Fixed scripts

src/python/latex_output.py

@@ -63,6 +63,45 @@
 }
 
 
+def check_fingerprint_consistency(result_df: pd.DataFrame, merge_tools: List[str]):
+    """Check if the fingerprints are consistent.
+
+    Args:
+        result_df: DataFrame containing the results of the merge tools
+        merge_tools: list of merge tools
+    """
+    for merge_tool1 in merge_tools:
+        for merge_tool2 in merge_tools:
+            if merge_tool1 == "gitmerge_resolve" or merge_tool2 == "gitmerge_resolve":
+                continue
+            # ignore adajcent
+            if (
+                merge_tool1 == "gitmerge_ort_adjacent"
+                or merge_tool2 == "gitmerge_ort_adjacent"
+            ):
+                continue
+            if merge_tool1 != merge_tool2:
+                # Check if fingerprints are the same
+                same_fingerprint_mask = (
+                    result_df[merge_tool1 + "_merge_fingerprint"]
+                    == result_df[merge_tool2 + "_merge_fingerprint"]
+                )
+
+                # Check if results are the same
+                same_result_mask = result_df[merge_tool1] == result_df[merge_tool2]
+
+                # Check if the fingerprints are the same but the results are different
+                inconsistent_mask = same_fingerprint_mask & ~same_result_mask
+                if inconsistent_mask.sum() > 0:
+                    logger.error(
+                        f"Inconsistency found between {merge_tool1} and {merge_tool2} in {inconsistent_mask.sum()} cases."
+                    )
+                    logger.error(result_df[inconsistent_mask])
+                assert (
+                    inconsistent_mask.sum() == 0
+                ), f"Inconsistency found between {merge_tool1} and {merge_tool2} in {inconsistent_mask.sum()} cases."
+
+
 def merge_tool_latex_name(name: str) -> str:
     """Return the LaTeX name of a merge tool.
     Args:
@@ -385,44 +424,56 @@ def merge_two_states(
         Path(plots_output_path).mkdir(parents=True, exist_ok=True)
         Path(tables_output_path).mkdir(parents=True, exist_ok=True)
 
+        check_fingerprint_consistency(result_df, merge_tools)
+
         # Figure Heat map diffing
-        result = np.zeros((len(merge_tools), len(merge_tools)))
-        with Progress(
-            SpinnerColumn(),
-            TextColumn("[progress.description]{task.description}"),
-            BarColumn(),
-            TimeElapsedColumn(),
-            TimeRemainingColumn(),
-        ) as progress:
-            task = progress.add_task("Processing merges...", total=len(result_df))
-            for _, row in result_df.iterrows():
-                progress.update(task, advance=1)
-                for idx, merge_tool1 in enumerate(merge_tools):
-                    for idx2, merge_tool2 in enumerate(merge_tools[(idx + 1) :]):
-                        if (
-                            row[merge_tool1] in MERGE_CORRECT_NAMES
-                            or row[merge_tool2] in MERGE_CORRECT_NAMES
-                        ):
-                            if (
-                                row[merge_tool1 + "_merge_fingerprint"]
-                                != row[merge_tool2 + "_merge_fingerprint"]
-                            ):
-                                result[idx][idx2 + idx + 1] += 1
-                                result[idx2 + idx + 1][idx] += 1
+        result = pd.DataFrame(
+            {
+                merge_tool: {merge_tool: 0 for merge_tool in merge_tools}
+                for merge_tool in merge_tools
+            }
+        )
+        for merge_tool1 in merge_tools:
+            for merge_tool2 in merge_tools:
+                # Mask for different fingerprints
+                mask_diff_fingerprint = (
+                    result_df[merge_tool1 + "_merge_fingerprint"]
+                    != result_df[merge_tool2 + "_merge_fingerprint"]
+                )
+
+                # Mask if one of the results is in correct or incorrect names
+                merge_name_flags1 = result_df[merge_tool1].isin(
+                    MERGE_CORRECT_NAMES + MERGE_INCORRECT_NAMES
+                )
+                merge_name_flags2 = result_df[merge_tool2].isin(
+                    MERGE_CORRECT_NAMES + MERGE_INCORRECT_NAMES
+                )
+                mask_merge_name = merge_name_flags1 | merge_name_flags2
+
+                if merge_tool1 == "intellimerge" and merge_tool2 == "gitmerge_ort":
+                    print(mask_diff_fingerprint)
+                    print(mask_merge_name)
+                    print((mask_diff_fingerprint & mask_merge_name).sum())
+
+                # Calculate the result
+                result.loc[merge_tool1, merge_tool2] = (
+                    mask_diff_fingerprint & mask_merge_name
+                ).sum()
+
+        # Transform the result into a numpy array
         _, ax = plt.subplots(figsize=(8, 6))
-        result = np.tril(result)
+        result_array = np.tril(result.to_numpy())
         latex_merge_tool = [
-            "\\mbox{" + merge_tool_latex_name(i) + "}" for i in merge_tools
+            "\\mbox{" + merge_tool_latex_name(i) + "}" for i in result.columns
         ]
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             heatmap = sns.heatmap(
-                result / len(result_df),
+                result_array,
                 annot=True,
                 ax=ax,
                 xticklabels=latex_merge_tool,  # type: ignore
                 yticklabels=latex_merge_tool,  # type: ignore
-                fmt=".3f",
                 mask=np.triu(np.ones_like(result, dtype=bool), k=1),
                 cmap="Blues",
                 annot_kws={"size": 6},

src/python/replay_merge.py

@@ -103,7 +103,7 @@ def merge_replay(
                 delete_workdir=False,
                 lazy_clone=False,
             )
-            repo.checkout(merge_data["left"])
+            repo.checkout(merge_data["left"], use_cache=False)
 
         workdir = Path(f"{repo_slug}-merge-input-right")
         if not (WORKDIR_DIRECTORY / workdir).exists():
@@ -114,7 +114,7 @@ def merge_replay(
                 delete_workdir=False,
                 lazy_clone=False,
             )
-            repo.checkout(merge_data["right"])
+            repo.checkout(merge_data["right"], use_cache=False)
 
         workdir = Path(f"{repo_slug}-merge-input-base")
         if not (WORKDIR_DIRECTORY / workdir).exists():
@@ -129,7 +129,7 @@ def merge_replay(
                 ["git", "merge-base", merge_data["left"], merge_data["right"]],
                 stdout=subprocess.PIPE,
             ).stdout.decode("utf-8")
-            repo.checkout(base_commit)
+            repo.checkout(base_commit, use_cache=False)
 
         workdir = Path(f"{repo_slug}-merge-input-programmer")
         if not (WORKDIR_DIRECTORY / workdir).exists():
@@ -140,7 +140,7 @@ def merge_replay(
                 delete_workdir=False,
                 lazy_clone=False,
             )
-            repo.checkout(merge_data["merge"])
+            repo.checkout(merge_data["merge"], use_cache=False)
 
         for merge_tool in MERGE_TOOL:
             progress.update(task, advance=1)
@@ -201,6 +201,18 @@ def merge_replay(
                 use_cache=False,
             )
             assert repo.local_repo_path.exists()
+            if merge_result in (MERGE_STATE.Merge_failed, MERGE_STATE.Merge_success):
+                # Run 'git diff --name-only --diff-filter=U' to get the files with conflicts
+                conflict_files = subprocess.run(
+                    ["git", "diff", "--name-only", "--diff-filter=U"],
+                    cwd=repo.local_repo_path,
+                    stdout=subprocess.PIPE,
+                ).stdout.decode("utf-8")
+                is_conflict = len(conflict_files) > 0
+                assert (
+                    is_conflict == (merge_result == MERGE_STATE.Merge_failed)
+                ), f"merge_replay: merge_result {merge_result} does not match conflict_files {conflict_files}"
+
             root_dir = Path("replay_logs")
             log_path = root_dir / Path(
                 "merges/"

