feat: migrate from flask to fastapi (#56)

Author: NoOPeEKS

.gitlab-ci.yml

@@ -33,13 +33,13 @@ build:
   stage: build
   script:
     - export REPO_NAME=$CI_REGISTRY_IMAGE/conjugador-elastic
-    - export IMAGE_NAME=$REPO_NAME:$CI_COMMIT_REF_NAME
+    - export IMAGE_NAME=$REPO_NAME:$CI_COMMIT_REF_SLUG
     - export DOCKER_PATH=docker/elasticsearch/Dockerfile
     - docker build -f $DOCKER_PATH --tag $IMAGE_NAME .
     - docker push $IMAGE_NAME
 
     - export REPO_NAME=$CI_REGISTRY_IMAGE/conjugador-webserver
-    - export IMAGE_NAME=$REPO_NAME:$CI_COMMIT_REF_NAME
+    - export IMAGE_NAME=$REPO_NAME:$CI_COMMIT_REF_SLUG
     - export DOCKER_PATH=docker/webserver/Dockerfile
     - docker build -f $DOCKER_PATH --tag $IMAGE_NAME .
     - docker push $IMAGE_NAME

README.md

@@ -7,7 +7,7 @@ content from the spelling dictionary.
 
 1. extract.py reads diccionary file and extracts the verbs in JSON format (into data/jsons)
 2. index_creation.py reads the jsons and creates multiple indices in Elasticsearch
-3. Flask application at web/ serves the content
+3. FastAPI application at web/ serves the content
 
 #  Git clone
 

docker/elasticsearch/Dockerfile

@@ -19,6 +19,7 @@ RUN uv python install $(cat .python-version) && \
 
 ENV discovery.type=single-node
 ENV xpack.security.enabled=false
+ENV logger.level=error
 ENV ES_JAVA_OPTS="-Xms512m -Xmx512m"
 
 # Start Elasticsearch, wait for health, run indexing, then stop
@@ -74,5 +75,6 @@ COPY --from=build /usr/share/elasticsearch/data /usr/share/elasticsearch/data
 RUN chown -R elasticsearch:elasticsearch /usr/share/elasticsearch/data
 ENV discovery.type=single-node
 ENV xpack.security.enabled=false
+ENV logger.level=error
 ENV ES_JAVA_OPTS="-Xms512m -Xmx512m"
 USER elasticsearch

docker/webserver/entrypoint.sh

@@ -1,3 +1,3 @@
 #!/bin/sh
 mkdir -p /var/log/conjugador/
-uv run gunicorn web.web_search:app -b 0.0.0.0:8000 --error-logfile /var/log/conjugador/gnuicorn.log --workers=2
+uv run uvicorn web.main:app --host 0.0.0.0 --port 8000 --log-level error 2>>/var/log/conjugador/uvicorn_error.log

pyproject.toml

@@ -5,11 +5,14 @@ description = "Conjugador is a web application that allows to search and display
 readme = "README.md"
 requires-python = ">=3.12"
 dependencies = [
-    "elasticsearch==8.15.1",
+    "aiofiles>=25.1.0",
+    "elasticsearch[async]==8.15.1",
+    "fastapi>=0.128.0",
     "flask==3.0.2",
     "gunicorn==22.0.0",
     "psutil>=7.1.3",
     "pyuca==1.2",
+    "uvicorn>=0.40.0",
 ]
 
 [dependency-groups]

uv.lock

@@ -0,0 +1,112 @@
+#!/usr/bin/env python3
+# -*- encoding: utf-8 -*-
+#
+# Copyright (c) 2019-2020 Jordi Mas i Hernandez <jmas@softcatala.org>
+#
+# This program is free software; you can redistribute it and/or
+# modify it under the terms of the GNU Lesser General Public
+# License as published by the Free Software Foundation; either
+# version 2.1 of the License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+# Lesser General Public License for more details.
+#
+# You should have received a copy of the GNU Lesser General Public
+# License along with this program; if not, write to the
+# Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+# Boston, MA 02111-1307, USA.
+
+import json
+import logging
+
+from elasticsearch import AsyncElasticsearch
+
+from indexer.firstletter import FirstLetter
+
+
+class Autocomplete:
+    """
+    Autocomplete a word based on the information on the Elasticsearch indices.
+
+    Args:
+        es_client (AsyncElasticsearch): The client to use for the connection.
+    """
+
+    def __init__(self, es_client: AsyncElasticsearch) -> None:
+        """
+        Initializes the Autocomplete class with a preconfigured ES client.
+
+        Args:
+            es_client (AsyncElasticsearch): The client to use for the connection.
+        """
+        self.es_client = es_client
+        self.letter = FirstLetter()
+
+    async def get_results(self, word: str) -> list[dict]:
+        """
+        Gets the results from the prepared query, based on the word to autocomplete.
+
+        Args:
+            word (str): The word from which to autocomplete.
+
+        Returns:
+            list[dict]: A list of dictionaries containing the results.
+        """
+        letter = self.letter.from_word(word)
+        index_name = f"autocomplete-{letter}"
+
+        if not await self.es_client.indices.exists(index=index_name):
+            return []
+
+        query = {
+            "query": {
+                "prefix": {"verb_form.keyword": {"value": word.lower()}}
+            },
+            "sort": [
+                {
+                    "autocomplete_sorting.keyword": {
+                        "order": "asc",
+                    },
+                },
+            ],
+            "size": 1000,
+            "_source": ["verb_form", "infinitive", "url"],
+        }
+
+        try:
+            resp = await self.es_client.search(index=index_name, body=query)
+            results = [hit["_source"] for hit in resp["hits"]["hits"]]
+        except Exception as e:
+            logging.error(f"Error searching index '{index_name}': {e}")
+            results = []
+
+        return results
+
+    async def get_json(self, word: str) -> tuple[str, int]:
+        """
+        Gets a stringified JSON for all the results found for the autocomplete word.
+
+        Args:
+            word (str): The word from which to autocomplete.
+
+        Returns:
+            tuple[str, int]: A tuple containing the stringified JSON and the status code.
+        """
+        OK = 200
+        status = OK
+        results = await self.get_results(word)
+
+        all_results = []
+        for result in results:
+            verb = {
+                "verb_form": result["verb_form"],
+                "infinitive": result["infinitive"],
+                "url": result["url"],
+            }
+            all_results.append(verb)
+
+        return json.dumps(
+            all_results, indent=4, separators=(",", ": ")
+        ), status

