debian
/
newlisp
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
newlisp/nl-string.c

2542 lines
58 KiB
C
Raw Permalink Blame History

/* nl-string.c
Copyright (C) 2016 Lutz Mueller
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include "newlisp.h"
#ifdef SUPPORT_UTF8
#include <wctype.h>
#endif
#define PCRE_STATIC
#include "pcre.h"
#include "protos.h"
extern SYMBOL * sysSymbol[];
extern CELL * countCell;
extern int bigEndian;
extern int pagesize;
#define OVECCOUNT (MAX_REGEX_EXP * 3) /* max sub expressions in PCRE */
void regexError(char * msg_1, int num, const char * msg_2);
pcre * pcreCachedCompile(char * pattern, int options);
CELL * cellTokenString(char * *source, size_t * srclen, char *separator, pcre *re);
#ifdef USE_WIN_UTF16PATH
int open_utf16(const char* a, int b, int c);
#endif
/* ---------------------------- string processing ----------------------- */
int my_strnicmp(char * s1, char * s2, ssize_t size)
{
while(toupper(*s1) == toupper(*s2))
{
if(--size <= 0) return(0);
s1++, s2++;
}
return(toupper(*s1) - toupper(*s2));
}
CELL * substring(char * string, ssize_t slen, ssize_t offset, ssize_t len)
{
if(offset < 0)
{
offset = slen + offset;
if(offset < 0)
return(errorProc(ERR_STRING_INDEX_INVALID));
}
else
offset = (offset > slen) ? slen : offset;
if(len < 0)
{
len = slen - offset + len;
if(len < 0) len = 0;
}
if(len == MAX_LONG)
len = slen - offset;
else
len = ((offset + len) > slen) ? (slen - offset) : len;
return(stuffStringN(string + offset, len));
}
#define SEARCH_SIZE 0x1000
CELL * p_search(CELL * params)
{
UINT fileHandle;
ssize_t bytesRead;
ssize_t position;
int flag = 0;
CELL * next;
#ifdef LFS
off_t filePosition;
off_t foundPosition;
double result;
#else
ssize_t filePosition;
ssize_t foundPosition;
#endif
char * searchString;
char * buffer;
INT options;
size_t len;
params = getInteger(params, (UINT *)&fileHandle);
params = getStringSize(params, &searchString, &len, TRUE);
if(len == 0) return(nilCell);
options = -1;
if(params != nilCell)
{
next = params->next;
params = evaluateExpression(params);
if(isNumber(params->type) || params->type == CELL_STRING)
/* 10.6.1 also accept string for options */
parseRegexOptions(params, (UINT *)&options, FALSE);
else
{
flag = (params->type != CELL_NIL);
if(next != nilCell)
/* 10.6.1 also accept string for options */
parseRegexOptions(next, (UINT *)&options, TRUE);
}
}
buffer = (char *)allocMemory(SEARCH_SIZE + 1);
if((filePosition = lseek((int)fileHandle, 0, SEEK_CUR)) == -1)
return(nilCell);
foundPosition = 0;
do
{
memset(buffer, 0, SEARCH_SIZE + 1);
bytesRead = read((int)fileHandle, buffer, SEARCH_SIZE);
if(options == -1)
position = searchBuffer(buffer, bytesRead, searchString, len, 1);
else
position = searchBufferRegex(buffer, 0, searchString, bytesRead , options, &len);
if(position != -1)
{
if(flag)
foundPosition = filePosition + position + len;
else
foundPosition = filePosition + position;
break;
}
filePosition = filePosition + bytesRead - len;
lseek((int)fileHandle, filePosition, SEEK_SET);
} while (bytesRead == SEARCH_SIZE);
freeMemory(buffer);
if(position == -1) return(nilCell);
lseek((int)fileHandle, foundPosition, SEEK_SET);
#ifdef LFS
result = foundPosition;
return(stuffFloat(result));
#else
return(stuffInteger(foundPosition));
#endif
}
/* returns a new string cell */
CELL * implicitIndexString(CELL * cell, CELL * params)
{
ssize_t index;
char * ptr;
#ifndef SUPPORT_UTF8
char str[2];
#else
ssize_t p;
#endif
ptr = (char*)cell->contents;
getInteger(params, (UINT *)&index);
stringCell = cell;
if(cell->aux == 1 && (index == 0 || index == -1))
return(copyCell(cell));
#ifndef SUPPORT_UTF8
index = adjustNegativeIndex(index, cell->aux - 1);
str[0] = *(ptr + index);
str[1] = 0;
stringIndexPtr = ptr + index;
return(stuffStringN(str, 1));
#else
index = adjustNegativeIndex(index, utf8_wlen(ptr, ptr + cell->aux));
ptr = utf8_index(ptr, index);
p = utf8_1st_len(ptr);
if(ptr + p > (char *)cell->contents + cell->aux)
return(errorProc(ERR_INVALID_UTF8));
stringIndexPtr = ptr;
return(stuffStringN(ptr, p));
#endif
}
/* only used from setf() for string insertion
uses the global stringIndexPtr in insertPtr
the insert point set by implicitIndexString()
modifies cellStr with new and returns new
*/
CELL * setNthStr(CELL * cellStr, CELL * new, char * insertPtr)
{
char * newStr;
char * oldStr;
size_t newLen, oldLen, len, offset;
char * str;
getStringSize(new, &newStr, &newLen, FALSE);
oldStr = (char*)cellStr->contents;
oldLen = cellStr->aux - 1;
if(oldLen == 0) return(copyCell(cellStr));
#ifdef SUPPORT_UTF8
len = utf8_1st_len(insertPtr);
#else
len = 1;
#endif
str = callocMemory(oldLen + newLen - len + 1);
offset = insertPtr - oldStr;
memcpy(str, oldStr, offset);
memcpy(str + offset, newStr, newLen);
memcpy(str + offset + newLen, oldStr + offset + len, oldLen - offset - len);
cellStr->contents = (UINT)str;
cellStr->aux = oldLen + newLen - len + 1;
freeMemory(oldStr);
return(new);
}
CELL * p_char(CELL * params)
{
char * string;
ssize_t offset;
size_t len;
int num;
CELL * datCell;
char buff[2];
if(params == nilCell)
return(errorProc(ERR_MISSING_ARGUMENT));
params = getEvalDefault(params, &datCell);
switch(datCell->type)
{
case CELL_STRING:
symbolCheck = NULL;
string = (char *)datCell->contents;
#ifndef SUPPORT_UTF8
len = (size_t)datCell->aux - 1;
if(len == 0) return(nilCell);
#else
len = utf8_wlen(string, string + datCell->aux);
if(len == 0)
{
if(datCell->aux - 1 == 0)
return(nilCell);
else
return(stuffInteger(0));
