jsfile.c

来自「一个基于alice开发的机器人」· C语言 代码 · 共 2,148 行 · 第 1/5 页

{
    char *tmp;
    char *path = oldpath;
    char *base, *dir, *current, *result;
    jsint c;
    jsint back = 0;
    unsigned int i = 0, j = strlen(path)-1;

    /* This is probably optional */
	/* Remove possible spaces in the beginning and end */
    while(i<strlen(path)-1 && path[i]==' ') i++;
	while(j>=0 && path[j]==' ') j--;

	tmp = JS_malloc(cx, j-i+2);
	strncpy(tmp, &path[i], j-i+1);
    tmp[j-i+1] = '\0';

    path = tmp;

    /* pipe support */
    if(js_filenameHasAPipe(path))
        return JS_strdup(cx, path);
    /* file:// support */
    if(!strncmp(path, URL_PREFIX, strlen(URL_PREFIX)))
        return js_canonicalPath(cx, &path[strlen(URL_PREFIX)-1]);

    if (!js_isAbsolute(path))
        path = js_absolutePath(cx, path);
    else
        path = JS_strdup(cx, path);

    result = JS_strdup(cx, "");

    current = path;

    base = js_fileBaseName(cx, current);
    dir = js_fileDirectoryName(cx, current);

    /* TODO: MAC -- not going to work??? */
    while (strcmp(dir, current)) {
        if (!strcmp(base, "..")) {
            back++;
        } else
        if(!strcmp(base, ".")){
            /* ??? */
        } else {
            if (back>0)
                back--;
            else {
                tmp = result;
                result = JS_malloc(cx, strlen(base)+1+strlen(tmp)+1);
                if (!result) {
                    JS_free(cx, dir);
                    JS_free(cx, base);
                    JS_free(cx, current);
                    return NULL;
                }
                strcpy(result, base);
                c = strlen(result);
                if (*tmp) {
                    result[c] = FILESEPARATOR;
                    result[c+1] = '\0';
                    strcat(result, tmp);
                }
                JS_free(cx, tmp);
            }
        }
        JS_free(cx, current);
        JS_free(cx, base);
        current = dir;
        base =  js_fileBaseName(cx, current);
        dir = js_fileDirectoryName(cx, current);
    }

    tmp = result;
    result = JS_malloc(cx, strlen(dir)+1+strlen(tmp)+1);
    if (!result) {
        JS_free(cx, dir);
        JS_free(cx, base);
        JS_free(cx, current);
        return NULL;
    }
    strcpy(result, dir);
    c = strlen(result);
    if (tmp[0]!='\0') {
        if ((result[c-1]!=FILESEPARATOR)&&(result[c-1]!=FILESEPARATOR2)) {
            result[c] = FILESEPARATOR;
            result[c+1] = '\0';
        }
        strcat(result, tmp);
    }
    JS_free(cx, tmp);
    JS_free(cx, dir);
    JS_free(cx, base);
    JS_free(cx, current);

    return result;
}

/* -------------------------- Text conversion ------------------------------- */
/* The following is ripped from libi18n/unicvt.c and include files.. */

/*
 * UTF8 defines and macros
 */
#define ONE_OCTET_BASE          0x00    /* 0xxxxxxx */
#define ONE_OCTET_MASK          0x7F    /* x1111111 */
#define CONTINUING_OCTET_BASE   0x80    /* 10xxxxxx */
#define CONTINUING_OCTET_MASK   0x3F    /* 00111111 */
#define TWO_OCTET_BASE          0xC0    /* 110xxxxx */
#define TWO_OCTET_MASK          0x1F    /* 00011111 */
#define THREE_OCTET_BASE        0xE0    /* 1110xxxx */
#define THREE_OCTET_MASK        0x0F    /* 00001111 */
#define FOUR_OCTET_BASE         0xF0    /* 11110xxx */
#define FOUR_OCTET_MASK         0x07    /* 00000111 */
#define FIVE_OCTET_BASE         0xF8    /* 111110xx */
#define FIVE_OCTET_MASK         0x03    /* 00000011 */
#define SIX_OCTET_BASE          0xFC    /* 1111110x */
#define SIX_OCTET_MASK          0x01    /* 00000001 */

#define IS_UTF8_1ST_OF_1(x) (( (x)&~ONE_OCTET_MASK  ) == ONE_OCTET_BASE)
#define IS_UTF8_1ST_OF_2(x) (( (x)&~TWO_OCTET_MASK  ) == TWO_OCTET_BASE)
#define IS_UTF8_1ST_OF_3(x) (( (x)&~THREE_OCTET_MASK) == THREE_OCTET_BASE)
#define IS_UTF8_1ST_OF_4(x) (( (x)&~FOUR_OCTET_MASK ) == FOUR_OCTET_BASE)
#define IS_UTF8_1ST_OF_5(x) (( (x)&~FIVE_OCTET_MASK ) == FIVE_OCTET_BASE)
#define IS_UTF8_1ST_OF_6(x) (( (x)&~SIX_OCTET_MASK  ) == SIX_OCTET_BASE)
#define IS_UTF8_2ND_THRU_6TH(x) \
                    (( (x)&~CONTINUING_OCTET_MASK  ) == CONTINUING_OCTET_BASE)
#define IS_UTF8_1ST_OF_UCS2(x) \
            IS_UTF8_1ST_OF_1(x) \
            || IS_UTF8_1ST_OF_2(x) \
            || IS_UTF8_1ST_OF_3(x)


