nutrition_ai 3.2.12

nutrition_ai: ^3.2.12 copied to clipboard

• Bumped the version up to match other platform versions.

• Added a new function updateProductLocalization to customize results for API calls.

/// Sets the `Passio-Product-Localization` header to control language and region
/// for product search results. This affects localized product names, descriptions,
/// and enables region-specific semantic search tools.
///
/// **Parameters**
/// - [localizationCode]: Accepts BCP47 language tags or ISO 639-1 codes
///   (e.g., `"en"` for English, `"fr"` for French, `"de"` for German).
///
/// **Returns**
/// A `bool` indicating whether the localization setting was applied successfully.
///
/// ### Format
/// Examples of valid codes:
/// - `en-US` — English (United States)
/// - `en-GB` — English (United Kingdom)
/// - `ja-JP` — Japanese (Japan)
/// - `es-MX` — Spanish (Mexico)
/// - `fr-FR` — French (France)
/// - `de-DE` — German (Germany)
/// Or just language codes: `en`, `ja`, `es`, `fr`, `de`.
///
/// ### What It Does
/// - **Regional Semantic Search**: Uses region-specific semantic search tools if available.
/// - **Translation**: Translates responses to the target language if different from source.
/// - **Regional Preferences**: Optimizes results for the specified region.
///
/// ### Example
/// ```dart
/// final success = await updateProductLocalization("en-US");
Future<bool> updateProductLocalization(String? localizationCode) async {
  return NutritionAIPlatform.instance.updateProductLocalization(localizationCode);
}

• Added a new function updateSearchConfiguration to customize search behavior.

/// Sets the `Passio-Search-Config` header to customize search behavior, data sources,
/// and filtering options. All search and food endpoints support this header.
///
/// **Parameters**
/// - [config]: A JSON string containing the search configuration.
///
/// ### Filter Configuration
/// **Region Filters**  
/// - Controls geographic filtering and boosting.
/// - Accepts a map with region codes as keys and priorities (1–10) as values.
/// - Priority 1 = highest preference, Priority 10 = lowest.
/// - Valid codes: ISO2 (`us`, `gb`, `jp`), BCP47 (`en-US`, `en-GB`), or language codes (`en`, `ja`, `es`).
/// - Example: `{"regionFilters": {"us": 1, "gb": 3, "ja": 1}}`
/// - `strictRegionMode = true`: Only shows items from specified regions (branded items only).
/// - `strictRegionMode = false`: Boosts matching regions in results.
///
/// **Source Filters**  
/// - Strict filtering by data source.
/// - Accepts an array of source names: `fdc`, `firebase`, `legacy`, `openfood`, `passio`, `passio-foodai`.
/// - Example: `{"sourceFilters": ["fdc", "passio"]}`
///
/// **Type Filters**  
/// - Filter by item category.
/// - Accepts an array: `generic`, `branded`, `recipe`.
/// - Example: `{"typeFilters": ["branded", "generic"]}`
///
/// **Strict Region Mode**  
/// - `true`: Removes items without country data (strict regional search).
/// - `false`: Keeps items without country data (default).
/// - `null` or omitted: Uses global default.
///
/// ### Data Source Configuration
/// **Data Source Configs**  
/// - `passio`: Main Passio DB (included: 100 results, max: 1000).
/// - `edamam_ingredients`: Edamam ingredient DB (included: 10, max: 50).
/// - `edamam_recipes`: Edamam recipe DB (included: 0, max: 25).
///
/// ### Concern Flags
/// **Ignore With Concern Flags**  
/// - Filters out items with specific data quality concerns.
/// - Example: `{"ignoreWithConcernFlags": [1002, 1100, 1211]}`
/// - Common codes:
///   - `1002`: Missing essential nutrients.
///   - `1100`: Invalid serving size.
///   - `1205`: Excessive sodium.
///   - `1211`: Sum of nutrients exceeded 100g.
///   - `1300`: Missing ingredients list.
///
/// ### Complete Example
/// ```json
/// {
///   "dataSourceConfigs": [
///     {"name": "passio", "requestResultCount": 100},
///     {"name": "edamam_ingredients", "requestResultCount": 10}
///   ],
///   "regionFilters": {"us": 1, "gb": 3},
///   "typeFilters": ["branded"],
///   "sourceFilters": ["fdc", "passio"],
///   "strictRegionMode": false,
///   "ignoreWithConcernFlags": [1002, 1100]
/// }
/// ```
///
/// **Notes**
/// - Filters apply after search results are retrieved.
/// - Caching is disabled when filters are active.
/// - Generic items (USDA reference foods) are exempt from region filtering by default.
/// - Filtering reduces token costs by returning fewer results.
Future<void> updateSearchConfiguration(String? config) async {
  return NutritionAIPlatform.instance.updateSearchConfiguration(config);
}

