Improved elm server error handling

Author: lauraluebbert

docs/src/en/updates.md

@@ -8,6 +8,8 @@
   `type="orthologs"` is now the default, removing the need to specify the `type` argument when calling orthologs
 - [`gget diamond`](diamond.md):  
   Now supports translated alignment of nucleotide sequences to amino acid reference sequences using the `--translated` flag.
+- [`gget elm`](elm.md):   
+  Improved server error handling.
 
 **Version ≥ 0.29.0** (Sep 25, 2024):  
 - **New modules:**

gget/gget_setup.py

@@ -5,9 +5,8 @@
 import subprocess
 import platform
 import uuid
-from platform import python_version
 
-from .utils import set_up_logger
+from .utils import set_up_logger, check_file_for_error_message
 
 logger = set_up_logger()
 
@@ -183,28 +182,48 @@ def setup(module, verbose=True, out=None):
 
         # Check if files are present
         if os.path.exists(elm_instances_fasta):
+            # Check that file does not just contain an error message
+            check_file_for_error_message(
+                elm_instances_fasta,
+                "ELM instances fasta file",
+                ELM_INSTANCES_FASTA_DOWNLOAD,
+            )
             if verbose:
                 logger.info(f"ELM sequences file present.")
         else:
             logger.error("ELM FASTA file missing.")
 
         if os.path.exists(elm_classes_tsv):
+            # Check that file does not just contain an error message
+            check_file_for_error_message(
+                elm_classes_tsv, "ELM classes tsv file", ELM_CLASSES_TSV_DOWNLOAD
+            )
             if verbose:
                 logger.info("ELM classes file present.")
         else:
             logger.error("ELM classes file missing.")
 
         if os.path.exists(elm_instances_tsv):
+            # Check that file does not just contain an error message
+            check_file_for_error_message(
+                elm_instances_tsv, "ELM instances tsv file", ELM_INSTANCES_TSV_DOWNLOAD
+            )
             if verbose:
                 logger.info("ELM instances file present.")
         else:
             logger.error("ELM instances file missing.")
 
         if os.path.exists(elm_intdomains_tsv):
+            # Check that file does not just contain an error message
+            check_file_for_error_message(
+                elm_intdomains_tsv,
+                "ELM interaction domains tsv file",
+                ELM_INTDOMAINS_TSV_DOWNLOAD,
+            )
             if verbose:
-                logger.info("ELM interactions domains file present.")
+                logger.info("ELM interaction domains file present.")
         else:
-            logger.error("ELM interactions domains file missing.")
+            logger.error("ELM interaction domains file missing.")
 
     elif module == "alphafold":
         if platform.system() == "Windows":

gget/utils.py

@@ -77,6 +77,31 @@ def get_latest_cosmic():
     return int(soup.find("div", class_="news").get("id").split("v")[-1])
 
 
+def check_file_for_error_message(filepath, filename, download_path):
+    with open(filepath, "r", encoding="utf-8") as file:
+        content = file.read().strip()
+
+    # Define common error indicators
+    error_keywords = [
+        "Internal Server Error",
+        "Bad Gateway",
+        "Service Unavailable",
+        "502 Bad Gateway",
+        "500 Internal Server Error",
+    ]
+
+    # Check if the file contains an error message and raise ValueError + print error message if so
+    if any(keyword in content for keyword in error_keywords):
+        raise ValueError(
+            f"""
+            The {filename} downloaded from {download_path} 
+            contains an error message instead of valid data.\n
+            Error message:\n{content}\n
+            Please try again. If the problem persists, please report it here: https://github.com/pachterlab/gget/issues/new?template=issue_report.yml
+            """
+        )
+
+
 def read_fasta(fasta):
     """
     Args: