(improvement)(common) Upgraded langchain4j to version 0.31. (#1174)

2025-12-15 22:46:49 +00:00 · 2024-06-20 17:31:19 +08:00
parent 1df1fe5ad6
commit 92db381b6f
37 changed files with 241 additions and 888 deletions
--- a/launchers/standalone/src/main/java/com/tencent/supersonic/StandaloneLauncher.java
+++ b/launchers/standalone/src/main/java/com/tencent/supersonic/StandaloneLauncher.java
@@ -1,20 +1,17 @@
 package com.tencent.supersonic;

-import dev.langchain4j.S2LangChain4jAutoConfiguration;
 import org.springframework.boot.SpringApplication;
 import org.springframework.boot.autoconfigure.SpringBootApplication;
 import org.springframework.boot.autoconfigure.data.mongo.MongoDataAutoConfiguration;
 import org.springframework.boot.autoconfigure.mongo.MongoAutoConfiguration;
-import org.springframework.context.annotation.Import;
 import org.springframework.scheduling.annotation.EnableAsync;
 import org.springframework.scheduling.annotation.EnableScheduling;
 import springfox.documentation.swagger2.annotations.EnableSwagger2;

-@SpringBootApplication(scanBasePackages = {"com.tencent.supersonic"},
+@SpringBootApplication(scanBasePackages = {"com.tencent.supersonic", "dev.langchain4j.model"},
        exclude = {MongoAutoConfiguration.class, MongoDataAutoConfiguration.class})
@EnableScheduling
@EnableAsync
-@Import(S2LangChain4jAutoConfiguration.class)
@EnableSwagger2
 public class StandaloneLauncher {

--- a/launchers/standalone/src/main/resources/META-INF/spring.factories
+++ b/launchers/standalone/src/main/resources/META-INF/spring.factories
@@ -88,3 +88,11 @@ com.tencent.supersonic.auth.api.authentication.adaptor.UserAdaptor=\

 dev.langchain4j.store.embedding.S2EmbeddingStore=\
    dev.langchain4j.store.embedding.InMemoryS2EmbeddingStore
+
+org.springframework.boot.autoconfigure.EnableAutoConfiguration=\
+    dev.langchain4j.spring.LangChain4jAutoConfig,\
+    dev.langchain4j.openai.spring.AutoConfig,\
+    dev.langchain4j.ollama.spring.AutoConfig,\
+    dev.langchain4j.azure.openai.spring.AutoConfig,\
+    dev.langchain4j.azure.aisearch.spring.AutoConfig,\
+    dev.langchain4j.anthropic.spring.AutoConfig
--- a/launchers/standalone/src/main/resources/application-local.yaml
+++ b/launchers/standalone/src/main/resources/application-local.yaml
@@ -17,7 +17,14 @@ spring:
    url: jdbc:h2:mem:semantic;DATABASE_TO_UPPER=false
    username: root
    password: semantic
-
+  autoconfigure:
+    exclude:
+      - spring.dev.langchain4j.spring.LangChain4jAutoConfig
+      - spring.dev.langchain4j.openai.spring.AutoConfig
+      - spring.dev.langchain4j.ollama.spring.AutoConfig
+      - spring.dev.langchain4j.azure.openai.spring.AutoConfig
+      - spring.dev.langchain4j.azure.aisearch.spring.AutoConfig
+      - spring.dev.langchain4j.anthropic.spring.AutoConfig
 mybatis:
  mapper-locations=classpath:mappers/custom/*.xml,classpath*:/mappers/*.xml