src/python/utils/build_inconsistent_merges.py

@@ -0,0 +1,107 @@
+# -*- coding: utf-8 -*-
+"""Check for inconsistent results between merge tools and output them to a CSV.
+
+usage: python3 check_inconsistencies.py --result_csv <path_to_result_csv> --output_csv <path_to_output_csv>
+
+This script reads a result CSV file, checks for inconsistencies between merge tools based on fingerprints,
+and writes the inconsistent results to a new CSV file.
+
+Arguments:
+- result_csv: path to the result CSV file
+- output_csv: path to the output CSV file where inconsistent results will be saved
+"""
+
+import argparse
+import pandas as pd
+from typing import List
+
+
+def check_fingerprint_consistency(
+    result_df: pd.DataFrame, merge_tools: List[str]
+) -> pd.DataFrame:
+    """Check if the fingerprints are consistent and return inconsistent results.
+
+    Args:
+        result_df: DataFrame containing the results of the merge tools
+        merge_tools: list of merge tools
+
+    Returns:
+        DataFrame with inconsistent results
+    """
+    inconsistencies = []
+
+    for merge_tool1 in merge_tools:
+        for merge_tool2 in merge_tools:
+            if merge_tool1 == "gitmerge_resolve" or merge_tool2 == "gitmerge_resolve":
+                continue
+            # ignore adjacent
+            if (
+                merge_tool1 == "gitmerge_ort_adjacent"
+                or merge_tool2 == "gitmerge_ort_adjacent"
+            ):
+                continue
+            if merge_tool1 != merge_tool2:
+                # Check if fingerprints are the same
+                same_fingerprint_mask = (
+                    result_df[merge_tool1 + "_merge_fingerprint"]
+                    == result_df[merge_tool2 + "_merge_fingerprint"]
+                )
+
+                # Check if results are the same
+                same_result_mask = result_df[merge_tool1] == result_df[merge_tool2]
+
+                # Check if the fingerprints are the same but the results are different
+                inconsistent_mask = same_fingerprint_mask & ~same_result_mask
+                if inconsistent_mask.sum() > 0:
+                    inconsistent_results = result_df[inconsistent_mask].copy()
+                    inconsistent_results["tool1"] = merge_tool1
+                    inconsistent_results["tool2"] = merge_tool2
+                    inconsistencies.append(result_df[inconsistent_mask])
+
+    if inconsistencies:
+        return pd.concat(inconsistencies).drop_duplicates()
+    else:
+        return pd.DataFrame()
+
+
+def main():
+    """Main function"""
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--result_csv",
+        type=str,
+        help="Path to the result CSV file",
+        default="results/combined/result.csv",
+    )
+    parser.add_argument(
+        "--output_csv",
+        type=str,
+        help="Path to the output CSV file for inconsistent results",
+        default="results/combined/inconsistent_results.csv",
+    )
+    args = parser.parse_args()
+
+    # Read the result CSV file
+    result_df = pd.read_csv(args.result_csv, index_col="idx")
+
+    # List of merge tools
+    merge_tools = [
+        col.split("_merge_fingerprint")[0]
+        for col in result_df.columns
+        if col.endswith("_merge_fingerprint")
+    ]
+
+    # Check for inconsistencies
+    inconsistent_results = check_fingerprint_consistency(result_df, merge_tools)
+
+    if not inconsistent_results.empty:
+        # Write the inconsistent results to the output CSV file
+        inconsistent_results.to_csv(args.output_csv, index_label="idx")
+        print(f"Inconsistent results have been saved to {args.output_csv}")
+        print("Number of inconsistencies found:", len(inconsistent_results))
+    else:
+        print("No inconsistencies found.")
+
+
+if __name__ == "__main__":
+    main()