}
#endif
if(params != nilCell)
{
params = getInteger(params, (UINT*)&offset);
#ifdef SUPPORT_UTF8
if(getFlag(params)) /* treat as offset into 8-bit array */
{
offset = adjustNegativeIndex(offset, (size_t)datCell->aux - 1);
return(stuffInteger((UINT)*((unsigned char *)string + (UINT)offset)));
}
#endif
}
else offset = 0;
offset = adjustNegativeIndex(offset, (size_t)len);
#ifndef SUPPORT_UTF8
return(stuffInteger((UINT)*((unsigned char *)string + (UINT)offset)));
case CELL_LONG:
buff[0] = (int)datCell->contents;
break;
#ifndef NEWLISP64
case CELL_INT64:
num = *(INT64 *)&datCell->aux;
buff[0] = num;
break;
case CELL_FLOAT:
num = *(double*)&datCell->aux;
buff[0] = num;
#else /* NEWLISP64 */
case CELL_FLOAT:
num = *(double*)&datCell->contents;
buff[0] = num;
#endif /* NEWLISP64 */
break;
#else /* SUPPORT_UTF8 */
while(offset--) string += utf8_1st_len(string);
utf8_wchar(string, &num);
return(stuffInteger(num));
#ifndef NEWLISP64
case CELL_FLOAT:
case CELL_LONG:
case CELL_INT64:
string = allocMemory(UTF8_MAX_BYTES + 1);
if(datCell->type == CELL_FLOAT)
num = *(double*)&datCell->aux;
else if(datCell->type == CELL_INT64)
num = *(INT64 *)&datCell->aux;
else
num = datCell->contents;
#else /* NEWLISP64 */
case CELL_FLOAT:
case CELL_LONG:
string = allocMemory(UTF8_MAX_BYTES + 1);
if(datCell->type == CELL_FLOAT)
num = *(double*)&datCell->contents;
else
num = (int)datCell->contents;
#endif /* NEWLISP 64 */
len = wchar_utf8(num, string);
datCell = stuffStringN(string, len);
free(string);
return(datCell);
#endif /* SUPPORT_UTF8 */
default:
buff[0] = 0;
}
return(stuffStringN(buff, 1));
}
CELL * p_explode(CELL * params)
{
char * string;
CELL * cell;
CELL * result;
ssize_t size;
ssize_t len = 1;
int flag = 0;
#ifdef SUPPORT_UTF8
int clen;
char * ptr;
#endif
params = getEvalDefault(params, &cell);
if(isList(cell->type))
return(explodeList((CELL*)cell->contents, params));
if(cell->type == CELL_STRING)
{
string = (char *)cell->contents;
size = cell->aux - 1;
}
else
return(errorProcExt(ERR_LIST_OR_STRING_EXPECTED, cell));
if(params != nilCell)
{
params = getInteger(params, (UINT *)&len);
flag = getFlag(params);
if(!flag && len > size) len = size;
}
result = cell = getCell(CELL_EXPRESSION);
if(size == 0 || len <= 0) return(result);
#ifndef SUPPORT_UTF8
if(flag && size < len) return(result);
cell->contents = (UINT)stuffStringN(string, len);
cell = (CELL *)cell->contents;
string += len;
while((size -= len) > 0)
{
if(flag && size < len) break;
cell->next = stuffStringN(string, (size >= len) ? len : size);
cell = cell->next;
string += len;
}
#else
size = utf8_wlen(string, string + size + 1);
ptr = utf8_index(string, len);
clen = ptr - string;
if(flag && size < len) return(result);
cell->contents = (UINT)stuffStringN(string, clen);
cell = (CELL *)cell->contents;
string += clen;
while((size -= len) > 0)
{
if(flag && size < len) break;
ptr = utf8_index(string, len);
clen = ptr - string;
cell->next = stuffStringN(string, clen);
cell = cell->next;
string += clen;
}
#endif
return(result);
}
#define STR_LOWER 0
#define STR_UPPER 1
#define STR_TITLE 2
CELL * strUpperLower(CELL * params, int type)
{
#ifndef SUPPORT_UTF8
char * string;
char * ptr;
#else
char * utf8str;
int * unicode;
int * ptr;
size_t size;
#endif
CELL * cell;
int option = FALSE;
#ifndef SUPPORT_UTF8
params = getString(params, &string);
option = getFlag(params);
cell = stuffString(string);
ptr = (void *)cell->contents;
if(type == STR_UPPER)
while(*ptr) { *ptr = toupper(*ptr); ptr++; }
else
{
if(type == STR_TITLE)
if(*ptr) { *ptr = toupper(*ptr); ptr++; }
if(((type == STR_TITLE) && option) || (type == STR_LOWER))
while(*ptr) { *ptr = tolower(*ptr); ptr++; }
}
#else
params = getStringSize(params, &utf8str, &size, TRUE);
option = getFlag(params);
size = utf8_wlen(utf8str, utf8str + size + 1);
unicode = allocMemory((size + 1) * sizeof(int));
size = utf8_wstr(unicode, utf8str, size);
ptr = unicode;
/* Note that on many platforms towupper/towlower
do not work correctly for non-ascii unicodes */
if(type == STR_UPPER)
while(*ptr) { *ptr = towupper(*ptr); ptr++; }
else
{
if(type == STR_TITLE)
if(*ptr) { *ptr = towupper(*ptr); ptr++; }
if(((type == STR_TITLE) && option) || (type == STR_LOWER))
while(*ptr) { *ptr = towlower(*ptr); ptr++; }
}
utf8str = allocMemory(size * UTF8_MAX_BYTES + 1);
size = wstr_utf8(utf8str, unicode, size * UTF8_MAX_BYTES + 1);
utf8str = reallocMemory(utf8str, size + 1);
free(unicode);
return(makeStringCell(utf8str, size));
#endif
return(cell);
}
CELL * p_upper(CELL * params) {return(strUpperLower(params, STR_UPPER));}
CELL * p_lower(CELL * params) {return(strUpperLower(params, STR_LOWER));}
CELL * p_title(CELL * params) {return(strUpperLower(params, STR_TITLE));}
char * getFormatType(char * fmt, int * type)
{
PARSE_FORMAT:
/* get % */
while(*fmt != '%' && *fmt != 0) fmt++;
if(*fmt == 0)
{
*type = 0;
return(fmt);
}
fmt++;
/* is it double %% for '%' */
if(*fmt == '%')
{
fmt++;
goto PARSE_FORMAT;
}
/* get optional decimal format spec */
/* this GNU extension does not work on FreeBSD, OpenBSD */
/* works on OSX 10.9, Linux and Unix with GNU libc or alike */
if(*fmt == '\'') fmt++;
/* get width spec */
/* force + before numbers */
if(*fmt == '+')
fmt++;
else
{
/* left align numbers or strings */
if(*fmt == '-')
{
fmt++;
/* force + before numbers */
if(*fmt == '+') fmt++;
}
}
while(isdigit((int)*fmt) && *fmt != 0) fmt++;
if(*fmt == 0) return(NULL);
/* its a float or string with dot precision*/
if(*fmt == '.')
