开扒php内核函数，第三篇 implode
- WBOYOriginal
- 2016-06-06 19:42:251035Durchsuche
一开始觉得implode挺容易实现，但是写着写着才发现是挺复杂的，不说啦 来看看implode的用法吧 1 ? php 2 $arr = array ('Hello','World!','Beautiful','Day!' ); 3 echo implode (" ", $arr ); 4 ? 上面会输出 Hello World! Beautiful Day! 下面的程序的我写
一开始觉得implode挺容易实现，但是写着写着才发现是挺复杂的，不说啦
来看看implode的用法吧
<span>1</span> <span>php
</span><span>2</span> <span>$arr</span> = <span>array</span>('Hello','World!','Beautiful','Day!'<span>);
</span><span>3</span> <span>echo</span> <span>implode</span>(" ",<span>$arr</span><span>);
</span><span>4</span> ?>
上面会输出 Hello World! Beautiful Day!
下面的程序的我写的
<span> 1</span> <span>/*</span><span>字符串翻转</span><span>*/</span>
<span> 2</span> <span>void</span> strover(<span>char</span> *<span> str){
</span><span> 3</span> <span>int</span> len =<span> strlen(str);
</span><span> 4</span> <span>//</span><span>int half = strlen(str)/2;</span>
<span> 5</span> <span>int</span><span> i,j;
</span><span> 6</span> <span>char</span><span> tmp;
</span><span> 7</span> j = len-<span>1</span><span>;
</span><span> 8</span> <span>for</span>(i=<span>0</span>;i){
<span> 9</span> tmp =<span> str[j];
</span><span>10</span> str[j] =<span> str[i];
</span><span>11</span> str[i] =<span> tmp;
</span><span>12</span> j--<span>;
</span><span>13</span> <span> }
</span><span>14</span>
<span>15</span>
<span>16</span> <span>}
</span><span>17</span>
<span>18</span>
<span>19</span>
<span>20</span>
<span>21</span>
<span>22</span> <span>/*</span>
<span>23</span> <span> 2进制转十进制 要处理正负数啊 涉及到负数啊
</span><span>24</span> <span> 字符串翻转
</span><span>25</span> <span>*/</span>
<span>26</span> <span>char</span> * bin2decimal(<span>int</span><span> number){
</span><span>27</span>
<span>28</span> <span>int</span> q = <span>0</span>; <span>//</span><span>商</span>
<span>29</span> <span>int</span> r = <span>0</span>;<span>//</span><span>余数</span>
<span>30</span> <span>int</span> i = <span>0</span><span>;
</span><span>31</span> <span>int</span> tmp =<span> number;
</span><span>32</span> <span>int</span> is_negative = <span>0</span><span>;
</span><span>33</span> <span>char</span> *<span> res;
</span><span>34</span> res = (<span>char</span> *)malloc(<span>sizeof</span>(<span>char</span>)*<span>5</span>+<span>1</span><span>);
</span><span>35</span> <span>if</span>(number>=<span>0</span><span>){
</span><span>36</span>
<span>37</span> }<span>else</span><span>{
</span><span>38</span> tmp = -<span>number;
</span><span>39</span> res[i++] = <span>'</span><span>-</span><span>'</span><span>;
</span><span>40</span> is_negative = <span>1</span><span>;
</span><span>41</span> <span> }
</span><span>42</span>
<span>43</span> <span>do</span><span>{
</span><span>44</span> q = tmp/<span>10</span><span>;
</span><span>45</span>
<span>46</span> r = tmp%<span>10</span><span>;
</span><span>47</span> <span>//</span><span> tmp = q;
</span><span>48</span> <span>//</span><span> c = hex_str[r];</span>
<span>49</span> res[i++] = <span>'</span><span>0</span><span>'</span>+<span>r;
</span><span>50</span> tmp =<span> q;
</span><span>51</span> }<span>while</span><span>(tmp);
</span><span>52</span>
<span>53</span> res[i] = <span>'</span><span>\0</span><span>'</span><span>;
