From 22cd6081e0b00c09746a36164e50b4332f7b3f38 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Arnoux?=
 <39793176+jpjarnoux@users.noreply.github.com>
Date: Mon, 16 Feb 2026 17:34:35 +0100
Subject: [PATCH] Add `resolve_path` utility and integrate it into
 `check_tsv_sanity`

Enhanced path resolution by adding the `resolve_path` function to handle both absolute and relative paths robustly. Updated `check_tsv_sanity` to utilize this function for improved error handling and path validation.
---
 panorama/utils.py | 34 ++++++++++++++++++++++++++++++++--
 1 file changed, 32 insertions(+), 2 deletions(-)

diff --git a/panorama/utils.py b/panorama/utils.py
index 8ac3e77..9aa8b22 100644
--- a/panorama/utils.py
+++ b/panorama/utils.py
@@ -187,6 +187,35 @@ def is_empty(filepath):
     return os.path.getsize(filepath) == 0
 
 
+def resolve_path(path: Path, relative_directory: Path) -> Path:
+    """
+    Try the path from cwd first, then relative to the file location
+
+    Args:
+        path (Path): path to solve
+        relative_directory (Path): directory to resolve relative paths from
+
+    Returns:
+          Path: resolved path
+    """
+    path = Path(path)
+
+    # If absolute, use as-is
+    if path.is_absolute():
+        return path
+
+    # Try from the current working directory first
+    if path.exists():
+        return path
+
+    # Try relative to the file location
+    relative = relative_directory / path
+    if relative.exists():
+        return relative
+
+    # Return the cwd version if neither exists (for error reporting)
+    return path
+
 def check_tsv_sanity(tsv_path: Path) -> Dict[str, Dict[str, Union[int, str, Path]]]:
     """
     Check if the given TSV file is readable for the next PANORAMA step.
@@ -232,10 +261,11 @@ def check_tsv_sanity(tsv_path: Path) -> Dict[str, Dict[str, Union[int, str, Path
         logging.getLogger("PANORAMA").error("\n" + err_df.to_string().replace("\n", "\n\t"))
         raise ValueError("There is a line with no path (second column)")
     else:
+        tsv["path"] = tsv["path"].map(lambda x: resolve_path(x, relative_directory=tsv_path.parent))
         if not tsv["path"].map(lambda x: Path(x).exists()).all():
             err_df = pd.DataFrame(
-                [tsv["path"], ~tsv["path"].map(lambda x: Path(x).exists())],
-                ["pangenome_path", "error"],
+                [tsv["path"], tsv["path"].map(lambda x: Path(x).exists())],
+                ["pangenome_path", "exist"],
             ).T
             logging.getLogger("PANORAMA").error("\n" + err_df.to_string().replace("\n", "\n\t"))
             raise FileNotFoundError("Unable to locate one or more pangenome in your file.}")