jsscan.c

java script test programing source code

            atom = TOKENBUF_TO_ATOM();
            if (!atom)
                goto error;
            tp->pos.end.lineno = (uint16)ts->lineno;
            tp->t_op = JSOP_STRING;
            tp->t_atom = atom;
            tt = TOK_XMLATTR;
            goto out;

          case '>':
            tt = TOK_XMLTAGC;
            goto out;

          case '/':
            if (MatchChar(ts, '>')) {
                tt = TOK_XMLPTAGC;
                goto out;
            }
            /* FALL THROUGH */

          bad_xml_char:
          default:
            js_ReportCompileErrorNumber(cx, ts, JSREPORT_TS | JSREPORT_ERROR,
                                        JSMSG_BAD_XML_CHARACTER);
            goto error;
        }
        /* NOTREACHED */
    }
#endif /* JS_HAS_XML_SUPPORT */

retry:
    do {
        c = GetChar(ts);
        if (c == '\n') {
            ts->flags &= ~TSF_DIRTYLINE;
            if (ts->flags & TSF_NEWLINES)
                break;
        }
    } while (JS_ISSPACE(c));

    tp = NewToken(ts, -1);
    if (c == EOF) {
        tt = TOK_EOF;
        goto out;
    }

    hadUnicodeEscape = JS_FALSE;
    if (JS_ISIDSTART(c) ||
        (c == '\\' &&
         (c = GetUnicodeEscape(ts),
          hadUnicodeEscape = JS_ISIDSTART(c)))) {
        INIT_TOKENBUF();
        for (;;) {
            ADD_TO_TOKENBUF(c);
            c = GetChar(ts);
            if (c == '\\') {
                c = GetUnicodeEscape(ts);
                if (!JS_ISIDENT(c))
                    break;
                hadUnicodeEscape = JS_TRUE;
            } else {
                if (!JS_ISIDENT(c))
                    break;
            }
        }
        UngetChar(ts, c);

        /*
         * Check for keywords unless we saw Unicode escape or parser asks
         * to ignore keywords.
         */
        if (!hadUnicodeEscape &&
            !(ts->flags & TSF_KEYWORD_IS_NAME) &&
            TOKENBUF_OK() &&
            (kw = FindKeyword(TOKENBUF_BASE(), TOKENBUF_LENGTH()))) {
            if (kw->tokentype == TOK_RESERVED) {
                if (!js_ReportCompileErrorNumber(cx, ts,
                                                 JSREPORT_TS |
                                                 JSREPORT_WARNING |
                                                 JSREPORT_STRICT,
                                                 JSMSG_RESERVED_ID,
                                                 kw->chars)) {
                    goto error;
                }
            } else if (kw->version <= JSVERSION_NUMBER(cx)) {
                tt = kw->tokentype;
                tp->t_op = (JSOp) kw->op;
                goto out;
            }
        }

        atom = TOKENBUF_TO_ATOM();
        if (!atom)
            goto error;
        tp->t_op = JSOP_NAME;
        tp->t_atom = atom;
        tt = TOK_NAME;
        goto out;
    }

    if (JS7_ISDEC(c) || (c == '.' && JS7_ISDEC(PeekChar(ts)))) {
        jsint radix;
        const jschar *endptr;
        jsdouble dval;

        radix = 10;
        INIT_TOKENBUF();

        if (c == '0') {
            ADD_TO_TOKENBUF(c);
            c = GetChar(ts);
            if (JS_TOLOWER(c) == 'x') {
                ADD_TO_TOKENBUF(c);
                c = GetChar(ts);
                radix = 16;
            } else if (JS7_ISDEC(c)) {
                radix = 8;
            }
        }

        while (JS7_ISHEX(c)) {
            if (radix < 16) {
                if (JS7_ISLET(c))
                    break;

                /*
                 * We permit 08 and 09 as decimal numbers, which makes our
                 * behaviour a superset of the ECMA numeric grammar.  We might
                 * not always be so permissive, so we warn about it.
                 */
                if (radix == 8 && c >= '8') {
                    if (!js_ReportCompileErrorNumber(cx, ts,
                                                     JSREPORT_TS |
                                                     JSREPORT_WARNING,
                                                     JSMSG_BAD_OCTAL,
                                                     c == '8' ? "08" : "09")) {
                        goto error;
                    }
                    radix = 10;
                }
            }
            ADD_TO_TOKENBUF(c);
            c = GetChar(ts);
        }