</span><span>54</span>
<span>55</span> strover(&<span>res[is_negative]);
</span><span>56</span> <span>return</span><span> res;
</span><span>57</span>
<span>58</span>
<span>59</span> <span>}
</span><span>60</span>
<span>61</span>
<span>62</span>
<span>63</span>
<span>64</span> <span>/*</span>
<span>65</span> <span> c语言真的太麻烦啦，传数组，但是无法知道数组的长度，只能够手动传入
</span><span>66</span> <span>*/</span>
<span>67</span> <span>char</span> * implode(<span>int</span> *number,<span>int</span> size,<span>char</span> *<span> dem){
</span><span>68</span> <span>int</span> i = <span>0</span><span>;
</span><span>69</span> <span>char</span>*<span> c;
</span><span>70</span> <span>//</span><span>c[1] = '\0';</span>
<span>71</span> <span>struct</span><span> simple_mem{
</span><span>72</span> <span>char</span> *<span> res;
</span><span>73</span> unsigned <span>int</span><span> len;
</span><span>74</span> unsigned <span>int</span><span> used;
</span><span>75</span> <span> }test_mem;
</span><span>76</span> test_mem.res = (<span>char</span> *)malloc(<span>sizeof</span>(<span>char</span>)*<span>20</span><span>);
</span><span>77</span> test_mem.len = <span>sizeof</span>(<span>char</span>)*<span>20</span><span>;
</span><span>78</span> test_mem.used = <span>0</span><span>;
</span><span>79</span> <span>for</span>(;isize;){
<span>80</span> c=<span> bin2decimal(number[i]);
</span><span>81</span> memcpy(test_mem.res+<span>test_mem.used,c,strlen(c));
</span><span>82</span>
<span>83</span> test_mem.used+=<span>strlen(c);
</span><span>84</span> <span>if</span>(++isize){
<span>85</span> memcpy(test_mem.res+<span>test_mem.used,dem,strlen(dem));
</span><span>86</span> test_mem.used+=<span>strlen(dem);
</span><span>87</span> <span> }
</span><span>88</span>
<span>89</span> <span> }
</span><span>90</span> test_mem.res[test_mem.used] = <span>'</span><span>\0</span><span>'</span><span>;
</span><span>91</span> printf(<span>"</span><span>%s</span><span>"</span><span>,test_mem.res);
</span><span>92</span>
<span>93</span>
<span>94</span> }
我们写的implode写的函数是针对整形数组，php的当然什么类型都支持啊，c语言也可以实现泛型，但毕竟比较麻烦的，上面的程序还是比较多问题的，优化的地方有很多，但是我们是抱着学习的态度来的
<span> 1</span> <span>int</span><span> main(){
</span><span> 2</span> <span>//</span><span>char * res = bin2hex("a");
</span><span> 3</span> <span>//</span><span>printf("hex a=%s",res);
</span><span> 4</span> <span>//</span><span>char * res = hex2bin("6578616d706c65206865782064617461");</span>
<span> 5</span> <span>int</span> integer[<span>3</span>] = {<span>1</span>,-<span>24</span>,<span>3</span><span>};
</span><span> 6</span> implode(integer,<span>sizeof</span>(integer)/<span>sizeof</span>(<span>int</span>),<span>"</span><span>*</span><span>"</span><span>);
</span><span> 7</span>
<span> 8</span> <span>//</span><span>bin2decimal(-1234);</span>
<span> 9</span>
<span>10</span> <span>return</span> <span>0</span><span>;
</span><span>11</span> }
先说说思路吧
1,主要是算法是2进制转10进制 字符串显示，当然我们要注意负数啦，还有字符串翻转
2 内存分配，因为我们没有限制数组的长度，所以我们要动态去分配，其实我们可以有一样可以确定的是整形的范围 0到65535 就是说一个整形最多占5个字符，
3 其他就没什么啦
来看看php的吧
<span> 1</span> <span>/*</span>
<span> 2</span> <span> * Convert num to its decimal format.