{
fmt++;
while(isdigit((int)*fmt) && *fmt !=0) fmt++;
if(*fmt == 0) return(NULL);
if(*fmt == 'f' || *fmt == 'g' || *fmt =='G' || *fmt == 'e' || *fmt == 'E')
{
*type = CELL_FLOAT;
return(++fmt);
}
if(*fmt == 's')
{
*type = CELL_STRING;
return(++fmt);
}
else return(NULL);
}
/* its a float without dot */
if(*fmt == 'f' || *fmt == 'g' || *fmt =='G' || *fmt == 'e' || *fmt == 'E')
{
*type = CELL_FLOAT;
return(++fmt);
}
/* its an integer or character */
#ifdef TRU64
if(*fmt == 'd' || *fmt == 'i' || *fmt == 'u' || *fmt == 'x' || *fmt == 'X' || *fmt == 'c' || *fmt == 'o')
#else
if(*fmt == 'd' || *fmt == 'u' || *fmt == 'x' || *fmt == 'X' || *fmt == 'c' || *fmt == 'o')
#endif
{
#ifdef TRU64
#ifndef NEWLISP64
*type = CELL_INT64;
#else
*type = CELL_LONG;
#endif
#else
*type = CELL_LONG;
#endif
return(++fmt);
}
/* its a string */
if(*fmt == 's')
{
*type = CELL_STRING;
return(++fmt);
}
/* supporting ld, li, lu, lx, lX formats */
if(*fmt == 'l' && (*(fmt + 1) == 'd' || *(fmt + 1) == 'i' || *(fmt + 1) == 'u' || *(fmt + 1) =='x' || *(fmt + 1) == 'X'))
{
#ifndef NEWLISP64
*type = CELL_INT64;
#else
*type = CELL_LONG;
#endif
return(fmt+2);
}
#ifndef TRUE64 /* UNIX except TRUE64 and WINDOWS MINGW suporting lld, llu, llx, llX formats */
if(*fmt == 'l' && *(fmt + 1) == 'l' && (*(fmt + 2) == 'd' || *(fmt + 2) == 'u' || *(fmt + 2) =='x' || *(fmt + 2) == 'X'))
{
#ifndef NEWLISP64
*type = CELL_INT64;
#else
*type = CELL_LONG;
#endif
return(fmt+3);
}
#endif
#ifdef WINDOWS /* on MINGW also support MS conventions */
if(memcmp(fmt, "I64", 3) == 0 &&
(*(fmt + 3) == 'd' || *(fmt + 3) == 'u' || *(fmt + 3) =='x' || *(fmt + 3) == 'X'))
{
#ifndef NEWLISP64
*type = CELL_INT64;
#else
*type = CELL_LONG;
#endif
return(fmt+4);
}
#endif
/* L and q seem not to be suported on most GCC although in the docs
if(*fmt == 'L' && (*(fmt + 1) == 'd' || *(fmt + 1) =='x' || *(fmt + 1) == 'X'))
{
*type = CELL_INT64;
return(fmt+3);
}
*/
/* its a wchar_t unicode string */
if(*fmt == 'l' && *(fmt + 1) == 's')
{
*type = CELL_STRING;
return(fmt+2);
}
return(NULL);
}
CELL * p_format(CELL * params)
{
char * format;
char * fmt;
char * nextfmt;
CELL * cell;
STREAM fmtStream = {NULL, NULL, 0, 14, 0};
int fType;
double floatNum;
UINT intNum;
#ifndef NEWLISP64
INT64 bigNum;
#endif
int evalFlag = TRUE;
char saveChar;
params = getString(params, &format);
fmt = format;
fmtStream.buffer = fmtStream.ptr = callocMemory(15);
while(params->type != CELL_NIL)
{
/* printf("entry>%s<\n", fmt); */
nextfmt = getFormatType(fmt, &fType);
/* printf("exit>%s<\n", nextfmt); */
if(nextfmt == NULL)
{
closeStrStream(&fmtStream);
return(errorProcExt2(ERR_FORMAT_STRING, stuffString(format)));
}
if(evalFlag)
cell = evaluateExpression(params);
else
cell = params;
if(cell->type == CELL_EXPRESSION && fmt == format)
{
params = (CELL *)cell->contents;
evalFlag = FALSE;
cell = params;
}
if(fType == 0)
{
closeStrStream(&fmtStream);
return(errorProcExt(ERR_NUM_ARGS, params));
}
saveChar = *nextfmt;
*nextfmt = 0;
if(fType == CELL_LONG)
{
if(isNumber(cell->type))
cell = getIntegerExt(cell, &intNum, FALSE);
else goto FORMAT_DATA_ERROR;
varPrintf((UINT)&fmtStream, fmt, intNum);
goto NEXT_FORMAT;
}
#ifndef NEWLISP64
if(fType == CELL_INT64)
{
if(isNumber(cell->type))
cell = getInteger64Ext(cell, &bigNum, TRUE);
else goto FORMAT_DATA_ERROR;
varPrintf((UINT)&fmtStream, fmt, bigNum);
goto NEXT_FORMAT;
}
#endif
if(fType == CELL_FLOAT)
{
if(cell->type == CELL_FLOAT)
#ifndef NEWLISP64
floatNum = *(double *)&cell->aux;
else if(cell->type == CELL_INT64)
floatNum = *(INT64 *)&cell->aux;
#else
floatNum = *(double *)&cell->contents;
#endif
else if(cell->type == CELL_LONG)
floatNum = (INT)cell->contents;
#ifdef BIGINT
else if(cell->type == CELL_BIGINT)
floatNum = bigintCellToFloat(cell);
#endif
else goto FORMAT_DATA_ERROR;
varPrintf((UINT)&fmtStream, fmt, floatNum);
goto NEXT_FORMAT;
}
if(fType != cell->type)
goto FORMAT_DATA_ERROR;
/* printf("stream>%s< with >%s<\n", (char *)cell->contents, fmt); */
varPrintf((UINT)&fmtStream, fmt, cell->contents);
NEXT_FORMAT:
*nextfmt = saveChar;
fmt = nextfmt;
params = params->next;
continue;
FORMAT_DATA_ERROR:
*nextfmt = saveChar;
closeStrStream(&fmtStream);
return(errorProcExt(ERR_FORMAT_DATA_TYPE, params));
}
getFormatType(fmt, &fType);
if(fType != 0)
{
closeStrStream(&fmtStream);
errorProcExt2(ERR_NUM_ARGS, stuffString(format));
}
varPrintf((UINT)&fmtStream, fmt);
cell = makeStringCell(fmtStream.buffer, fmtStream.position);
symbolCheck = NULL;
return(cell);
}
void openStrStream(STREAM * stream, size_t buffSize, int reopenFlag)
{
if(stream->buffer != NULL && reopenFlag)
{
if(buffSize == stream->size)
{
memset(stream->buffer, 0, buffSize);
stream->position = stream->handle = 0;
stream->ptr = stream->buffer;
return;
}
}
freeMemory(stream->buffer);
stream->buffer = stream->ptr = callocMemory(buffSize + 1);
stream->size = buffSize;
stream->position = stream->handle = 0;
}
void closeStrStream(STREAM * stream)
{
if(stream->buffer != NULL)
freeMemory(stream->buffer);
stream->buffer = stream->ptr = NULL;
stream->size = stream->position = 0;
if(stream->handle != 0)
{
close((int)stream->handle);
stream->handle = 0;
}
}
char * cellToString(CELL * cell, size_t * size, int quoteFlag)
{
static STREAM strStream = {NULL, NULL, 0, 0, 0};
if(cell->type == CELL_STRING)
{
*size = cell->aux -1;
return((char *)cell->contents);
}
openStrStream(&strStream, MAX_STRING, TRUE);
prettyPrintFlags |= PRETTYPRINT_STRING;
printCell(cell , quoteFlag, (UINT)&strStream);
prettyPrintFlags &= ~PRETTYPRINT_STRING;
*size = strStream.position;
return(strStream.buffer);
}
void writeStreamChar(STREAM * stream, char chr)
{
if(stream->position == stream->size)
{
stream->size += stream->size / 2;
stream->buffer = reallocMemory(stream->buffer, stream->size + 1);
memset(stream->buffer + stream->position, 0, stream->size - stream->position + 1);
stream->ptr = stream->buffer + stream->position;
}
*(stream->ptr++) = chr;
stream->position++;
}
void writeStreamStr(STREAM * stream, char * buff, size_t length)
{
size_t newPosition;
if(length == 0) length = strlen(buff);
newPosition = stream->position + length;
if(newPosition >= stream->size)
{
while(newPosition >= stream->size)
stream->size += stream->size / 2;
stream->buffer = reallocMemory(stream->buffer, stream->size + 1);
memset(stream->buffer + stream->position, 0, stream->size - stream->position + 1);
stream->ptr = stream->buffer + stream->position;
}
memcpy(stream->ptr, buff, length);
stream->ptr += length;
stream->position = newPosition;
}