#define MAX_UCS2            0xFFFF
#define DEFAULT_CHAR        0x003F  /* Default char is "?" */
#define BYTE_MASK           0xBF
#define BYTE_MARK           0x80


/* Function: one_ucs2_to_utf8_char
 *
 * Function takes one UCS-2 char and writes it to a UTF-8 buffer.
 * We need a UTF-8 buffer because we don't know before this
 * function how many bytes of utf-8 data will be written. It also
 * takes a pointer to the end of the UTF-8 buffer so that we don't
 * overwrite data. This function returns the number of UTF-8 bytes
 * of data written, or -1 if the buffer would have been overrun.
 */

#define LINE_SEPARATOR      0x2028
#define PARAGRAPH_SEPARATOR 0x2029
static int16 one_ucs2_to_utf8_char(unsigned char *tobufp,
        unsigned char *tobufendp, uint16 onechar)
{

     int16 numUTF8bytes = 0;

    if((onechar == LINE_SEPARATOR)||(onechar == PARAGRAPH_SEPARATOR))
    {
        strcpy((char*)tobufp, "\n");
        return strlen((char*)tobufp);;
    }

        if (onechar < 0x80) {               numUTF8bytes = 1;
        } else if (onechar < 0x800) {       numUTF8bytes = 2;
        } else if (onechar <= MAX_UCS2) {   numUTF8bytes = 3;
        } else { numUTF8bytes = 2;
                 onechar = DEFAULT_CHAR;
        }

        tobufp += numUTF8bytes;

        /* return error if we don't have space for the whole character */
        if (tobufp > tobufendp) {
            return(-1);
        }


        switch(numUTF8bytes) {

            case 3: *--tobufp = (onechar | BYTE_MARK) & BYTE_MASK; onechar >>=6;
                    *--tobufp = (onechar | BYTE_MARK) & BYTE_MASK; onechar >>=6;
                    *--tobufp = onechar |  THREE_OCTET_BASE;
                    break;

            case 2: *--tobufp = (onechar | BYTE_MARK) & BYTE_MASK; onechar >>=6;
                    *--tobufp = onechar | TWO_OCTET_BASE;
                    break;
            case 1: *--tobufp = (unsigned char)onechar;  break;
        }

        return(numUTF8bytes);
}

/*
 * utf8_to_ucs2_char
 *
 * Convert a utf8 multibyte character to ucs2
 *
 * inputs: pointer to utf8 character(s)
 *         length of utf8 buffer ("read" length limit)
 *         pointer to return ucs2 character
 *
 * outputs: number of bytes in the utf8 character
 *          -1 if not a valid utf8 character sequence
 *          -2 if the buffer is too short
 */
static int16
utf8_to_ucs2_char(const unsigned char *utf8p, int16 buflen, uint16 *ucs2p)
{
    uint16 lead, cont1, cont2;

    /*
     * Check for minimum buffer length
     */
    if ((buflen < 1) || (utf8p == NULL)) {
        return -2;
    }
    lead = (uint16) (*utf8p);

    /*
     * Check for a one octet sequence
     */
    if (IS_UTF8_1ST_OF_1(lead)) {
        *ucs2p = lead & ONE_OCTET_MASK;
        return 1;
    }

    /*
     * Check for a two octet sequence
     */
    if (IS_UTF8_1ST_OF_2(*utf8p)) {
        if (buflen < 2)
            return -2;
        cont1 = (uint16) *(utf8p+1);
        if (!IS_UTF8_2ND_THRU_6TH(cont1))
            return -1;
        *ucs2p =  (lead & TWO_OCTET_MASK) << 6;
        *ucs2p |= cont1 & CONTINUING_OCTET_MASK;
        return 2;
    }

    /*
     * Check for a three octet sequence
     */
    else if (IS_UTF8_1ST_OF_3(lead)) {
        if (buflen < 3)
            return -2;
        cont1 = (uint16) *(utf8p+1);
        cont2 = (uint16) *(utf8p+2);
        if (   (!IS_UTF8_2ND_THRU_6TH(cont1))
            || (!IS_UTF8_2ND_THRU_6TH(cont2)))
            return -1;
        *ucs2p =  (lead & THREE_OCTET_MASK) << 12;
        *ucs2p |= (cont1 & CONTINUING_OCTET_MASK) << 6;
        *ucs2p |= cont2 & CONTINUING_OCTET_MASK;
        return 3;
    }
    else { /* not a valid utf8/ucs2 character */
        return -1;
    }
}