</span><span> 3</span> <span> * Return value:
</span><span> 4</span> <span> * - a pointer to a string containing the number (no sign)
</span><span> 5</span> <span> * - len contains the length of the string
</span><span> 6</span> <span> * - is_negative is set to TRUE or FALSE depending on the sign
</span><span> 7</span> <span> * of the number (always set to FALSE if is_unsigned is TRUE)
</span><span> 8</span> <span> *
</span><span> 9</span> <span> * The caller provides a buffer for the string: that is the buf_end argument
</span><span>10</span> <span> * which is a pointer to the END of the buffer + 1 (i.e. if the buffer
</span><span>11</span> <span> * is declared as buf[ 100 ], buf_end should be &buf[ 100 ])
</span><span>12</span> <span>*/</span>
<span>13</span> <span>/*</span><span> char * ap_php_conv_10() {{{ </span><span>*/</span>
<span>14</span> <span>char</span> *<span> ap_php_conv_10(register wide_int num, register bool_int is_unsigned,
</span><span>15</span> register bool_int * is_negative, <span>char</span> *buf_end, register <span>int</span> *<span>len)
</span><span>16</span> <span>{
</span><span>17</span> register <span>char</span> *p =<span> buf_end;
</span><span>18</span> <span> register u_wide_int magnitude;
</span><span>19</span>
<span>20</span> <span>if</span><span> (is_unsigned) {
</span><span>21</span> magnitude =<span> (u_wide_int) num;
</span><span>22</span> *is_negative =<span> FALSE;
</span><span>23</span> } <span>else</span><span> {
</span><span>24</span> *is_negative = (num 0<span>);
</span><span>25</span>
<span>26</span> <span>/*</span>
<span>27</span> <span> * On a 2's complement machine, negating the most negative integer
</span><span>28</span> <span> * results in a number that cannot be represented as a signed integer.
</span><span>29</span> <span> * Here is what we do to obtain the number's magnitude:
</span><span>30</span> <span> * a. add 1 to the number
</span><span>31</span> <span> * b. negate it (becomes positive)
</span><span>32</span> <span> * c. convert it to unsigned
</span><span>33</span> <span> * d. add 1
</span><span>34</span> <span>*/</span>
<span>35</span> <span>if</span> (*<span>is_negative) {
</span><span>36</span> wide_int t = num + <span>1</span><span>;
</span><span>37</span> magnitude = ((u_wide_int) - t) + <span>1</span><span>;
</span><span>38</span> } <span>else</span><span> {
</span><span>39</span> magnitude =<span> (u_wide_int) num;
</span><span>40</span> <span> }
</span><span>41</span> <span> }
</span><span>42</span>
<span>43</span> <span>/*</span>
<span>44</span> <span> * We use a do-while loop so that we write at least 1 digit
</span><span>45</span> <span>*/</span>
<span>46</span> <span>do</span><span> {
</span><span>47</span> register u_wide_int new_magnitude = magnitude / <span>10</span><span>;
</span><span>48</span>
<span>49</span> *--p = (<span>char</span>)(magnitude - new_magnitude * <span>10</span> + <span>'</span><span>0</span><span>'</span><span>);
</span><span>50</span> magnitude =<span> new_magnitude;
</span><span>51</span> <span> }
</span><span>52</span> <span>while</span><span> (magnitude);
</span><span>53</span>
<span>54</span> *len = buf_end -<span> p;
</span><span>55</span> <span>return</span><span> (p);
</span><span>56</span> }
> php5ts_debug.dll!ap_php_conv_10(__int64 num=-278, int is_unsigned=0, int * is_negative=0x00c3e154, char * buf_end=0x00c3e9c0, int * len=0x00c3ea64) 行320 C