/* creates a memory buffer and reads byte into it until
the [/text] is found
*/
#define LLEN 7
#define LIMIT "[/text]"
char * readStreamText(STREAM * stream, int * size)
{
STREAM outStream = {NULL, NULL, 0, 0, 0};
ssize_t findPos = -1;
size_t searchLen;
char * result;
openStrStream(&outStream, MAX_STRING, 0);
while(findPos == -1)
{
if((searchLen = strlen(stream->ptr)) < LLEN)
break;
findPos = searchBuffer(stream->ptr, searchLen, LIMIT, LLEN, TRUE);
if(findPos != -1)
{
if(findPos > 0)
writeStreamStr(&outStream, stream->ptr, findPos);
stream->ptr += findPos + LLEN;
result = allocMemory(outStream.position + 1);
memcpy(result, outStream.buffer, outStream.position);
*size = outStream.position;
*(result + outStream.position) = 0;
closeStrStream(&outStream);
return(result);
}
writeStreamStr(&outStream, stream->ptr, searchLen - LLEN);
stream->position += (stream->ptr - stream->buffer);
stream->position += searchLen - LLEN;
if(stream->handle == 0) /* its not a file */
{
stream->buffer = stream->ptr;
break;
}
else
{
if(searchLen == LLEN) break;
lseek(stream->handle, stream->position, SEEK_SET);
memset(stream->buffer, 0, stream->size + 1);
if(read(stream->handle, stream->buffer, stream->size) > 0)
stream->ptr = stream->buffer;
else
{
*stream->ptr = 0;
break;
}
}
}
closeStrStream(&outStream);
return(NULL);
}
/* this is only used for reading with getToken()
stream->size does not reflect the real size of
the buffer as in makeStreamFromFile()
*/
int makeStreamFromString(STREAM * stream, char * str)
{
int len = strlen(str);
stream->handle = stream->position = 0;
stream->buffer = stream->ptr = str;
/* make getToken work to the end of str */
stream->size = len + 4 * MAX_STRING;
return(len);
}
int makeStreamFromFile(STREAM * stream, char * fileName, size_t size, size_t offset)
{
int result;
if((stream->handle = openFile(fileName, "r", NULL)) == -1)
return(0);
stream->ptr = stream->buffer = (char *)callocMemory(size + 1);
if(offset != 0)
lseek((int)stream->handle, offset, SEEK_SET);
stream->position = offset;
stream->size = size;
/* load first buffer */
if((result = read(stream->handle, stream->buffer, size)) < 0)
{
closeStrStream(stream);
return(0);
}
return(TRUE);
}
ssize_t searchBuffer(char * buffer, size_t length, char * string, size_t size, int caseFlag)
{
size_t position = 0;
if(caseFlag == FALSE)
{
while(position < length)
{
if(toupper(*buffer) == toupper(*string))
if(my_strnicmp(buffer, string, size) == 0) break;
position++;
buffer++;
}
}
else /* case sensitive */
{
while(position < length)
{
if(*buffer == *string)
if(memcmp(buffer, string, size) == 0) break;
position++;
buffer++;
}
}
if(position == length) return(-1);
return(position);
}
/* eliminated in v.8.4.1
CELL * patternMatchS(char * pattern, char * string)
{
CELL * cell;
CELL * match;
CELL * star, * plus;
char * start;
int starLen, len;
start = string; starLen = len = 0;
MATCH:
switch(*pattern)
{
case '\\':
++pattern;
if(*pattern != *string) return(nilCell);
break;
case 0:
if(*string != 0) return(nilCell);
return(stuffStringN(start, len));
case '?':
if(*string == 0) return(nilCell);
break;
case '#':
if(*string == 0) return(nilCell);
if(!isDigit((unsigned char)*string)) return(nilCell);
break;
case '+':
if(*(pattern + 1) == 0)
{
if(*string == 0)
plus = stuffString("");
else if(*(string + 1) == 0)
plus = stuffStringN(string, 1);
else return(nilCell);
if(len == 0) return(plus);
cell = stuffStringN(start, len);
cell->next = plus;
return(cell);
}
if((match = patternMatchS(pattern+1, string)) != nilCell)
plus = stuffString("");
else if((match = patternMatchS(pattern+1, string+1)) != nilCell)
plus = stuffStringN(string, 1);
else return(nilCell);
if(len == 0) cell = plus;
else
{
cell = stuffStringN(start, len);
cell->next = plus;
}
plus->next = match;
return(cell);
case '*':
if(*(pattern + 1) == 0)
{
if(len == 0) return(stuffString(string));
cell = stuffStringN(start, len);
cell->next = stuffString(string);
return(cell);
}
if((match = patternMatchS(pattern+1, string)) != nilCell)
{
star = stuffStringN(start+len, starLen);
if(len == 0) cell = star;
else
{
cell = stuffStringN(start, len);
cell->next = star;
}
star->next = match;
return(cell);
}
if(*string != 0) ++string, ++starLen;
else return(nilCell);
goto MATCH;
default:
if(*pattern != *string) return(nilCell);
break;
}
++pattern;
++string;
++len;
goto MATCH;
}
*/
CELL * p_integer(CELL * params)
{
char * intString;
INT64 num;
INT base;
CELL * deflt, * cell;
INT64 result;
deflt = getEvalDefault(params, &cell);
symbolCheck = NULL;
if(cell->type == CELL_STRING)
intString = (char *)cell->contents;
else if(isNumber(cell->type))
{
#ifdef BIGINT
if(cell->type == CELL_BIGINT)
num = bigintToInt64(cell);
else
#endif
getInteger64Ext(cell, &num, FALSE);
return(stuffInteger64(num));
}
else
return(copyCell(evaluateExpression(deflt)));
while(isspace((int)*intString)) intString++;
if(deflt->next != nilCell)
getInteger(deflt->next, (UINT *)&base);
else base = 0;
if(base == 16)
{
if(!isxdigit((unsigned char)*intString))
{
if(*intString != '-' && *intString != '+')
goto INT_DEFAULT;
if(!isxdigit((unsigned char)*(intString+1)))
goto INT_DEFAULT;
}
}
else if(!isDigit((unsigned char)*intString))
{
if(*intString != '-' && *intString != '+')
goto INT_DEFAULT;
if(!isDigit((unsigned char)*(intString+1)))
goto INT_DEFAULT;
}
else if(*intString == '0' && (*(intString + 1) == 'b' || *(intString + 1) == 'B'))
{
intString += 2;
base = 2;
}
#ifdef TRU64
result = strtoul(intString, NULL, base);
#else
result = strtoull(intString,(char **)0, base);
#endif
return(stuffInteger64(result));
INT_DEFAULT:
return(copyCell(evaluateExpression(deflt)));
}
CELL * p_float(CELL * params)
{
char * fltString;
double value;
CELL * deflt;
CELL * cell;
deflt = getEvalDefault(params, &cell);
symbolCheck = NULL;
if(cell->type == CELL_STRING)
fltString = (char *)cell->contents;
else if(isNumber(cell->type))
{
#ifdef BIGINT
if(cell->type == CELL_BIGINT)
value = bigintCellToFloat(cell);
else
#endif
getFloat(cell, &value);
return(stuffFloat(value));
}
else
return(copyCell(evaluateExpression(deflt)));
while(isspace((int)*fltString)) fltString++;
if(!isDigit((unsigned char)*fltString))
{
if(*fltString != '-' && *fltString != '+' && *fltString != lc_decimal_point)
return(copyCell(evaluateExpression(deflt)));
if(!isDigit((unsigned char)*(fltString+1)))
return(copyCell(evaluateExpression(deflt)));
}
value = atof(fltString);
return( stuffFloat(value) );
}
CELL * p_symbol(CELL * params)
{
char * token;
char number[32];
SYMBOL * context;
SYMBOL * sPtr;
CELL * cell;
cell = evaluateExpression(params);
switch(cell->type)
{
case CELL_LONG:
snprintf(number, 30, "%"PRIdPTR, cell->contents);
token = number;
break;
#ifndef NEWLISP64
case CELL_INT64:
token = number;
snprintf(number, 30, "%"PRId64, *(INT64 *)&cell->aux);
break;
#endif /* NEWLISP64 */
case CELL_FLOAT:
#ifndef NEWLISP64
snprintf(number, 30, "%1.10g",*(double *)&cell->aux);
#else
snprintf(number, 30, "%1.10g",*(double *)&cell->contents);
#endif
token = number;
break;
case CELL_STRING:
token = (char*)cell->contents;
if((cell->aux - 1) > MAX_SYMBOL)
errorProcExt(ERR_STRING_TOO_LONG, cell);
break;
case CELL_SYMBOL:
sPtr = (SYMBOL*)cell->contents;
token = sPtr->name;
break;
default:
return(errorProcExt(ERR_NUMBER_OR_STRING_EXPECTED, params));
}
if((params = params->next) == nilCell)
context = currentContext;
else if((context = getCreateContext(params, TRUE)) == NULL)
return(errorProcExt(ERR_SYMBOL_OR_CONTEXT_EXPECTED, params));
if(params->next != nilCell)
{
cell = evaluateExpression(params->next);
if(isNil(cell))
{
sPtr = lookupSymbol(token, context);
if(sPtr == NULL) return(nilCell);
return(stuffSymbol(sPtr));
}
}
sPtr = translateCreateSymbol(token, CELL_NIL, context, TRUE);
return(stuffSymbol(sPtr));
}
CELL * p_symbolSource(CELL * params)
{
STREAM strStream = {NULL, NULL, 0, 0, 0};
CELL * cell;
openStrStream(&strStream, MAX_STRING, 0);
serializeSymbols(params, (UINT)&strStream);
cell = stuffString(strStream.buffer);
closeStrStream(&strStream);
return(cell);
}
CELL * p_string(CELL * params)
{
CELL * cell;
STREAM strStream = {NULL, NULL, 0, 14, 0};
strStream.buffer = strStream.ptr = callocMemory(15);
prettyPrintFlags |= PRETTYPRINT_STRING;
while (params != nilCell)
{
cell = evaluateExpression(params);
if(cell->type == CELL_STRING) /* speed optimization for strings */
writeStreamStr(&strStream, (char *)cell->contents, 0);
else
printCell(cell , FALSE, (UINT)&strStream);
params = params->next;
}
prettyPrintFlags &= ~PRETTYPRINT_STRING;
/* cell = stuffStringN(strStream.buffer, strStream.position); */
cell = makeStringCell(strStream.buffer, strStream.position);
return(cell);
}
UINT getAddress(CELL * params)
{
UINT num;
if(params == nilCell)
{
errorProc(ERR_MISSING_ARGUMENT);
return(0);
}
getEvalDefault(params, &params);
#ifndef NEWLISP64
if(params->type == CELL_INT64)
{
num = *(INT64 *)&params->aux;
return(num);
}
else if(params->type == CELL_FLOAT)
{
num = *(INT64 *)&params->aux;
return(num);
}
#else
if(params->type == CELL_FLOAT)
{
num = *(double *)&params->contents;
return(num);
}
#endif
return(params->contents);
}
CELL * p_getChar(CELL * params)
{
return(stuffInteger((UINT)*(char*)getAddress(params)));
}
CELL * p_getString(CELL * params)
{
UINT size, len;
UINT address;
UINT offset = 0;
int inc = 1;
char * limit;
char zeros[4] = {0,0,0,0};
address = getAddress(params);
if(address == 0)
errorProc(ERR_CANNOT_CONVERT_NULL);
/* (get-string <address>) ; get zero terminated ASCII or UTF-8 */
if(params->next == nilCell)
return(stuffString((char *)address));
params = getInteger(params->next, &size);
/* (get-string <address> <int-bytes>) ; get no bytes binary data */
if(params == nilCell)
return(stuffStringN((char *)address, size));
getStringSize(params, &limit, (size_t *)&len, TRUE);
/* (get-string <address> <int-bytes> <str-limit>) ; get until limit or bytes read */
if(len == 4 || len == 2) if(memcmp(limit, zeros, len) == 0) inc = len;
while(offset < size)
{
if(offset < (size - len) &&
memcmp((char *)(address + offset), limit, len) == 0) break;
offset += inc;
}
/* printf("size: %ld, len: %ld, inc: %d, offset: %ld\n", size, len, inc, offset); */
return(stuffStringN((char *)address, offset));
}
CELL * p_getInteger(CELL * params)
{
return(stuffInteger(*(int *)getAddress(params)));
}
CELL * p_getLong(CELL * params)
{
return(stuffInteger64(*(INT64 *)getAddress(params)));
}
CELL * p_getFloat(CELL * params)
{
return(stuffFloat(*(double*)getAddress(params)));
}
CELL * p_address(CELL * params)
{
if(params == nilCell)
return(errorProc(ERR_MISSING_ARGUMENT));
getEvalDefault(params, &params);
switch(params->type)
{
case CELL_LONG:
return(stuffInteger((UINT)&params->contents));
#ifndef NEWLISP64
case CELL_INT64:
case CELL_FLOAT:
return(stuffInteger((UINT)&params->aux));
#else
case CELL_FLOAT:
return(stuffInteger((UINT)&params->contents));
#endif
default:
break;
}
return(stuffInteger(params->contents));
}
CELL * p_copyMemory(CELL * params)
{
UINT toAddress, fromAddress, nBytes;
CELL * cell;
cell = evaluateExpression(params);
params = params->next;
fromAddress = getAddress(cell);
cell = evaluateExpression(params);
toAddress = getAddress(cell);
getInteger(params->next, &nBytes);
memcpy((char*)toAddress, (char*)fromAddress, nBytes);
return(stuffInteger(nBytes));
}
CELL * cellTokenString(char * * source, size_t * srclen, char * separator, pcre * re)
{
char * start;
char * src;
ssize_t len, slen;
int ovector[OVECCOUNT];
int rc;
if(*source == NULL) return(stuffString(""));
start = src = *source;
len = 0;
slen = strlen(separator);
if(re == NULL)
{
while(*src != 0)
{
if(*src == *separator)
{
if(strncmp(src, separator, slen) == 0)
{
*source = src+slen;
if(**source == 0) /* last token is separator */
*source = NULL;
return(stuffStringN(start, len));
}
}
src++;
len++;
}
*source = src;
if(len == 0) return(NULL);
return(stuffStringN(start, len));
}
else if(*src != 0)
{
len = *srclen;
rc = pcre_exec(re, NULL, src, len, 0, 0, ovector, OVECCOUNT);
/* matching failed */
if (rc == -1)
{
*source = src + len;
return(stuffStringN(src, len));
}
/* error in pcre_exec() */
if (rc < 0)
regexError("error", rc, "when executing");
*source = src + ovector[1];
*srclen -= ovector[1];
if(**source == 0) /* last token is separator */
*source = NULL;
if(ovector[1] - ovector[0] == 0)
return(NULL);
return(stuffStringN(src, ovector[0]));
}
return(NULL);
}
CELL * p_parse(CELL * params)
{
CELL * cell;
CELL * newCell;
CELL * result;
char * string;
char * separator;
char token[MAX_STRING];
STREAM stream = {NULL, NULL, 0, 0, 0};
int tklen;
size_t srclen;
/* PCRE stuff */
INT options = 0;
pcre *re = NULL;
params = getStringSize(params, &string, &srclen, TRUE);
if(params == nilCell) separator = NULL;
else
{
params = getString(params, &separator);
if(params != nilCell)
{
/* 10.6.1 also accept string for options */
parseRegexOptions(params, (UINT *)&options, TRUE);
/* Compile the regular expression in separator */
re = pcreCachedCompile(separator, options);
}
else re = NULL;
}
if(separator == NULL)
makeStreamFromString(&stream, string);
result = cell = getCell(CELL_EXPRESSION);
