文件上传-内部地址转第三方地址

场景

将系统内部的文件通过远程地址上传到第三方系统拿到第三方地址,我这里第三方还没有开放接口,苦逼地只能F12从第三方平台尝试他们接口的协议情况.记录下来是因为,这种虽然不是什么高深的技术用法,但是这种涉及基础的灵活使用总记不住,因为很多是固定写法,每次都要调不少时间,这次记录下,给我和大家参考下.

代码

 /**
     * 两个参数都不能为空，并且fileName必须要有后缀并且和文件的实际后缀一致
     *
     * @param ehrUrl
     * @param fileName
     * @return
     */
    public String getMokaUrlByEhrUrl(String ehrUrl, String fileName) {
        if (StringUtils.isAnyBlank(ehrUrl, fileName)) {
            throw new BizProcessException("url和fileName都不能为空！");
        }
        String suffix = FileUtil.getSuffix(fileName);
        if (StringUtils.isBlank(suffix)) {
            throw new BizProcessException("文件名中的后缀不能为空!");
        }
        try {
            //优先从库里查-尽量减少对moka外部接口的请求
            EhrFileMessage ehrFileMessage = ehrFileMessageMapper.selectOne(
                Wrappers.lambdaQuery(EhrFileMessage.class).select(EhrFileMessage::getMokaUrl)
                    .eq(EhrFileMessage::getEhrUrl, ehrUrl).last("limit 1"));
            if (ehrFileMessage != null) {
                return ehrFileMessage.getMokaUrl();
            }
            URIBuilder uriBuilder = new URIBuilder("https://xxx.com/api/workflow/v1/form/uploadAttachment");
            uriBuilder.addParameter("bus", "20");
            uriBuilder.addParameter("loginType", "10");
            URI uri = uriBuilder.build();
            int i = ehrUrl.lastIndexOf("/");
            String encode = URLEncoder.encode(ehrUrl.substring(i + 1, ehrUrl.length()));
            String ehrUrlEncode = ehrUrl.substring(0, i + 1) + encode;
            URL url = new URL(ehrUrlEncode);
            InputStream inputStream = url.openStream();
            byte[] bytes = inputStream.readAllBytes();
            CloseableHttpClient httpClient = HttpClients.createDefault();
            HttpPost httpPost = new HttpPost(uri);
            MultipartEntityBuilder builder = MultipartEntityBuilder.create();
            builder.addBinaryBody("file", bytes, ContentType.DEFAULT_BINARY, fileName);
            HttpEntity multipart = builder.build();
            httpPost.setEntity(multipart);
            httpPost.setHeader("Cookie", "moka-uid=" + this.getMokaToken());
            CloseableHttpResponse response = httpClient.execute(httpPost);
            Integer statusCode =
                Optional.ofNullable(response).map(CloseableHttpResponse::getStatusLine).map(StatusLine::getStatusCode)
                    .orElse(500);
            if (statusCode == 200 && response.getEntity() != null) {
                JSONObject jsonObject = JSON.parseObject(EntityUtils.toString(response.getEntity()));
                String mokaUrl = jsonObject.getJSONObject("data").getString("url");
                if (StringUtils.isNotBlank(mokaUrl)) {
                    EhrFileMessage ehrFileMessageNew = new EhrFileMessage();
                    ehrFileMessageNew.setEhrUrl(ehrUrl);
                    ehrFileMessageNew.setMokaUrl(mokaUrl);
                    ehrFileMessageNew.setFileName(fileName);
                    ehrFileMessageMapper.insert(ehrFileMessageNew);
                }
                return mokaUrl;
            } else {
                log.error("enter MokaProxyService.getMokaUrlByEhrUrl,response:{}", JSON.toJSONString(response));
            }
        } catch (Exception e) {
            log.error("enter MokaProxyService.getMokaUrlByEhrUrl,error:{}", e);
        }
        return null;
    }

    public String getMokaToken() {
        try {
            Object o = crawlerRedissonClient.getMap("ResumeHR:mokahr").get("moka-uid");
            if (o != null) {
                return o.toString();
            } else {
                return null;
            }
        } catch (Exception e) {
            log.error("enter MokaProxyService.getMokaToken,error:{}", e);
        }
        return null;
    }

我是和爬虫工程师一起做的,因为没有开放接口,这种请求toc接口是需要登录信息的,这边是我们通过redis对接的,他定时更新cookie;
这里遇到过的问题是,流的处理问题,遇到过文件损坏只支持pdf.但是这种 byte[] bytes = inputStream.readAllBytes();非拼接的方法解决了,还有文件的名字的后缀要规范;
后面的优化是,做了个字典表,防止请求过多,别人还以为是被攻击了呢.

本次redis不变的临时链接redis的示例

    @PostConstruct
    public void init() {
        Config config = new Config();
        config.setCodec(StringCodec.INSTANCE);
        config.setTransportMode(TransportMode.NIO);
        config.useSingleServer().setAddress("redis://xxx:6379").setDatabase(3).setPassword("xxx");
        crawlerRedissonClient =  Redisson.create(config);
    }