• Added support for Google Play's 16KB compatibility requirement.

Added APIs #

• Added a function to fetch all nutrients of a food item as a raw json string called fetchNutrientsJSON.
• Added field concerns to PassioFoodDataInfo, representing a list of concern flag ids.

Added APIs #

• Added a new function generateMealPlan to generate meal plan from request string.

Future<PassioResult<PassioGeneratedMealPlan>> generateMealPlan(String request)

• Added a new function generateMealPlanPreview to generate meal plan preview from request string.

Future<PassioResult<PassioGeneratedMealPlan>> generateMealPlanPreview(String request)

• Added a new function recognizeSpeechRemoteWithGrouping to recognize speech input and group recognized food items by meal.

Future<PassioResult<PassioRecognitionResult>> recognizeSpeechRemoteWithGrouping(String text)

• Added a new function recognizeImageRemoteWithGrouping to recognize food items from an image and group them by meal.

Future<PassioResult<PassioRecognitionResult>> recognizeImageRemoteWithGrouping(Uint8List bytes, {PassioImageResolution resolution = PassioImageResolution.res_512, String? message})

Removed APIs #

• Removed transformCGRectForm function
• Removed startNutritionFactsDetection function
• Removed stopNutritionFactsDetection function
• Removed PassioNutritionFacts class

• Fixed iOS build issues for smoother project setup.

Added APIs #

• Added a new function fetchUltraProcessingFoodRating that provides the ultra-processing food rating for a given food item:

Future<PassioResult<PassioUPFRating>> fetchUltraProcessingFoodRating(PassioFoodItem foodItem) async {
    return NutritionAIPlatform.instance.fetchUltraProcessingFoodRating(foodItem);
}

class PassioUPFRating {
  final String chainOfThought;
  final List<String> highlightedIngredients;
  final int? rating;
}

• Added field concerns to the PassioFoodMetadata object

final List<int>? concerns;

Deprecated APIs #

• Deprecated the following local recognition functions, which will be removed in a future release. Use recognizeImageRemote instead:
  • startFoodDetection
  • stopFoodDetection
  • startNutritionFactsDetection
  • stopNutritionFactsDetection

Removed APIs #

• Removed volume detection related functionality from the SDK.

Added APIs #

• Added two new parameters to the PassioConfiguration class: proxyUrl and proxyHeaders. If supplied, the SDK will use the proxyUrl value as a base url when executing networking calls, and will append the provided proxyHeaders to the request itself. When proxyUrl is provided, the SDK operates in remoteOnly state.

Added APIs #

• Added a new search function that provides contextually connected items to the search term:

Future<PassioSearchResponse> searchForFoodSemantic(String term) async {
  return NutritionAIPlatform.instance.searchForFoodSemantic(term);
}

• Added a new api for a recipe builder feature, that gives a list of predicted next ingredients based on the current list of ingredients in the recipe:

Future<List<PassioFoodDataInfo>> predictNextIngredients(List<String> currentIngredients) async {
  return NutritionAIPlatform.instance.predictNextIngredients(currentIngredients);
}

Refactored APIs #

• Refactored PassioNutritionFacts: renamed servingSizeQuantity to servingQuantity and servingSizeUnitName to servingUnit; removed servingSize; added weightQuantity and weightUnit.
• Refactored fetchTagsFor and fetchInflammatoryEffectData to accept refCode as a parameter instead of passioID.

Removed APIs #

• Removed function detectFoodIn. Instead, recognizeImageRemote should be used

• Fixed Android build issues for smoother project setup.

⚠️ Breaking Changes #

• Configuration for Android is no longer required. Please remove any code related to finding the AAR file from your top build.gradle file (Project: android).

Refactored APIs #

• The fetchFoodItemForDataInfo method now includes optional servingQuantity and servingUnit parameters to provide details based on the specified serving size:

Future<PassioFoodItem?> fetchFoodItemForDataInfo(PassioFoodDataInfo passioFoodDataInfo, {double? servingQuantity, String? servingUnit})

Added APIs #

• Added the reportFoodItem function to report incorrect food items:

Future<PassioResult<bool>> reportFoodItem({String refCode = '', String productCode = '', List<String>? notes})

• Added the submitUserCreatedFood function to submit user-created food items:

Future<PassioResult<bool>> submitUserCreatedFood(PassioFoodItem item)

Fixed #

• Resolved an issue with the iOS build process that was causing build failures in certain environments.

3.2.0+1 #

Fixed #

• Resolved an issue with the iOS build process that was causing build failures in certain environments.

Added APIs #

• Added function recognizeNutritionFactsRemote that can parse the nutrition facts table from an image and return a PassioFoodItem with the scanned nutrients

Future<PassioFoodItem?> recognizeNutritionFactsRemote(Uint8List bytes, {PassioImageResolution resolution = PassioImageResolution.res_1080})

• Added support for localized content. Using updateLanguage with a two digit ISO 639-1 language code will transform the food names and serving sizes in the SDK responses.

Future<bool> updateLanguage(String languageCode)

• Added remoteOnly flag to the PassioConfiguration class. With this flag enabled, the SDK won't download the files needed for local recognition. In this case only remote recognition is possible.

class PassioConfiguration{
  ...
  final bool remoteOnly;
}

Refactored APIs #

• recognizeImageRemote can now scan barcodes and nutrition facts. The PassioAdvisorFoodInfo class has been augmented to handle these responses

class PassioAdvisorFoodInfo {
  final PassioFoodDataInfo? foodDataInfo;
  final PassioFoodItem? packagedFoodItem;
  final String portionSize;
  final PassioFoodResultType resultType;
  final String recognisedName;
  final double weightGrams;
}

enum PassioFoodResultType {
  foodItem,
  barcode,
  nutritionFacts,
}

• Property tags was added to PassioFoodDataInfo

class PassioFoodDataInfo {
    ...
    final List<String>? tags;
}

• Vitamin A RAE was added to the PassioNutrients class.

class PassioNutrients {
  ...
  final UnitMass? _vitaminARAE;
}

• Improved scanning performance on Android

Added APIs #

• Added fiber value to PassioSearchNutritionPreview

Added APIs #

• With the new key comes the new pricing model using tokens. Although all token usage can be seen in the portal, the SDK itself offers a listener to count the tokens being used by the SDK.

void setAccountListener(PassioAccountListener? listener)

abstract interface class PassioAccountListener {
    void onTokenBudgetUpdate(PassioTokenBudget tokenBudget);
}

class PassioTokenBudget {
    final String apiName;
    final int budgetCap;
    final int periodUsage;
    final int tokensUsed;
}

• Control Flashlight: Manage the flashlight with:

/// [enabled] A required parameter to turn the flashlight on or off.
/// [level] An optional parameter specifying the intensity level of the flashlight, with a default value of 1. Note that the level parameter is only applicable on iOS for now.
Future<void> enableFlashlight({required bool enabled, double level = 1})

• Get Camera Zoom Levels: Retrieve the minimum and maximum zoom levels with:

Future<PassioCameraZoomLevel> getMinMaxCameraZoomLevel()

class PassioCameraZoomLevel {
  final double? minZoomLevel;
  final double? maxZoomLevel;
}