while(string != NULL)
{
if(separator != NULL)
{
if((newCell = cellTokenString(&string, &srclen, separator, re)) == NULL)
break;
}
else
{
if(getToken(&stream, token, &tklen) == TKN_EMPTY) break;
newCell = stuffString(token);
}
if(cell == result)
result->contents = (UINT)newCell;
else
cell->next = newCell;
cell = newCell;
if(string == NULL)
newCell->next = stuffString("");
}
return(result);
}
#define PACK_NONE 0
#define PACK_CHAR 1
#define PACK_BYTE 2
#define PACK_INT 3
#define PACK_UNSIGNED_INT 4
#define PACK_LONG 5
#define PACK_UNSIGNED_LONG 6
#define PACK_LONG_LONG 7
#define PACK_UNSIGNED_LONG_LONG 8
#define PACK_FLOAT 9
#define PACK_DOUBLE 10
#define PACK_STRING 11
#define PACK_NULL 12
#define PACK_BIG_ENDIAN 13
#define PACK_LITTLE_ENDIAN 14
void swapEndian(char * data, int n)
{
char tmp[8];
int i;
i = n;
while(i) { tmp[n - i] = data[i - 1]; i--; }
memcpy(data, &tmp[0], n);
}
CELL * p_pack(CELL * params)
{
char * format;
char * source;
char * packed;
char * pPtr;
CELL * cell;
ssize_t length, ln;
int len, type;
int listFlag = 0;
int endianSwitch = 0;
char chrV;
unsigned char byteV;
short int shortV;
unsigned short int uint16V; /* 16 bit */
unsigned int uint32V; /* 32 bit */
unsigned long long uint64V; /* 64 bit */
float floatV;
double doubleV;
#ifdef FFI
cell = evaluateExpression(params);
params = params->next;
if(cell->type == CELL_IMPORT_FFI)
return(packFFIstruct(cell, params)); /* nl-import.c */
if(cell->type != CELL_STRING)
return(errorProc(ERR_INVALID_PARAMETER));
format = (char *)cell->contents;
#else
params = getString(params, &format);
#endif
source = format;
length = 0;
while((source = parsePackFormat(source, &len, &type)) != NULL)
length += len;
if(length == 0) return(stuffString(""));
pPtr = packed = allocMemory(length);
source = format;
length = 0;
while((source = parsePackFormat(source, &len, &type)) != NULL)
{
if(type == PACK_NULL)
{
memset(pPtr, 0, len);
pPtr += len;
length += len;
continue;
}
else if(type == PACK_LITTLE_ENDIAN || type == PACK_BIG_ENDIAN)
{
endianSwitch = ((type == PACK_BIG_ENDIAN) != bigEndian);
continue;
}
if(params->type == CELL_NIL) break;
if(listFlag)
cell = params;
else
cell = evaluateExpression(params);
/* accept data in a list */
if(isList(cell->type))
{
cell = (CELL *)cell->contents;
params = cell;
listFlag = 1;
}
#ifndef NEWLISP64
if(cell->type == CELL_FLOAT || cell->type == CELL_INT64)
uint64V = *(INT64 *)&cell->aux;
else /* CELL_LONG and CELL_STRING */
uint64V = cell->contents;
#else
uint64V = cell->contents;
#endif
if(type < PACK_FLOAT && cell->type == CELL_FLOAT)
uint64V = *(double *)&cell->aux;
switch(type)
{
case PACK_NONE:
break;
case PACK_BYTE:
byteV = (char)uint64V;
memcpy(pPtr, &byteV, 1);
break;
case PACK_CHAR:
chrV = (char)uint64V;
memcpy(pPtr, &chrV, 1);
break;
case PACK_INT:
shortV = (short int)uint64V;
memcpy(pPtr, &shortV, 2);
if(endianSwitch) swapEndian(pPtr, 2);
break;
case PACK_UNSIGNED_INT:
uint16V = (unsigned short int)uint64V;
memcpy(pPtr, &uint16V, 2);
if(endianSwitch) swapEndian(pPtr, 2);
break;
case PACK_LONG:
case PACK_UNSIGNED_LONG:
uint32V = (unsigned int)uint64V;
memcpy(pPtr, &uint32V, 4);
if(endianSwitch) swapEndian(pPtr, 4);
break;
case PACK_LONG_LONG:
case PACK_UNSIGNED_LONG_LONG:
memcpy(pPtr, &uint64V, 8);
if(endianSwitch) swapEndian(pPtr, 8);
break;
case PACK_FLOAT:
case PACK_DOUBLE:
if(cell->type == CELL_FLOAT)
doubleV = *(double *)&uint64V;
else doubleV = (double)uint64V;
if(type == PACK_FLOAT)
{
floatV = doubleV;
memcpy(pPtr, &floatV, 4);
if(endianSwitch) swapEndian(pPtr, 4);
}
else
{
memcpy(pPtr, &doubleV, 8);
if(endianSwitch) swapEndian(pPtr, 8);
}
break;
case PACK_STRING:
if(cell->type == CELL_STRING)
{
ln = cell->aux - 1;
if(len <= ln)
memcpy(pPtr, (void *)cell->contents, len);
else
{
memcpy(pPtr, (void*)cell->contents, ln);
memset(pPtr + ln, 0, len - ln);
}
}
else memset(pPtr, 0, len);
break;
default:
break;
}
pPtr += len;
params = params->next;
length += len;
}
cell = stuffStringN(packed, length);
free(packed);
return(cell);
}
char * parsePackFormat(char * format, int * length, int * type)
{
*length = 0;
while(*format == ' ') format++;
if(*format == 0) return(NULL);
switch(*format)
{
case '<':
*type = PACK_LITTLE_ENDIAN;
format++;
break;
case '>':
*type = PACK_BIG_ENDIAN;
format++;
break;
case 'b':
*length = 1;
*type = PACK_BYTE;
format++;
break;
case 'c':
*length = 1;
*type = PACK_CHAR;
format++;
break;
case 's':
case 'n':
*type = (*format == 's') ? PACK_STRING : PACK_NULL;
format++;
if(isdigit((int)*format) )
{
*length = atol(format);
while(isdigit((int)*format)) format++;
}
else *length = 1;
break;
case 'd':
case 'u':
*type = (*format == 'd') ? PACK_INT : PACK_UNSIGNED_INT;
*length = 2;
format++;
break;
case 'l':
if(*(format + 1) == 'd' || *(format + 1) == 'u')
{
*length = 4;
*type = (*(format + 1) == 'd') ? PACK_LONG : PACK_UNSIGNED_LONG;
format += 2;
}
else if(*(format + 1) == 'f')
{
*length = 8;
*type = PACK_DOUBLE;
format += 2;
}
else
{
*type = PACK_NONE;
format++;
}
break;
case 'L':
if(*(format + 1) == 'd' || *(format + 1) == 'u')
{
*length = 8;
*type = (*(format + 1) == 'd') ? PACK_LONG_LONG : PACK_UNSIGNED_LONG_LONG;
format += 2;
}
else
{
*type = PACK_NONE;
format++;
}
break;
case 'f':
*length = 4;
*type = PACK_FLOAT;
format++;
break;
default:
*type = PACK_NONE;
format++;
break;
}
return(format);
}
CELL * p_unpack(CELL * params)
{
CELL * cell;
char * format;
char * source;
char * pPtr;
CELL * result;
CELL * next = NULL;
size_t length, maxlen;
int len;
int type;
int endianSwitch = 0;
char chrV;
unsigned char byteV;
short shortV;
unsigned short uint16V;
int int32V;
unsigned int uint32V;
unsigned long long int uint64V;
float floatV;
double doubleV;
cell = evaluateExpression(params->next);
if(cell->type == CELL_STRING)
{
pPtr = (char *)cell->contents;
maxlen = cell->aux - 1;
}
else
{
getIntegerExt(cell, (void*)&pPtr, FALSE);
maxlen = MAX_LONG;
}
#ifdef FFI
cell = evaluateExpression(params);
if(cell->type == CELL_IMPORT_FFI)
return(unpackFFIstruct(cell, pPtr)); /* nl-import.c */
if(cell->type != CELL_STRING)
return(errorProc(ERR_INVALID_PARAMETER));
format = (char *)cell->contents;
#else
params = getString(params, &format);
#endif
length = 0;
source = format;
result = getCell(CELL_EXPRESSION);
next = NULL;
while( (source = parsePackFormat(source, &len, &type))!= NULL)
{
if(length + len > maxlen) break;
if(type == PACK_LITTLE_ENDIAN || type == PACK_BIG_ENDIAN)
{
endianSwitch = ((type == PACK_BIG_ENDIAN) != bigEndian);
continue;
}
switch(type)
{
case PACK_NULL:
pPtr += len;
length += len;
