Oracle的varcha2和MySQL的varchar是有区别的,Oracle放的是字节,所以如果Oracle的编码集是GBK,可以用如下语句查看,那么,varcha2(100)就只能存放50个汉字,但是MySQL的却是100个,因为Mysql是按照字符来算的。
select userenv('language') from dual;
也因为这样,我们有时候会发现,当数据长度超过的时候会报长度超过限制的异常,此时我们可能会直接在java中用字符串截取的方式来截取。但是这样子要注意,我们必须按字节来截取,当然是Oracle的情况下,可以用如下方法来操作(java其实是不需要这样子处理的,直接扩展字段即可,但是有些业务场景扩展不了就只能截取了):
/*** 根据指定的字节长度截取字符串* @param orignal : 原始字符串* @param count : 要截取的字节长度* @return*/public static String substringByByteLength(String orignal, int count){if(orignal != null && !"".equals(orignal)){int tmp = count;try {orignal = new String(orignal.getBytes("GBK"),"GBK");System.out.println(orignal.getBytes("GBK").length);if(tmp > 0 && tmp < orignal.getBytes("GBK").length){System.out.println("截取前的值:"+orignal);StringBuffer buff = new StringBuffer();char c;for(int i = 0; i < tmp; i++){c = orignal.charAt(i);if((buff.toString() + String.valueOf(c)).getBytes("GBK").length <= count ){buff.append(c);}if(isChinese(c)){--tmp; //遇到中文汉字,截取字节总数减1}}String str = buff.toString();System.out.println("截取后的值:"+str);return str;}return orignal;} catch (UnsupportedEncodingException e) {e.printStackTrace();}}return "";}public static boolean isChinese(char ch) {//获取此字符的UniCodeBlockCharacter.UnicodeBlock ub = Character.UnicodeBlock.of(ch);// GENERAL_PUNCTUATION 判断中文的“号// CJK_SYMBOLS_AND_PUNCTUATION 判断中文的。号// HALFWIDTH_AND_FULLWIDTH_FORMS 判断中文的,号if (ub == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS || ub == Character.UnicodeBlock.CJK_COMPATIBILITY_IDEOGRAPHS|| ub == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_A || ub == Character.UnicodeBlock.CJK_UNIFIED_IDEOGRAPHS_EXTENSION_B|| ub == Character.UnicodeBlock.CJK_SYMBOLS_AND_PUNCTUATION || ub == Character.UnicodeBlock.HALFWIDTH_AND_FULLWIDTH_FORMS|| ub == Character.UnicodeBlock.GENERAL_PUNCTUATION) {return true;}return false;}
