str.lower -> str.casefold; support for plural acronyms (issue #80)

CHANGELOG.md

@@ -16,6 +16,7 @@ Also adds a new ALFPath class to replace alf path functions.
 - ALF cache table generation has lower memory footprint
 - setup in silent mode now uses defaults if base url matches default one
 - bugfix: error downloading from http server with keep_uuids=True
+- one.alf.spec.readableALF and one.alf.spec._dromedary preserve plural acronyms, e.g. 'ROIs'
 
 ### Added
 

one/alf/spec.py

@@ -246,21 +246,22 @@ def _dromedary(string) -> str:
     >>> _dromedary('motion_energy') == 'motionEnergy'
     >>> _dromedary('passive_RFM') == 'passive RFM'
     >>> _dromedary('FooBarBaz') == 'fooBarBaz'
+    >>> _dromedary('mpci ROIs') == 'mpciROIs'
 
     See Also
     --------
     readableALF
     """
     def _capitalize(x):
-        return x if x.isupper() else x.capitalize()
+        return x if re.match(r'^[A-Z]+s?$', x) else x.capitalize()
     if not string:  # short circuit on None and ''
         return string
     first, *other = re.split(r'[_\s]', string)
     if len(other) == 0:
         # Already camel/Pascal case, ensure first letter lower case
         return first[0].lower() + first[1:]
     # Convert to camel case, preserving all-uppercase elements
-    first = first if first.isupper() else first.casefold()
+    first = first if re.match(r'^[A-Z]+s?$', first) else first.lower()
     return ''.join([first, *map(_capitalize, other)])
 
 
@@ -486,10 +487,10 @@ def readableALF(name: str, capitalize: bool = False) -> str:
     """
     words = []
     i = 0
-    matches = re.finditer(r'[A-Z](?=[a-z0-9])|(?<=[a-z0-9])[A-Z]', name)
+    matches = re.finditer(r'[A-Z](?=[a-rt-z0-9])|(?<=[a-z0-9])[A-Z]', name)
     for j in map(re.Match.start, matches):
         words.append(name[i:j])
         i = j
     words.append(name[i:])
-    display_str = ' '.join(map(lambda s: s if s.isupper() else s.lower(), words))
+    display_str = ' '.join(map(lambda s: s if re.match(r'^[A-Z]+s?$', s) else s.lower(), words))
     return display_str[0].upper() + display_str[1:] if capitalize else display_str

one/api.py

@@ -1926,13 +1926,13 @@ def list_aggregates(self, relation: str, identifier: str = None,
             r'^[\w\/]+(?=aggregates\/)', '', n=1, regex=True)
         # The relation is the first part after 'aggregates', i.e. the second part
         records['relation'] = records['rel_path'].map(
-            lambda x: x.split('aggregates')[-1].split('/')[1].lower())
-        records = records[records['relation'] == relation.lower()]
+            lambda x: x.split('aggregates')[-1].split('/')[1].casefold())
+        records = records[records['relation'] == relation.casefold()]
 
         def path2id(p) -> str:
             """Extract identifier from relative path."""
             parts = alfiles.rel_path_parts(p)[0].split('/')
-            idx = list(map(str.lower, parts)).index(relation.lower()) + 1
+            idx = list(map(str.casefold, parts)).index(relation.casefold()) + 1
             return '/'.join(parts[idx:])
 
         records['identifier'] = records['rel_path'].map(path2id)

one/params.py

@@ -153,7 +153,7 @@ def setup(client=None, silent=False, make_default=None, username=None, cache_dir
             if par[k] and len(par[k]) >= 2 and par[k][0] in quotes and par[k][-1] in quotes:
                 warnings.warn('Do not use quotation marks with input answers', UserWarning)
                 ans = input('Strip quotation marks from response? [Y/n]:').strip() or 'y'
-                if ans.lower()[0] == 'y':
+                if ans.casefold()[0] == 'y':
                     par[k] = par[k].strip(quotes)
             if k == 'ALYX_URL':
                 client = par[k]
@@ -185,17 +185,17 @@ def setup(client=None, silent=False, make_default=None, username=None, cache_dir
             answer = input(
                 'Warning: the directory provided is already a cache for another URL.  '
                 'This may cause conflicts.  Would you like to change the cache location? [Y/n]')
-            if answer and answer[0].lower() == 'n':
+            if answer and answer[0].casefold() == 'n':
                 break
             cache_dir = input(prompt) or cache_dir  # Prompt for another directory
 
         if make_default is None:
             answer = input('Would you like to set this URL as the default one? [Y/n]')
-            make_default = (answer or 'y')[0].lower() == 'y'
+            make_default = (answer or 'y')[0].casefold() == 'y'
 