• Set Camera Zoom Level: Adjust the camera zoom level:

/// [zoomLevel] Sets the desired zoom level. Ensure it’s within the range obtained from ```getMinMaxCameraZoomLevel()```.
Future<void> setCameraZoomLevel({required double zoomLevel})

Refactored APIs #

• Updated the fetchFoodItemForDataInfo method to include an optional weightGrams parameter:

/// [weightGrams] An optional parameter that provides information specific to the specified weight in grams.
Future<PassioFoodItem?> fetchFoodItemForDataInfo(PassioFoodDataInfo passioFoodDataInfo,{double? weightGrams})

• The NutritionAdvisor no longer needs a separate key for configuration. The PassioSDK needs to be configured in order to call the NutritionAdvisor APIs. Also, any previous NutritionAdvisor keys won't work with this version, a new key needs to be obtained at https://www.passio.ai/

Added APIs #

• Added PassioImageResolution resolution as a parameter of the recognizeImageRemote function. This enables the caller to set the target resolution of the image uploaded to the server. Smaller resolutions will result in faster response times, while higher resolutions should provide more accurate results.

enum PassioImageResolution {
  res_512,
  res_1080,
  full,
}

• Added a function to retrieve possible hidden ingredients for a given food name.

Future<PassioResult<List<PassioAdvisorFoodInfo>>> fetchHiddenIngredients(String foodName)

• Added a function to retrieve possible visual alternatives for a given food name.

Future<PassioResult<List<PassioAdvisorFoodInfo>>> fetchVisualAlternatives(String foodName)

• Added a function to retrieve possible ingredients if a more complex food for a given food name.

Future<PassioResult<List<PassioAdvisorFoodInfo>>> fetchPossibleIngredients(String foodName)

New Features #

Nutrition AI Advisor

• To access the API of the Advisor use NutritionAdvisor.instance
• The method Future<PassioResult<PassioAdvisorResponse>> fetchIngredients(PassioAdvisorResponse response) works only with the response from sendMessage(). It returns data if the PassioAdvisorResponse contains tools.

Future<PassioResult> configure(String key)

Future<PassioResult> initConversation()

Future<PassioResult<PassioAdvisorResponse>> sendMessage(String message)

Future<PassioResult<PassioAdvisorResponse>> sendImage(Uint8List bytes)

Future<PassioResult<PassioAdvisorResponse>> fetchIngredients(PassioAdvisorResponse response)

Nutrition Facts Detection

void startNutritionFactsDetection(NutritionFactsRecognitionListener listener)

Future<void> stopNutritionFactsDetection()

abstract interface class NutritionFactsRecognitionListener {
  void onNutritionFactsRecognized(PassioNutritionFacts? nutritionFacts, String? text);
}

Refactored APIs #

• The FoodDetectionConfiguration object no longer has the detectNutritionFacts parameter.

Added APIs #

• Added speech recognition API that retrieves a list of recognized foods from the input text query.

Future<List<PassioSpeechRecognitionModel>> recognizeSpeechRemote(String text)

• Added LLM image detection, that works remotely through Passio's backend.

Future<List<PassioAdvisorFoodInfo>> recognizeImageRemote(Uint8List bytes)

• Added a function to retrieve a PassioFoodItem for a v2 PassioID

Future<PassioFoodItem?> fetchFoodItemLegacy(PassioID passioID)

Added APIs #

• Added two functions to fetch a list of meal plans and fetch a meal plan for a certain day.

Future<List<PassioMealPlan>> fetchMealPlans()
Future<List<PassioMealPlanItem>> fetchMealPlanForDay(String mealPlanLabel, int day)

• Added refCode as an attribute to the PassioFoodItem and PassioIngredient classes.
• Added method to fetch a food item using just the refCode attribute

Future<PassioFoodItem?> fetchFoodItemForRefCode(String refCode)

Refactored APIs #

• Refactored PassioSearchNutritionPreview

class PassioSearchNutritionPreview {
  final int calories;
  final double carbs;
  final double fat;
  final double protein;
  final double servingQuantity;
  final String servingUnit;
  final double weightQuantity;
  final String weightUnit;
}

• Renamed PassioSearchResult to PassioFoodDataInfo.
• Renamed fetchFoodItemForSearchResult to fetchFoodItemForDataInfo.
• Renamed MealTime to PassioMealTime.

