Improve encoding and stability (#43)

Extends PR #42

pyalex/api.py

@@ -23,31 +23,44 @@ def __setattr__(self, key, value):
 config = AlexConfig(
     email=None,
     api_key=None,
-    user_agent="pyalex/" + __version__,
+    user_agent="pyalex/{__version__}",
     openalex_url="https://api.openalex.org",
     max_retries=0,
     retry_backoff_factor=0.1,
     retry_http_codes=[429, 500, 503],
 )
 
 
+def _quote_oa_value(v):
+    """Prepare a value for the OpenAlex API.
+
+    Applies URL encoding to strings and converts booleans to lowercase strings.
+    """
+
+    # workaround for bug https://groups.google.com/u/1/g/openalex-users/c/t46RWnzZaXc
+    if isinstance(v, bool):
+        return str(v).lower()
+
+    if isinstance(v, str):
+        return quote_plus(v)
+
+    return v
+
+
 def _flatten_kv(d, prefix=""):
     if isinstance(d, dict):
         t = []
         for k, v in d.items():
             if isinstance(v, list):
-                t.extend([f"{prefix}.{k}:{i}" for i in v])
+                t.extend([f"{prefix}.{k}:{_quote_oa_value(i)}" for i in v])
             else:
                 new_prefix = f"{prefix}.{k}" if prefix else f"{k}"
                 x = _flatten_kv(v, prefix=new_prefix)
                 t.append(x)
 
         return ",".join(t)
     else:
-        # workaround for bug https://groups.google.com/u/1/g/openalex-users/c/t46RWnzZaXc
-        d = str(d).lower() if isinstance(d, bool) else d
-
-        return f"{prefix}:{d}"
+        return f"{prefix}:{_quote_oa_value(d)}"
 
 
 def _params_merge(params, add_params):
@@ -199,10 +212,11 @@ def _get_multi_items(self, record_list):
 
     def _full_collection_name(self):
         if self.params is not None and "q" in self.params.keys():
-            base_url = config.openalex_url + "/autocomplete/"
-            return base_url + self.__class__.__name__.lower()
+            return (
+                f"{config.openalex_url}/autocomplete/{self.__class__.__name__.lower()}"
+            )
         else:
-            return config.openalex_url + "/" + self.__class__.__name__.lower()
+            return f"{config.openalex_url}/{self.__class__.__name__.lower()}"
 
     def __getattr__(self, key):
         if key == "groupby":
@@ -223,7 +237,7 @@ def __getitem__(self, record_id):
             return self._get_multi_items(record_id)
 
         return self._get_from_url(
-            self._full_collection_name() + "/" + record_id, return_meta=False
+            f"{self._full_collection_name()}/{record_id}", return_meta=False
         )
 
     @property
@@ -236,15 +250,14 @@ def url(self):
             if v is None:
                 pass
             elif isinstance(v, list):
-                v_quote = [quote_plus(q) for q in v]
-                l_params.append(k + "=" + ",".join(v_quote))
+                l_params.append("{}={}".format(k, ",".join(map(_quote_oa_value, v))))
             elif k in ["filter", "sort"]:
-                l_params.append(k + "=" + quote_plus(_flatten_kv(v)))
+                l_params.append(f"{k}={_flatten_kv(v)}")
             else:
-                l_params.append(k + "=" + quote_plus(str(v)))
+                l_params.append(f"{k}={_quote_oa_value(v)}")
 
         if l_params:
-            return self._full_collection_name() + "?" + "&".join(l_params)
+            return "{}?{}".format(self._full_collection_name(), "&".join(l_params))
 
         return self._full_collection_name()
 
@@ -464,7 +477,7 @@ class autocompletes(BaseOpenAlex):
 
     def __getitem__(self, key):
         return self._get_from_url(
-            config.openalex_url + "/autocomplete" + "?q=" + key, return_meta=False
+            f"{config.openalex_url}/autocomplete?q={key}", return_meta=False
         )
 
 

tests/test_pyalex.py

@@ -139,7 +139,7 @@ def test_works_multifilter():
 
 
 def test_works_url():
-    url = "https://api.openalex.org/works?filter=publication_year%3A2020%2Cis_oa%3Atrue"
+    url = "https://api.openalex.org/works?filter=publication_year:2020,is_oa:true"
 
     assert url == Works().filter(publication_year=2020, is_oa=True).url
     assert url == Works().filter(publication_year=2020).filter(is_oa=True).url
@@ -258,7 +258,7 @@ def test_random_publishers():
 
 def test_and_operator():
     # https://github.com/J535D165/pyalex/issues/11
-    url = "https://api.openalex.org/works?filter=institutions.country_code%3Atw%2Cinstitutions.country_code%3Ahk%2Cinstitutions.country_code%3Aus%2Cpublication_year%3A2022"
+    url = "https://api.openalex.org/works?filter=institutions.country_code:tw,institutions.country_code:hk,institutions.country_code:us,publication_year:2022"
 
     assert (
         url
@@ -288,12 +288,12 @@ def test_and_operator():
 
 
 def test_sample():
-    url = "https://api.openalex.org/works?filter=publication_year%3A2020%2Cis_oa%3Atrue&sample=50"
+    url = "https://api.openalex.org/works?filter=publication_year:2020,is_oa:true&sample=50"
     assert url == Works().filter(publication_year=2020, is_oa=True).sample(50).url
 
 
 def test_sample_seed():
-    url = "https://api.openalex.org/works?filter=publication_year%3A2020%2Cis_oa%3Atrue&sample=50&seed=535"  # noqa
+    url = "https://api.openalex.org/works?filter=publication_year:2020,is_oa:true&sample=50&seed=535"  # noqa
     assert (
         url
         == Works().filter(publication_year=2020, is_oa=True).sample(50, seed=535).url
@@ -332,3 +332,22 @@ def test_autocomplete():
 
 def test_filter_urlencoding():
     assert Works().filter(doi="10.1207/s15327809jls0703&4_2").count() == 1
+    assert (
+        Works()["https://doi.org/10.1207/s15327809jls0703&4_2"]["id"]
+        == "https://openalex.org/W4238483711"
+    )
+
+
+@pytest.mark.skip("This test is not working due to inconsistencies in the API.")
+def test_urlencoding_list():
+    assert (
+        Works()
+        .filter(
+            doi=[
+                "https://doi.org/10.1207/s15327809jls0703&4_2",
+                "https://doi.org/10.1001/jama.264.8.944b",
+            ]
+        )
+        .count()
+        == 2
+    )