PHP 原生的similar_text()函数、levenshtein()函数对中文汉字支持不好，我自己写了一个
similar_text()中文汉字版
<span> 1</span> <?<span>php
</span><span> 2</span> <span>//</span><span>拆分字符串 </span>
<span> 3</span> <span>function</span> split_str(<span>$str</span><span>) {
</span><span> 4</span> <span>preg_match_all</span>("/./u", <span>$str</span>, <span>$arr</span><span>);
</span><span> 5</span> <span>return</span> <span>$arr</span>[0<span>];
</span><span> 6</span> <span> }
</span><span> 7</span>
<span> 8</span> <span>//</span><span>相似度检测 </span>
<span> 9</span> <span>function</span> similar_text_cn(<span>$str1</span>, <span>$str2</span><span>) {
</span><span>10</span> <span>$arr_1</span> = <span>array_unique</span>(split_str(<span>$str1</span><span>));
</span><span>11</span> <span>$arr_2</span> = <span>array_unique</span>(split_str(<span>$str2</span><span>));
</span><span>12</span> <span>$similarity</span> = <span>count</span>(<span>$arr_2</span>) - <span>count</span>(<span>array_diff</span>(<span>$arr_2</span>, <span>$arr_1</span><span>));
</span><span>13</span>
<span>14</span> <span>return</span> <span>$similarity</span><span>;
</span><span>15</span> } levenshtein()中文汉字版
<span> 1</span> <?<span>php
</span><span> 2</span> <span>//</span><span>拆分字符串 </span>
<span> 3</span> <span>function</span> mbStringToArray(<span>$string</span>, <span>$encoding</span> = 'UTF-8'<span>) {
</span><span> 4</span> <span>$arrayResult</span> = <span>array</span><span>();
</span><span> 5</span>
<span> 6</span> <span>while</span> (<span>$iLen</span> = mb_strlen(<span>$string</span>, <span>$encoding</span><span>)) {
</span><span> 7</span> <span>array_push</span>(<span>$arrayResult</span>, mb_substr(<span>$string</span>, 0, 1, <span>$encoding</span><span>));
</span><span> 8</span> <span>$string</span> = mb_substr(<span>$string</span>, 1, <span>$iLen</span>, <span>$encoding</span><span>);
</span><span> 9</span> <span> }
</span><span>10</span>
<span>11</span> <span>return</span> <span>$arrayResult</span><span>;
</span><span>12</span> <span> }
</span><span>13</span>
<span>14</span> <span>//</span><span>编辑距离 </span>
<span>15</span> <span>function</span> levenshtein_cn(<span>$str1</span>, <span>$str2</span>, <span>$costReplace</span> = 1, <span>$encoding</span> = 'UTF-8'<span>) {
</span><span>16</span> <span>$count_same_letter</span> = 0<span>;
</span><span>17</span> <span>$d</span> = <span>array</span><span>();
</span><span>18</span>
<span>19</span> <span>$mb_len1</span> = mb_strlen(<span>$str1</span>, <span>$encoding</span><span>);
</span><span>20</span> <span>$mb_len2</span> = mb_strlen(<span>$str2</span>, <span>$encoding</span><span>);
</span><span>21</span>
<span>22</span> <span>$mb_str1</span> = mbStringToArray(<span>$str1</span>, <span>$encoding</span><span>);
</span><span>23</span> <span>$mb_str2</span> = mbStringToArray(<span>$str2</span>, <span>$encoding</span><span>);
</span><span>24</span>
<span>25</span> <span>for</span> (<span>$i1</span> = 0; <span>$i1</span> <= <span>$mb_len1</span>; <span>$i1</span>++<span>) {
</span><span>26</span> <span>$d</span>[<span>$i1</span>] = <span>array</span><span>();
</span><span>27</span> <span>$d</span>[<span>$i1</span>][0] = <span>$i1</span><span>;
</span><span>28</span> <span> }
</span><span>29</span>
<span>30</span> <span>for</span> (<span>$i2</span> = 0; <span>$i2</span> <= <span>$mb_len2</span>; <span>$i2</span>++<span>) {
</span><span>31</span> <span>$d</span>[0][<span>$i2</span>] = <span>$i2</span><span>;
</span><span>32</span> <span> }
</span><span>33</span>
