From 988dfa44e689eb6be4e95b91fc5d3438f40a433b Mon Sep 17 00:00:00 2001
From: Vip-Augus <850366301@qq.com>
Date: Mon, 10 Jun 2019 20:06:04 +0800
Subject: [PATCH] =?UTF-8?q?=E6=89=8B=E6=8A=8A=E6=89=8B=E6=95=99=E4=BD=A0?=
 =?UTF-8?q?=E5=86=99=E7=94=9F=E4=BA=A7=E7=8E=AF=E5=A2=83=20OOM=20=E4=BB=A3?=
 =?UTF-8?q?=E7=A0=81=20=E6=9C=AC=E4=BA=BA=E5=AE=9E=E4=B9=A0=E6=9C=9F?=
 =?UTF-8?q?=E5=86=99=E8=BF=87=E7=9A=84=20OOM=20=E4=BB=A3=E7=A0=81=3D-=3D?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 java/OOM.java | 85 +++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 85 insertions(+)
 create mode 100644 java/OOM.java

diff --git a/java/OOM.java b/java/OOM.java
new file mode 100644
index 0000000..92e891a
--- /dev/null
+++ b/java/OOM.java
@@ -0,0 +1,85 @@
+// 手把手教你写生产环境 OOM 代码，具体分析请看这个：https://juejin.im/post/5ce9067df265da1bc4142f18
+public Table<CustomerAccountTitle, Assistant, List<SubsidiaryLedgerRow>> queryMoreAssistantSubsidiaryLedger(AssistantSubsidiaryLedgerQueryParam queryParam) {
+    
+    Table<CustomerAccountTitle, Assistant, List<SubsidiaryLedgerRow>> resultTable = HashBasedTable.create();
+    // 这个 list 有 2w 条数据
+    for (Assistant assistant : assistants) {
+        
+        //这个 list 从数据库拿，4w 条数据，数据库交互，结合外面的看，2w * 4w，然后加上基础方法 baseQueryAssistantSubsidiaryLedgerRowsMap 的循环，线上直接 OOM 了
+        List<AssistantTitleBalance> balances = balanceService.selectFromYear(customerId, accountSetId, PeriodUtils.getYear(fixedBeginPeriod));
+        Map<CustomerAccountTitle, List<SubsidiaryLedgerRow>> rowMap = baseQueryAssistantSubsidiaryLedgerRowsMap(
+                titles,
+                AssistantType.fromCode(queryParam.getAssistantType()),
+                assistant.getId(),
+                fixedBeginPeriod,
+                endPeriod,
+                null,
+                balances);
+        for (Map.Entry<CustomerAccountTitle, List<SubsidiaryLedgerRow>> temp : rowMap.entrySet()) {
+            resultTable.put(temp.getKey(), assistant, temp.getValue());
+        }
+    }
+    return resultTable;
+}
+
+// 这是个基础方法
+private Map<CustomerAccountTitle, List<SubsidiaryLedgerRow>> baseQueryAssistantSubsidiaryLedgerRowsMap(
+        List<CustomerAccountTitle> titles,
+        AssistantType assistantType,
+        Integer assistantId,
+        String beginPeriod,
+        String endPeriod,
+        BookType ledgerType,
+        List<AssistantTitleBalance> balances) {
+    Map<CustomerAccountTitle, List<SubsidiaryLedgerRow>> resultMap = Maps.newHashMap();
+
+    List<String> titleCodes = Lists.newArrayList(Lists.transform(titles, new Function<CustomerAccountTitle, String>() {
+        @Override
+        public String apply(CustomerAccountTitle input) {
+            return input.getCode();
+        }
+    }));
+
+    for (AssistantTitleBalance balance : balances) {
+        for (String titleCode : titleCodes) {
+            if (!balance.getAccountTitleCode().startsWith(titleCode)) {
+                continue;
+            }
+            List<Balance> tempBalances = balanceTable.get(titleCode, year);
+            if (tempBalances == null) {
+                tempBalances = new ArrayList();
+                balanceTable.put(titleCode, year, tempBalances);
+            }
+            tempBalances.add(balance);
+        }
+    }
+    Table<String, String, List<AccountEntry>> entryTable = HashBasedTable.create();
+    // 这里就就恐怖了，每次查询有 2w 条数据，数据库交互
+    Map<Integer, AccountDocument> documentMap = documentService.selectFromYear(beginPeriod);
+    for (AccountDocument document : documentMap.values()) {
+        //  内层遍历，每次 10 +
+        for (AccountEntry entry : document.getEntryList()) {
+            // 再遍历，最外层有 200+
+            for (String titleCode : titleCodes) {
+                if (!entry.isUseAssistant() || AssistantType.fromCode(entry.getAssistantType()) != assistantType || !assistantId.equals(entry.getAssistantId()) || !entry.getAccountTitleCode().startsWith(titleCode)) {
+                    continue;
+                }
+                List<AccountEntry> entries = entryTable.get(titleCode, document.getAccountPeriod());
+                if (entries == null) {
+                    entries = new ArrayList();
+                    entryTable.put(titleCode, document.getAccountPeriod(), entries);
+                }
+                entries.add(entry);
+            }
+        }
+    }
+    List<String> periods = PeriodUtils.getPeriodsUpTo(beginPeriod, endPeriod);
+    for (CustomerAccountTitle title : titles) {
+        if (ledgerType != null && AccountTitleUtils.getBookType(title) == ledgerType) {
+            resultMap.put(title, buildSubsidiaryLedgerRows(periods, title, balanceTable.row(title.getCode()), entryTable.row(title.getCode()), documentMap, ledgerType, PeriodUtils.getIntegerYear(beginPeriod)));
+        } else if (ledgerType == null) {
+            resultMap.put(title, buildSubsidiaryLedgerRows(periods, title, balanceTable.row(title.getCode()), entryTable.row(title.getCode()), documentMap, AccountTitleUtils.getBookType(title), PeriodUtils.getIntegerYear(beginPeriod)));
+        }
+    }
+    return resultMap;
+}
\ No newline at end of file
-- 
Gitee