Initial service commit

.env

@@ -0,0 +1,20 @@
+# Password for the 'elastic' user (at least 6 characters)
+# This needs to be set in the environment variables
+# ELASTIC_PASSWORD=
+
+# Version of Elastic products
+STACK_VERSION=8.3.3
+
+# Set the cluster name
+CLUSTER_NAME=docker-cluster
+
+# Set to 'basic' or 'trial' to automatically start the 30-day trial
+LICENSE=basic
+
+# Port to expose Elasticsearch HTTP API to the host
+ES_PORT=9200
+#ES_PORT=127.0.0.1:9200
+
+# Increase or decrease based on the available host memory (in bytes)
+# 1GB works well, 2GB and above leads to lower latency
+MEM_LIMIT=2147483648

.gitignore

@@ -102,7 +102,6 @@ celerybeat.pid
 *.sage.py
 
 # Environments
-.env
 .venv
 env/
 venv/
@@ -127,3 +126,7 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
+
+# Pycharm
+.idea/
+.DS_Store

Dockerfile

@@ -0,0 +1,7 @@
+# Instructions from https://fastapi.tiangolo.com/deployment/docker/
+FROM python:3.9
+WORKDIR /code
+COPY ./requirements.txt /code/requirements.txt
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+COPY ./app /code/app
+CMD ["uvicorn", "app.api:app", "--proxy-headers", "--host", "0.0.0.0", "--port", "8000"]

README.md

@@ -1,2 +1,55 @@
 # openfoodfacts-search
 Open Food Facts Search API V3 using ElasticSearch - https://wiki.openfoodfacts.org/Search_API_V3
+
+This API is currently in development. It is not serving any production traffic. The [Work Plan](https://wiki.openfoodfacts.org/Search_API_V3#Work_Plan) will be updated as development continues
+
+### Organization
+The main file is `api.py`, and the Product schema is in `models/product.py`. 
+
+The `scripts/` directory contains various scripts for manual validation, constructing the product schema, importing, etc.
+
+### Running locally
+Firstly, make sure your environment is configured:
+```commandline
+export ELASTIC_PASSWORD=PASSWORD_HERE
+```
+
+Then start docker:
+```console
+docker-compose up -d
+```
+
+Docker spins up:
+- Two elasticsearch nodes
+- [Elasticvue](https://elasticvue.com/)
+- The search service on port 8000
+
+You will then need to import from CSV (see instructions below).
+
+### Development
+For development, you have two options for running the service:
+1. Docker
+2. Locally
+
+To develop on docker, make the changes you need, then build the image and compose by running:
+```console
+docker build -t off_search_image .
+docker-compose up -d
+```
+
+However, this tends to be slower than developing locally.
+
+To develop locally, create a venv, install dependencies, then run the service:
+```console
+virtualenv .
+source venv/bin/activate
+pip install -r requirements.txt
+uvicorn app.api:app --reload --port=8001
+```
+Note that it's important to use port 8001, as port 8000 will be used by the docker version of the search service. 
+
+
+### Helpful commands:
+To import data from the [CSV export](https://world.openfoodfacts.org/data):
+```console
+python scripts/perform_import.py --filename=/path/to/file.csv

app/api.py

@@ -0,0 +1,45 @@
+from elasticsearch_dsl import Q
+from fastapi import FastAPI, HTTPException
+
+from app.models.product import Product
+from app.models.request import AutocompleteRequest, SearchRequest
+from app.utils import connection, constants, response
+
+app = FastAPI()
+connection.get_connection()
+
+
+# TODO: Remove this commented out code, so that it's not confusing about where the current GET API is served
+# (retaining temporarily as a proof of concept)
+# @app.get("/{barcode}")
+# def get_product(barcode: str):
+#     results = Product.search().query("match", code=barcode).execute()
+#     results_dict = [r.to_dict() for r in results]
+#
+#     if not results_dict:
+#         raise HTTPException(status_code=404, detail="Barcode not found")
+#
+#     product = results_dict[0]
+#     return product
+
+@app.post("/autocomplete")
+def autocomplete(request: AutocompleteRequest):
+    # TODO: This function needs unit testing
+    if not request.search_fields:
+        request.search_fields = constants.AUTOCOMPLETE_FIELDS
+    for field in request.search_fields:
+        if field not in constants.AUTOCOMPLETE_FIELDS:
+            raise HTTPException(status_code=400, detail="Invalid field: {}".format(field))
+
+    match_queries = []
+    for field in request.search_fields:
+        match_queries.append(Q('match', **{field: request.text}))
+
+    results = Product.search().query('bool', should=match_queries).extra(size=request.get_num_results()).execute()
+    resp = response.create_response(results, request)
+    return resp
+
+
+@app.post("/search")
+def search(request: SearchRequest):
+    raise NotImplementedError()