        if (radix == 10 && (c == '.' || JS_TOLOWER(c) == 'e')) {
            if (c == '.') {
                do {
                    ADD_TO_TOKENBUF(c);
                    c = GetChar(ts);
                } while (JS7_ISDEC(c));
            }
            if (JS_TOLOWER(c) == 'e') {
                ADD_TO_TOKENBUF(c);
                c = GetChar(ts);
                if (c == '+' || c == '-') {
                    ADD_TO_TOKENBUF(c);
                    c = GetChar(ts);
                }
                if (!JS7_ISDEC(c)) {
                    js_ReportCompileErrorNumber(cx, ts,
                                                JSREPORT_TS | JSREPORT_ERROR,
                                                JSMSG_MISSING_EXPONENT);
                    goto error;
                }
                do {
                    ADD_TO_TOKENBUF(c);
                    c = GetChar(ts);
                } while (JS7_ISDEC(c));
            }
        }

        /* Put back the next char and NUL-terminate tokenbuf for js_strto*. */
        UngetChar(ts, c);
        ADD_TO_TOKENBUF(0);

        if (!TOKENBUF_OK())
            goto error;
        if (radix == 10) {
            if (!js_strtod(cx, TOKENBUF_BASE(), &endptr, &dval)) {
                js_ReportCompileErrorNumber(cx, ts,
                                            JSREPORT_TS | JSREPORT_ERROR,
                                            JSMSG_OUT_OF_MEMORY);
                goto error;
            }
        } else {
            if (!js_strtointeger(cx, TOKENBUF_BASE(), &endptr, radix, &dval)) {
                js_ReportCompileErrorNumber(cx, ts,
                                            JSREPORT_TS | JSREPORT_ERROR,
                                            JSMSG_OUT_OF_MEMORY);
                goto error;
            }
        }
        tp->t_dval = dval;
        tt = TOK_NUMBER;
        goto out;
    }

    if (c == '"' || c == '\'') {
        qc = c;
        INIT_TOKENBUF();
        while ((c = GetChar(ts)) != qc) {
            if (c == '\n' || c == EOF) {
                UngetChar(ts, c);
                js_ReportCompileErrorNumber(cx, ts,
                                            JSREPORT_TS | JSREPORT_ERROR,
                                            JSMSG_UNTERMINATED_STRING);
                goto error;
            }
            if (c == '\\') {
                switch (c = GetChar(ts)) {
                  case 'b': c = '\b'; break;
                  case 'f': c = '\f'; break;
                  case 'n': c = '\n'; break;
                  case 'r': c = '\r'; break;
                  case 't': c = '\t'; break;
                  case 'v': c = '\v'; break;

                  default:
                    if ('0' <= c && c < '8') {
                        int32 val = JS7_UNDEC(c);

                        c = PeekChar(ts);
                        if ('0' <= c && c < '8') {
                            val = 8 * val + JS7_UNDEC(c);
                            GetChar(ts);
                            c = PeekChar(ts);
                            if ('0' <= c && c < '8') {
                                int32 save = val;
                                val = 8 * val + JS7_UNDEC(c);
                                if (val <= 0377)
                                    GetChar(ts);
                                else
                                    val = save;
                            }
                        }

                        c = (jschar)val;
                    } else if (c == 'u') {
                        jschar cp[4];
                        if (PeekChars(ts, 4, cp) &&
                            JS7_ISHEX(cp[0]) && JS7_ISHEX(cp[1]) &&
                            JS7_ISHEX(cp[2]) && JS7_ISHEX(cp[3])) {
                            c = (((((JS7_UNHEX(cp[0]) << 4)
                                    + JS7_UNHEX(cp[1])) << 4)
                                  + JS7_UNHEX(cp[2])) << 4)
                                + JS7_UNHEX(cp[3]);
                            SkipChars(ts, 4);
                        }
                    } else if (c == 'x') {
                        jschar cp[2];
                        if (PeekChars(ts, 2, cp) &&
                            JS7_ISHEX(cp[0]) && JS7_ISHEX(cp[1])) {
                            c = (JS7_UNHEX(cp[0]) << 4) + JS7_UNHEX(cp[1]);
                            SkipChars(ts, 2);
                        }
                    } else if (c == '\n' && JS_VERSION_IS_ECMA(cx)) {
                        /* ECMA follows C by removing escaped newlines. */
                        continue;
                    }
                    break;
                }
            }
            ADD_TO_TOKENBUF(c);
        }
        atom = TOKENBUF_TO_ATOM();
        if (!atom)
            goto error;
        tp->pos.end.lineno = (uint16)ts->lineno;
        tp->t_op = JSOP_STRING;
        tp->t_atom = atom;
        tt = TOK_STRING;
        goto out;
    }