/* ----------------------------- Helper functions --------------------------- */
/* Ripped off from lm_win.c .. */
/* where is strcasecmp?.. for now, it's case sensitive..
 *
 * strcasecmp is in strings.h, but on windows it's called _stricmp...
 * will need to #ifdef this
*/

static int32
js_FileHasOption(JSContext *cx, const char *oldoptions, const char *name)
{
    char *comma, *equal, *current;
    char *options = JS_strdup(cx, oldoptions);
    int32 found = 0;

	current = options;
    for (;;) {
        comma = strchr(current, ',');
        if (comma) *comma = '\0';
        equal = strchr(current, '=');
        if (equal) *equal = '\0';
        if (strcmp(current, name) == 0) {
            if (!equal || strcmp(equal + 1, "yes") == 0)
                found = 1;
            else
                found = atoi(equal + 1);
        }
        if (equal) *equal = '=';
        if (comma) *comma = ',';
        if (found || !comma)
            break;
        current = comma + 1;
    }
    JS_free(cx, options);
    return found;
}

/* empty the buffer */
static void
js_ResetBuffers(JSFile * file)
{
    file->charBufferUsed = JS_FALSE;
    file->nbBytesInBuf = 0;
    file->linebuffer = NULL;    /* TODO: check for mem. leak? */
}

/* Reset file attributes */
static void
js_ResetAttributes(JSFile * file){
	file->mode = file->type = 0;
    file->isOpen = JS_FALSE;
    file->handle = NULL;
    file->nativehandle = NULL;
    file->hasRandomAccess = JS_TRUE; /* innocent until proven guilty */
	file->hasAutoflush = JS_FALSE;
    file->isNative = JS_FALSE;
    file->isPipe = JS_FALSE;

	js_ResetBuffers(file);
}

static JSBool
js_FileOpen(JSContext *cx, JSObject *obj, JSFile *file, char *mode){
    JSString *type, *mask;
    jsval v[2];
    jsval rval;

    type =  JS_InternString(cx, asciistring);
    mask =  JS_NewStringCopyZ(cx, mode);
    v[0] = STRING_TO_JSVAL(mask);
    v[1] = STRING_TO_JSVAL(type);

    if (!file_open(cx, obj, 2, v, &rval)) {
        return JS_FALSE;
    }
    return JS_TRUE;
}

/* Buffered version of PR_Read. Used by js_FileRead */
static int32
js_BufferedRead(JSFile * f, char *buf, int32 len)
{
    int32 count = 0;

    while (f->nbBytesInBuf>0&&len>0) {
        buf[0] = f->byteBuffer[0];
        f->byteBuffer[0] = f->byteBuffer[1];
        f->byteBuffer[1] = f->byteBuffer[2];
        f->nbBytesInBuf--;
        len--;
        buf+=1;
        count++;
    }

    if (len>0) {
        count+= (!f->isNative)?
                    PR_Read(f->handle, buf, len):
                    fread(buf, 1, len, f->nativehandle);
    }
    return count;
}

static int32
js_FileRead(JSContext *cx, JSFile * file, jschar*buf, int32 len, int32 mode)
{
    unsigned char*aux;
    int32 count, i;
    jsint remainder;
    unsigned char utfbuf[3];

    if (file->charBufferUsed) {
        buf[0] = file->charBuffer;
        buf++;
        len--;
        file->charBufferUsed = JS_FALSE;
    }

    switch (mode) {
    case ASCII:
        aux = (unsigned char*)JS_malloc(cx, len);
        if (!aux) {
        return 0;
        }
        count = js_BufferedRead(file, aux, len);
        if (count==-1) {
        JS_free(cx, aux);
        return 0;
        }
        for (i = 0;i<len;i++) {
        buf[i] = (jschar)aux[i];
        }
        JS_free(cx, aux);
        break;
    case UTF8:
        remainder = 0;
        for (count = 0;count<len;count++) {
            i = js_BufferedRead(file, utfbuf+remainder, 3-remainder);
            if (i<=0) {
                return count;
            }
            i = utf8_to_ucs2_char(utfbuf, (int16)i, &buf[count] );
            if (i<0) {
                return count;
            } else {
                if (i==1) {
                    utfbuf[0] = utfbuf[1];
                    utfbuf[1] = utfbuf[2];
                    remainder = 2;
                } else
                if (i==2) {
                    utfbuf[0] = utfbuf[2];
                    remainder = 1;
                } else
                if (i==3)
                    remainder = 0;
            }
        }
        while (remainder>0) {
            file->byteBuffer[file->nbBytesInBuf] = utfbuf[0];
            file->nbBytesInBuf++;
            utfbuf[0] = utfbuf[1];
            utfbuf[1] = utfbuf[2];
            remainder--;
        }
        break;
    case UCS2: