haoqg commented on issue #8309: URL: https://github.com/apache/shardingsphere/issues/8309#issuecomment-733615552
5.0 是用的SPI机制，假如需要使用自定义分库或者分表策略的话，以官方的mod为例子，我们需要这么干，首先 在项目里的resources 目录下新建一个文件夹 META-INF/services，文件夹里新建一个文件org.apache.shardingsphere.sharding.spi.ShardingAlgorithm  在此文件里把自己的类加上 比如，我这里新建一个类， public final class BatchIdShardingAlgorithm implements StandardShardingAlgorithm<Comparable<?>>, ShardingAutoTableAlgorithm { private static final String SHARDING_COUNT_KEY = "sharding-count"; private Properties props = new Properties(); private int shardingCount; public BatchIdShardingAlgorithm() { } @Override public void init() { this.shardingCount = this.getShardingCount(); } private int getShardingCount() { return 0; } @Override public String doSharding(Collection<String> availableTargetNames, PreciseShardingValue<Comparable<?>> shardingValue) { return "usoa_detail_flow_0"; } @Override public Collection<String> doSharding(Collection<String> availableTargetNames, RangeShardingValue<Comparable<?>> shardingValue) { return this.isContainAllTargets(shardingValue) ? availableTargetNames : this.getAvailableTargetNames(availableTargetNames, shardingValue); } private boolean isContainAllTargets(RangeShardingValue<Comparable<?>> shardingValue) { return !shardingValue.getValueRange().hasUpperBound() || shardingValue.getValueRange().hasLowerBound() && this.getLongValue(shardingValue.getValueRange().upperEndpoint()) - this.getLongValue(shardingValue.getValueRange().lowerEndpoint()) >= (long)(this.shardingCount - 1); } private Collection<String> getAvailableTargetNames(Collection<String> availableTargetNames, RangeShardingValue<Comparable<?>> shardingValue) { Collection<String> result = new LinkedHashSet(availableTargetNames.size()); for(long i = this.getLongValue(shardingValue.getValueRange().lowerEndpoint()); i <= this.getLongValue(shardingValue.getValueRange().upperEndpoint()); ++i) { Iterator var6 = availableTargetNames.iterator(); while(var6.hasNext()) { String each = (String)var6.next(); if (each.endsWith(String.valueOf(i % (long)this.shardingCount))) { result.add(each); } } } return result; } private long getLongValue(Comparable<?> value) { return Long.parseLong(value.toString()); } @Override public int getAutoTablesAmount() { return this.shardingCount; } @Override public String getType() { return "BATCH-ID"; } @Override public Collection<String> getAllPropertyKeys() { return Collections.singletonList("sharding-count"); } @Override @Generated public Properties getProps() { return this.props; } @Override @Generated public void setProps(Properties props) { this.props = props; } @Generated public void setShardingCount(int shardingCount) { this.shardingCount = shardingCount; } } 然后就可以使用自定义的策略了 Properties tableShardingAlgorithmrProps = new Properties(); shardingRuleConfig.getShardingAlgorithms().put("tableShardingAlgorithm", new ShardingSphereAlgorithmConfiguration("BATCH-ID", tableShardingAlgorithmrProps)); 大概就是这么个机制， 其实源码里就是这么干的，有图为证  ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