<span>34</span> <span>for</span> (<span>$i1</span> = 1; <span>$i1</span> <= <span>$mb_len1</span>; <span>$i1</span>++<span>) {
</span><span>35</span> <span>for</span> (<span>$i2</span> = 1; <span>$i2</span> <= <span>$mb_len2</span>; <span>$i2</span>++<span>) {
</span><span>36</span> <span>//</span><span> $cost = ($str1[$i1 - 1] == $str2[$i2 - 1]) ? 0 : 1; </span>
<span>37</span> <span>if</span> (<span>$mb_str1</span>[<span>$i1</span> - 1] === <span>$mb_str2</span>[<span>$i2</span> - 1<span>]) {
</span><span>38</span> <span>$cost</span> = 0<span>;
</span><span>39</span> <span>$count_same_letter</span>++<span>;
</span><span>40</span> } <span>else</span><span> {
</span><span>41</span> <span>$cost</span> = <span>$costReplace</span>; <span>//</span><span>替换 </span>
<span>42</span> <span> }
</span><span>43</span>
<span>44</span> <span>$d</span>[<span>$i1</span>][<span>$i2</span>] = <span>min</span>(<span>$d</span>[<span>$i1</span> - 1][<span>$i2</span>] + 1, <span>//</span><span>插入 </span>
<span>45</span> <span>$d</span>[<span>$i1</span>][<span>$i2</span> - 1] + 1, <span>//</span><span>删除 </span>
<span>46</span> <span>$d</span>[<span>$i1</span> - 1][<span>$i2</span> - 1] + <span>$cost</span><span>);
</span><span>47</span> <span> }
</span><span>48</span> <span> }
</span><span>49</span>
<span>50</span> <span>return</span> <span>$d</span>[<span>$mb_len1</span>][<span>$mb_len2</span><span>];
</span><span>51</span> <span>//</span><span>return array('distance' => $d[$mb_len1][$mb_len2], 'count_same_letter' => $count_same_letter); </span>
<span>52</span> } 最长公共子序列LCS()
<span> 1</span> <?<span>php
</span><span> 2</span> <span>//</span><span>最长公共子序列英文版 </span>
<span> 3</span> <span>function</span> LCS_en(<span>$str_1</span>, <span>$str_2</span><span>) {
</span><span> 4</span> <span>$len_1</span> = <span>strlen</span>(<span>$str_1</span><span>);
</span><span> 5</span> <span>$len_2</span> = <span>strlen</span>(<span>$str_2</span><span>);
</span><span> 6</span> <span>$len</span> = <span>$len_1</span> > <span>$len_2</span> ? <span>$len_1</span> : <span>$len_2</span><span>;
</span><span> 7</span>
<span> 8</span> <span>$dp</span> = <span>array</span><span>();
</span><span> 9</span> <span>for</span> (<span>$i</span> = 0; <span>$i</span> <= <span>$len</span>; <span>$i</span>++<span>) {
</span><span>10</span> <span>$dp</span>[<span>$i</span>] = <span>array</span><span>();
</span><span>11</span> <span>$dp</span>[<span>$i</span>][0] = 0<span>;
</span><span>12</span> <span>$dp</span>[0][<span>$i</span>] = 0<span>;
</span><span>13</span> <span> }
</span><span>14</span>
<span>15</span> <span>for</span> (<span>$i</span> = 1; <span>$i</span> <= <span>$len_1</span>; <span>$i</span>++<span>) {
</span><span>16</span> <span>for</span> (<span>$j</span> = 1; <span>$j</span> <= <span>$len_2</span>; <span>$j</span>++<span>) {
</span><span>17</span> <span>if</span> (<span>$str_1</span>[<span>$i</span> - 1] == <span>$str_2</span>[<span>$j</span> - 1<span>]) {
</span><span>18</span> <span>$dp</span>[<span>$i</span>][<span>$j</span>] = <span>$dp</span>[<span>$i</span> - 1][<span>$j</span> - 1] + 1<span>;
</span><span>19</span> } <span>else</span><span> {
</span><span>20</span> <span>$dp</span>[<span>$i</span>][<span>$j</span>] = <span>$dp</span>[<span>$i</span> - 1][<span>$j</span>] > <span>$dp</span>[<span>$i</span>][<span>$j</span> - 1] ? <span>$dp</span>[<span>$i</span> - 1][<span>$j</span>] : <span>$dp</span>[<span>$i</span>][<span>$j</span> - 1<span>];
</span><span>21</span> <span> }
</span><span>22</span> <span> }
