util.c

Apache HTTP Server 是一个功能强大的灵活的与HTTP/1.1相兼容的web服务器.这里给出的是Apache HTTP服务器的源码。

        register int c;
        register size_t i = 0;

        buf[0] = '\0';
        /* skip leading whitespace */
        do {
            c = cfp->getch(cfp->param);
        } while (c == '\t' || c == ' ');

        if (c == EOF)
            return 1;
        
        if(bufsize < 2) {
            /* too small, assume caller is crazy */
            return 1;
        }

        while (1) {
            if ((c == '\t') || (c == ' ')) {
                buf[i++] = ' ';
                while ((c == '\t') || (c == ' '))
                    c = cfp->getch(cfp->param);
            }
            if (c == CR) {
                /* silently ignore CR (_assume_ that a LF follows) */
                c = cfp->getch(cfp->param);
            }
            if (c == LF) {
                /* increase line number and return on LF */
                ++cfp->line_number;
            }
            if (c == EOF || c == 0x4 || c == LF || i >= (bufsize - 2)) {
                /* 
                 *  check for line continuation
                 */
                if (i > 0 && buf[i-1] == '\\') {
                    i--;
                    if (!(i > 0 && buf[i-1] == '\\')) {
                        /* line is continued */
                        c = cfp->getch(cfp->param);
                        continue;
                    }
                    /* else nothing needs be done because
                     * then the backslash is escaped and
                     * we just strip to a single one
                     */
                }
                /* blast trailing whitespace */
                while (i > 0 && apr_isspace(buf[i - 1]))
                    --i;
                buf[i] = '\0';
#ifdef DEBUG_CFG_LINES
                ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, NULL,
                             "Read config: %s", buf);
#endif
                return 0;
            }
            buf[i] = c;
            ++i;
            c = cfp->getch(cfp->param);
        }
    }
}

/* Size an HTTP header field list item, as separated by a comma.
 * The return value is a pointer to the beginning of the non-empty list item
 * within the original string (or NULL if there is none) and the address
 * of field is shifted to the next non-comma, non-whitespace character.
 * len is the length of the item excluding any beginning whitespace.
 */
AP_DECLARE(const char *) ap_size_list_item(const char **field, int *len)
{
    const unsigned char *ptr = (const unsigned char *)*field;
    const unsigned char *token;
    int in_qpair, in_qstr, in_com;

    /* Find first non-comma, non-whitespace byte */

    while (*ptr == ',' || apr_isspace(*ptr))
        ++ptr;

    token = ptr;

    /* Find the end of this item, skipping over dead bits */

    for (in_qpair = in_qstr = in_com = 0;
         *ptr && (in_qpair || in_qstr || in_com || *ptr != ',');
         ++ptr) {

        if (in_qpair) {
            in_qpair = 0;
        }
        else {
            switch (*ptr) {
                case '\\': in_qpair = 1;      /* quoted-pair         */
                           break;
                case '"' : if (!in_com)       /* quoted string delim */
                               in_qstr = !in_qstr;
                           break;
                case '(' : if (!in_qstr)      /* comment (may nest)  */
                               ++in_com;
                           break;
                case ')' : if (in_com)        /* end comment         */
                               --in_com;
                           break;
                default  : break;
            }
        }
    }

    if ((*len = (ptr - token)) == 0) {
        *field = (const char *)ptr;
        return NULL;
    }

    /* Advance field pointer to the next non-comma, non-white byte */

    while (*ptr == ',' || apr_isspace(*ptr))
        ++ptr;

    *field = (const char *)ptr;
    return (const char *)token;
}

/* Retrieve an HTTP header field list item, as separated by a comma,
 * while stripping insignificant whitespace and lowercasing anything not in
 * a quoted string or comment.  The return value is a new string containing
 * the converted list item (or NULL if none) and the address pointed to by
 * field is shifted to the next non-comma, non-whitespace.
 */
AP_DECLARE(char *) ap_get_list_item(apr_pool_t *p, const char **field)
{
    const char *tok_start;
    const unsigned char *ptr;
    unsigned char *pos;
    char *token;
    int addspace = 0, in_qpair = 0, in_qstr = 0, in_com = 0, tok_len = 0;

    /* Find the beginning and maximum length of the list item so that
     * we can allocate a buffer for the new string and reset the field.
     */
    if ((tok_start = ap_size_list_item(field, &tok_len)) == NULL) {
        return NULL;
    }
    token = apr_palloc(p, tok_len + 1);

    /* Scan the token again, but this time copy only the good bytes.
     * We skip extra whitespace and any whitespace around a '=', '/',
     * or ';' and lowercase normal characters not within a comment,
     * quoted-string or quoted-pair.
     */
    for (ptr = (const unsigned char *)tok_start, pos = (unsigned char *)token;
         *ptr && (in_qpair || in_qstr || in_com || *ptr != ',');
         ++ptr) {

        if (in_qpair) {
            in_qpair = 0;
            *pos++ = *ptr;
        }
        else {
            switch (*ptr) {
                case '\\': in_qpair = 1;
                           if (addspace == 1)
                               *pos++ = ' ';
                           *pos++ = *ptr;
                           addspace = 0;
                           break;
                case '"' : if (!in_com)
                               in_qstr = !in_qstr;
                           if (addspace == 1)
                               *pos++ = ' ';
                           *pos++ = *ptr;
                           addspace = 0;
                           break;
                case '(' : if (!in_qstr)
                               ++in_com;
                           if (addspace == 1)
                               *pos++ = ' ';
                           *pos++ = *ptr;
                           addspace = 0;
                           break;
                case ')' : if (in_com)
                               --in_com;
                           *pos++ = *ptr;
                           addspace = 0;
                           break;
                case ' ' :
                case '\t': if (addspace)
                               break;
                           if (in_com || in_qstr)
                               *pos++ = *ptr;
                           else
                               addspace = 1;
                           break;
                case '=' :
                case '/' :
                case ';' : if (!(in_com || in_qstr))
                               addspace = -1;
                           *pos++ = *ptr;
                           break;
                default  : if (addspace == 1)
                               *pos++ = ' ';
                           *pos++ = (in_com || in_qstr) ? *ptr
                                                        : apr_tolower(*ptr);
                           addspace = 0;
                           break;
            }
        }
    }
    *pos = '\0';

    return token;
}

/* Find an item in canonical form (lowercase, no extra spaces) within
 * an HTTP field value list.  Returns 1 if found, 0 if not found.
 * This would be much more efficient if we stored header fields as
 * an array of list items as they are received instead of a plain string.
 */
AP_DECLARE(int) ap_find_list_item(apr_pool_t *p, const char *line,
                                  const char *tok)
{
    const unsigned char *pos;
    const unsigned char *ptr = (const unsigned char *)line;
    int good = 0, addspace = 0, in_qpair = 0, in_qstr = 0, in_com = 0;

    if (!line || !tok)
        return 0;

    do {  /* loop for each item in line's list */

        /* Find first non-comma, non-whitespace byte */

        while (*ptr == ',' || apr_isspace(*ptr))
            ++ptr;

        if (*ptr)
            good = 1;  /* until proven otherwise for this item */
        else
            break;     /* no items left and nothing good found */