web/models/__init__.py web/conjugador/__init__.pyweb/models/__init__.py renamed to web/conjugador/__init__.py

@@ -21,7 +21,7 @@
 import json
 import logging
 
-from elasticsearch import Elasticsearch
+from elasticsearch import AsyncElasticsearch
 from pyuca import Collator
 
 
@@ -30,85 +30,68 @@ class IndexLetter:
     Search a letter in the Elasticsearch index.
 
     Args:
-        letter (str): The letter to search for.
-        es_url (str | None): The url to connect to an Elasticsearch instance.
+        es_client (AsyncElasticsearch): The client to use for the ES connection.
     """
 
-    DEFAULT_ES_HOST = "http://localhost:9200"
-
-    def __init__(self, letter: str, es_url: str | None = None) -> None:
+    def __init__(self, es_client: AsyncElasticsearch) -> None:
         """
-        Initializes the IndexLetter class with a letter to look for.
+        Initializes the IndexLetter class with a preconfigured ES client.
 
         Args:
-            letter (str): The letter to search for.
-            es_url (str | None): The url to connect to an Elasticsearch instance.
+            es_client (AsyncElasticsearch): The client to use for the ES connection.
         """
-        if not es_url:
-            es_url = self.DEFAULT_ES_HOST
-
-        self.letter = letter
-        self.es_client = Elasticsearch(es_url)
+        self.es_client = es_client
         self.index_name = "letter-index"
         self.collator = Collator()
-        self.num_results = 0
-        self.results = []
-
-    def get_num_results(self) -> int:
-        """
-        Retrieves the number of results found.
-
-        Returns:
-            int: Num of results.
-        """
-        return self.num_results
 
-    def get_results(self) -> list[dict]:
+    async def get_results(self, letter: str) -> list[dict]:
         """
         Gets the results from the prepared query, based on the letter.
 
+        Args:
+            letter (str): The letter of the index to check.
+
         Returns:
             list[dict]: A list of dicts containing the results.
         """
-        if not self.es_client.indices.exists(index=self.index_name):
-            self.results = []
-            self.num_results = 0
-            return self.results
+        if not await self.es_client.indices.exists(index=self.index_name):
+            return []
 
         query = {
-            "query": {"term": {"index_letter.keyword": self.letter}},
+            "query": {"term": {"index_letter.keyword": letter}},
             "collapse": {"field": "verb_form.keyword"},
             "size": 10000,
             "_source": ["verb_form", "infinitive"],
         }
 
         try:
-            response = self.es_client.search(index=self.index_name, body=query)
-            hits = response["hits"]["hits"]
-            self.results = [hit["_source"] for hit in hits]
-            self.results.sort(
-                key=lambda x: self.collator.sort_key(x["verb_form"])
+            response = await self.es_client.search(
+                index=self.index_name, body=query
             )
-            self.num_results = len(self.results)
+            hits = response["hits"]["hits"]
+            results = [hit["_source"] for hit in hits]
+            results.sort(key=lambda x: self.collator.sort_key(x["verb_form"]))
 
         except Exception as e:
             logging.error(f"Error searching index {self.index_name}: {e}")
-            self.results = []
-            self.num_results = 0
+            results = []
 
-        return self.results
+        return results
 
-    def get_json(self) -> tuple[str, int]:
+    async def get_json(self, letter: str) -> tuple[str, int]:
         """
-        Gets a stringified JSON for all the results found for the initialized
-        letter.
+        Gets a stringified JSON for all the results found for the given letter.
+
+        Args:
+            letter (str): The letter of the index to check.
 
         Returns:
             tuple[str, int]: A tuple containing the stringified JSON and the status code.
         """
         OK = 200
         status = OK
-        results = self.get_results()
+        results = await self.get_results(letter)
+
         all_results = []
         for result in results:
             verb = {"verb_form": result["verb_form"]}