php5ts_debug.dll!format_converter(buf_area * odp=0x00c3eb9c, const char * fmt=0x105d799e, char * ap=0x00c3ecc0) 行869 + 0x34 字节 C
php5ts_debug.dll!strx_printv(int * ccp=0x00c3eca0, char * buf=0x00c3ee90, unsigned int len=12, const char * format=0x105d799c, char * ap=0x00c3ecbc) 行1213 + 0x11 字节 C
php5ts_debug.dll!ap_php_slprintf(char * buf=0x00c3ee90, unsigned int len=12, const char * format=0x105d799c, ...) 行1229 + 0x19 字节 C
php5ts_debug.dll!php_implode(_zval_struct * delim=0x030dffd8, _zval_struct * arr=0x030dff88, _zval_struct * return_value=0x030e0028, void * * * tsrm_ls=0x00353040) 行1154 + 0x1b 字节 C
php5ts_debug.dll!zif_implode(int ht=2, _zval_struct * return_value=0x030e0028, _zval_struct * * return_value_ptr=0x00000000, _zval_struct * this_ptr=0x00000000, int return_value_used=1, void * * * tsrm_ls=0x00353040) 行1250 + 0x15 字节 C
php5ts_debug.dll!zend_do_fcall_common_helper_SPEC(_zend_execute_data * execute_data=0x030c20d8, void * * * tsrm_ls=0x00353040) 行643 + 0x62 字节 C
php5ts_debug.dll!ZEND_DO_FCALL_SPEC_CONST_HANDLER(_zend_execute_data * execute_data=0x030c20d8, void * * * tsrm_ls=0x00353040) 行2234 C
php5ts_debug.dll!execute(_zend_op_array * op_array=0x030dfa40, void * * * tsrm_ls=0x00353040) 行410 + 0x11 字节 C
php5ts_debug.dll!zend_execute_scripts(int type=8, void * * * tsrm_ls=0x00353040, _zval_struct * * retval=0x00000000, int file_count=3, ...) 行1329 + 0x21 字节 C
php5ts_debug.dll!php_execute_script(_zend_file_handle * primary_file=0x00c3fcf4, void * * * tsrm_ls=0x00353040) 行2502 + 0x1b 字节 C
php.exe!do_cli(int argc=2, char * * argv=0x00352fa0, void * * * tsrm_ls=0x00353040) 行989 + 0x10 字节 C
php.exe!main(int argc=2, char * * argv=0x00352fa0) 行1365 + 0x11 字节 C
调用堆栈如上
<span>1</span> <span>do</span><span> {
</span><span>2</span> register u_wide_int new_magnitude = magnitude / <span>10</span><span>;
</span><span>3</span>
<span>4</span> *--p = (<span>char</span>)(magnitude - new_magnitude * <span>10</span> + <span>'</span><span>0</span><span>'</span><span>);
</span><span>5</span> magnitude =<span> new_magnitude;
</span><span>6</span> <span> }
</span><span>7</span> <span>while</span> (magnitude);
关键是这段代码，作者没有像我们 用取余去计算，而是 把它乘，举个例子吧
magnitude = 283
new_magnitude = 283/10 = 28
*--p = 283 - 28*10+'0' = '3'
magnitude = new_magnitude = 28
然后继续上面的步骤啦
取余考虑和乘法考虑那个高，不知道作者的想法是怎样的，有时间用汇编证明一下，那个用的指令比较多
第二个的就是 作者用了倒序字符复制 *--p,这就要读读内存的代码啦
<strong>*len = buf_end -</strong><span><strong> p; 字符串长度可以这样计算的，指针的作用就是不错啊<br></strong><br>回到调用的地方如下<br><br></span>
<span> 1</span> s = ap_php_conv_10(i_num, (*fmt) == <span>'</span><span>u</span><span>'</span>, &<span>is_negative,
</span><span> 2</span> &num_buf[NUM_BUF_SIZE], &<span>s_len);
</span><span> 3</span> <span> FIX_PRECISION(adjust_precision, precision, s, s_len);
</span><span> 4</span>
<span> 5</span> <span>if</span> (*fmt != <span>'</span><span>u</span><span>'</span><span>) {
</span><span> 6</span> <span>if</span><span> (is_negative) {
</span><span> 7</span> prefix_char = <span>'</span><span>-</span><span>'</span><span>;
</span><span> 8</span> } <span>else</span> <span>if</span><span> (print_sign) {
</span><span> 9</span> prefix_char = <span>'</span><span>+</span><span>'</span><span>;
</span><span>10</span> } <span>else</span> <span>if</span><span> (print_blank) {
</span><span>11</span> prefix_char = <span>'</span> <span>'</span><span>;
</span><span>12</span> <span> }