        /* We skip extra whitespace and any whitespace around a '=', '/',
         * or ';' and lowercase normal characters not within a comment,
         * quoted-string or quoted-pair.
         */
        for (pos = (const unsigned char *)tok;
             *ptr && (in_qpair || in_qstr || in_com || *ptr != ',');
             ++ptr) {

            if (in_qpair) {
                in_qpair = 0;
                if (good)
                    good = (*pos++ == *ptr);
            }
            else {
                switch (*ptr) {
                    case '\\': in_qpair = 1;
                               if (addspace == 1)
                                   good = good && (*pos++ == ' ');
                               good = good && (*pos++ == *ptr);
                               addspace = 0;
                               break;
                    case '"' : if (!in_com)
                                   in_qstr = !in_qstr;
                               if (addspace == 1)
                                   good = good && (*pos++ == ' ');
                               good = good && (*pos++ == *ptr);
                               addspace = 0;
                               break;
                    case '(' : if (!in_qstr)
                                   ++in_com;
                               if (addspace == 1)
                                   good = good && (*pos++ == ' ');
                               good = good && (*pos++ == *ptr);
                               addspace = 0;
                               break;
                    case ')' : if (in_com)
                                   --in_com;
                               good = good && (*pos++ == *ptr);
                               addspace = 0;
                               break;
                    case ' ' :
                    case '\t': if (addspace || !good)
                                   break;
                               if (in_com || in_qstr)
                                   good = (*pos++ == *ptr);
                               else
                                   addspace = 1;
                               break;
                    case '=' :
                    case '/' :
                    case ';' : if (!(in_com || in_qstr))
                                   addspace = -1;
                               good = good && (*pos++ == *ptr);
                               break;
                    default  : if (!good)
                                   break;
                               if (addspace == 1)
                                   good = (*pos++ == ' ');
                               if (in_com || in_qstr)
                                   good = good && (*pos++ == *ptr);
                               else
                                   good = good && (*pos++ == apr_tolower(*ptr));
                               addspace = 0;
                               break;
                }
            }
        }
        if (good && *pos)
            good = 0;          /* not good if only a prefix was matched */

    } while (*ptr && !good);

    return good;
}


/* Retrieve a token, spacing over it and returning a pointer to
 * the first non-white byte afterwards.  Note that these tokens
 * are delimited by semis and commas; and can also be delimited
 * by whitespace at the caller's option.
 */

AP_DECLARE(char *) ap_get_token(apr_pool_t *p, const char **accept_line,
                                int accept_white)
{
    const char *ptr = *accept_line;
    const char *tok_start;
    char *token;
    int tok_len;

    /* Find first non-white byte */

    while (*ptr && apr_isspace(*ptr))
        ++ptr;

    tok_start = ptr;

    /* find token end, skipping over quoted strings.
     * (comments are already gone).
     */

    while (*ptr && (accept_white || !apr_isspace(*ptr))
           && *ptr != ';' && *ptr != ',') {
        if (*ptr++ == '"')
            while (*ptr)
                if (*ptr++ == '"')
                    break;
    }

    tok_len = ptr - tok_start;
    token = apr_pstrndup(p, tok_start, tok_len);

    /* Advance accept_line pointer to the next non-white byte */

    while (*ptr && apr_isspace(*ptr))
        ++ptr;

    *accept_line = ptr;
    return token;
}


/* find http tokens, see the definition of token from RFC2068 */
AP_DECLARE(int) ap_find_token(apr_pool_t *p, const char *line, const char *tok)
{
    const unsigned char *start_token;
    const unsigned char *s;

    if (!line)
        return 0;

    s = (const unsigned char *)line;
    for (;;) {
        /* find start of token, skip all stop characters, note NUL
         * isn't a token stop, so we don't need to test for it
         */
        while (TEST_CHAR(*s, T_HTTP_TOKEN_STOP)) {
            ++s;
        }
        if (!*s) {
            return 0;
        }
        start_token = s;
        /* find end of the token */
        while (*s && !TEST_CHAR(*s, T_HTTP_TOKEN_STOP)) {
            ++s;
        }
        if (!strncasecmp((const char *)start_token, (const char *)tok,
                         s - start_token)) {
            return 1;
        }
        if (!*s) {
            return 0;
        }
    }
}


