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。

/* Copyright 1999-2005 The Apache Software Foundation or its licensors, as
 * applicable.
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

/*
 * util.c: string utility things
 * 
 * 3/21/93 Rob McCool
 * 1995-96 Many changes by the Apache Software Foundation
 * 
 */

/* Debugging aid:
 * #define DEBUG            to trace all cfg_open*()/cfg_closefile() calls
 * #define DEBUG_CFG_LINES  to trace every line read from the config files
 */

#include "apr.h"
#include "apr_strings.h"
#include "apr_lib.h"

#define APR_WANT_STDIO
#define APR_WANT_STRFUNC
#include "apr_want.h"

#if APR_HAVE_UNISTD_H
#include <unistd.h>
#endif
#if APR_HAVE_NETDB_H
#include <netdb.h>              /* for gethostbyname() */
#endif

#define CORE_PRIVATE

#include "ap_config.h"
#include "apr_base64.h"
#include "httpd.h"
#include "http_main.h"
#include "http_log.h"
#include "http_protocol.h"
#include "http_config.h"
#include "util_ebcdic.h"

#ifdef HAVE_PWD_H
#include <pwd.h>
#endif
#ifdef HAVE_GRP_H
#include <grp.h>
#endif

/* A bunch of functions in util.c scan strings looking for certain characters.
 * To make that more efficient we encode a lookup table.  The test_char_table
 * is generated automatically by gen_test_char.c.
 */
#include "test_char.h"

/* we assume the folks using this ensure 0 <= c < 256... which means
 * you need a cast to (unsigned char) first, you can't just plug a
 * char in here and get it to work, because if char is signed then it
 * will first be sign extended.
 */
#define TEST_CHAR(c, f)        (test_char_table[(unsigned)(c)] & (f))

/* Win32/NetWare/OS2 need to check for both forward and back slashes
 * in ap_getparents() and ap_escape_url.
 */
#ifdef CASE_BLIND_FILESYSTEM
#define IS_SLASH(s) ((s == '/') || (s == '\\'))
#else
#define IS_SLASH(s) (s == '/')
#endif


/*
 * Examine a field value (such as a media-/content-type) string and return
 * it sans any parameters; e.g., strip off any ';charset=foo' and the like.
 */
AP_DECLARE(char *) ap_field_noparam(apr_pool_t *p, const char *intype)
{
    const char *semi;

    if (intype == NULL) return NULL;

    semi = ap_strchr_c(intype, ';');
    if (semi == NULL) {
        return apr_pstrdup(p, intype);
    } 
    else {
        while ((semi > intype) && apr_isspace(semi[-1])) {
            semi--;
        }
        return apr_pstrndup(p, intype, semi - intype);
    }
}

AP_DECLARE(char *) ap_ht_time(apr_pool_t *p, apr_time_t t, const char *fmt,
                              int gmt)
{
    apr_size_t retcode;
    char ts[MAX_STRING_LEN];
    char tf[MAX_STRING_LEN];
    apr_time_exp_t xt;

    if (gmt) {
        const char *f;
        char *strp;

        apr_time_exp_gmt(&xt, t);
        /* Convert %Z to "GMT" and %z to "+0000";
         * on hosts that do not have a time zone string in struct tm,
         * strftime must assume its argument is local time.
         */
        for(strp = tf, f = fmt; strp < tf + sizeof(tf) - 6 && (*strp = *f)
            ; f++, strp++) {
            if (*f != '%') continue;
            switch (f[1]) {
            case '%':
                *++strp = *++f;
                break;
            case 'Z':
                *strp++ = 'G';
                *strp++ = 'M';
                *strp = 'T';
                f++;
                break;
            case 'z': /* common extension */
                *strp++ = '+';
                *strp++ = '0';
                *strp++ = '0';
                *strp++ = '0';
                *strp = '0';
                f++;
                break;
            }
        }
        *strp = '\0';
        fmt = tf;
    }
    else {
        apr_time_exp_lt(&xt, t);
    }

    /* check return code? */
    apr_strftime(ts, &retcode, MAX_STRING_LEN, fmt, &xt);
    ts[MAX_STRING_LEN - 1] = '\0';
    return apr_pstrdup(p, ts);
}

/* Roy owes Rob beer. */
/* Rob owes Roy dinner. */

/* These legacy comments would make a lot more sense if Roy hadn't
 * replaced the old later_than() routine with util_date.c.
 *
 * Well, okay, they still wouldn't make any sense.
 */

/* Match = 0, NoMatch = 1, Abort = -1
 * Based loosely on sections of wildmat.c by Rich Salz
 * Hmmm... shouldn't this really go component by component?
 */
AP_DECLARE(int) ap_strcmp_match(const char *str, const char *expected)
{
    int x, y;

    for (x = 0, y = 0; expected[y]; ++y, ++x) {
        if ((!str[x]) && (expected[y] != '*'))
            return -1;
        if (expected[y] == '*') {
            while (expected[++y] == '*');
            if (!expected[y])
                return 0;
            while (str[x]) {
                int ret;
                if ((ret = ap_strcmp_match(&str[x++], &expected[y])) != 1)
                    return ret;
            }
            return -1;
        }
        else if ((expected[y] != '?') && (str[x] != expected[y]))
            return 1;
    }
    return (str[x] != '\0');
}

