From 69a284de472a249781b8a86c2d68b1bb0267f1b9 Mon Sep 17 00:00:00 2001
From: Bouwe Andela <b.andela@esciencecenter.nl>
Date: Mon, 16 Jan 2023 11:46:55 +0100
Subject: [PATCH] Fix issues with searching for files on ESGF (#1863)

---
 esmvalcore/esgf/_download.py   | 10 +++++++---
 tests/unit/esgf/test_search.py | 25 +++++++++++++++++++++++++
 2 files changed, 32 insertions(+), 3 deletions(-)

diff --git a/esmvalcore/esgf/_download.py b/esmvalcore/esgf/_download.py
index ce4a0e8fc3..9dd10fb397 100644
--- a/esmvalcore/esgf/_download.py
+++ b/esmvalcore/esgf/_download.py
@@ -224,7 +224,7 @@ def same_file(result):
             # cmip5.output1.ICHEC.EC-EARTH.historical
             # .mon.atmos.Amon.r1i1p1.v20121115
             variable = file.name.split('_')[0]
-            if 'variable' in facets and facets['variable'] == variable:
+            if 'variable' not in facets or facets['variable'] == variable:
                 files.append(file)
             else:
                 logger.debug(
@@ -279,8 +279,12 @@ def _get_facets(results):
         facets = {
             'project': project,
         }
-        for idx, key in enumerate(keys):
-            facets[key] = values[idx]
+        if len(keys) == len(values):
+            for idx, key in enumerate(keys):
+                facets[key] = values[idx]
+        else:
+            logger.debug("Wrong dataset_id_template_ %s for dataset %s",
+                         template, dataset_id)
         # The dataset_id does not contain the short_name for all projects,
         # so get it from the filename if needed:
         if 'short_name' not in facets:
diff --git a/tests/unit/esgf/test_search.py b/tests/unit/esgf/test_search.py
index ef38dd952d..52959f2b8a 100644
--- a/tests/unit/esgf/test_search.py
+++ b/tests/unit/esgf/test_search.py
@@ -372,6 +372,31 @@ def test_select_by_time_nodate():
     assert result == files
 
 
+def test_invalid_dataset_id_template():
+    dataset_id = (
+        'obs4MIPs.IUP.XCH4_CRDP3.xch4.mon.v100')
+    dataset_id_template = (
+        '%(project)s.%(institute)s.%(source_id)s.%(time_frequency)s'
+    )
+    filenames = ['xch4_ghgcci_l3_v100_200301_201412.nc']
+    results = [
+        FileResult(
+            json={
+                'title': filename,
+                'dataset_id': dataset_id + '|esgf.ceda.ac.uk',
+                'dataset_id_template_': [dataset_id_template],
+                'project': ['obs4MIPs'],
+                'size': 100,
+                'source_id': 'XCH4_CRDP3',
+            },
+            context=None,
+        ) for filename in filenames
+    ]
+    file = ESGFFile(results)
+
+    assert file.name == filenames[0]
+
+
 def test_search_unknown_project():
     project = 'Unknown'
     msg = (f"Unable to download from ESGF, because project {project} is not on"