</span><span>13</span> <span> }
</span><span>14</span> <span>break</span>;
num_buf[NUM_BUF_SIZE] 这个东西长度为2048，不知道为什么要分配这么多的内存<br><br>上面的判断就是看看是不是负数，然后就 赋给修饰符<br><br>
if (prefix_char != NUL) {
*--s = prefix_char;
s_len++;
}
应该很容易吧
下面来看下一层的调用
<span> 1</span> <span>while</span> (zend_hash_get_current_data_ex(Z_ARRVAL_P(arr), (<span>void</span> **) &tmp, &pos) ==<span> SUCCESS) {
</span><span> 2</span> <span>switch</span> ((*tmp)-><span>type) {
</span><span> 3</span> <span>case</span><span> IS_STRING:
</span><span> 4</span> smart_str_appendl(&<span>implstr, Z_STRVAL_PP(tmp), Z_STRLEN_PP(tmp));
</span><span> 5</span> <span>break</span><span>;
</span><span> 6</span>
<span> 7</span> <span>case</span><span> IS_LONG: {
</span><span> 8</span> <span>char</span> stmp[MAX_LENGTH_OF_LONG + <span>1</span><span>];
</span><span> 9</span> str_len = slprintf(stmp, <span>sizeof</span>(stmp), <span>"</span><span>%ld</span><span>"</span><span>, Z_LVAL_PP(tmp));
</span><span>10</span> smart_str_appendl(&<span>implstr, stmp, str_len);
</span><span>11</span> <span> }
</span><span>12</span> <span>break</span><span>;
</span><span>13</span>
<span>14</span> <span>case</span><span> IS_BOOL:
</span><span>15</span> <span>if</span> (Z_LVAL_PP(tmp) == <span>1</span><span>) {
</span><span>16</span> smart_str_appendl(&implstr, <span>"</span><span>1</span><span>"</span>, <span>sizeof</span>(<span>"</span><span>1</span><span>"</span>)-<span>1</span><span>);
</span><span>17</span> <span> }
</span><span>18</span> <span>break</span><span>;
</span><span>19</span>
<span>20</span> <span>case</span><span> IS_NULL:
</span><span>21</span> <span>break</span><span>;
</span><span>22</span>
<span>23</span> <span>case</span><span> IS_DOUBLE: {
</span><span>24</span> <span>char</span> *<span>stmp;
</span><span>25</span> str_len = spprintf(&stmp, <span>0</span>, <span>"</span><span>%.*G</span><span>"</span>, (<span>int</span><span>) EG(precision), Z_DVAL_PP(tmp));
</span><span>26</span> smart_str_appendl(&<span>implstr, stmp, str_len);
</span><span>27</span> <span> efree(stmp);
</span><span>28</span> <span> }
</span><span>29</span> <span>break</span><span>;
</span><span>30</span>
<span>31</span> <span>case</span><span> IS_OBJECT: {
</span><span>32</span> <span>int</span><span> copy;
</span><span>33</span> <span> zval expr;
</span><span>34</span> zend_make_printable_zval(*tmp, &expr, &<span>copy);
</span><span>35</span> smart_str_appendl(&<span>implstr, Z_STRVAL(expr), Z_STRLEN(expr));
</span><span>36</span> <span>if</span><span> (copy) {
</span><span>37</span> zval_dtor(&<span>expr);
</span><span>38</span> <span> }
</span><span>39</span> <span> }
</span><span>40</span> <span>break</span><span>;
</span><span>41</span>
<span>42</span> <span>default</span><span>:
</span><span>43</span> tmp_val = **<span>tmp;
</span><span>44</span> zval_copy_ctor(&<span>tmp_val);
</span><span>45</span> convert_to_string(&<span>tmp_val);
</span><span>46</span> smart_str_appendl(&<span>implstr, Z_STRVAL(tmp_val), Z_STRLEN(tmp_val));
</span><span>47</span> zval_dtor(&<span>tmp_val);
</span><span>48</span> <span>break</span><span>;
</span><span>49</span>
<span>50</span> }
我们是在这段代码
case IS_LONG: {
char stmp[MAX_LENGTH_OF_LONG + 1];
str_len = slprintf(stmp, sizeof(stmp), "%ld", Z_LVAL_PP(tmp));
smart_str_appendl(&implstr, stmp, str_len);
}
+ &implstr 0x00c3ef04 {c=0x030e0100 "1.5-" len=4 a=78 } smart_str *