         # Verify setup pars
         answer = input('Are the above settings correct? [Y/n]')
-        if answer and answer.lower()[0] == 'n':
+        if answer and answer.casefold()[0] == 'n':
             print('SETUP ABANDONED.  Please re-run.')
             return par_current
     else:

one/registration.py

@@ -73,7 +73,7 @@ def get_dataset_type(filename, dtypes):
             if dt.name == obj_attr:
                 dataset_types.append(dt)
         # Check whether pattern matches filename
-        elif fnmatch(filename.name.lower(), dt.filename_pattern.lower()):
+        elif fnmatch(filename.name.casefold(), dt.filename_pattern.casefold()):
             dataset_types.append(dt)
     n = len(dataset_types)
     if n == 0:

one/remote/aws.py

@@ -211,9 +211,9 @@ def get_s3_from_alyx(alyx, repo_name=REPO_DEFAULT):
       returned resource will use an unsigned signature.
     """
     session_keys, bucket_name = get_aws_access_keys(alyx, repo_name)
-    no_creds = not any(filter(None, (v for k, v in session_keys.items() if 'key' in k.lower())))
+    no_creds = not any(filter(None, (v for k, v in session_keys.items() if 'key' in k.casefold())))
     session = boto3.Session(**session_keys)
-    if no_creds and 'public' in bucket_name.lower():
+    if no_creds and 'public' in bucket_name.casefold():
         config = Config(signature_version=UNSIGNED)
     else:
         config = None

one/remote/globus.py

@@ -241,7 +241,7 @@ def get_token(client_id, refresh_tokens=True):
     fields = ('refresh_token', 'access_token', 'expires_at_seconds')
     print('To get a new token, go to this URL and login: {0}'.format(authorize_url))
     auth_code = input('Enter the code you get after login here (press "c" to cancel): ').strip()
-    if auth_code and auth_code.lower() != 'c':
+    if auth_code and auth_code.casefold() != 'c':
         token_response = client.oauth2_exchange_code_for_tokens(auth_code)
         globus_transfer_data = token_response.by_resource_server['transfer.api.globus.org']
         return {k: globus_transfer_data.get(k) for k in fields}

one/tests/alf/test_alf_spec.py

@@ -128,13 +128,15 @@ def test_dromedary(self):
         self.assertEqual(alf_spec._dromedary('passive_RFM'), 'passiveRFM')
         self.assertEqual(alf_spec._dromedary('ROI Motion Energy'), 'ROIMotionEnergy')
         self.assertEqual(alf_spec._dromedary(''), '')
+        self.assertEqual(alf_spec._dromedary('mpci ROIs'), 'mpciROIs')
 
     def test_readable_ALF(self):
         """Test for one.alf.spec.readableALF function."""
         self.assertEqual(alf_spec.readableALF('DAQData'), 'DAQ data')
         self.assertEqual(alf_spec.readableALF('ROIMotion'), 'ROI motion')
         self.assertEqual(alf_spec.readableALF('blueChipTime'), 'blue chip time')
         self.assertEqual(alf_spec.readableALF('someROIDataset'), 'some ROI dataset')
+        self.assertEqual(alf_spec.readableALF('someROIsDataset'), 'some ROIs dataset')
         self.assertEqual(alf_spec.readableALF('fooBAR'), 'foo BAR')
         self.assertEqual(alf_spec.readableALF('fooBAR', capitalize=True), 'Foo BAR')
 

one/tests/test_one.py

@@ -1972,14 +1972,14 @@ def test_setup(self):
         url = TEST_DB_1['base_url']
 
         def mock_input(prompt):
-            if prompt.lower().startswith('warning'):
+            if prompt.casefold().startswith('warning'):
                 if not getattr(mock_input, 'conflict_warn', False):    # Checks both responses
                     mock_input.conflict_warn = True
                     return 'y'
                 return 'n'
-            elif 'download cache' in prompt.lower():
+            elif 'download cache' in prompt.casefold():
                 return Path(self.tempdir.name).joinpath('downloads').as_posix()
-            elif 'url' in prompt.lower():
+            elif 'url' in prompt.casefold():
                 return url
             else:
                 return 'mock_input'

one/tests/test_params.py

@@ -28,9 +28,9 @@ def setUp(self) -> None:
 
     def _mock_input(self, prompt, **kwargs):
         """Stub function for builtins.input"""
-        if prompt.lower().startswith('warning'):
+        if prompt.casefold().startswith('warning'):
             return 'n'
-        elif 'url' in prompt.lower():
+        elif 'url' in prompt.casefold():
             return self.url
         else:
             for k, v in kwargs.items():

one/util.py

@@ -472,11 +472,11 @@ def autocomplete(term, search_terms) -> str:
     """
     Validate search term and return complete name, e.g. autocomplete('subj') == 'subject'.
     """
-    term = term.lower()
+    term = term.casefold()
     # Check if term already complete
     if term in search_terms:
         return term
-    full_key = (x for x in search_terms if x.lower().startswith(term))
+    full_key = (x for x in search_terms if x.casefold().startswith(term))
     key_ = next(full_key, None)
     if not key_:
         raise ValueError(f'Invalid search term "{term}", see `one.search_terms()`')

one/webclient.py

@@ -102,7 +102,7 @@ def wrapper_decorator(alyx_client, *args, expires=None, clobber=False, **kwargs)
             The REST response JSON either from cached file or directly from remote.
         """
         expires = expires or alyx_client.default_expiry
-        mode = (alyx_client.cache_mode or '').lower()
+        mode = (alyx_client.cache_mode or '').casefold()
         if args[0].__name__ != mode and mode != '*':
             return method(alyx_client, *args, **kwargs)
         # Check cache