continue;
case PACK_BYTE:
memcpy(&byteV, pPtr, 1);
cell = makeCell(CELL_LONG, byteV);
break;
case PACK_CHAR:
memcpy(&chrV, pPtr, 1);
cell = makeCell(CELL_LONG, chrV);
break;
case PACK_INT:
memcpy(&shortV, pPtr, 2);
if(endianSwitch) swapEndian((char*)&shortV, 2);
cell = makeCell(CELL_LONG, shortV);
break;
case PACK_UNSIGNED_INT:
memcpy(&uint16V, pPtr, 2);
if(endianSwitch) swapEndian((char*)&uint16V, 2);
cell = makeCell(CELL_LONG, uint16V);
break;
case PACK_LONG:
memcpy(&int32V, pPtr, 4);
if(endianSwitch) swapEndian((char *)&int32V, 4);
#ifndef NEWLISP64
cell = getCell(CELL_INT64);
*(INT64 *)&cell->aux = int32V;
#else
cell = getCell(CELL_LONG);
*(INT *)&cell->contents = int32V;
#endif
break;
case PACK_UNSIGNED_LONG:
memcpy(&uint32V, pPtr, 4);
if(endianSwitch) swapEndian((char*)&uint32V, 4);
#ifndef NEWLISP64
cell = getCell(CELL_INT64);
*(INT64 *)&cell->aux = uint32V;
#else
cell = makeCell(CELL_LONG, uint32V);
#endif
break;
case PACK_LONG_LONG:
case PACK_UNSIGNED_LONG_LONG:
memcpy(&uint64V, pPtr, 8);
if(endianSwitch) swapEndian((char*)&uint64V, 8);
#ifndef NEWLISP64
cell = getCell(CELL_INT64);
memcpy(&cell->aux, &uint64V, 8);
#else
cell = makeCell(CELL_LONG, uint64V);
#endif
break;
case PACK_FLOAT:
memcpy(&floatV, pPtr, 4);
if(endianSwitch) swapEndian((char*)&floatV, 4);
doubleV = floatV;
cell = stuffFloat(doubleV);
break;
case PACK_DOUBLE:
memcpy(&doubleV, pPtr, 8);
if(endianSwitch) swapEndian((char*)&doubleV, 8);
cell = stuffFloat(doubleV);
break;
case PACK_STRING:
if(pPtr == NULL)
errorProc(ERR_CANNOT_CONVERT_NULL);
cell = stuffStringN(pPtr, len);
break;
default:
cell = getCell(CELL_NIL);
break;
}
pPtr += len; length += len;
if(next == NULL)
result->contents = (UINT)cell;
else
next->next = cell;
next = cell;
}
return(result);
}
CELL * p_trim(CELL * params)
{
char * str;
char * ptr;
size_t left = 0, right = 0, len;
char * trimChr;
int lchr, rchr;
#ifdef SUPPORT_UTF8
int * wstr;
int * wptr;
CELL * result;
#endif
params = getStringSize(params, &str, &len, TRUE);
if(params == nilCell)
{
ptr = str;
while((unsigned char)*str <= 32 && *str != 0) str++, left++;
if(left == len)
return(stuffString(""));
str = ptr + len -1;
while((unsigned char)*str <= 32) str--, right++;
return(stuffStringN(ptr + left, len - left - right));
}
#ifndef SUPPORT_UTF8
ptr = str;
#else
len = utf8_wlen(str, str + len + 1);
wptr = wstr = allocMemory((len + 1) * sizeof(int));
len = utf8_wstr(wstr, str, len);
#endif
if(len == 0)
return(stuffString(str));
params = getString(params, &trimChr);
#ifndef SUPPORT_UTF8
lchr = *trimChr;
#else
utf8_wchar(trimChr, &lchr);
#endif
if(params != nilCell)
{
getString(params, &trimChr);
#ifndef SUPPORT_UTF8
rchr = *trimChr;
#else
utf8_wchar(trimChr, &rchr);
#endif
}
else rchr = lchr;
#ifndef SUPPORT_UTF8
while(*str == lchr) str++, left++;
#else
while(*wstr == lchr) wstr++, left++;
#endif
if(left == len)
return(stuffString(""));
#ifndef SUPPORT_UTF8
str = ptr + len - 1;
while(*str == rchr && right < len) str--, ++right;
return(stuffStringN(ptr + left, len - left - right));
#else
wstr = wptr + len - 1;
while(*wstr == rchr && right < len) wstr--, ++right;
*(wptr + len - right) = 0;
str = allocMemory((len - left - right + 1) * UTF8_MAX_BYTES);
len = wstr_utf8(str, wptr + left, (len - left - right) * UTF8_MAX_BYTES);
free(wptr);
result = stuffStringN(str, len);
free(str);
return(result);
#endif
}
/* ------- PRCE Perl Compatible Regular Expressions for all platforms ------- */
/* see also http://www.pcre.org/ */
/* currently PCRE v.5.0 from 13-Sep-2004 -> pcre.h */
/* additional regex option */
#define REPLACE_ONCE 0x8000
#define PRECOMPILED 0x10000
void regexError(char * msg_1, int num, const char * msg_2);
/* PCRE options from manual not allowed from newLISP regex options in pcre_compile() */
#define PCRE_NOT_COMPILE (PCRE_NOTBOL | PCRE_NOTEOL | PCRE_NOTEMPTY)
/* PCRE options from manual allowed for pcre_exec() */
#define PCRE_EXEC_OPTIONS (PCRE_ANCHORED | PCRE_NOTBOL | PCRE_NOTEOL | PCRE_NOTEMPTY)
CELL * p_regex(CELL * params)
{
pcre *re;
int ovector[OVECCOUNT];
int rc, idx;
char * pattern;
char * subject;
#ifdef SUPPORT_UTF8
char * string;
#endif
INT options = 0;
UINT offset = 0;
int len;
size_t size;
CELL * cell, * result, * strCell;
params = getString(params, &pattern);
params = getStringSize(params, &subject, &size, TRUE);
if(params != nilCell)
{
/* 10.6.1 also accept string for options */
params = parseRegexOptions(params, (UINT *)&options, TRUE);
if(params != nilCell)
getInteger(params, &offset);
}
/* Compile the regular expression in the first argument */
re = pcreCachedCompile(pattern, (int)options);
/* Compilation succeeded: match the subject in the second argument */
rc = pcre_exec(
re, /* the compiled pattern */
NULL, /* no extra data - we didn't study the pattern */
subject, /* the subject string */
size, /* the length of the subject */
(int)offset, /* start at offset 0 in the subject */
(int)options & PCRE_EXEC_OPTIONS,
ovector, /* output vector for substring information */
OVECCOUNT); /* number of elements in the output vector */
/* Matching failed */
if (rc == -1)
return(nilCell);
/* error in pcre_exec() */
if (rc < 0)
regexError("error", rc, "when executing");
/* Match succeded */
/* Show substrings stored in the output vector */
result = cell = getCell(CELL_EXPRESSION);
for(idx = 0; idx < rc; idx++)
{
len = ovector[2*idx+1] - ovector[2*idx];
strCell = stuffStringN(subject + ovector[2*idx], len);
deleteList((CELL*)sysSymbol[idx]->contents);
sysSymbol[idx]->contents = (UINT)copyCell(strCell);
if(idx == 0)
{
cell->contents = (UINT)strCell;
cell = (CELL *)cell->contents;
}
else
{
cell->next = strCell;
cell = cell->next;
}
#ifdef SUPPORT_UTF8
if(2048 & options)
cell->next = stuffInteger(utf8_wlen(subject, subject + ovector[2*idx] + 1));
else
#endif
cell->next = stuffInteger(ovector[2*idx]);
cell = cell->next;
#ifdef SUPPORT_UTF8
if(2048 & options)
{
string = (char *)strCell->contents;
cell->next = stuffInteger(utf8_wlen(string, string + len + 1));
}
else
#endif
cell->next = stuffInteger(len);
cell = cell->next;
}
return(result);
}
CELL * p_regexComp(CELL * params)
{
char * pattern;
INT options = 0;
pcre * re;
int rc;
size_t size;
CELL * result;
params = getString(params, &pattern);
if(params != nilCell)
/* 10.6.1 also accept string for options */
params = parseRegexOptions(params, (UINT *)&options, TRUE);
/* Compile the regular expression in the first argument */
re = pcreCachedCompile(pattern, (int)options);
rc = pcre_fullinfo(re, NULL, PCRE_INFO_SIZE, &size);
if(rc < 0) return(nilCell);