+ stmp 0x00c3ee90 "-278" char [12]
str_len 4 int
+ tmp 0x030e0924 _zval_struct * *
php数字默认类型是长整形的，从上面可知道，stmp="-278",strlen = 4,
我们接下来看看implstr是这样处理的，首先他的结构是这样的
<span>1</span> typedef <span>struct</span><span> {
</span><span>2</span> <span>char</span> *<span>c; 指向一段内存
</span><span>3</span> <span> size_t len; 已经用了多小
</span><span>4</span> <span> size_t a; 总共有多小
</span><span>5</span> } smart_str;
smart_str_appendl 的定义是这样的<br><br>
<span>1</span> <span>#define</span> smart_str_appendl_ex(dest, src, nlen, what) do { \
<span>2</span> <span> register size_t __nl; \
</span><span>3</span> smart_str *__dest = (smart_str *<span>) (dest); \
</span><span>4</span> <span> \
</span><span>5</span> <span> smart_str_alloc4(__dest, (nlen), (what), __nl); \
</span><span>6</span> memcpy(__dest->c + __dest-><span>len, (src), (nlen)); \
</span><span>7</span> __dest->len =<span> __nl; \
</span><span>8</span> } <span>while</span> (<span>0</span>)
复制字符串用了memcpy
smart_str_alloc4这个定义如下
<span> 1</span> <span>#define</span> smart_str_alloc4(d, n, what, newlen) do { \
<span> 2</span> <span>if</span> (!(d)-><span>c) { \
</span><span> 3</span> (d)->len = <span>0</span><span>; \
</span><span> 4</span> newlen =<span> (n); \
</span><span> 5</span> (d)->a = newlen SMART_STR_START_SIZE \
<span> 6</span> ?<span> SMART_STR_START_SIZE \
</span><span> 7</span> : newlen +<span> SMART_STR_PREALLOC; \
</span><span> 8</span> <span> SMART_STR_DO_REALLOC(d, what); \
</span><span> 9</span> } <span>else</span><span> { \
</span><span>10</span> newlen = (d)->len +<span> (n); \
</span><span>11</span> <span>if</span> (newlen >= (d)-><span>a) { \
</span><span>12</span> (d)->a = newlen +<span> SMART_STR_PREALLOC; \
</span><span>13</span> <span> SMART_STR_DO_REALLOC(d, what); \
</span><span>14</span> <span> } \
</span><span>15</span> <span> } \
</span><span>16</span> } <span>while</span> (<span>0</span>)
这个很清楚啦流程啦
如果implstr 没有分配过的，那么闲分配一段内存
如果implstr分配过，并且当前的空间不够容纳新的字符 在这基础上扩展啦 SMART_STR_PREALLOC =78 不知道为什么是78
看下定义吧
1 #define SMART_STR_DO_REALLOC(d, what) \
2 (d)->c = SMART_STR_REALLOC((d)->c, (d)->a + 1, (what))
1 #define SMART_STR_REALLOC(a,b,c) perealloc((a),(b),(c))
1 #define perealloc(ptr, size, persistent) ((persistent)?__zend_realloc((ptr), (size)):erealloc((ptr), (size)))
<span>1</span> inline <span>static</span> <span>void</span> * __zend_realloc(<span>void</span> *<span>p, size_t len)
</span><span>2</span> <span>{
</span><span>3</span> p =<span> realloc(p, len);
</span><span>4</span> <span>if</span><span> (p) {
</span><span>5</span> <span>return</span><span> p;
</span><span>6</span> <span> }
</span><span>7</span> fprintf(stderr, <span>"</span><span>Out of memory\n</span><span>"</span><span>);
</span><span>8</span> exit(<span>1</span><span>);
</span><span>9</span> }
最终是调用了 c语言的realloc函数，这样就大概明白了吧
到最后加上分割符号
1 if (++i != numelems) {
2 smart_str_appendl(&implstr, Z_STRVAL_P(delim), Z_STRLEN_P(delim));
3 }
上面是分析了整数的implode，至于浮点数，对象，字符串 的implode大家可以用上面的方法去研究下
Stellungnahme：
Der Inhalt dieses Artikels wird freiwillig von Internetnutzern beigesteuert und das Urheberrecht liegt beim ursprünglichen Autor. Diese Website übernimmt keine entsprechende rechtliche Verantwortung. Wenn Sie Inhalte finden, bei denen der Verdacht eines Plagiats oder einer Rechtsverletzung besteht, wenden Sie sich bitte an admin@php.cn
Vorheriger Artikel：php全局变量和正则表达式错误解决方案Nächster Artikel：php上传文件代码