
	377	// Need to take into account all of the possible Unicode
	378	// line ending sequences.
	379	//
	380	//------------------------------------------------------------------------------------------
	381	void nextLine(int startPos) {
	382	if (startPos == 0) {
	383	lineNum = 0;
	384	} else {
	385	lineNum++;
	386	}
	387	lineStart = lineEnd = startPos;
	388
	389	for (;;) {
	390	if (lineEnd >= fileLen) {
	391	return;
	392	}
	393	UChar c = ucharBuf[lineEnd];
	394	lineEnd++;
	395	if (c == 0x0a \|\| // Line Feed
	396	c == 0x0c \|\| // Form Feed
	397	c == 0x0d \|\| // Carriage Return
	398	c == 0x85 \|\| // Next Line
	399	c == 0x2028 \|\| // Line Separator
	400	c == 0x2029) // Paragraph separator
	401	{
	402	break;
	403	}
	404	}
	405
	406	// Check for CR/LF sequence, and advance over the LF if we're in the middle of one.
	407	if (lineEnd < fileLen &&
	408	ucharBuf[lineEnd-1] == 0x0d &&
	409	ucharBuf[lineEnd] == 0x0a)
	410	{
	411	lineEnd++;
	412	}
	413	}
	414
	415
	416	//------------------------------------------------------------------------------------------
	417	//
	418	// printMatch Called when a matching line has been located.
	419	// Print out the line from the file with the match, after
	420	// converting it back to the default code page.
	421	//
	422	//------------------------------------------------------------------------------------------
	423	void printMatch() {
	424	char buf[2000];
	425	UErrorCode status = U_ZERO_ERROR;
	426
	427	// If we haven't already created a converter for output, do it now.
	428	if (outConverter == 0) {
	429	outConverter = ucnv_open(NULL, &status);
	430	if (U_FAILURE(status)) {
	431	fprintf(stderr, "ugrep: Error opening default converter: \"%s\"\n",
	432	u_errorName(status));
	433	exit(-1);
	434	}
	435	};
	436
	437	// Convert the line to be printed back to the default 8 bit code page.
	438	// If the line is too long for our buffer, just truncate it.
	439	ucnv_fromUChars(outConverter,
440	buf, // destination buffer for conversion
441	sizeof(buf), // capacity of destination buffer
442	&ucharBuf[lineStart], // Input to conversion
443	lineEnd-lineStart, // number of UChars to convert
444	&status);
445	buf[sizeof(buf)-1] = 0; // Add null for use in case of too long lines.
446	// The converter null-terminates its output unless
447	// the buffer completely fills.
448
449	if (displayFileName) {
450	printf("%s:", fileName);
451	}
452	if (displayLineNum) {
453	printf("%d:", lineNum);
454	}
455	printf("%s", buf);
456	}
457