</span><span>23</span> <span> }
</span><span>24</span>
<span>25</span> <span>return</span> <span>$dp</span>[<span>$len_1</span>][<span>$len_2</span><span>];
</span><span>26</span> <span> }
</span><span>27</span>
<span>28</span> <span>//</span><span>拆分字符串 </span>
<span>29</span> <span>function</span> mbStringToArray(<span>$string</span>, <span>$encoding</span> = 'UTF-8'<span>) {
</span><span>30</span> <span>$arrayResult</span> = <span>array</span><span>();
</span><span>31</span>
<span>32</span> <span>while</span> (<span>$iLen</span> = mb_strlen(<span>$string</span>, <span>$encoding</span><span>)) {
</span><span>33</span> <span>array_push</span>(<span>$arrayResult</span>, mb_substr(<span>$string</span>, 0, 1, <span>$encoding</span><span>));
</span><span>34</span> <span>$string</span> = mb_substr(<span>$string</span>, 1, <span>$iLen</span>, <span>$encoding</span><span>);
</span><span>35</span> <span> }
</span><span>36</span>
<span>37</span> <span>return</span> <span>$arrayResult</span><span>;
</span><span>38</span> <span> }
</span><span>39</span>
<span>40</span> <span>//</span><span>最长公共子序列中文版 </span>
<span>41</span> <span>function</span> LCS_cn(<span>$str1</span>, <span>$str2</span>, <span>$encoding</span> = 'UTF-8'<span>) {
</span><span>42</span> <span>$mb_len1</span> = mb_strlen(<span>$str1</span>, <span>$encoding</span><span>);
</span><span>43</span> <span>$mb_len2</span> = mb_strlen(<span>$str2</span>, <span>$encoding</span><span>);
</span><span>44</span>
<span>45</span> <span>$mb_str1</span> = mbStringToArray(<span>$str1</span>, <span>$encoding</span><span>);
</span><span>46</span> <span>$mb_str2</span> = mbStringToArray(<span>$str2</span>, <span>$encoding</span><span>);
</span><span>47</span>
<span>48</span> <span>$len</span> = <span>$mb_len1</span> > <span>$mb_len2</span> ? <span>$mb_len1</span> : <span>$mb_len2</span><span>;
</span><span>49</span>
<span>50</span> <span>$dp</span> = <span>array</span><span>();
</span><span>51</span> <span>for</span> (<span>$i</span> = 0; <span>$i</span> <= <span>$len</span>; <span>$i</span>++<span>) {
</span><span>52</span> <span>$dp</span>[<span>$i</span>] = <span>array</span><span>();
</span><span>53</span> <span>$dp</span>[<span>$i</span>][0] = 0<span>;
</span><span>54</span> <span>$dp</span>[0][<span>$i</span>] = 0<span>;
</span><span>55</span> <span> }
</span><span>56</span>
<span>57</span> <span>for</span> (<span>$i</span> = 1; <span>$i</span> <= <span>$mb_len1</span>; <span>$i</span>++<span>) {
</span><span>58</span> <span>for</span> (<span>$j</span> = 1; <span>$j</span> <= <span>$mb_len2</span>; <span>$j</span>++<span>) {
</span><span>59</span> <span>if</span> (<span>$mb_str1</span>[<span>$i</span> - 1] == <span>$mb_str2</span>[<span>$j</span> - 1<span>]) {
</span><span>60</span> <span>$dp</span>[<span>$i</span>][<span>$j</span>] = <span>$dp</span>[<span>$i</span> - 1][<span>$j</span> - 1] + 1<span>;
</span><span>61</span> } <span>else</span><span> {
</span><span>62</span> <span>$dp</span>[<span>$i</span>][<span>$j</span>] = <span>$dp</span>[<span>$i</span> - 1][<span>$j</span>] > <span>$dp</span>[<span>$i</span>][<span>$j</span> - 1] ? <span>$dp</span>[<span>$i</span> - 1][<span>$j</span>] : <span>$dp</span>[<span>$i</span>][<span>$j</span> - 1<span>];
</span><span>63</span> <span> }
</span><span>64</span> <span> }
</span><span>65</span> <span> }
</span><span>66</span>
<span>67</span> <span>return</span> <span>$dp</span>[<span>$mb_len1</span>][<span>$mb_len2</span><span>];
</span><span>68</span> }