@@ -67,42 +74,6 @@ s2:
    names: S2VisitsDemo,S2ArtistDemo
    enableLLM: true

-  langchain4j:
-    #1.chat-model
-    chat-model:
-      provider: open_ai
-      openai:
-        # Replace with your LLM configs
-        # Note: The default API key `demo` is provided by langchain4j community
-        #       which limits 1000 tokens per request.
-        base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
-        api-key: ${OPENAI_API_KEY:demo}
-        model-name: ${OPENAI_MODEL_NAME:gpt-3.5-turbo}
-        temperature: ${OPENAI_TEMPERATURE:0.0}
-        timeout: ${OPENAI_TIMEOUT:PT60S}
-    #2.embedding-model
-    #2.1 in_memory(default)
-    embedding-model:
-      provider: in_process
-    #      inProcess:
-    #        modelPath: /data/model.onnx
-    #        vocabularyPath: /data/onnx_vocab.txt
-    #        shibing624/text2vec-base-chinese
-    #2.2 open_ai
-    #  embedding-model:
-    #    provider: open_ai
-    #    openai:
-    #      api-key: api_key
-    #      modelName: all-minilm-l6-v2.onnx
-
-    #2.2 hugging_face
-    #  embedding-model:
-    #    provider: hugging_face
-    #    hugging-face:
-    #      access-token: hg_access_token
-    #      model-id: sentence-transformers/all-MiniLM-L6-v2
-    #      timeout: 1h
-
 # swagger配置
 swagger:
  title: 'SuperSonic平台接口文档'
@@ -115,3 +86,20 @@ swagger:
    email:
    url: ''
  version: 3.0
+
+
+langchain4j:
+  open-ai:
+    chat-model:
+      # Replace with your LLM configs
+      # Note: The default API key `demo` is provided by langchain4j community
+      #       which limits 1000 tokens per request.
+      base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
+      api-key: ${OPENAI_API_KEY:demo}
+      model-name: ${OPENAI_MODEL_NAME:gpt-3.5-turbo}
+      temperature: ${OPENAI_TEMPERATURE:0.0}
+      timeout: ${OPENAI_TIMEOUT:PT60S}
+# java.lang.RuntimeException: dev.ai4j.openai4j.OpenAiHttpException: Too many requests
+    embedding-model:
+      base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
+      api-key: ${OPENAI_API_KEY:demo}
--- a/launchers/standalone/src/test/java/com/tencent/supersonic/chat/MetricTest.java
+++ b/launchers/standalone/src/test/java/com/tencent/supersonic/chat/MetricTest.java
@@ -40,7 +40,6 @@ public class MetricTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));

        expectedParseInfo.getDimensionFilters().add(DataUtils.getFilter("user_name",
                FilterOperatorEnum.EQUALS, "alice", "用户", 2L));
@@ -74,7 +73,6 @@ public class MetricTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));
        expectedParseInfo.setDateInfo(DataUtils.getDateConf(DateConf.DateMode.RECENT, unit, period, startDay, endDay));
        expectedParseInfo.setQueryType(QueryType.METRIC);

@@ -103,7 +101,6 @@ public class MetricTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));
        expectedParseInfo.getDimensions().add(DataUtils.getSchemaElement("部门"));

        expectedParseInfo.setDateInfo(DataUtils.getDateConf(DateConf.DateMode.RECENT, unit, period, startDay, endDay));
@@ -124,7 +121,6 @@ public class MetricTest extends BaseTest {
        expectedParseInfo.setAggType(NONE);
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));
        List<String> list = new ArrayList<>();
        list.add("alice");
        list.add("lucy");
@@ -171,7 +167,6 @@ public class MetricTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));
        expectedParseInfo.getDimensions().add(DataUtils.getSchemaElement("部门"));

        expectedParseInfo.setDateInfo(DataUtils.getDateConf(DateConf.DateMode.RECENT, unit, period, startDay, endDay));
@@ -197,7 +192,6 @@ public class MetricTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));
        expectedParseInfo.getDimensionFilters().add(DataUtils.getFilter("user_name",
                FilterOperatorEnum.EQUALS, "alice", "用户", 2L));

--- a/launchers/standalone/src/test/java/com/tencent/supersonic/chat/MultiTurnsTest.java
+++ b/launchers/standalone/src/test/java/com/tencent/supersonic/chat/MultiTurnsTest.java
@@ -29,7 +29,6 @@ public class MultiTurnsTest extends BaseTest {

        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问次数"));
        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("人均访问次数"));
-        expectedParseInfo.getMetrics().add(DataUtils.getSchemaElement("访问用户数"));

        expectedParseInfo.getDimensionFilters().add(DataUtils.getFilter("user_name",
                FilterOperatorEnum.EQUALS, "alice", "用户", 2L));
--- a/launchers/standalone/src/test/java/com/tencent/supersonic/headless/BaseTest.java
+++ b/launchers/standalone/src/test/java/com/tencent/supersonic/headless/BaseTest.java
@@ -14,7 +14,7 @@ import com.tencent.supersonic.headless.api.pojo.request.SemanticQueryReq;
 import com.tencent.supersonic.headless.api.pojo.response.SemanticQueryResp;
 import com.tencent.supersonic.headless.server.service.QueryService;
 import com.tencent.supersonic.util.DataUtils;
-import org.apache.commons.collections4.CollectionUtils;
+import org.apache.commons.collections.CollectionUtils;
 import org.springframework.beans.factory.annotation.Autowired;