AP_DECLARE(int) ap_strcasecmp_match(const char *str, const char *expected)
{
    int x, y;

    for (x = 0, y = 0; expected[y]; ++y, ++x) {
        if (!str[x] && expected[y] != '*')
            return -1;
        if (expected[y] == '*') {
            while (expected[++y] == '*');
            if (!expected[y])
                return 0;
            while (str[x]) {
                int ret;
                if ((ret = ap_strcasecmp_match(&str[x++], &expected[y])) != 1)
                    return ret;
            }
            return -1;
        }
        else if (expected[y] != '?'
                 && apr_tolower(str[x]) != apr_tolower(expected[y]))
            return 1;
    }
    return (str[x] != '\0');
}

/* We actually compare the canonical root to this root, (but we don't
 * waste time checking the case), since every use of this function in 
 * httpd-2.0 tests if the path is 'proper', meaning we've already passed
 * it through apr_filepath_merge, or we haven't.
 */
AP_DECLARE(int) ap_os_is_path_absolute(apr_pool_t *p, const char *dir) 
{
    const char *newpath;
    const char *ourdir = dir;
    if (apr_filepath_root(&newpath, &dir, 0, p) != APR_SUCCESS
            || strncmp(newpath, ourdir, strlen(newpath)) != 0) {
        return 0;
    }
    return 1;
}

AP_DECLARE(int) ap_is_matchexp(const char *str)
{
    register int x;

    for (x = 0; str[x]; x++)
        if ((str[x] == '*') || (str[x] == '?'))
            return 1;
    return 0;
}

/*
 * Here's a pool-based interface to POSIX regex's regcomp().
 * Note that we return regex_t instead of being passed one.
 * The reason is that if you use an already-used regex_t structure,
 * the memory that you've already allocated gets forgotten, and
 * regfree() doesn't clear it. So we don't allow it.
 */

static apr_status_t regex_cleanup(void *preg)
{
    regfree((regex_t *) preg);
    return APR_SUCCESS;
}

AP_DECLARE(regex_t *) ap_pregcomp(apr_pool_t *p, const char *pattern,
                                   int cflags)
{
    regex_t *preg = apr_palloc(p, sizeof(regex_t));

    if (regcomp(preg, pattern, cflags)) {
        return NULL;
    }

    apr_pool_cleanup_register(p, (void *) preg, regex_cleanup,
                              apr_pool_cleanup_null);

    return preg;
}

AP_DECLARE(void) ap_pregfree(apr_pool_t *p, regex_t * reg)
{
    regfree(reg);
    apr_pool_cleanup_kill(p, (void *) reg, regex_cleanup);
}

/*
 * Similar to standard strstr() but we ignore case in this version.
 * Based on the strstr() implementation further below.
 */
AP_DECLARE(char *) ap_strcasestr(const char *s1, const char *s2)
{
    char *p1, *p2;
    if (*s2 == '\0') {
        /* an empty s2 */
        return((char *)s1);
    }
    while(1) {
        for ( ; (*s1 != '\0') && (apr_tolower(*s1) != apr_tolower(*s2)); s1++);
        if (*s1 == '\0') {
            return(NULL);
        }
        /* found first character of s2, see if the rest matches */
        p1 = (char *)s1;
        p2 = (char *)s2;
        for (++p1, ++p2; apr_tolower(*p1) == apr_tolower(*p2); ++p1, ++p2) {
            if (*p1 == '\0') {
                /* both strings ended together */
                return((char *)s1);
            }
        }
        if (*p2 == '\0') {
            /* second string ended, a match */
            break;
        }
        /* didn't find a match here, try starting at next character in s1 */
        s1++;
    }
    return((char *)s1);
}

/*
 * Returns an offsetted pointer in bigstring immediately after
 * prefix. Returns bigstring if bigstring doesn't start with
 * prefix or if prefix is longer than bigstring while still matching.
 * NOTE: pointer returned is relative to bigstring, so we
 * can use standard pointer comparisons in the calling function
 * (eg: test if ap_stripprefix(a,b) == a)
 */
AP_DECLARE(const char *) ap_stripprefix(const char *bigstring,
                                        const char *prefix)
{
    const char *p1;

    if (*prefix == '\0')
        return bigstring;

    p1 = bigstring;
    while (*p1 && *prefix) {
        if (*p1++ != *prefix++)
            return bigstring;
    }
    if (*prefix == '\0')
        return p1;

    /* hit the end of bigstring! */
    return bigstring;
}

/* 
 * Apache stub function for the regex libraries regexec() to make sure the
 * whole regex(3) API is available through the Apache (exported) namespace.
 * This is especially important for the DSO situations of modules.
 * DO NOT MAKE A MACRO OUT OF THIS FUNCTION!
 */
AP_DECLARE(int) ap_regexec(regex_t *preg, const char *string,
                           size_t nmatch, regmatch_t pmatch[], int eflags)
{
    return regexec(preg, string, nmatch, pmatch, eflags);
}

