Makefile

# Makefile
include .env

# Setting env variables
export $(shell sed 's/=.*//' .env)

# Define variables
PY_FILES = $(shell find . -name "*.py" -not -path "./hf_cache/*")

REINDEX_ES_DEFAULT=false
REINIT_DB_DEFAULT=false
DEFACTO_DEFAULT=true
REINIT_GRAFANA_DEFAULT=false
RECREATE_DASHBOARDS_DEFAULT=false


# Export the variable for all targets
.EXPORT_ALL_VARIABLES: 
PYDEVD_DISABLE_FILE_VALIDATION=1

# Target to create a conda environment and install dependencies
create_local_env:
	conda create -y -n $(ENV_NAME) python=$(PYTHON_VERSION)
	conda run -n $(ENV_NAME) pip install -r requirements.txt
	conda run -n $(ENV_NAME) pip install -r requirements-dev.txt
	conda run -n $(ENV_NAME) python -m ipykernel install --user --name=$(ENV_NAME) --display-name "Python ($(ENV_NAME))"

# Target to remove the Jupyter kernel and delete the conda environment
remove_local_env:
	@if jupyter kernelspec list | grep -q $(ENV_NAME); then \
		jupyter kernelspec remove -f $(ENV_NAME); \
	else \
		echo "Jupyter kernel '$(ENV_NAME)' not found."; \
	fi
	@if conda info --envs | grep -q $(ENV_NAME); then \
		conda env remove -n $(ENV_NAME); \
	else \
		echo "Conda environment '$(ENV_NAME)' not found."; \
	fi

# Docker-compose
compose:
	docker-compose up -d

# Format py files
format_py:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) isort $(PY_FILES)
	@PYTHONPATH=./ conda run -n $(ENV_NAME) black $(PY_FILES)

# Linting py files
lint_py:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) pylint $(PY_FILES) --exit-zero

# Running unit_tests
unit_tests:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) pytest ./tests

# Running integration_tests
integration_tests:
	@./integration_tests/connectivity_check.sh $(keep-containers-running)

# Download ollama models specified as CHAT_MODEL & EMBED_MODEL
setup_ollama:
	@./scripts/setup_ollama.sh

# Cache $ASR_MODEL in hf_cache dir
cache_asr_model:
	@PYTHONPATH=./ conda run --no-capture-output -n $(ENV_NAME) \
		python ./scripts/cache_asr_model.py $(ASR_MODEL)

# Cache $PODCAST_DATASET in hf_cache dir
cache_dataset:
	@PYTHONPATH=./ conda run --no-capture-output -n $(ENV_NAME) \
		python ./scripts/cache_dataset.py $(PODCAST_DATASET)

# Start prefect server and worker
prefect_start_server:
	@if ! curl -s -o /dev/null localhost:4200; then \
		echo "Prefect server is not running. Starting the server..."; \
		conda run -n $(ENV_NAME) prefect server start & \
		for i in $$(seq 1 10); do \
			if curl -s -o /dev/null localhost:4200; then \
				echo "\033[0;32mPrefect server started successfully.\033[0m"; \
				exit 0; \
			else \
				echo "Waiting for Prefect server to start... ($$i/10)"; \
				sleep 5; \
			fi; \
		done; \
		echo "\033[0;31mFailed to start the Prefect server after multiple attempts.\033[0m"; \
	else \
		echo "\033[0;32mPrefect server is already running.\033[0m"; \
	fi

# Stop prefect server
prefect_stop_server:
	@if curl -s -o /dev/null localhost:4200; then \
		echo "Prefect server is running. Stopping the server..."; \
		kill $$(ps aux | grep "prefect server" | grep -v grep | awk '{print $$2}') 2>/dev/null || true; \
		for i in $$(seq 1 10); do \
			if ! curl -s -o /dev/null localhost:4200; then \
				echo "\033[0;32mPrefect server terminated successfully.\033[0m"; \
				exit 0; \
			else \
				echo "Waiting for Prefect server to terminate... ($$i/10)"; \
				sleep 2; \
			fi; \
		done; \
		echo "\033[0;31mFailed to terminate the Prefect server after multiple attempts.\033[0m"; \
	else \
		echo "\033[0;32mPrefect server is not running.\033[0m"; \
	fi

# Start a new prefect worker
prefect_start_worker:
	@conda run -n $(ENV_NAME) prefect worker start --pool ${WORK_POOL_NAME} &
	@sleep 2
	@echo "\033[0;32mWorker started successfully\033[0m"
	@count=$$(ps aux | grep "prefect worker start --pool $(WORK_POOL_NAME)" | grep -v grep | wc -l); \
		echo "Total number of active workers: $$((count / 2))"

# Kill all prefect workers
prefect_kill_workers:
	@./scripts/kill_prefect_workers.sh

# Re-initialize prefect db & work-pool
reinit_prefect:
	@./scripts/reinit_prefect.sh

# Deploy or redeploy prefect flows
redeploy_flows:
	@echo "Re-deploying prefect flows ..."
	@PYTHONPATH=./ conda run -n $(ENV_NAME) python scripts/redeploy_flows.py \
		--reindex_es "$(REINDEX_ES_DEFAULT)" \
		--reinit_db "$(REINIT_DB_DEFAULT)" \
		--defacto "$(DEFACTO_DEFAULT)" \
		--reinit_grafana "$(REINIT_GRAFANA_DEFAULT)" \
		--recreate_dashboards "$(RECREATE_DASHBOARDS_DEFAULT)"
	@echo "\033[0;32mSuccessfully redeployed prefect flows.\033[0m"

# Re-indexing es with defacto mode
reindex_es_defacto:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) \
		prefect deployment run setup_es/ad-hoc \
		-p reindex_es=true \
		-p defacto=true

# Re-initilizing app backend db
reinit_db:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) \
		prefect deployment run init_db/ad-hoc \
		-p reinit_db=true

# Re-setting up Grafana data source & dashboards
resetup_grafana:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) \
		prefect deployment run setup_grafana/ad-hoc \
		$(if $(reinit_grafana),-p reinit_grafana=$(reinit_grafana)) \
		-p recreate_dashboards=true

# Get number of docs in es index for a passed id (not _id)
es_count_filtered:
	@if [ -z "$(ID)" ]; then \
		curl -X GET "http://${ELASTIC_SETUP_HOST}:${ELASTIC_PORT}/${ES_INDEX_NAME}/_count" \
		-H "Content-Type: application/json" \
		-d '{}'; \
	else \
		curl -X GET "http://${ELASTIC_SETUP_HOST}:${ELASTIC_PORT}/${ES_INDEX_NAME}/_count" \
		-H "Content-Type: application/json" \
		-d '{"query":{"terms":{"_id":["$(ID)"]}}}'; \
	fi

# Run process_new_episodes once (this wouldn't affect its weekly schedule)
process_new_episodes_run:
	@PYTHONPATH=./ conda run -n $(ENV_NAME) \
		prefect deployment run process_new_episodes/midnight-every-sunday

.PHONY: integration_tests