ZuulFilter的兩種初始化過程我們在前面已經(jīng)分析過柱徙，這一節(jié)我們也直奔主題缓屠，講講ZuulFilter初始化之后的調(diào)用過程，看看整個(gè)調(diào)用過程中Zuul是如何處理的护侮，都經(jīng)過了哪些步驟敌完？下面我們就以routing filters執(zhí)行過程進(jìn)行分析！

ZuulServletFilter

public class ZuulServletFilter implements Filter {
    @Override
    public void doFilter(ServletRequest servletRequest, ServletResponse servletResponse, FilterChain filterChain) throws IOException, ServletException {
        try {
            ......
            try {
                routing();
            } catch (ZuulException e) {
                error(e);
                postRouting();
                return;
            }
            ......
        } catch (Throwable e) {
            error(new ZuulException(e, 500, "UNCAUGHT_EXCEPTION_FROM_FILTER_" + e.getClass().getName()));
        } finally {
            RequestContext.getCurrentContext().unset();
        }
    }
    //routing filters執(zhí)行階段羊初，請求下游服務(wù)在此階段執(zhí)行
    void routing() throws ZuulException {
        zuulRunner.route();
    }
}

ZuulRunner

public class ZuulRunner {
    ......
    //初始化滨溉，默認(rèn)直接使用HttpServletRequest輸入流，輸出則使用HttpServletResponseWrapper包裝流
    public void init(HttpServletRequest servletRequest, HttpServletResponse servletResponse) {
        RequestContext ctx = RequestContext.getCurrentContext();
        if (bufferRequests) {
            ctx.setRequest(new HttpServletRequestWrapper(servletRequest));
        } else {
            ctx.setRequest(servletRequest);
        }
        ctx.setResponse(new HttpServletResponseWrapper(servletResponse));
    }
    ......
    public void route() throws ZuulException {
        FilterProcessor.getInstance().route();
    }
    ......
}

ZuulRunner這個(gè)對象類中我們看見了有HttpServletRequest长赞、HttpServletResponse晦攒、HttpServletRequestWrapper、HttpServletResponseWrapper四種類型
HttpServletRequest中參數(shù)是無法修改的得哆，HttpServletResponse中輸出流是無法讀取的脯颜，但是往往我們有很多需求是需要進(jìn)行請求參數(shù)的修改及響應(yīng)輸出流讀取的，這個(gè)時(shí)候就需要我們使用HttpServletRequestWrapper贩据、HttpServletResponseWrapper這兩個(gè)包裝類栋操，
HttpServletRequestWrapper作為HttpServletRequest的包裝類，主要職責(zé)就是可以替換HttpServletRequest中的參數(shù)
HttpServletResponseWrapper作為HttpServletResponse的包裝類饱亮，主要職責(zé)就是可以讀取HttpServletResponse中的輸出流矾芙！

FilterProcessor#route()

public class FilterProcessor {
    ......
    public void route() throws ZuulException {
        try {
            runFilters("route");
        } catch (ZuulException e) {
            throw e;
        } catch (Throwable e) {
            throw new ZuulException(e, 500, "UNCAUGHT_EXCEPTION_IN_ROUTE_FILTER_" + e.getClass().getName());
        }
    }
    ......
    public Object runFilters(String sType) throws Throwable {
        if (RequestContext.getCurrentContext().debugRouting()) {
            Debug.addRoutingDebug("Invoking {" + sType + "} type filters");
        }
        boolean bResult = false;
        //根據(jù)filter類型獲取所有相關(guān)類型的ZuulFilter
        List<ZuulFilter> list = FilterLoader.getInstance().getFiltersByType(sType);
        if (list != null) {
            for (int i = 0; i < list.size(); i++) {
                ZuulFilter zuulFilter = list.get(i);
                //執(zhí)行ZuulFilter#runFilter()調(diào)用IZuulFilter#run()方法
                Object result = processZuulFilter(zuulFilter);
                if (result != null && result instanceof Boolean) {
                    bResult |= ((Boolean) result);
                }
            }
        }
        return bResult;
    }
    ......
}

從片段中我們也比較清晰的得知調(diào)用過程為：
1.根據(jù)filter類型獲取所有相關(guān)類型的ZuulFilter
2.執(zhí)行ZuulFilter的run()方法
那么我們看看如何獲取List<ZuulFilter>這個(gè)集合對象的。