result = stuffStringN((char *)re, size);
return(result);
}
ssize_t searchBufferRegex(char * string, size_t offset,
char * pattern, size_t length, int options, size_t * len)
{
pcre *re;
int ovector[OVECCOUNT];
int rc, idx;
CELL * cell;
options &= ~REPLACE_ONCE ; /* turn custom bits off for PCRE */
/* Compile the regular expression in the first argument */
re = pcreCachedCompile(pattern, options);
/* Compilation succeeded: match the subject in the second argument */
rc = pcre_exec(re, NULL, string, length, offset, options & PCRE_EXEC_OPTIONS, ovector, OVECCOUNT);
/* matching failed */
if (rc == -1)
return(-1);
/* error in pcre_exec() */
if (rc < 0)
regexError("error", rc, "when executing");
for(idx = 0; idx < rc; idx++)
{
cell = stuffStringN(string + ovector[2*idx], ovector[2*idx+1] - ovector[2*idx]);
deleteList((CELL*)sysSymbol[idx]->contents);
sysSymbol[idx]->contents = (UINT)cell;
}
if(len != NULL)
*len = ovector[1] - ovector[0];
return((UINT)ovector[0]);
}
pcre * pcreCachedCompile(char * pattern, int options)
{
const char * error;
int errOffset;
static char * cPattern = NULL;
static pcre * re = NULL;
static int cacheOptions = -1;
UINT len;
if(options & 0x10000) return((pcre *)pattern);
if((cPattern == NULL) || (strcmp(cPattern, pattern) != 0) || (options != cacheOptions))
{
cacheOptions = options;
if(cPattern != NULL) freeMemory(cPattern);
len = strlen(pattern);
cPattern = (char *)allocMemory(len + 1);
memcpy(cPattern, pattern, len + 1);
if(re != NULL) (pcre_free)(re);
options = options & ~PCRE_NOT_COMPILE;
re = pcre_compile(pattern, options, &error, &errOffset, NULL);
/* Compilation failed: print the error message and exit */
if (re == NULL)
{
freeMemory(cPattern);
cPattern = NULL;
regexError("offset", errOffset, error);
}
}
return(re);
}
void regexError(char * msg_1, int number, const char * msg_2)
{
CELL * cell;
char * errorBuff = malloc(256);
snprintf(errorBuff, 256, "%s %d %s", msg_1, number, msg_2);
cell = stuffString(errorBuff);
free(errorBuff);
fatalError(ERR_REGEX, cell, 0);
}
/* replace string with or without (options = -1) regular expressions */
typedef struct {
int offset;
int length;
char * repStr;
int repLen;
void * next;
} REGEX;
void freeRegex(REGEX * regex);
char * replaceString
(char * keyStr, size_t keyLen, char * buff, size_t buffLen, CELL * exprCell,
UINT * cnt, int options, size_t * newLen)
{
ssize_t oldLen, findPos;
size_t count, offset;
size_t repLen;
char * newBuff;
CELL * cell;
REGEX * start_rx = NULL, * end_rx = NULL;
UINT * resultStackIdxSave;
int errNo;
int bias = 0;
*newLen = oldLen = 0;
*cnt = 0;
offset = count = 0;
/* save all found string and fill sys variables $0, $1 ... etc */
while(offset <= buffLen)
{
if(options == -1)
findPos = searchBuffer(buff + offset, buffLen - offset, keyStr, keyLen, TRUE);
else
{
findPos = searchBufferRegex(buff, offset, keyStr, buffLen, options, &keyLen);
itSymbol->contents = sysSymbol[0]->contents;
}
if(findPos == -1) break;
if(options != -1) findPos -= offset;
if(count == 0)
start_rx = end_rx = (REGEX *)callocMemory(sizeof(REGEX));
else
{
end_rx->next = callocMemory(sizeof(REGEX));
end_rx = end_rx->next;
}
end_rx->next = NULL;
end_rx->offset = findPos + offset; /* pos where pattern found */
resultStackIdxSave = resultStackIdx;
countCell->contents++;
if((cell = evaluateExpressionSafe(exprCell, &errNo)) == NULL)
{
freeRegex(start_rx);
longjmp(errorJump, errNo);
}
end_rx->length = keyLen; /* length of pattern found */
if(cell->type == CELL_STRING)
{
repLen = end_rx->repLen = cell->aux-1;
end_rx->repStr = (char *)allocMemory(repLen + 1); /* replacement string */
memcpy(end_rx->repStr, (char *)cell->contents, cell->aux);
}
else /* if replacement expression is not string leave old content */
{
repLen = end_rx->repLen = keyLen;
end_rx->repStr = (char *)callocMemory(repLen + 1);
memcpy(end_rx->repStr, buff + offset + findPos, keyLen);
}
cleanupResults(resultStackIdxSave);
count++;
offset += findPos + keyLen; /* next start for search */
oldLen += keyLen; /* space occupied by old content */
*newLen += repLen; /* space needed by new replacement content */
bias = (keyLen == 0);
offset += bias;
if(options != -1)
if(options & REPLACE_ONCE) break;
}
if(count == 0) return(NULL);
*newLen = buffLen - oldLen + *newLen;
newBuff = callocMemory(*newLen + 1);
end_rx = start_rx;
/* count is now offset into the new buffer */
*cnt = count;
count = start_rx->offset;
memcpy(newBuff, buff, count);
while(start_rx != NULL) /* replace */
{
repLen = start_rx->repLen;
memcpy(newBuff + count, start_rx->repStr, repLen);
count += repLen;
freeMemory(start_rx->repStr);
end_rx = start_rx->next;
if(end_rx != NULL) /* copy from buffer */
{
repLen = end_rx->offset - start_rx->offset - start_rx->length;
memcpy(newBuff + count, buff + start_rx->offset + start_rx->length, repLen);
count += repLen;
}
freeMemory(start_rx);
start_rx = end_rx;
}
/*
printf("count %d buffLen %d offset %d bias %d str %s lencpy %d\n",
count, buffLen, offset, bias, buff + offset - bias, buffLen - offset + bias);
*/
memcpy(newBuff + count, buff + offset - bias, buffLen - offset + bias);
itSymbol->contents = (UINT)nilCell;
return(newBuff);
}
void freeRegex(REGEX * regex)
{
REGEX * oldRegex;
while(regex != NULL)
{
if(regex->repStr != NULL)
freeMemory(regex->repStr);
oldRegex = regex;
regex = regex->next;
free(oldRegex);
}
}
/* support strings in the integer-options field of:
directory, ends-with, find, find-all, parse, regex, replace, search, starts-with
instead of: (find "xyz" buff (| 1 4 2048))
example: (find "x.z" "abcX\nZ" "is") => 3
for PCRE_CASELESS, PCRE_DOTALL
*/
CELL* parseRegexOptions(CELL* params, UINT * options, int evalFlag)
{
CELL* cell;
int i;
char* buffer;
if (evalFlag)
cell = evaluateExpression(params);
else
cell = params;
*options = 0;
if (cell->type == CELL_STRING)
{
buffer = (char*)cell->contents;
for (i = 0; i < cell->aux - 1; i ++) {
switch (buffer[i]) {
case 'i': *options |= 1; break; /* PCRE_CASELESS */
case 'm': *options |= 2; break; /* PCRE_MULTILINE */
case 's': *options |= 4; break; /* PCRE_DOTALL */
case 'x': *options |= 8; break; /* PCRE_EXTENDED */
case 'A': *options |= 16; break; /* PCRE_ANCHORED */
case 'D': *options |= 32; break; /* PCRE_DOLLAR_ENDONLY */
case 'U': *options |= 512; break; /* PCRE_UNGREEDY */
case 'u': *options |= 2048; break; /* PCRE_UTF8 */
default: return(errorProcExt(ERR_INVALID_OPTION, cell));
}
}
} else { getIntegerExt(cell, options, FALSE); }
return (params->next);
}
/* eof */
Reference in New Issue View Git Blame Copy Permalink