[ Index ]

PHP Cross Reference of WordPress Trunk (Updated Daily)

Search

title

Body

[close]

/wp-includes/php-ai-client/src/ -> AiClient.php (summary)

(no description)

File Size: 513 lines (24 kb)
Included or required:0 times
Referenced: 0 times
Includes or requires: 0 files

Defines 1 class

AiClient:: (22 methods):
  defaultRegistry()
  setEventDispatcher()
  getEventDispatcher()
  setCache()
  getCache()
  isConfigured()
  prompt()
  input()
  generateResult()
  generateTextResult()
  generateImageResult()
  convertTextToSpeechResult()
  generateSpeechResult()
  generateVideoResult()
  generateEmbeddingResult()
  generateEmbedding()
  generateEmbeddings()
  message()
  validateModelOrConfigParameter()
  getConfiguredPromptBuilder()
  getConfiguredEmbeddingBuilder()
  applyModelOrConfig()


Class: AiClient  - X-Ref

Main AI Client class providing both fluent and traditional APIs for AI operations.

This class serves as the primary entry point for AI operations, offering:
- Fluent API for easy-to-read chained method calls
- Traditional API for array-based configuration (WordPress style)
- Integration with provider registry for model discovery
- Support for three model specification approaches

All model requirements analysis and capability matching is handled
automatically by the PromptBuilder, which provides intelligent model
discovery based on prompt content and configuration.

## Model Specification Approaches

### 1. Specific Model Instance
Use a specific ModelInterface instance when you know exactly which model to use:
```php
$model = $registry->getProvider('openai')->getModel('gpt-4');
$result = AiClient::generateTextResult('What is PHP?', $model);
```

### 2. ModelConfig for Auto-Discovery
Use ModelConfig to specify requirements and let the system discover the best model:
```php
$config = new ModelConfig();
$config->setTemperature(0.7);
$config->setMaxTokens(150);

$result = AiClient::generateTextResult('What is PHP?', $config);
```

### 3. Automatic Discovery (Default)
Pass null or omit the parameter for intelligent model discovery based on prompt content:
```php
// System analyzes prompt and selects appropriate model automatically
$result = AiClient::generateTextResult('What is PHP?');
$imageResult = AiClient::generateImageResult('A sunset over mountains');
```

## Fluent API Examples
```php
// Fluent API with automatic model discovery
$result = AiClient::prompt('Generate an image of a sunset')
->usingTemperature(0.7)
->generateImageResult();

// Fluent API with specific model
$result = AiClient::prompt('What is PHP?')
->usingModel($specificModel)
->usingTemperature(0.5)
->generateTextResult();

// Fluent API with model configuration
$result = AiClient::prompt('Explain quantum physics')
->usingModelConfig($config)
->generateTextResult();
```

defaultRegistry()   X-Ref
Gets the default provider registry instance.

return: ProviderRegistry The default provider registry.

setEventDispatcher(?EventDispatcherInterface $dispatcher)   X-Ref
Sets the event dispatcher for prompt lifecycle events.

The event dispatcher will be used to dispatch BeforeGenerateResultEvent and
AfterGenerateResultEvent during prompt generation.

param: EventDispatcherInterface|null $dispatcher The event dispatcher, or null to disable.
return: void

getEventDispatcher()   X-Ref
Gets the event dispatcher for prompt lifecycle events.

return: EventDispatcherInterface|null The event dispatcher, or null if not set.

setCache(?CacheInterface $cache)   X-Ref
Sets the PSR-16 cache for storing and retrieving cached data.

The cache can be used to store AI responses and other data to avoid
redundant API calls and improve performance.

param: CacheInterface|null $cache The PSR-16 cache instance, or null to disable caching.
return: void

getCache()   X-Ref
Gets the PSR-16 cache instance.

return: CacheInterface|null The cache instance, or null if not set.

isConfigured($availabilityOrIdOrClassName)   X-Ref
Checks if a provider is configured and available for use.

Supports multiple input formats for developer convenience:
- ProviderAvailabilityInterface: Direct availability check
- string (provider ID): e.g., AiClient::isConfigured('openai')
- string (class name): e.g., AiClient::isConfigured(OpenAiProvider::class)

When using string input, this method leverages the ProviderRegistry's centralized
dependency management, ensuring HttpTransporter and authentication are properly
injected into availability instances.

param: ProviderAvailabilityInterface|string|class-string<ProviderInterface> $availabilityOrIdOrClassName
return: bool True if the provider is configured and available, false otherwise.

prompt($prompt = null, ?ProviderRegistry $registry = null)   X-Ref
Creates a new prompt builder for fluent API usage.

Returns a PromptBuilder instance configured with the specified or default registry.
The traditional API methods in this class delegate to PromptBuilder
for all generation logic.

param: Prompt $prompt Optional initial prompt content.
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: PromptBuilder The prompt builder instance.

input($input = null, ?ProviderRegistry $registry = null)   X-Ref
Creates an embedding builder for fluent embedding generation.