Deprecated APIs #

• Removed fetchFoodItemForSuggestion. Now fetchFoodItemForDataInfo is used.

Added APIs #

• Added API to fetch suggestion for a certain meal time. To fetch the full nutritional data for a suggestion item use fetchFoodItemForSuggestion.

enum MealTime {
  breakfast,
  lunch,
  dinner,
  snack,
}

Future<List<PassioSearchResult>> fetchSuggestions(MealTime mealTime)

Future<PassioFoodItem?> fetchFoodItemForSuggestion(PassioSearchResult suggestion)

Added Micronutrients: #

• Zinc
• Selenium
• Folic acid
• Chromium
• Vitamin-K Phylloquinone
• Vitamin-K Menaquinone4
• Vitamin-K Dihydrophylloquinone

Refactored APIs #

• Renamed fetchSearchResult to fetchFoodItemForSearchResult.

• Version 3 of the Passio SDK introduces major changes to the nutritional data class and the search functionality. The SDK no longer supports offline work, there is no more local database.

Deprecated APIs #

• lookupPassioAttributesFor has been removed because it was querying the local database.
• PassioIDAttributes, PassioFoodItemData and PassioFoodRecipe have been removed. The new data model that will handle nutritional data is called PassioFoodItem

Refactored APIs #

• searchForFood now returns PassioSearchResponse. In PassioSearchResponse you will get list of PassioSearchResult and a list of search options. The PassioSearchResult represent a specific food item associated with the search term.
• fetchPassioIDAttributesForBarcode and fetchPassioIDAttributesForPackagedFood have been replaced with fetchFoodItemForProductCode than now returns a PassioFoodItem result
• DetectedCandidate now has an attribute called foodName
• FoodRecognitionListener method onRecognitionResults can now return nullable FoodCandidates
• fetchNutrientsFor has been renamed to fetchInflammatoryEffectData, and PassioNutrient has been renamed to InflammatoryEffectData

Added APIs #

• fetchSearchResult returns a PassioFoodItem object for a given PassioSearchResult
• fetchFoodItemForPassioID returns a PassioFoodItem object for a given passioID corresponding to a result from the visual detection
• Added class PassioSearchResult that represents a result from the searchForFood function
• Added class PassioFoodItem that represent a food item from the Passio Nutritional database. It has a list of PassioIngredients, with their respective PassioFoodAmounts and PassioNutrients

• Added fetchNutrientsFor method to retrieve a map of nutrients for a 100 grams of a specific food item.

• No API changes

• Add PlatformImage parameter to the recognitionResults of the FoodRecognitionListener. It represents the image that was analyzed by the camera recognition system.

• Fixed iOS camera not stopping issue

• Fixed Android camera not stopping issue

• Removed app bar from PassioPreview
• Fixed issue with food items that don't have an origin
• Fixed Android issue with caching PassioFoodItemData

• Clean up of the example app code
• Added distribution github repository with the source code and issue tracker

• Fix issue with PassioSDKError

• Add detectFoodIn to run detection from a single image.
• Add fetchTagsFor to fetch the list of tags for a food item.
• Add iconURLFor to get the url for downloading an icon for a food item.
• Add transformCGRectForm to calculate the bouding box of scanned food item.
• Add volume detection using VolumeDetectionMode
• Add android:required="false" to camera manifest features in Android.
• Refactor lookupIconsFor to return a list of PassioFoodIcons objects.
• Refactor PassioPreview widget to remove all unnecessary widgets.

• Add missing .aar dependency

• Alpha release of the Nutrition AI SDK for Flutter.
• Visual, barcode and packaged food detection using the camera of the device.
• Fetching nutritional data of the scanned food.
• Searching by food name.
• Fetching the icon for the food item.