From e1a590eec70a169df335ba2d78ba43d035b3aa70 Mon Sep 17 00:00:00 2001
From: Jess <20195932+wrongkindofdoctor@users.noreply.github.com>
Date: Mon, 30 Sep 2024 16:51:20 -0400
Subject: [PATCH] Fix issues in pp query (#692)

* fix hr -> 1hr freq conversion in pp query
try using regex string contains standard_name in query

* add check for parameter type to xr_parser approximate_attribute_value

* remove regex from pp query standard_name

* add check that bounds is populated in cf.assessor, then check coord attrs and only run coord bounds check if bounda s are not None in xr_parser
---
 src/preprocessor.py | 25 ++++++++++-----------
 src/xr_parser.py    | 54 ++++++++++++++++++++++++++++++---------------
 2 files changed, 48 insertions(+), 31 deletions(-)

diff --git a/src/preprocessor.py b/src/preprocessor.py
index 188c90faf..388667ca4 100644
--- a/src/preprocessor.py
+++ b/src/preprocessor.py
@@ -12,7 +12,6 @@
 from src.util import datelabel as dl
 import cftime
 import intake
-import math
 import numpy as np
 import xarray as xr
 import collections
@@ -922,31 +921,31 @@ def query_catalog(self,
 
         for case_name, case_d in case_dict.items():
             # path_regex = re.compile(r'(?i)(?<!\\S){}(?!\\S+)'.format(case_name))
-            path_regex = re.compile(r'({})'.format(case_name))
-            # path_regex = '*' + case_name + '*'
+            path_regex = [re.compile(r'({})'.format(case_name))]
 
             for var in case_d.varlist.iter_vars():
                 realm_regex = var.realm + '*'
-                var_id = var.translation.name
-                standard_name = var.translation.standard_name
-                if var.translation.convention == 'no_translation':
-                    date_range = var.T.range
-                    var_id = var.name
-                    standard_name = var.standard_name
+                date_range = var.T.range
+                var_id = var.name
+                standard_name = var.standard_name
+                if var.translation.convention is not None:
+                    var_id = var.translation.name
+                    standard_name = var.translation.standard_name
+                    date_range = var.translation.T.range
                 if var.is_static:
                     date_range = None
                     freq = "fx"
                 else:
-                    date_range = var.translation.T.range
                     freq = var.T.frequency
-                    if freq == 'hr':
-                        freq = '1hr'
                 if not isinstance(freq, str):
                     freq = freq.format_local()
+                if freq == 'hr':
+                    freq = '1hr'
+
                 # define initial query dictionary with variable settings requirements that do not change if
                 # the variable is translated
                 case_d.query['frequency'] = freq
-                case_d.query['path'] = [path_regex]
+                case_d.query['path'] = path_regex
                 case_d.query['realm'] = realm_regex
                 case_d.query['standard_name'] = standard_name
                 case_d.query['variable_id'] = var_id
diff --git a/src/xr_parser.py b/src/xr_parser.py
index 40e4dce82..371f746b3 100644
--- a/src/xr_parser.py
+++ b/src/xr_parser.py
@@ -722,7 +722,13 @@ def approximate_attribute_value(self, our_name: str, ds_name: str) -> bool:
         """Determine if the dataset attribute value is an approximate match to the expected attribute value"""
         exclude = ["with", "on", "in", "of", "at", "near"]
         our_name_split = [i for i in our_name.split('_') if i not in exclude]
-        ds_name_split = [i for i in ds_name.split('_') if i not in exclude]
+        if isinstance(ds_name, str):
+            ds_name_split = [i for i in ds_name.split('_') if i not in exclude]
+        elif isinstance(ds_name, list):
+            for n in ds_name:
+                print(n)
+            ds_name_split = [i for i in ds_name[0].split('_')]
+
         isect = set(our_name_split).intersection(ds_name_split)
 
         if len(isect) >= len(our_name_split) - 2 and len(isect) > 0:
@@ -868,6 +874,8 @@ def reconcile_attr(self, our_var, ds_var, our_attr_name, ds_attr_name=None,
         """Compare attribute of a :class:`~src.data_model.DMVariable` (*our_var*)
         with what's set in the xarray.Dataset (*ds_var*).
         """
+        if ds_var is None:
+            return
         if ds_attr_name is None:
             ds_attr_name = our_attr_name
         our_attr = getattr(our_var, our_attr_name)
@@ -1063,28 +1071,38 @@ def reconcile_coord_bounds(self, our_coord, ds, ds_coord_name):
         expectations based on the model's convention (*our_var*), for the bounds
         on the dimension coordinate *our_coord*.
         """
-        try:
+        if len(ds.cf.bounds) > 0:
             bounds = ds.cf.get_bounds(ds_coord_name)
-        except KeyError:
-            # cf accessor could't find associated bounds variable
+        elif hasattr(ds[ds_coord_name], 'attrs'):
+            if ds[ds_coord_name].attrs.get('bounds', None):
+                bounds = ds[ds_coord_name].bounds
+                if isinstance(bounds, str):
+                    our_coord.bounds_var = None
+                    return
+            else:
+                our_coord.bounds_var = None
+                return
+        else:
+            # cf accessor couldn't find associated bounds variable
+            bounds = None
             our_coord.bounds_var = None
             return
-
         # Inherit standard_name from our_coord if not present (regardless of
         # skip_std_name), overwriting metadata on bounds if different
-        self.reconcile_attr(our_coord, bounds, 'standard_name',
-                            fill_ours=False, fill_ds=True, overwrite_ours=False
-                            )
-        # Inherit units from our_coord if not present (regardless of skip_units),
-        # overwriting metadata on bounds if different
-        self.reconcile_attr(our_coord, bounds, 'units',
-                            comparison_func=units.units_equal,
-                            fill_ours=False, fill_ds=True, overwrite_ours=False
-                            )
-        if our_coord.name != bounds.name:
-            self.log.debug("Updating %s for '%s' to value '%s' from dataset.",
-                           'bounds', our_coord.name, bounds.name)
-        our_coord.bounds_var = bounds
+        if bounds is not None:
+            self.reconcile_attr(our_coord, bounds, 'standard_name',
+                                fill_ours=False, fill_ds=True, overwrite_ours=False
+                                )
+            # Inherit units from our_coord if not present (regardless of skip_units),
+            # overwriting metadata on bounds if different
+            self.reconcile_attr(our_coord, bounds, 'units',
+                                comparison_func=units.units_equal,
+                                fill_ours=False, fill_ds=True, overwrite_ours=False
+                                )
+            if our_coord.name != bounds.name:
+                self.log.debug("Updating %s for '%s' to value '%s' from dataset.",
+                               'bounds', our_coord.name, bounds.name)
+                our_coord.bounds_var = bounds
 
     def reconcile_dimension_coords(self, our_var, ds):
         """Reconcile name, standard_name and units attributes between the