src/python/utils/delete_intellimerge_keys_from_cache.py

@@ -0,0 +1,62 @@
+# -*- coding: utf-8 -*-
+"""Delete the keys containing 'imports' in the JSON files in the given directory."""
+
+import os
+import json
+
+
+def count_import_keys(directory):
+    """Count the number of keys containing 'imports' in the JSON files in the given directory."""
+    count = 0
+    for root, _, files in os.walk(directory):
+        json_files = [f for f in files if f.endswith(".json")]
+        for json_file in json_files:
+            file_path = os.path.join(root, json_file)
+            with open(file_path, "r", encoding="utf-8") as file:
+                data = json.load(file)
+
+            # Count keys containing 'adjacent'
+            keys_to_delete = [key for key in data if "intellimerge" in key]
+            count += len(keys_to_delete)
+    return count
+
+
+def delete_import_keys(directory):
+    """Delete the keys containing 'imports' in the JSON files in the given directory."""
+    total_deleted = 0
+    for root, _, files in os.walk(directory):
+        json_files = [f for f in files if f.endswith(".json")]
+        for json_file in json_files:
+            file_path = os.path.join(root, json_file)
+            with open(file_path, "r", encoding="utf-8") as file:
+                data = json.load(file)
+
+            # Record keys to delete
+            keys_to_delete = [key for key in data.keys() if "intellimerge" in key]
+            if keys_to_delete:
+                for key in keys_to_delete:
+                    del data[key]
+                    total_deleted += 1
+
+            # Save the modified data back to file
+            with open(file_path, "w", encoding="utf-8") as file:
+                json.dump(data, file, indent=4)
+
+    return total_deleted
+
+
+def main():
+    """Main function."""
+    directory = "cache"
+    potential_deletions = count_import_keys(directory)
+    print(f"Potential deletions: {potential_deletions}")
+    confirm = input("Do you want to proceed with deleting these keys? (yes/no): ")
+    if confirm.lower() == "yes":
+        total_deleted = delete_import_keys(directory)
+        print(f"Total keys deleted: {total_deleted}")
+    else:
+        print("Operation cancelled.")
+
+
+if __name__ == "__main__":
+    main()

src/scripts/merge_tools/gitmerge_ort_imports.sh

@@ -12,9 +12,16 @@ if "$MERGE_SCRIPTS_DIR"/gitmerge.sh "$clone_dir" "$branch1" "$branch2" "$strateg
 fi
 
 cd "$clone_dir" || exit 1
-if ! "$MERGE_SCRIPTS_DIR"/resolve-import-conflicts; then
-  echo "Conflict"
-  exit 1
+"$MERGE_SCRIPTS_DIR"/resolve-import-conflicts;
+
+# Detect conflicts using Git commands
+conflict_files=$(git diff --name-only --diff-filter=U)
+
+if [ -n "$conflict_files" ]; then
+    echo "Conflict detected in the following files:"
+    echo "$conflict_files"
+    exit 1
 fi
 
+echo "No conflicts detected."
 exit 0

src/scripts/merge_tools/gitmerge_ort_imports_ignorespace.sh

@@ -12,9 +12,16 @@ if "$MERGE_SCRIPTS_DIR"/gitmerge.sh "$clone_dir" "$branch1" "$branch2" "$strateg
 fi
 
 cd "$clone_dir" || exit 1
-if ! "$MERGE_SCRIPTS_DIR"/resolve-import-conflicts; then
-  echo "Conflict"
-  exit 1
+"$MERGE_SCRIPTS_DIR"/resolve-import-conflicts;
+
+# Detect conflicts using Git commands
+conflict_files=$(git diff --name-only --diff-filter=U)
+
+if [ -n "$conflict_files" ]; then
+    echo "Conflict detected in the following files:"
+    echo "$conflict_files"
+    exit 1
 fi
 
+echo "No conflicts detected."
 exit 0