Minor fixes (#319)

.gitignore

@@ -18,6 +18,7 @@ output.txt
 replay_logs/
 my.secrets
 
+repos-small-test
 output/
 merge_repo/
 merges_small_valid_subsamples/

Makefile

@@ -33,11 +33,13 @@ check-python-style:
 # This target deletes files that are not committed to version control.
 clean:
 	${MAKE} clean-workdir
-	rm -rf repos
+	rm -rf repos-small-test
 	rm -rf scratch
 	rm -rf results/small
 	rm -rf .valid_merges_counters
 
+clean-repos:
+	rm -rf repos
 # This target deletes files in the cache, which is commited to version control.
 clean-cache:
 	rm -rf cache
@@ -95,24 +97,19 @@ copy-paper:
 	rsync -av --exclude='*.csv' results ../AST-Merging-Evaluation-Paper/
 	find  ../AST-Merging-Evaluation-Paper/ -type d -empty -delete
 
-# Update cache
-update-cache-results:
-	python3 src/python/cache_merger.py
-	make compress-cache
-
 # As of 2023-07-31, this takes 5-20 minutes to run, depending on your machine.
 small-test:
 	${MAKE} clean-test-cache clean
-	./run_small.sh --include_trivial_merges --no_timing
+	AST_REPOS_PATH=repos-small-test ./run_small.sh --include_trivial_merges --no_timing
 	${MAKE} compress-small-cache
 	${MAKE} small-test-diff
 	rm -rf results/small
-	./run_small.sh --include_trivial_merges --no_timing
+	AST_REPOS_PATH=repos-small-test ./run_small.sh --include_trivial_merges --no_timing
 	${MAKE} small-test-diff
 
 small-test-without-cleaning:
 	${MAKE} clean-test-cache
-	./run_small.sh --include_trivial_merges --no_timing
+	AST_REPOS_PATH=repos-small-test ./run_small.sh --include_trivial_merges --no_timing
 	${MAKE} small-test-diff
 
 update-figures:
@@ -147,7 +144,7 @@ clean-workdir:
 
 clean-local:
 	${MAKE} clean-workdir
-	rm -rf repos
+	rm -rf repos-small-test
 
 check-merges-reproducibility:
 	@echo "Running replay_merge sequentially for each idx..."

README.md

@@ -222,6 +222,6 @@ To run style checking run `make style`.
 
 To investigate differences between two mergers:
 
-* edit file `src/python/select_from_results.py` to reflect the differences you are interested in.
-* run `src/python/select_from_results.py` to create a .csv database containing only the differences.
+* edit file `src/python/utils/select_from_results.py` to reflect the differences you are interested in.
+* run `src/python/utils/select_from_results.py` to create a .csv database containing only the differences.
 * run `src/python/replay_merge.py --idx INDEX` (maybe add `-test`) for the index of the merge you are interested in.

run.sh

@@ -146,17 +146,12 @@ mkdir -p "$OUT_DIR"
 if [ -d "$CACHE_DIR" ]; then
     find "$CACHE_DIR" -name "*.lock" -delete
 fi
-if [ -d "repos" ]; then
-    find "repos/locks" -name "*.lock" -delete
+REPOS_PATH=${AST_REPOS_PATH:-repos}
+if [ -d "$REPOS_PATH" ]; then
+    find "$REPOS_PATH" -name "*.lock" -delete
 fi
 
-# Check if .workdir exists and delete it
-if [ -d .workdir ]; then
-    chmod -R +w .workdir
-    rm -rf .workdir
-fi
-
-python3 src/python/delete_cache_placeholders.py \
+python3 src/python/utils/delete_cache_placeholders.py \
     --cache_dir "$CACHE_DIR"
 
 python3 src/python/write_head_hashes.py \

src/python/delete_cache_entry_of_merge_tests.py

@@ -1,5 +1,5 @@
 # -*- coding: utf-8 -*-
-"""Delete the keys containing 'imports' in the JSON files in the given directory."""
+"""Delete the keys containing the SHA of the merge tests from the cache."""
 
 import pandas as pd
 from pathlib import Path

src/python/repo.py

@@ -239,7 +239,7 @@ def copy_repo(self) -> None:
         if not self.repo_path.exists():
             self.clone_repo()
         if self.local_repo_path.exists():
-            return
+            shutil.rmtree(self.local_repo_path, ignore_errors=True)
         self.workdir.mkdir(parents=True, exist_ok=True)
         shutil.copytree(
             self.repo_path,