AP_DECLARE(int) ap_find_last_token(apr_pool_t *p, const char *line,
                                   const char *tok)
{
    int llen, tlen, lidx;

    if (!line)
        return 0;

    llen = strlen(line);
    tlen = strlen(tok);
    lidx = llen - tlen;

    if (lidx < 0 ||
        (lidx > 0 && !(apr_isspace(line[lidx - 1]) || line[lidx - 1] == ',')))
        return 0;

    return (strncasecmp(&line[lidx], tok, tlen) == 0);
}

AP_DECLARE(char *) ap_escape_shell_cmd(apr_pool_t *p, const char *str)
{
    char *cmd;
    unsigned char *d;
    const unsigned char *s;

    cmd = apr_palloc(p, 2 * strlen(str) + 1);        /* Be safe */
    d = (unsigned char *)cmd;
    s = (const unsigned char *)str;
    for (; *s; ++s) {

#if defined(OS2) || defined(WIN32)
        /* 
         * Newlines to Win32/OS2 CreateProcess() are ill advised.
         * Convert them to spaces since they are effectively white
         * space to most applications
         */
        if (*s == '\r' || *s == '\n') {
             *d++ = ' ';
             continue;
         }
#endif

        if (TEST_CHAR(*s, T_ESCAPE_SHELL_CMD)) {
            *d++ = '\\';
        }
        *d++ = *s;
    }
    *d = '\0';

    return cmd;
}

static char x2c(const char *what)
{
    register char digit;

#if !APR_CHARSET_EBCDIC
    digit = ((what[0] >= 'A') ? ((what[0] & 0xdf) - 'A') + 10
             : (what[0] - '0'));
    digit *= 16;
    digit += (what[1] >= 'A' ? ((what[1] & 0xdf) - 'A') + 10
              : (what[1] - '0'));
#else /*APR_CHARSET_EBCDIC*/
    char xstr[5];
    xstr[0]='0';
    xstr[1]='x';
    xstr[2]=what[0];
    xstr[3]=what[1];
    xstr[4]='\0';
    digit = apr_xlate_conv_byte(ap_hdrs_from_ascii,
                                0xFF & strtol(xstr, NULL, 16));
#endif /*APR_CHARSET_EBCDIC*/
    return (digit);
}

/*
 * Unescapes a URL.
 * Returns 0 on success, non-zero on error
 * Failure is due to
 *   bad % escape       returns HTTP_BAD_REQUEST
 *
 *   decoding %00 -> \0  (the null character)
 *   decoding %2f -> /   (a special character)
 *                      returns HTTP_NOT_FOUND
 */
AP_DECLARE(int) ap_unescape_url(char *url)
{
    register int badesc, badpath;
    char *x, *y;

    badesc = 0;
    badpath = 0;
    /* Initial scan for first '%'. Don't bother writing values before
     * seeing a '%' */
    y = strchr(url, '%');
    if (y == NULL) {
        return OK;
    }
    for (x = y; *y; ++x, ++y) {
        if (*y != '%')
            *x = *y;
        else {
            if (!apr_isxdigit(*(y + 1)) || !apr_isxdigit(*(y + 2))) {
                badesc = 1;
                *x = '%';
            }
            else {
                *x = x2c(y + 1);
                y += 2;
                if (IS_SLASH(*x) || *x == '\0')
                    badpath = 1;
            }
        }
    }
    *x = '\0';
    if (badesc)
        return HTTP_BAD_REQUEST;
    else if (badpath)
        return HTTP_NOT_FOUND;
    else
        return OK;
}

AP_DECLARE(int) ap_unescape_url_keep2f(char *url)
{
    register int badesc, badpath;
    char *x, *y;

    badesc = 0;
    badpath = 0;
    /* Initial scan for first '%'. Don't bother writing values before
     * seeing a '%' */
    y = strchr(url, '%');
    if (y == NULL) {
        return OK;
    }
    for (x = y; *y; ++x, ++y) {
        if (*y != '%') {
            *x = *y;
        }
        else {
            if (!apr_isxdigit(*(y + 1)) || !apr_isxdigit(*(y + 2))) {
                badesc = 1;