Merge pull request #97 from CatchTheTornado/hotfix/2025-01-19/duplicated-unify-and-missing-libglib20

Author: pkarw

dev.Dockerfile

@@ -7,6 +7,8 @@ RUN echo 'Acquire::http::Pipeline-Depth 0;\nAcquire::http::No-Cache true;\nAcqui
 RUN apt-get clean && rm -rf /var/lib/apt/lists/* \
     && apt-get update --fix-missing \
     && apt-get install -y \
+        libglib2.0-0 \
+        libglib2.0-dev \
         libgl1-mesa-glx \
         poppler-utils \
         libmagic1 \

dev.gpu.Dockerfile

@@ -17,6 +17,7 @@ RUN apt-get update && \
     python3-pip \
     libgl1 \
     libglib2.0-0 \
+    libglib2.0-dev \
     curl \
     gnupg2 \
     ca-certificates \

run.sh

@@ -53,7 +53,18 @@ echo "Your ENV settings loaded from .env.localhost file: "
 printenv
 
 CELERY_BIN="$(pwd)/.venv/bin/celery"
-CELERY_PID=$(pgrep -f "$CELERY_BIN")
+CELERY_PIDS=$(pgrep -f "$CELERY_BIN")
+
+if [ -n "$CELERY_PIDS" ]; then
+  echo "Killing existing Celery processes from $CELERY_BIN with PIDs:"
+  echo "$CELERY_PIDS"
+  for PID in $CELERY_PIDS; do
+    kill "$PID" || echo "Failed to kill process with PID: $PID"
+  done
+else
+  echo "No running Celery process found."
+fi
+
 REDIS_PORT=6379 # will move it to .envs in near future
 
 if lsof -i :$REDIS_PORT | grep LISTEN >/dev/null; then

text_extract_api/files/file_formats/file_format.py

@@ -178,15 +178,6 @@ def convertible_to() -> Dict[Type["FileFormat"], Callable[[Type["FileFormat"]],
         """
         return {}
 
-    def unify(self) -> Type["FileFormat"]:
-        """
-        Converts the file to a universal type for that format (e.g., JPEG from png, bmp, tiff, etc.).
-        Default implementation returns a new instance of the current format.
-
-        :return: Unified format as a new self instance.
-        """
-        return self
-
     @classmethod
     def default_iterator_file_format(cls) -> Type["FileFormat"]:
         if not cls.is_pageable():