    switch (c) {
      case '\n': tt = TOK_EOL; goto eol_out;
      case ';':  tt = TOK_SEMI; break;
      case '[':  tt = TOK_LB; break;
      case ']':  tt = TOK_RB; break;
      case '{':  tt = TOK_LC; break;
      case '}':  tt = TOK_RC; break;
      case '(':  tt = TOK_LP; break;
      case ')':  tt = TOK_RP; break;
      case ',':  tt = TOK_COMMA; break;
      case '?':  tt = TOK_HOOK; break;

      case '.':
#if JS_HAS_XML_SUPPORT
        if (MatchChar(ts, c))
            tt = TOK_DBLDOT;
        else
#endif
            tt = TOK_DOT;
        break;

      case ':':
#if JS_HAS_XML_SUPPORT
        if (MatchChar(ts, c)) {
            tt = TOK_DBLCOLON;
            break;
        }
#endif
        /*
         * Default so compiler can modify to JSOP_GETTER if 'p getter: v' in an
         * object initializer, likewise for setter.
         */
        tp->t_op = JSOP_NOP;
        tt = TOK_COLON;
        break;

      case '|':
        if (MatchChar(ts, c)) {
            tt = TOK_OR;
        } else if (MatchChar(ts, '=')) {
            tp->t_op = JSOP_BITOR;
            tt = TOK_ASSIGN;
        } else {
            tt = TOK_BITOR;
        }
        break;

      case '^':
        if (MatchChar(ts, '=')) {
            tp->t_op = JSOP_BITXOR;
            tt = TOK_ASSIGN;
        } else {
            tt = TOK_BITXOR;
        }
        break;

      case '&':
        if (MatchChar(ts, c)) {
            tt = TOK_AND;
        } else if (MatchChar(ts, '=')) {
            tp->t_op = JSOP_BITAND;
            tt = TOK_ASSIGN;
        } else {
            tt = TOK_BITAND;
        }
        break;

      case '=':
        if (MatchChar(ts, c)) {
            tp->t_op = MatchChar(ts, c) ? JSOP_NEW_EQ : (JSOp)cx->jsop_eq;
            tt = TOK_EQOP;
        } else {
            tp->t_op = JSOP_NOP;
            tt = TOK_ASSIGN;
        }
        break;

      case '!':
        if (MatchChar(ts, '=')) {
            tp->t_op = MatchChar(ts, '=') ? JSOP_NEW_NE : (JSOp)cx->jsop_ne;
            tt = TOK_EQOP;
        } else {
            tp->t_op = JSOP_NOT;
            tt = TOK_UNARYOP;
        }
        break;

#if JS_HAS_XML_SUPPORT
      case '@':
        tt = TOK_AT;
        break;
#endif

      case '<':
#if JS_HAS_XML_SUPPORT
        /*
         * After much testing, it's clear that Postel's advice to protocol
         * designers ("be liberal in what you accept, and conservative in what
         * you send") invites a natural-law repercussion for JS as "protocol":
         *
         * "If you are liberal in what you accept, others will utterly fail to
         *  be conservative in what they send."
         *
         * Which means you will get <!-- comments to end of line in the middle
         * of .js files, and after if conditions whose then statements are on
         * the next line, and other wonders.  See at least the following bugs:
         * https://bugzilla.mozilla.org/show_bug.cgi?id=309242
         * https://bugzilla.mozilla.org/show_bug.cgi?id=309712
         * https://bugzilla.mozilla.org/show_bug.cgi?id=310993
         *
         * So without JSOPTION_XML, we never scan an XML comment or CDATA
         * literal.  We always scan <! as the start of an HTML comment hack
         * to end of line, used since Netscape 2 to hide script tag content
         * from script-unaware browsers.
         */
        if ((ts->flags & TSF_OPERAND) &&
            (JS_HAS_XML_OPTION(cx) || PeekChar(ts) != '!')) {
            /* Check for XML comment or CDATA section. */
            if (MatchChar(ts, '!')) {
                INIT_TOKENBUF();

                /* Scan XML comment. */
                if (MatchChar(ts, '-')) {
                    if (!MatchChar(ts, '-'))
                        goto bad_xml_markup;
                    while ((c = GetChar(ts)) != '-' || !MatchChar(ts, '-')) {
                        if (c == EOF)
                            goto bad_xml_markup;
                        ADD_TO_TOKENBUF(c);
                    }
                    tt = TOK_XMLCOMMENT;
                    tp->t_op = JSOP_XMLCOMMENT;
                    goto finish_xml_markup;
                }

                /* Scan CDATA section. */
                if (MatchChar(ts, '[')) {
                    jschar cp[6];