Accepts a single input or a list of inputs; each input is embedded independently. Chain an
embedding generation method such as generateEmbedding() or generateEmbeddings() to produce
one vector per input.

A model must be specified on the returned builder via usingModel() or usingProviderModel().
Embedding vectors are only comparable to other vectors from the same model, so no model is
selected automatically.

param: EmbeddingInput|list<EmbeddingInput>|null $input Optional initial input(s) to embed.
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: EmbeddingBuilder The embedding builder instance.

generateResult($prompt, $modelOrConfig, ?ProviderRegistry $registry = null)   X-Ref
Generates content using a unified API that automatically detects model capabilities.

When no model is provided, this method delegates to PromptBuilder for intelligent
model discovery based on prompt content and configuration. When a model is provided,
it infers the capability from the model's interfaces and delegates to the capability-based method.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig $modelOrConfig Specific model to use, or model configuration
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

generateTextResult($prompt, $modelOrConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates text using the traditional API approach.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig Optional specific model to use,
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

generateImageResult($prompt, $modelOrConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates an image using the traditional API approach.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig Optional specific model to use,
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

convertTextToSpeechResult($prompt, $modelOrConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Converts text to speech using the traditional API approach.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig Optional specific model to use,
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

generateSpeechResult($prompt, $modelOrConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates speech using the traditional API approach.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig Optional specific model to use,
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

generateVideoResult($prompt, $modelOrConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates a video using the traditional API approach.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig Optional specific model to use,
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: GenerativeAiResult The generation result.

generateEmbeddingResult($input, $model, ?ModelConfig $modelConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates embeddings using the traditional API approach.

A model is required. Embedding vectors are only comparable to other vectors from the same
model, so no model is selected automatically.

param: EmbeddingInput|list<EmbeddingInput> $input The input(s) to embed.
param: ModelInterface|ProviderModelTuple $model The model to use.
param: ModelConfig|null $modelConfig Optional model configuration to apply.
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: EmbeddingResult The embedding result.

generateEmbedding($input, $model, ?ModelConfig $modelConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates an embedding using the traditional API approach.

A model is required. Embedding vectors are only comparable to other vectors from the same
model, so no model is selected automatically.

param: EmbeddingInput $input The input to embed.
param: ModelInterface|ProviderModelTuple $model The model to use.
param: ModelConfig|null $modelConfig Optional model configuration to apply.
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: Embedding The generated embedding vector.

generateEmbeddings(array $inputs, $model, ?ModelConfig $modelConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Generates embeddings for a list of inputs using the traditional API approach.

A model is required. Embedding vectors are only comparable to other vectors from the same
model, so no model is selected automatically.

param: list<EmbeddingInput> $inputs The inputs to embed.
param: ModelInterface|ProviderModelTuple $model The model to use.
param: ModelConfig|null $modelConfig Optional model configuration to apply.
param: ProviderRegistry|null $registry Optional custom registry. If null, uses default.
return: list<Embedding> The generated embedding vectors.

message(?string $text = null)   X-Ref
Creates a new message builder for fluent API usage.

This method will be implemented once MessageBuilder is available.
MessageBuilder will provide a fluent interface for constructing complex
messages with multiple parts, attachments, and metadata.

param: string|null $text Optional initial message text.
return: object MessageBuilder instance (type will be updated when MessageBuilder is available).

validateModelOrConfigParameter($modelOrConfig)   X-Ref
Validates that parameter is ModelInterface, ModelConfig, or null.

param: mixed $modelOrConfig The parameter to validate.
return: void

getConfiguredPromptBuilder($prompt, $modelOrConfig, ?ProviderRegistry $registry = null)   X-Ref
Configures PromptBuilder based on model/config parameter type.

param: Prompt $prompt The prompt content.
param: ModelInterface|ModelConfig|null $modelOrConfig The model or config parameter.
param: ProviderRegistry|null $registry Optional custom registry to use.
return: PromptBuilder Configured prompt builder.

getConfiguredEmbeddingBuilder($input, $model, ?ModelConfig $modelConfig = null, ?ProviderRegistry $registry = null)   X-Ref
Configures an EmbeddingBuilder with the required model and optional configuration.

param: EmbeddingInput|list<EmbeddingInput> $input The input(s) to embed.
param: mixed $model The model to use, expected to be a ModelInterface instance or a
param: ModelConfig|null $modelConfig Optional model configuration to apply.
param: ProviderRegistry|null $registry Optional custom registry to use.
return: EmbeddingBuilder Configured embedding builder.

applyModelOrConfig($builder, $modelOrConfig)   X-Ref
Applies a model or model configuration to a builder.

Works with any builder that exposes the shared model resolution methods
(see {@see \WordPress\AiClient\Builders\Traits\ModelResolutionTrait}).

param: T $builder The builder to configure.
param: ModelInterface|ModelConfig|null $modelOrConfig Specific model, model configuration,
return: T The configured builder.



Generated : Sun Nov 22 08:20:01 2026 Cross-referenced by PHPXref