 import java.util.ArrayList;
--- a/launchers/standalone/src/test/resources/META-INF/spring.factories
+++ b/launchers/standalone/src/test/resources/META-INF/spring.factories
@@ -1,3 +1,5 @@
+### headless-chat SPIs
+
 com.tencent.supersonic.headless.chat.mapper.SchemaMapper=\
    com.tencent.supersonic.headless.chat.mapper.EmbeddingMapper, \
    com.tencent.supersonic.headless.chat.mapper.KeywordMapper, \
@@ -9,20 +11,20 @@ com.tencent.supersonic.headless.chat.parser.SemanticParser=\
    com.tencent.supersonic.headless.chat.parser.llm.LLMSqlParser, \
    com.tencent.supersonic.headless.chat.parser.QueryTypeParser

-com.tencent.supersonic.chat.server.parser.ChatParser=\
-    com.tencent.supersonic.chat.server.parser.NL2PluginParser, \
-    com.tencent.supersonic.chat.server.parser.MultiTurnParser, \
-    com.tencent.supersonic.chat.server.parser.NL2SQLParser
-
-com.tencent.supersonic.chat.server.executor.ChatExecutor=\
-    com.tencent.supersonic.chat.server.executor.PluginExecutor, \
-    com.tencent.supersonic.chat.server.executor.SqlExecutor
-
 com.tencent.supersonic.headless.chat.corrector.SemanticCorrector=\
    com.tencent.supersonic.headless.chat.corrector.SchemaCorrector, \
    com.tencent.supersonic.headless.chat.corrector.TimeCorrector, \
    com.tencent.supersonic.headless.chat.corrector.GrammarCorrector

+com.tencent.supersonic.headless.chat.knowledge.file.FileHandler=\
+    com.tencent.supersonic.headless.chat.knowledge.file.FileHandlerImpl
+
+com.tencent.supersonic.headless.chat.parser.llm.DataSetResolver=\
+  com.tencent.supersonic.headless.chat.parser.llm.HeuristicDataSetResolver
+
+
+### headless-core SPIs
+
 com.tencent.supersonic.headless.core.parser.converter.HeadlessConverter=\
    com.tencent.supersonic.headless.core.parser.converter.DefaultDimValueConverter,\
    com.tencent.supersonic.headless.core.parser.converter.SqlVariableParseConverter,\
@@ -41,18 +43,24 @@ com.tencent.supersonic.headless.core.parser.SqlParser=\
 com.tencent.supersonic.headless.core.cache.QueryCache=\
    com.tencent.supersonic.headless.core.cache.DefaultQueryCache

+
+### headless-server SPIs
+
 com.tencent.supersonic.headless.server.processor.ResultProcessor=\
    com.tencent.supersonic.headless.server.processor.ParseInfoProcessor, \
    com.tencent.supersonic.headless.server.processor.SqlInfoProcessor

-com.tencent.supersonic.headless.chat.parser.llm.DataSetResolver=\
-  com.tencent.supersonic.headless.chat.parser.llm.HeuristicDataSetResolver

-com.tencent.supersonic.auth.authentication.interceptor.AuthenticationInterceptor=\
-    com.tencent.supersonic.auth.authentication.interceptor.DefaultAuthenticationInterceptor
+### chat-server SPIs

-com.tencent.supersonic.auth.api.authentication.adaptor.UserAdaptor=\
-    com.tencent.supersonic.auth.authentication.adaptor.DefaultUserAdaptor
+com.tencent.supersonic.chat.server.parser.ChatParser=\
+    com.tencent.supersonic.chat.server.parser.NL2PluginParser, \
+    com.tencent.supersonic.chat.server.parser.MultiTurnParser,\
+    com.tencent.supersonic.chat.server.parser.NL2SQLParser
+
+com.tencent.supersonic.chat.server.executor.ChatExecutor=\
+    com.tencent.supersonic.chat.server.executor.PluginExecutor, \
+    com.tencent.supersonic.chat.server.executor.SqlExecutor

 com.tencent.supersonic.chat.server.plugin.recognize.PluginRecognizer=\
    com.tencent.supersonic.chat.server.plugin.recognize.embedding.EmbeddingRecallRecognizer
@@ -67,5 +75,16 @@ com.tencent.supersonic.chat.server.processor.execute.ExecuteResultProcessor=\
    com.tencent.supersonic.chat.server.processor.execute.DimensionRecommendProcessor,\
    com.tencent.supersonic.chat.server.processor.execute.MetricRatioProcessor