FilterLoader#getFiltersByType()

public class FilterLoader {
    ......
    public List<ZuulFilter> getFiltersByType(String filterType) {
        //如果hashFiltersByType緩存對象中已經(jīng)有當(dāng)前filterType類型的ZuulFilter集合則使用緩存
        List<ZuulFilter> list = hashFiltersByType.get(filterType);
        if (list != null) return list;
        list = new ArrayList<ZuulFilter>();
        //獲取全局對象filterRegistry中保存的所有ZuulFilter
        Collection<ZuulFilter> filters = filterRegistry.getAllFilters();
        //添加到list對象中
        for (Iterator<ZuulFilter> iterator = filters.iterator(); iterator.hasNext(); ) {
            ZuulFilter filter = iterator.next();
            if (filter.filterType().equals(filterType)) {
                list.add(filter);
            }
        }
        //根據(jù)filterOrder優(yōu)先級排序近上，默認(rèn)從小到大
        Collections.sort(list);
        //添加到hashFiltersByType對象中
        hashFiltersByType.putIfAbsent(filterType, list);
        return list;
    }
    ......
}

通過filterType類型獲取對應(yīng)的List<ZuulFilter>集合對象剔宪，其中筆者剛開始以為List<ZuulFilter> list = hashFiltersByType.get(filterType);這個(gè)地方有問題，因?yàn)閱为?dú)看這個(gè)地方會(huì)有一個(gè)問題，就是這個(gè)地方獲取出來如果有數(shù)據(jù)歼跟，那么就直接返回了和媳，那比如我動(dòng)態(tài)增加了一個(gè)Groovy文件并且類型相同的ZuulFilter，那這個(gè)地方看起來好想是有問題的哈街。
所以帶著疑問留瞳，我去仔細(xì)看了看初始化Groovy文件并初始化為ZuulFilter對象后添加到全局filterRegistry的過程，在FilterLoader#putFilter()這個(gè)方法中骚秦，有兩行非常重要的代碼她倘，List<ZuulFilter> list = hashFiltersByType.get(filter.filterType());
hashFiltersByType.remove(filter.filterType());這里如果有相同類型的，那么就直接就刪除了filterType對應(yīng)的List<ZuulFilter>集合數(shù)據(jù)作箍，在后面重新賦值硬梁，所以疑問迎刃而解！

FilterProcessor#processZuulFilter()

public class FilterProcessor {
    ......
    //執(zhí)行ZuulFilter的run()邏輯
    public Object processZuulFilter(ZuulFilter filter) throws ZuulException {
        ......
        try {
            ......
            //重點(diǎn):執(zhí)行ZuulFilter#runFilter()內(nèi)部會(huì)調(diào)用run()方法
            ZuulFilterResult result = filter.runFilter();
            ExecutionStatus s = result.getStatus();
            execTime = System.currentTimeMillis() - ltime;
            //失敗胞得、成功的邏輯
            switch (s) {
                case FAILED:
                    t = result.getException();
                    ctx.addFilterExecutionSummary(filterName, ExecutionStatus.FAILED.name(), execTime);
                    break;
                case SUCCESS:
                    o = result.getResult();
                    ctx.addFilterExecutionSummary(filterName, ExecutionStatus.SUCCESS.name(), execTime);
                    if (bDebug) {
                        Debug.addRoutingDebug("Filter {" + filterName + " TYPE:" + filter.filterType() + " ORDER:" + filter.filterOrder() + "} Execution time = " + execTime + "ms");
                        Debug.compareContextState(filterName, copy);
                    }
                    break;
                default:
                    break;
            }
            if (t != null) throw t;
            usageNotifier.notify(filter, s);
            return o;
        } catch (Throwable e) {
           ......
        }
    }
    ......
}

在processZuulFilter這個(gè)方法體中荧止，會(huì)記錄每個(gè)ZuulFilter的耗時(shí)時(shí)間，以及執(zhí)行過程中的成功阶剑、失敗跃巡、異常的狀態(tài)都會(huì)通知給Netflix Servo 實(shí)現(xiàn)監(jiān)控信息采集(JMX -Java Management Extensions標(biāo)準(zhǔn))，監(jiān)控信息不過多分析(PS:開啟JMX牧愁，然后使用JConsole或Visual VM進(jìn)行預(yù)覽)

ZuulFilter#runFilter()

public abstract class ZuulFilter implements IZuulFilter, Comparable<ZuulFilter> {
    ......
    public ZuulFilterResult runFilter() {
        ZuulFilterResult zr = new ZuulFilterResult();
        //默認(rèn)false素邪，及所有ZuulFilter都有效
        if (!isFilterDisabled()) {
            //是否滿足過濾條件
            if (shouldFilter()) {
                Tracer t = TracerFactory.instance().startMicroTracer("ZUUL::" + this.getClass().getSimpleName());
                try {
                    //重點(diǎn):實(shí)現(xiàn)了ZuulFilter都會(huì)重寫這個(gè)方法，執(zhí)行具體的業(yè)務(wù)邏輯方法
                    Object res = run();
                    zr = new ZuulFilterResult(res, ExecutionStatus.SUCCESS);
                } catch (Throwable e) {
                    t.setName("ZUUL::" + this.getClass().getSimpleName() + " failed");
                    zr = new ZuulFilterResult(ExecutionStatus.FAILED);
                    zr.setException(e);
                } finally {
                    t.stopAndLog();
                }
            } else {
                zr = new ZuulFilterResult(ExecutionStatus.SKIPPED);
            }
        }
        return zr;
    }
    ......
}

分析到這里猪半，執(zhí)行run()方法兔朦，因?yàn)槲覀兪欠治鰎oute 類型的filterType，并且是調(diào)用下游服務(wù)的ZuulFilter磨确，所以我們最終會(huì)執(zhí)行到RibbonRoutingFilter這個(gè)實(shí)現(xiàn)類沽甥，那這個(gè)實(shí)現(xiàn)類都做了什么事情呢？我們接著往下俐填！

RibbonRoutingFilter

public class RibbonRoutingFilter extends ZuulFilter {
    ......
    //配置的routes路由為使用serviceId方式安接，非url配置方式
    @Override
    public boolean shouldFilter() {
        RequestContext ctx = RequestContext.getCurrentContext();
        return (ctx.getRouteHost() == null && ctx.get(SERVICE_ID_KEY) != null
                && ctx.sendZuulResponse());
    }

    @Override
    public Object run() {
        RequestContext context = RequestContext.getCurrentContext();
        //添加忽略的頭信息,zuul的ignored-headers配置,會(huì)過濾這些頭信息參數(shù)
        this.helper.addIgnoredHeaders();
        try {
            //構(gòu)建Ribbon命令執(zhí)行的上下文，保存相關(guān)參數(shù)的對象
            RibbonCommandContext commandContext = buildCommandContext(context);
            //調(diào)用下游服務(wù)英融，返回結(jié)果
            ClientHttpResponse response = forward(commandContext);
            setResponse(response);
            return response;
        }
        catch (ZuulException ex) {
            throw new ZuulRuntimeException(ex);
        }
        catch (Exception ex) {
            throw new ZuulRuntimeException(ex);
        }
    }
    ......
}

這里我們簡單解釋下在run()方法中經(jīng)過的步驟RibbonRoutingFilter(routing filters)->Hystrix->Ribbon-> Response盏檐。如果想了解routing filters->Hystrix->Ribbon-> Response整個(gè)詳細(xì)過程請參閱:Spring Cloud Hystrix 分析(四)之Zuul集成

1. 構(gòu)建Ribbon命令執(zhí)行的上下文，保存相關(guān)參數(shù)的對象驶悟，封裝RibbonCommandContext對象(頭信息胡野、請求參數(shù)信息)
2. 封裝HttpClientRibbonCommand(RibbonCommand)對象，設(shè)置負(fù)載均衡客戶端痕鳍，而且內(nèi)部繼承HystrixExecutable硫豆，可以使用Hystrix熔斷功能
3. 執(zhí)行命令龙巨，最終調(diào)用到HystrixCommand#execute()
4. AbstractRibbonCommand#run()方法中this.client.executeWithLoadBalancer(request, config);通過負(fù)載均衡客戶端進(jìn)行下游服務(wù)的調(diào)用并返回結(jié)果

文章到這里也接近尾聲，本節(jié)主要講解了route類型ZuulFilter的整個(gè)調(diào)用過程熊响，其他類型的也基本差不多的旨别，都是相同類型的ZuulFilter按照從小到大的優(yōu)先級進(jìn)行執(zhí)行，如果文章對你有所幫助汗茄，就點(diǎn)贊關(guān)注吧秸弛！