jsscan.c

java script test programing source code

    if (!ENSURE_STRING_BUFFER(sb, 1))
        return;
    bp = sb->ptr;
    *bp++ = c;
    *bp = 0;
    sb->ptr = bp;
}

#if JS_HAS_XML_SUPPORT

void
js_RepeatChar(JSStringBuffer *sb, jschar c, uintN count)
{
    jschar *bp;

    if (!STRING_BUFFER_OK(sb) || count == 0)
        return;
    if (!ENSURE_STRING_BUFFER(sb, count))
        return;
    for (bp = sb->ptr; count; --count)
        *bp++ = c;
    *bp = 0;
    sb->ptr = bp;
}

void
js_AppendCString(JSStringBuffer *sb, const char *asciiz)
{
    size_t length;
    jschar *bp;

    if (!STRING_BUFFER_OK(sb) || *asciiz == '\0')
        return;
    length = strlen(asciiz);
    if (!ENSURE_STRING_BUFFER(sb, length))
        return;
    for (bp = sb->ptr; length; --length)
        *bp++ = (jschar) *asciiz++;
    *bp = 0;
    sb->ptr = bp;
}

void
js_AppendJSString(JSStringBuffer *sb, JSString *str)
{
    size_t length;
    jschar *bp;

    if (!STRING_BUFFER_OK(sb))
        return;
    length = JSSTRING_LENGTH(str);
    if (length == 0 || !ENSURE_STRING_BUFFER(sb, length))
        return;
    bp = sb->ptr;
    js_strncpy(bp, JSSTRING_CHARS(str), length);
    bp += length;
    *bp = 0;
    sb->ptr = bp;
}

static JSBool
GetXMLEntity(JSContext *cx, JSTokenStream *ts)
{
    ptrdiff_t offset, length, i;
    int32 c, d;
    JSBool ispair;
    jschar *bp, digit;
    char *bytes;
    JSErrNum msg;

    /* Put the entity, including the '&' already scanned, in ts->tokenbuf. */
    offset = PTRDIFF(ts->tokenbuf.ptr, ts->tokenbuf.base, jschar);
    FastAppendChar(&ts->tokenbuf, '&');
    while ((c = GetChar(ts)) != ';') {
        if (c == EOF || c == '\n') {
            js_ReportCompileErrorNumber(cx, ts,
                                        JSREPORT_TS | JSREPORT_ERROR,
                                        JSMSG_END_OF_XML_ENTITY);
            return JS_FALSE;
        }
        FastAppendChar(&ts->tokenbuf, (jschar) c);
    }

    /* Let length be the number of jschars after the '&', including the ';'. */
    length = PTRDIFF(ts->tokenbuf.ptr, ts->tokenbuf.base, jschar) - offset;
    bp = ts->tokenbuf.base + offset;
    c = d = 0;
    ispair = JS_FALSE;
    if (length > 2 && bp[1] == '#') {
        /* Match a well-formed XML Character Reference. */
        i = 2;
        if (length > 3 && JS_TOLOWER(bp[i]) == 'x') {
            if (length > 9)     /* at most 6 hex digits allowed */
                goto badncr;
            while (++i < length) {
                digit = bp[i];
                if (!JS7_ISHEX(digit))
                    goto badncr;
                c = (c << 4) + JS7_UNHEX(digit);
            }
        } else {
            while (i < length) {
                digit = bp[i++];
                if (!JS7_ISDEC(digit))
                    goto badncr;
                c = (c * 10) + JS7_UNDEC(digit);
                if (c < 0)
                    goto badncr;
            }
        }

        if (0x10000 <= c && c <= 0x10FFFF) {
            /* Form a surrogate pair (c, d) -- c is the high surrogate. */
            d = 0xDC00 + (c & 0x3FF);
            c = 0xD7C0 + (c >> 10);
            ispair = JS_TRUE;
        } else {
            /* Enforce the http://www.w3.org/TR/REC-xml/#wf-Legalchar WFC. */
            if (c != 0x9 && c != 0xA && c != 0xD &&
                !(0x20 <= c && c <= 0xD7FF) &&
                !(0xE000 <= c && c <= 0xFFFD)) {
                goto badncr;
            }
        }
    } else {
        /* Try to match one of the five XML 1.0 predefined entities. */
        switch (length) {
          case 3:
            if (bp[2] == 't') {
                if (bp[1] == 'l')
                    c = '<';
                else if (bp[1] == 'g')
                    c = '>';
            }
            break;
          case 4:
            if (bp[1] == 'a' && bp[2] == 'm' && bp[3] == 'p')
                c = '&';
            break;
          case 5:
            if (bp[3] == 'o') {
                if (bp[1] == 'a' && bp[2] == 'p' && bp[4] == 's')
                    c = '\'';
                else if (bp[1] == 'q' && bp[2] == 'u' && bp[4] == 't')
                    c = '"';
            }
            break;
        }
        if (c == 0) {
            msg = JSMSG_UNKNOWN_XML_ENTITY;
            goto bad;
        }
    }