+### auth-authentication SPIs
+
+com.tencent.supersonic.auth.authentication.interceptor.AuthenticationInterceptor=\
+    com.tencent.supersonic.auth.authentication.interceptor.DefaultAuthenticationInterceptor
+
+com.tencent.supersonic.auth.api.authentication.adaptor.UserAdaptor=\
+    com.tencent.supersonic.auth.authentication.adaptor.DefaultUserAdaptor
+
+
+### common SPIs
+
 dev.langchain4j.store.embedding.S2EmbeddingStore=\
    dev.langchain4j.store.embedding.InMemoryS2EmbeddingStore
--- a/launchers/standalone/src/test/resources/application-local.yaml
+++ b/launchers/standalone/src/test/resources/application-local.yaml
@@ -17,7 +17,14 @@ spring:
    url: jdbc:h2:mem:semantic;DATABASE_TO_UPPER=false
    username: root
    password: semantic
-
+  autoconfigure:
+    exclude:
+      - spring.dev.langchain4j.spring.LangChain4jAutoConfig
+      - spring.dev.langchain4j.openai.spring.AutoConfig
+      - spring.dev.langchain4j.ollama.spring.AutoConfig
+      - spring.dev.langchain4j.azure.openai.spring.AutoConfig
+      - spring.dev.langchain4j.azure.aisearch.spring.AutoConfig
+      - spring.dev.langchain4j.anthropic.spring.AutoConfig
 mybatis:
  mapper-locations=classpath:mappers/custom/*.xml,classpath*:/mappers/*.xml

@@ -36,7 +43,11 @@ logging:
    dev.ai4j.openai4j: DEBUG

 s2:
+  pyllm:
+    url: http://127.0.0.1:9092
+
  parser:
+    url: ${s2.pyllm.url}
    strategy: ONE_PASS_SELF_CONSISTENCY
    exemplar-recall:
      number: 10
@@ -50,6 +61,14 @@ s2:
  corrector:
    additional:
      information: true
+    date: true
+  functionCall:
+    url: ${s2.pyllm.url}
+
+  embedding:
+    url: ${s2.pyllm.url}
+    persistent:
+      path: /tmp

  demo:
    names: S2VisitsDemo,S2ArtistDemo
@@ -59,24 +78,6 @@ s2:
    cache:
      enable: false

-  langchain4j:
-    #1.chat-model
-    chat-model:
-      provider: open_ai
-      openai:
-        # Replace with your LLM configs
-        # Note: The default API key `demo` is provided by langchain4j community
-        #       which limits 1000 tokens per request.
-        base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
-        api-key: ${OPENAI_API_KEY:demo}
-        model-name: ${OPENAI_MODEL_NAME:gpt-3.5-turbo}
-        temperature: ${OPENAI_TEMPERATURE:0.0}
-        timeout: ${OPENAI_TIMEOUT:PT60S}
-    #2.embedding-model
-    #2.1 in_memory(default)
-    embedding-model:
-      provider: in_process
-
 # swagger配置
 swagger:
  title: 'SuperSonic平台接口文档'
@@ -88,4 +89,21 @@ swagger:
    name:
    email:
    url: ''
-  version: 3.0
+  version: 3.0
+
+
+langchain4j:
+  open-ai:
+    chat-model:
+      # Replace with your LLM configs
+      # Note: The default API key `demo` is provided by langchain4j community
+      #       which limits 1000 tokens per request.
+      base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
+      api-key: ${OPENAI_API_KEY:demo}
+      model-name: ${OPENAI_MODEL_NAME:gpt-3.5-turbo}
+      temperature: ${OPENAI_TEMPERATURE:0.0}
+      timeout: ${OPENAI_TIMEOUT:PT60S}
+# java.lang.RuntimeException: dev.ai4j.openai4j.OpenAiHttpException: Too many requests
+#    embedding-model:
+#      base-url: ${OPENAI_API_BASE:https://api.openai.com/v1}
+#      api-key: ${OPENAI_API_KEY:demo}