app/models/product.py

@@ -0,0 +1,241 @@
+from elasticsearch_dsl import Document, Date, Double, Keyword, Text, Integer
+
+from app.utils import constants
+from app.utils.analyzers import autocomplete
+
+
+class Product(Document):
+    """
+    This should mirror the fields here: /~https://github.com/openfoodfacts/openfoodfacts-server/blob/main/html/data/data-fields.txt
+    Use scripts/generate_product_from_data_fields.py to regenerate from data-fields.txt, but be careful for manual
+    adjustments
+    """
+
+    class Index:
+        name = constants.INDEX_ALIAS
+        settings = {
+            "number_of_shards": 4,
+        }
+
+    # barcode of the product (can be EAN-13 or internal codes for some food stores), for products without a barcode, Open Food Facts assigns a number starting with the 200 reserved prefix
+    code = Keyword()
+    # url of the product page on Open Food Facts
+    url = Keyword()
+    # contributor who first added the product
+    creator = Keyword()
+    # date that the product was added (UNIX timestamp format)
+    created_t = Integer()
+    # date that the product was added (iso8601 format: yyyy-mm-ddThh:mn:ssZ)
+    created_datetime = Date()
+    # date that the product page was last modified
+    last_modified_t = Integer()
+    last_modified_datetime = Date()
+    # name of the product
+    product_name = Text(analyzer='snowball', fields={'autocomplete': Text(analyzer=autocomplete), 'raw': Keyword()})
+    generic_name = Keyword()
+    # quantity and unit
+    quantity = Keyword()
+    # shape, material
+    packaging = Keyword()
+    packaging_tags = Text(multi=True)
+    brands = Text(analyzer='snowball', fields={'autocomplete': Text(analyzer=autocomplete), 'raw': Keyword()})
+    brands_tags = Text(multi=True)
+    categories = Text(analyzer='snowball', fields={'autocomplete': Text(analyzer=autocomplete), 'raw': Keyword()})
+    categories_tags = Text(multi=True)
+    categories_fr = Keyword()
+    # origins of ingredients
+    origins = Keyword()
+    origins_tags = Text(multi=True)
+    # places where manufactured or transformed
+    manufacturing_places = Keyword()
+    manufacturing_places_tags = Text(multi=True)
+    labels = Keyword()
+    labels_tags = Text(multi=True)
+    labels_fr = Keyword()
+    emb_codes = Keyword()
+    emb_codes_tags = Text(multi=True)
+    # coordinates corresponding to the first packaging code indicated
+    first_packaging_code_geo = Keyword()
+    cities = Keyword()
+    cities_tags = Text(multi=True)
+    purchase_places = Keyword()
+    stores = Keyword()
+    # list of countries where the product is sold
+    countries = Keyword()
+    countries_tags = Text(multi=True)
+    countries_fr = Keyword()
+    ingredients_text = Keyword()
+    traces = Keyword()
+    traces_tags = Text(multi=True)
+    # serving size in g
+    serving_size = Keyword()
+    # indicates if the nutrition facts are indicated on the food label
+    no_nutriments = Keyword()
+    # number of food additives
+    additives_n = Keyword()
+    additives = Keyword()
+    additives_tags = Text(multi=True)
+    ingredients_from_palm_oil_n = Keyword()
+    ingredients_from_palm_oil = Keyword()
+    ingredients_from_palm_oil_tags = Text(multi=True)
+    ingredients_that_may_be_from_palm_oil_n = Keyword()
+    ingredients_that_may_be_from_palm_oil = Keyword()
+    ingredients_that_may_be_from_palm_oil_tags = Text(multi=True)
+    # nutrition grade ('a' to 'e'). see https://fr.openfoodfacts.org/nutriscore
+    nutrition_grade_fr = Keyword()
+    main_category = Keyword()
+    main_category_fr = Keyword()
+    image_url = Keyword()
+    image_small_url = Keyword()
+    energy_100g = Double()
+    energy_kj_100g = Double()
+    energy_kcal_100g = Double()
+    proteins_100g = Double()
+    casein_100g = Double()
+    serum_proteins_100g = Double()
+    nucleotides_100g = Double()
+    carbohydrates_100g = Double()
+    sugars_100g = Double()
+    sucrose_100g = Double()
+    glucose_100g = Double()
+    fructose_100g = Double()
+    lactose_100g = Double()
+    maltose_100g = Double()
+    maltodextrins_100g = Double()
+    starch_100g = Double()
+    polyols_100g = Double()
+    fat_100g = Double()
+    saturated_fat_100g = Double()
+    butyric_acid_100g = Double()
+    caproic_acid_100g = Double()
+    caprylic_acid_100g = Double()
+    capric_acid_100g = Double()
+    lauric_acid_100g = Double()
+    myristic_acid_100g = Double()
+    palmitic_acid_100g = Double()