    /* If we matched, retract ts->tokenbuf and store the entity's value. */
    *bp++ = (jschar) c;
    if (ispair)
        *bp++ = (jschar) d;
    *bp = 0;
    ts->tokenbuf.ptr = bp;
    return JS_TRUE;

badncr:
    msg = JSMSG_BAD_XML_NCR;
bad:
    /* No match: throw a TypeError per ECMA-357 10.3.2.1 step 8(a). */
    bytes = js_DeflateString(cx, bp + 1,
                             PTRDIFF(ts->tokenbuf.ptr, bp, jschar) - 1);
    if (bytes) {
        js_ReportCompileErrorNumber(cx, ts, JSREPORT_TS | JSREPORT_ERROR,
                                    msg, bytes);
        JS_free(cx, bytes);
    }
    return JS_FALSE;
}

#endif /* JS_HAS_XML_SUPPORT */

JSTokenType
js_PeekToken(JSContext *cx, JSTokenStream *ts)
{
    JSTokenType tt;

    if (ts->lookahead != 0) {
        tt = ts->tokens[(ts->cursor + ts->lookahead) & NTOKENS_MASK].type;
    } else {
        tt = js_GetToken(cx, ts);
        js_UngetToken(ts);
    }
    return tt;
}

JSTokenType
js_PeekTokenSameLine(JSContext *cx, JSTokenStream *ts)
{
    JSTokenType tt;

    if (!ON_CURRENT_LINE(ts, CURRENT_TOKEN(ts).pos))
        return TOK_EOL;
    ts->flags |= TSF_NEWLINES;
    tt = js_PeekToken(cx, ts);
    ts->flags &= ~TSF_NEWLINES;
    return tt;
}

/*
 * We have encountered a '\': check for a Unicode escape sequence after it,
 * returning the character code value if we found a Unicode escape sequence.
 * Otherwise, non-destructively return the original '\'.
 */
static int32
GetUnicodeEscape(JSTokenStream *ts)
{
    jschar cp[5];
    int32 c;

    if (PeekChars(ts, 5, cp) && cp[0] == 'u' &&
        JS7_ISHEX(cp[1]) && JS7_ISHEX(cp[2]) &&
        JS7_ISHEX(cp[3]) && JS7_ISHEX(cp[4]))
    {
        c = (((((JS7_UNHEX(cp[1]) << 4)
                + JS7_UNHEX(cp[2])) << 4)
              + JS7_UNHEX(cp[3])) << 4)
            + JS7_UNHEX(cp[4]);
        SkipChars(ts, 5);
        return c;
    }
    return '\\';
}

static JSToken *
NewToken(JSTokenStream *ts, ptrdiff_t adjust)
{
    JSToken *tp;

    ts->cursor = (ts->cursor + 1) & NTOKENS_MASK;
    tp = &CURRENT_TOKEN(ts);
    tp->ptr = ts->linebuf.ptr + adjust;
    tp->pos.begin.index = ts->linepos +
                          PTRDIFF(tp->ptr, ts->linebuf.base, jschar) -
                          ts->ungetpos;
    tp->pos.begin.lineno = tp->pos.end.lineno = (uint16)ts->lineno;
    return tp;
}

JSTokenType
js_GetToken(JSContext *cx, JSTokenStream *ts)
{
    JSTokenType tt;
    int32 c, qc;
    JSToken *tp;
    JSAtom *atom;
    JSBool hadUnicodeEscape;
    const struct keyword *kw;

#define INIT_TOKENBUF()     (ts->tokenbuf.ptr = ts->tokenbuf.base)
#define TOKENBUF_LENGTH()   PTRDIFF(ts->tokenbuf.ptr, ts->tokenbuf.base, jschar)
#define TOKENBUF_OK()       STRING_BUFFER_OK(&ts->tokenbuf)
#define TOKENBUF_TO_ATOM()  (TOKENBUF_OK()                                    \
                             ? js_AtomizeChars(cx,                            \
                                               TOKENBUF_BASE(),               \
                                               TOKENBUF_LENGTH(),             \
                                               0)                             \
                             : NULL)
#define ADD_TO_TOKENBUF(c)  FastAppendChar(&ts->tokenbuf, (jschar) (c))

/* The following 4 macros should only be used when TOKENBUF_OK() is true. */
#define TOKENBUF_BASE()     (ts->tokenbuf.base)
#define TOKENBUF_CHAR(i)    (ts->tokenbuf.base[i])
#define TRIM_TOKENBUF(i)    (ts->tokenbuf.ptr = ts->tokenbuf.base + i)
#define NUL_TERM_TOKENBUF() (*ts->tokenbuf.ptr = 0)

    /* Check for a pushed-back token resulting from mismatching lookahead. */
    while (ts->lookahead != 0) {
        JS_ASSERT(!(ts->flags & TSF_XMLTEXTMODE));
        ts->lookahead--;
        ts->cursor = (ts->cursor + 1) & NTOKENS_MASK;
        tt = CURRENT_TOKEN(ts).type;
        if (tt != TOK_EOL || (ts->flags & TSF_NEWLINES))
            return tt;
    }

    /* If there was a fatal error, keep returning TOK_ERROR. */
    if (ts->flags & TSF_ERROR)
        return TOK_ERROR;

#if JS_HAS_XML_SUPPORT
    if (ts->flags & TSF_XMLTEXTMODE) {
        tt = TOK_XMLSPACE;      /* veto if non-space, return TOK_XMLTEXT */
        tp = NewToken(ts, 0);
        INIT_TOKENBUF();
        qc = (ts->flags & TSF_XMLONLYMODE) ? '<' : '{';

        while ((c = GetChar(ts)) != qc && c != '<' && c != EOF) {
            if (c == '&' && qc == '<') {
                if (!GetXMLEntity(cx, ts))
                    goto error;
                tt = TOK_XMLTEXT;
                continue;
            }

            if (!JS_ISXMLSPACE(c))
                tt = TOK_XMLTEXT;
            ADD_TO_TOKENBUF(c);
        }
        UngetChar(ts, c);

        if (TOKENBUF_LENGTH() == 0) {
            atom = NULL;
        } else {
            atom = TOKENBUF_TO_ATOM();
            if (!atom)
                goto error;
        }
        tp->pos.end.lineno = (uint16)ts->lineno;
        tp->t_op = JSOP_STRING;
        tp->t_atom = atom;
        goto out;
    }

    if (ts->flags & TSF_XMLTAGMODE) {
        tp = NewToken(ts, 0);
        c = GetChar(ts);
        if (JS_ISXMLSPACE(c)) {
            do {
                c = GetChar(ts);
            } while (JS_ISXMLSPACE(c));
            UngetChar(ts, c);
            tt = TOK_XMLSPACE;
            goto out;
        }

        if (c == EOF) {
            tt = TOK_EOF;
            goto out;
        }

        INIT_TOKENBUF();
        if (JS_ISXMLNSSTART(c)) {
            JSBool sawColon = JS_FALSE;

            ADD_TO_TOKENBUF(c);
            while ((c = GetChar(ts)) != EOF && JS_ISXMLNAME(c)) {
                if (c == ':') {
                    int nextc;

                    if (sawColon ||
                        (nextc = PeekChar(ts),
                         ((ts->flags & TSF_XMLONLYMODE) || nextc != '{') &&
                         !JS_ISXMLNAME(nextc))) {
                        js_ReportCompileErrorNumber(cx, ts,
                                                    JSREPORT_TS |
                                                    JSREPORT_ERROR,
                                                    JSMSG_BAD_XML_QNAME);
                        goto error;
                    }
                    sawColon = JS_TRUE;
                }

                ADD_TO_TOKENBUF(c);
            }

            UngetChar(ts, c);
            atom = TOKENBUF_TO_ATOM();
            if (!atom)
                goto error;
            tp->t_op = JSOP_STRING;
            tp->t_atom = atom;
            tt = TOK_XMLNAME;
            goto out;
        }

        switch (c) {
          case '{':
            if (ts->flags & TSF_XMLONLYMODE)
                goto bad_xml_char;
            tt = TOK_LC;
            goto out;

          case '=':
            tt = TOK_ASSIGN;
            goto out;

          case '"':
          case '\'':
            qc = c;
            while ((c = GetChar(ts)) != qc) {
                if (c == EOF) {
                    js_ReportCompileErrorNumber(cx, ts,
                                                JSREPORT_TS | JSREPORT_ERROR,
                                                JSMSG_UNTERMINATED_STRING);
                    goto error;
                }

                /*
                 * XML attribute values are double-quoted when pretty-printed,
                 * so escape " if it is expressed directly in a single-quoted
                 * attribute value.
                 */
                if (c == '"' && !(ts->flags & TSF_XMLONLYMODE)) {
                    JS_ASSERT(qc == '\'');
                    js_AppendCString(&ts->tokenbuf, js_quot_entity_str);
                    continue;
                }

                if (c == '&' && (ts->flags & TSF_XMLONLYMODE)) {
                    if (!GetXMLEntity(cx, ts))
                        goto error;
                    continue;
                }

                ADD_TO_TOKENBUF(c);
            }