【原创】Jmeter正则表达式提取json中多个关联值

使用jmeter时response中返回为json格式如下

{
    "return_code": 0,
    "return_msg": "ok",
    "data": [
        {
            "name": "武汉隆安置业有限公司",
            "id": 4,
            "product": "[{"value":"7953AC76-A85A-E011-B371-001D096CF989","text":"\u9e7f\u9e23\u82d1"}]"
        },
        {
            "name": "湖北省建工房地产开发有限公司",
            "id": 10,
            "product": "[{"value":"8C53AC76-A85A-E011-B371-001D096CF989","text":"\u5f90\u4e1c\u96c5\u82d1"}]"
        }
    ]
}

需要将其中的"id": 4" id": 10,取出（4，10）用来后续断言，由于每次返回id的组数可能都不同，故无法使用$$模板号来取

此时正则表达式匹配数字选择-1来匹配所有符合的关联值

查看结果树时发现看不到所有id的值，通过添加后置处理器Debug PostProcessor查看提取的所有结果

当前提取出的结果：

custom_id_1=4
custom_id_1_g=1
custom_id_1_g0="id":4,"product"
custom_id_1_g1=4
custom_id_2=10
custom_id_2_g=1
custom_id_2_g0="id":10,"product"
custom_id_2_g1=10
custom_id_matchNr=2

 可以通过逻辑控制器-for each控制器逐一取出关联值

并且赋值给数组tmp，其中${custom_id_matchNr}为匹配到的关联值个数

最后打印验证下去取出的值是否正确

4与10被成功取出，证明tmp即为最终需要的数组

================================================================================================================================================

Jmeter正则实现探究：

当前提取出的结果：

custom_id_1=4
custom_id_1_g=1
custom_id_1_g0="id":4,"product"
custom_id_1_g1=4
custom_id_2=10
custom_id_2_g=1
custom_id_2_g0="id":10,"product"
custom_id_2_g1=10
custom_id_matchNr=2

可以发现取出的内容并不只是一个数组

下面还有分组 group1  group2

查看源码实现

    private void saveGroups(MatchResult result, String namep, JMeterVariables vars) {
        if (result != null) {
            for (int x = 0; x < result.groups(); x++) {
                vars.put(namep + "_g" + x, result.group(x)); //$NON-NLS-1$
            }
        }
    }

List<MatchResult> collectAllMatches = new ArrayList<>();
        try {
            PatternMatcher matcher = JMeterUtils.getMatcher();
            PatternMatcherInput input = new PatternMatcherInput(textToMatch);
            while (matcher.contains(input, searchPattern)) {
                MatchResult match = matcher.getMatch();
                collectAllMatches.add(match);
            }
        } finally {
            if (name.length() > 0){
                vars.put(name + "_matchNr", Integer.toString(collectAllMatches.size())); //$NON-NLS-1$
            }
        }

        if (collectAllMatches.size() == 0) {
            return defaultValue;
        }

        if (valueIndex.equals(ALL)) {
            StringBuilder value = new StringBuilder();
            Iterator<MatchResult> it = collectAllMatches.iterator();
            boolean first = true;
            while (it.hasNext()) {
                if (!first) {
                    value.append(between);
                } else {
                    first = false;
                }
                value.append(generateResult(it.next(), name, tmplt, vars));
            }
            return value.toString();
        } else if (valueIndex.equals(RAND)) {
            MatchResult result = collectAllMatches.get(ThreadLocalRandom.current().nextInt(collectAllMatches.size()));
            return generateResult(result, name, tmplt, vars);
        } else {
            try {
                int index = Integer.parseInt(valueIndex) - 1;
                MatchResult result = collectAllMatches.get(index);
                return generateResult(result, name, tmplt, vars);
            } catch (NumberFormatException e) {
                float ratio = Float.parseFloat(valueIndex);
                MatchResult result = collectAllMatches
                        .get((int) (collectAllMatches.size() * ratio + .5) - 1);
                return generateResult(result, name, tmplt, vars);
            } catch (IndexOutOfBoundsException e) {
                return defaultValue;
            }
        }

custom_id_1_g=1

custom_id_2_g=1  // 代表group的长度

custom_id_2_g0="id":10,"product"  //匹配的内容
custom_id_2_g1=10  //提取的内容