[FEATURE]: Custom output parsing for AI services (fixes TODO) #1118
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This change gives users the ability to provide custom output parsers to AI services. -- Context -- Previously, users were constrained to the set of data types that were supported by langchain4j (most primitives, string lists, string sets, enums, and JSON POJOs). Generic lists and sets were not supported. -- Change -- Users can now provide a custom ParserProvider when creating an AI service via the builder, or can provide individual OutputParsers or ParserFactory's on each service method with the @parser annotation. This enables users to have full control of the output parsing process, while at the same time they continue to benefit from all the existing output parsers that come bundled with langchain4j. In addition to replacing the default parser provider, users can override existing parser instances (e.g. for boolean or int), or add parsers for their own custom types (e.g. parsing instances of Foo). Finally, users can add their own ParserFactory instances which will be called if no output parser instance is found, allowing the user to construct a parser instance on the fly for the requried type. This change also adds support for generic List<> and Set<> types, where the element can be any supported parsable type (e.g. a List<Integer> works by default). Furthermore, a List<Foo> works by parsing each line using the registered parser for Foo (either a custom parser, or falling back to the default JsonParser). -- Checklist -- Before submitting this PR, please check the following points: - [X] I have added unit and integration tests for my change - [X] All unit and integration tests in the module I have added/changed are green - [X] All unit and integration tests in the [core](https://github.com/langchain4j/langchain4j/tree/main/langchain4j-core) and [main](https://github.com/langchain4j/langchain4j/tree/main/langchain4j) modules are green - [x] I have added/updated the [documentation](https://github.com/langchain4j/langchain4j/tree/main/docs/docs) - [ ] I have added an example in the [examples repo](https://github.com/langchain4j/langchain4j-examples) (only for "big" features)
ashimoon
force-pushed
the
custom-output-parsing
branch
from
fc21103
to
20a9b1f
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This change gives users the ability to provide custom output parsers to AI services.
-- Context --
Previously, users were constrained to the set of data types that were supported by langchain4j (most primitives, string lists, string sets, enums, and JSON POJOs).
-- Change --
Users can now provide a custom ParserProvider when creating an AI service via the builder, or can provide individual OutputParsers on each service method with the @parser annotation.
This enables users to have full control of the output parsing process, while at the same time they continue to benefit from all the existing output parsers that come bundled with langchain4j.
-- Checklist --
Before submitting this PR, please check the following points:
main modules are green
Issue
Fixes TODO for supporting custom parsing: https://github.com/langchain4j/langchain4j/blob/main/langchain4j/src/main/java/dev/langchain4j/service/DefaultAiServices.java#L100
General checklist
Checklist for adding new model integration
Checklist for adding new embedding store integration
{NameOfIntegration}EmbeddingStoreITthat extends from eitherEmbeddingStoreITorEmbeddingStoreWithFilteringITChecklist for changing existing embedding store integration
{NameOfIntegration}EmbeddingStoreworks correctly with the data persisted using the latest released version of LangChain4j