refactor: Make language-tool a separate service

Dockerfile

@@ -27,15 +27,7 @@ COPY . .
 RUN mkdir -p src/ctk_functions/data/signatures
 COPY --from=unzipper /files/*.png /home/site/wwwroot/src/ctk_functions/data/signatures
 
-RUN apt-get clean; apt-get -y update && \
-    mkdir -p /usr/share/man/man1/ && \
-    apt-get install -y openjdk-17-jdk && \
-    apt-get install -y openjdk-17-jre && \
-    update-alternatives --config java && \
-    update-alternatives --config javac && \
-    uv sync --frozen --no-cache && \
-    uv run python -c 'import spacy; spacy.load("en_core_web_sm")' && \
-    uv run python -c \
-      'import language_tool_python; language_tool_python.LanguageTool ("en-US")'
+RUN uv sync --frozen --no-cache && \
+    uv run python -c 'import spacy; spacy.load("en_core_web_sm")'
 
 CMD ["uv", "run", "fastapi", "run", "src/ctk_functions/app.py", "--port", "8000",  "--host", "0.0.0.0"]

src/ctk_functions/core/config.py

@@ -26,6 +26,8 @@ class Settings(pydantic_settings.BaseSettings):
     AZURE_OPENAI_LLM_DEPLOYMENT: pydantic.SecretStr
     AZURE_OPENAI_ENDPOINT: pydantic.SecretStr
 
+    LANGUAGE_TOOL_URL: str | None = None
+
     LOGGER_VERBOSITY: int = logging.INFO
     LOGGER_PHI_LOGGING_LEVEL: int = pydantic.Field(1, lt=logging.DEBUG)
     LOG_PHI: bool = pydantic.Field(

src/ctk_functions/routers/intake/intake_processing/utils/language_utils.py

@@ -7,8 +7,10 @@
 from docx import document
 from docx.text import paragraph
 
+from ctk_functions.core import config
 from ctk_functions.text import corrections
 
+settings = config.get_settings()
 NLP = spacy.load("en_core_web_sm", enable=["parser"])
 
 # c.f. https://community.languagetool.org/rule/list?lang=en for a list of rules.
@@ -46,6 +48,7 @@ def __init__(
         self.document = doc
         self.correcter = corrections.LanguageCorrecter(
             enabled_rules or DEFAULT_LANGUAGE_RULES,
+            settings.LANGUAGE_TOOL_URL,
         )
 
     def correct(self) -> None:

src/ctk_functions/routers/language_tool/controller.py

@@ -1,8 +1,11 @@
 """Functions for converting files between different formats."""
 
+from ctk_functions.core import config
 from ctk_functions.routers.language_tool import schemas
 from ctk_functions.text import corrections
 
+settings = config.get_settings()
+
 
 def language_tool(body: schemas.PostLanguageToolRequest) -> str:
     """Corrects the grammar of the input text.
@@ -13,5 +16,5 @@ def language_tool(body: schemas.PostLanguageToolRequest) -> str:
     Returns:
         The corrected text.
     """
-    correcter = corrections.LanguageCorrecter(body.rules)
+    correcter = corrections.LanguageCorrecter(body.rules, settings.LANGUAGE_TOOL_URL)
     return correcter.run(body.text)

src/ctk_functions/text/corrections.py

@@ -14,15 +14,18 @@ class LanguageCorrecter:
     def __init__(
         self,
         enabled_rules: Iterable[str],
+        remote_server: str | None = None,
     ) -> None:
         """Initializes the language tool.
 
         Args:
             enabled_rules: The rules to enable for the correction.
+            remote_server: The remote server to connect to.
 
         """
         self.language_tool = language_tool_python.LanguageTool(
-            "en-US",
+            language="en-US",
+            remote_server=remote_server,
         )
         self.language_tool.enabled_rules = set(enabled_rules)
         self.language_tool.enabled_rules_only = True