AP_DECLARE(size_t) ap_regerror(int errcode, const regex_t *preg, char *errbuf,
                               size_t errbuf_size)
{
    return regerror(errcode, preg, errbuf, errbuf_size);
}


/* This function substitutes for $0-$9, filling in regular expression
 * submatches. Pass it the same nmatch and pmatch arguments that you
 * passed ap_regexec(). pmatch should not be greater than the maximum number
 * of subexpressions - i.e. one more than the re_nsub member of regex_t.
 *
 * input should be the string with the $-expressions, source should be the
 * string that was matched against.
 *
 * It returns the substituted string, or NULL on error.
 *
 * Parts of this code are based on Henry Spencer's regsub(), from his
 * AT&T V8 regexp package.
 */

AP_DECLARE(char *) ap_pregsub(apr_pool_t *p, const char *input,
                              const char *source, size_t nmatch,
                              regmatch_t pmatch[])
{
    const char *src = input;
    char *dest, *dst;
    char c;
    size_t no;
    int len;

    if (!source)
        return NULL;
    if (!nmatch)
        return apr_pstrdup(p, src);

    /* First pass, find the size */

    len = 0;

    while ((c = *src++) != '\0') {
        if (c == '&')
            no = 0;
        else if (c == '$' && apr_isdigit(*src))
            no = *src++ - '0';
        else
            no = 10;

        if (no > 9) {                /* Ordinary character. */
            if (c == '\\' && (*src == '$' || *src == '&'))
                c = *src++;
            len++;
        }
        else if (no < nmatch && pmatch[no].rm_so < pmatch[no].rm_eo) {
            len += pmatch[no].rm_eo - pmatch[no].rm_so;
        }

    }

    dest = dst = apr_pcalloc(p, len + 1);

    /* Now actually fill in the string */

    src = input;

    while ((c = *src++) != '\0') {
        if (c == '&')
            no = 0;
        else if (c == '$' && apr_isdigit(*src))
            no = *src++ - '0';
        else
            no = 10;

        if (no > 9) {                /* Ordinary character. */
            if (c == '\\' && (*src == '$' || *src == '&'))
                c = *src++;
            *dst++ = c;
        }
        else if (no < nmatch && pmatch[no].rm_so < pmatch[no].rm_eo) {
            len = pmatch[no].rm_eo - pmatch[no].rm_so;
            memcpy(dst, source + pmatch[no].rm_so, len);
            dst += len;
        }

    }
    *dst = '\0';

    return dest;
}

/*
 * Parse .. so we don't compromise security
 */
AP_DECLARE(void) ap_getparents(char *name)
{
    char *next;
    int l, w, first_dot;

    /* Four paseses, as per RFC 1808 */
    /* a) remove ./ path segments */
    for (next = name; *next && (*next != '.'); next++) {
    }
    
    l = w = first_dot = next - name;
    while (name[l] != '\0') {
        if (name[l] == '.' && IS_SLASH(name[l + 1])
            && (l == 0 || IS_SLASH(name[l - 1])))
            l += 2;
        else
            name[w++] = name[l++];
    }

    /* b) remove trailing . path, segment */
    if (w == 1 && name[0] == '.')
        w--;
    else if (w > 1 && name[w - 1] == '.' && IS_SLASH(name[w - 2]))
        w--;
    name[w] = '\0';

    /* c) remove all xx/../ segments. (including leading ../ and /../) */
    l = first_dot;

    while (name[l] != '\0') {
        if (name[l] == '.' && name[l + 1] == '.' && IS_SLASH(name[l + 2])
            && (l == 0 || IS_SLASH(name[l - 1]))) {
            register int m = l + 3, n;

            l = l - 2;
            if (l >= 0) {
                while (l >= 0 && !IS_SLASH(name[l]))
                    l--;
                l++;
            }
            else
                l = 0;
            n = l;
            while ((name[n] = name[m]))
                (++n, ++m);
        }
        else
            ++l;
    }

    /* d) remove trailing xx/.. segment. */
    if (l == 2 && name[0] == '.' && name[1] == '.')
        name[0] = '\0';
    else if (l > 2 && name[l - 1] == '.' && name[l - 2] == '.'
             && IS_SLASH(name[l - 3])) {
        l = l - 4;
        if (l >= 0) {
            while (l >= 0 && !IS_SLASH(name[l]))
                l--;
            l++;
        }
        else
            l = 0;
        name[l] = '\0';
    }
}

AP_DECLARE(void) ap_no2slash(char *name)
{
    char *d, *s;

    s = d = name;

#ifdef HAVE_UNC_PATHS
    /* Check for UNC names.  Leave leading two slashes. */
    if (s[0] == '/' && s[1] == '/')
        *d++ = *s++;
#endif

    while (*s) {
        if ((*d++ = *s) == '/') {
            do {
                ++s;
            } while (*s == '/');
        }
        else {
            ++s;
        }
    }
    *d = '\0';
}