+    stearic_acid_100g = Double()
+    arachidic_acid_100g = Double()
+    behenic_acid_100g = Double()
+    lignoceric_acid_100g = Double()
+    cerotic_acid_100g = Double()
+    montanic_acid_100g = Double()
+    melissic_acid_100g = Double()
+    monounsaturated_fat_100g = Double()
+    polyunsaturated_fat_100g = Double()
+    omega_3_fat_100g = Double()
+    alpha_linolenic_acid_100g = Double()
+    eicosapentaenoic_acid_100g = Double()
+    docosahexaenoic_acid_100g = Double()
+    omega_6_fat_100g = Double()
+    linoleic_acid_100g = Double()
+    arachidonic_acid_100g = Double()
+    gamma_linolenic_acid_100g = Double()
+    dihomo_gamma_linolenic_acid_100g = Double()
+    omega_9_fat_100g = Double()
+    oleic_acid_100g = Double()
+    elaidic_acid_100g = Double()
+    gondoic_acid_100g = Double()
+    mead_acid_100g = Double()
+    erucic_acid_100g = Double()
+    nervonic_acid_100g = Double()
+    trans_fat_100g = Double()
+    cholesterol_100g = Double()
+    fiber_100g = Double()
+    sodium_100g = Double()
+    # % vol of alcohol
+    alcohol_100g = Double()
+    vitamin_a_100g = Double()
+    vitamin_d_100g = Double()
+    vitamin_e_100g = Double()
+    vitamin_k_100g = Double()
+    vitamin_c_100g = Double()
+    vitamin_b1_100g = Double()
+    vitamin_b2_100g = Double()
+    vitamin_pp_100g = Double()
+    vitamin_b6_100g = Double()
+    vitamin_b9_100g = Double()
+    vitamin_b12_100g = Double()
+    # also known as Vitamine B8
+    biotin_100g = Double()
+    # also known as Vitamine B5
+    pantothenic_acid_100g = Double()
+    silica_100g = Double()
+    bicarbonate_100g = Double()
+    potassium_100g = Double()
+    chloride_100g = Double()
+    calcium_100g = Double()
+    phosphorus_100g = Double()
+    iron_100g = Double()
+    magnesium_100g = Double()
+    zinc_100g = Double()
+    copper_100g = Double()
+    manganese_100g = Double()
+    fluoride_100g = Double()
+    selenium_100g = Double()
+    chromium_100g = Double()
+    molybdenum_100g = Double()
+    iodine_100g = Double()
+    caffeine_100g = Double()
+    taurine_100g = Double()
+    # pH (no unit)
+    ph_100g = Double()
+    # % of fruits, vegetables and nuts (excluding potatoes, yams, manioc)
+    fruits_vegetables_nuts_100g = Double()
+    # carbon footprint (as indicated on the packaging of some products)
+    carbon_footprint_100g = Double()
+    # Nutri-Score - Nutrition score derived from the UK FSA score and adapted for the French market (formula defined by the team of Professor Hercberg)
+    nutrition_score_fr_100g = Double()
+    # nutrition score defined by the UK Food Standards Administration (FSA)
+    nutrition_score_uk_100g = Double()
+    countries_en = Keyword()
+    pnns_groups_1 = Keyword()
+    pnns_groups_2 = Keyword()
+    states = Keyword()
+    states_tags = Text(multi=True)
+    states_en = Keyword()
+    ecoscore_grade = Keyword()
+    image_nutrition_url = Keyword()
+    image_nutrition_small_url = Keyword()
+    origins_en = Keyword()
+    ingredients_tags = Text(multi=True)
+    image_ingredients_url = Keyword()
+    image_ingredients_small_url = Keyword()
+    salt_100g = Double()
+    fruits_vegetables_nuts_estimate_100g = Double()
+    fruits_vegetables_nuts_estimate_from_ingredients_100g = Double()
+    fruits_vegetables_nuts_dried_100g = Double()
+    categories_en = Keyword()
+    nutriscore_score = Keyword()
+    nutriscore_grade = Keyword()
+    food_groups = Keyword()
+    food_groups_tags = Text(multi=True)
+    food_groups_en = Keyword()
+    ecoscore_score = Keyword()
+    main_category_en = Keyword()
+    additives_en = Keyword()
+    nova_group = Keyword()
+    labels_en = Keyword()
+    allergens = Keyword()
+    packaging_en = Keyword()
+    packaging_text = Keyword()
+    serving_quantity = Keyword()
+    carbon_footprint_from_meat_or_fish_100g = Double()
+    energy_from_fat_100g = Double()
+    folates_100g = Double()
+    soluble_fiber_100g = Double()
+    insoluble_fiber_100g = Double()
+    phylloquinone_100g = Double()
+    cocoa_100g = Double()
+    choline_100g = Double()
+    inositol_100g = Double()
+    collagen_meat_protein_ratio_100g = Double()
+    beta_carotene_100g = Double()
+    chlorophyl_100g = Double()
+    glycemic_index_100g = Double()
+    water_hardness_100g = Double()
+    beta_glucan_100g = Double()
+    carnitine_100g = Double()
+    traces_en = Keyword()
+    brand_owner = Keyword()
+    abbreviated_product_name = Keyword()
+    allergens_en = Keyword()