fix: install en_core_web_sm at runtime instead of via poetry (#60)

### Summary of Changes

The previous setup worked perfectly until we tried to publish the
package to PyPI:

```
Invalid value for requires_dist. Error: Can't have direct dependency: 'en_core_web_sm @ https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.5.0/en_core_web_sm-3.5.0-py3-none-any.whl'
```

So, we must now instead guard every usage of `en_core_web_sm` with this
block to ensure it is downloaded before it is loaded:

```py
try:
    nlp = spacy.load("en_core_web_sm")
except OSError:
    spacy.cli.download("en_core_web_sm")
    nlp = spacy.load("en_core_web_sm")
```

pyproject.toml

@@ -24,9 +24,6 @@ numpydoc = "^1.5"
 scipy = "^1.10.1"
 spacy = "^3.5.1"
 
-[tool.poetry.dependencies.en_core_web_sm]
-url = "https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-3.5.0/en_core_web_sm-3.5.0-py3-none-any.whl"
-
 [tool.poetry.group.dev.dependencies]
 pytest = "^7.2.1"
 pytest-cov = "^4.0.0"

tests/processing/dependencies/test_get_dependency.py

@@ -1,4 +1,5 @@
 import spacy
+import spacy.cli
 from library_analyzer.processing.api.model import (
     Parameter,
     ParameterAssignment,
@@ -18,18 +19,22 @@
     extract_lefts_and_rights,
 )
 
-nlp = spacy.load("en_core_web_sm")
+try:
+    nlp = spacy.load("en_core_web_sm")
+except OSError:
+    spacy.cli.download("en_core_web_sm")
+    nlp = spacy.load("en_core_web_sm")
 
 
-def test_extract_lefts_and_rights():
+def test_extract_lefts_and_rights() -> None:
     # string from https://spacy.io/usage/linguistic-features#navigating
     doc = nlp("Autonomous cars shift insurance liability toward manufacturers")
     doc_head_token = doc[2]
     extracted_lefts_and_rights = extract_lefts_and_rights(doc_head_token)
     assert extracted_lefts_and_rights == doc.text.split()
 
 
-def test_extract_action():
+def test_extract_action() -> None:
     action_is_ignored = nlp(
         "this parameter is ignored when fit_intercept is set to False."
     )
@@ -66,7 +71,7 @@ def test_extract_action():
     assert action == Action(action=", X is assumed to be a kernel matrix")
 
 
-def test_extract_condition():
+def test_extract_condition() -> None:
     condition_is_none = nlp(
         "If func is None , then func will be the identity function."
     )
@@ -94,7 +99,7 @@ def test_extract_condition():
     assert condition == Condition(condition="If metric is a string")
 
 
-def test_extract_dependencies_from_docstring_pattern_adverbial_clause():
+def test_extract_dependencies_from_docstring_pattern_adverbial_clause() -> None:
     param_docstring_nlp = nlp("ignored when probability is False")
     dependent_param = Parameter(
         id_="sklearn/sklearn.linear_model/LogisticRegression/random_state",