Add lots of random fixes

.github/workflows/cd.yml

@@ -22,9 +22,13 @@ jobs:
           pip install setuptools wheel hatch
           hatch env create
 
+      - name: Make sure we didn't forget anything in pre-commit
+        run: |
+          hatch run pre-commit run --all
+
       - name: Test with pytest
         run: |
-          hatch run test:run
+          hatch run pytest
 
       - name: Build and publish
         env:

.pre-commit-config.yaml

@@ -5,19 +5,16 @@ repos:
       - id: trailing-whitespace
       - id: end-of-file-fixer
       - id: check-yaml
-  - repo: https://github.com/psf/black
-    rev: 22.10.0
-    hooks:
-      - id: black
-        language_version: python3
-  - repo: https://github.com/pycqa/isort
-    rev: 5.12.0
-    hooks:
-      - id: isort
-        name: isort (python)
-        args: [ "--profile", "black" ]
-  - repo: https://github.com/charliermarsh/ruff-pre-commit
+  - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: 'v0.0.254'
+    rev: v0.3.3
     hooks:
+      # Run the linter.
       - id: ruff
+        args: [ --fix ]
+      # Run the formatter.
+      - id: ruff-format
+  - repo: https://github.com/pre-commit/mirrors-mypy
+    rev: 'v1.9.0'
+    hooks:
+    -   id: mypy

pyproject.toml

@@ -17,20 +17,27 @@ dependencies = [
     "bibtexparser~=1.4.0",
     "networkx~=3.0",
     "typer[all]~=0.9.0",
+    "xlsxwriter~=3.2.0",
 ]
 
 [project.optional-dependencies]
 dev = [
-    "black~=22.10.0",
     "pytest~=7.2.0",
-    "isort~=5.12.0",
     "pre-commit~=2.20.0",
-    "ruff~=0.0.254",
+    "ruff~=0.3.3",
+    "mypy~=1.9.0",
 ]
 
-[tool.ruff]
+[project.scripts]
+bibx = "bibx.cli:app"
+
+[tool.ruff.lint]
+select = ["I", "E", "F", "W"]
 ignore = ["E501"]
 
+[tool.mypy]
+mypy_path = "./stubs/"
+
 [build-system]
 requires = ["hatchling"]
 build-backend = "hatchling.build"
@@ -40,17 +47,3 @@ path = "src/bibx/__init__.py"
 
 [tool.hatch.envs.default]
 features = ["dev"]
-
-[tool.hatch.envs.test]
-dependencies = [
-    "coverage[toml]",
-    "pytest",
-    "pytest-cov",
-]
-[tool.hatch.envs.test.scripts]
-run-coverage = "pytest --cov-config=pyproject.toml --cov=bibx --cov=tests"
-run = "run-coverage --no-cov"
-
-[tool.hatch.envs.docs.scripts]
-build = "mkdocs build --clean --strict"
-serve = "mkdocs serve --dev-addr localhost:8000"

src/bibx/_entities/collection.py

@@ -98,7 +98,7 @@ def cited_by_year(self) -> Dict[int, int]:
             cited_items_per_year[year] = 0
 
         for article in self.articles:
-            if article.times_cited is None:
+            if article.times_cited is None or article.year is None:
                 continue
             if article.year in cited_items_per_year:
                 cited_items_per_year[article.year] += article.times_cited

src/bibx/_entities/collection_builders/base.py

@@ -4,5 +4,4 @@
 
 
 class CollectionBuilder(Protocol):
-    def build(self) -> Collection:
-        ...
+    def build(self) -> Collection: ...

src/bibx/_entities/collection_builders/scopus_bib.py

@@ -32,9 +32,11 @@ def _article_from_entry(self, entry: dict) -> Article:
         if "author" not in entry or "year" not in entry:
             raise MissingCriticalInformation()
         if "note" in entry:
-            times_cited = re.search(r"cited By (\d+)", entry["note"], re.IGNORECASE)
-            if times_cited:
-                times_cited = int(times_cited.groups()[0])
+            match = re.search(r"cited By (\d+)", entry["note"], re.IGNORECASE)
+            if match:
+                times_cited = int(match.groups()[0])
+            else:
+                times_cited = None
         else:
             times_cited = None
         return Article(

src/bibx/_entities/collection_builders/scopus_ris.py

@@ -18,7 +18,7 @@ def _size(file) -> int:
     return size
 
 
-def _int_or_nothing(raw: List[str]) -> Optional[int]:
+def _int_or_nothing(raw: Optional[List[str]]) -> Optional[int]:
     if not raw:
         return None
     try:
@@ -102,9 +102,11 @@ def _article_form_reference(cls, scopusref: str) -> Article:
         return Article(
             authors=[f"{first_name} {last_name.replace(' ', '').replace('.', '')}"],
             year=int(year),
-            journal=journal.strip().replace(".", "").upper()
-            if not journal.isspace()
-            else None,
+            journal=(
+                journal.strip().replace(".", "").upper()
+                if not journal.isspace()
+                else None
+            ),
             volume=volume_info.get("volume"),
             page=volume_info.get("page"),
             doi=doi,

src/bibx/_entities/collection_builders/wos.py

@@ -281,7 +281,7 @@ def _get_articles_from_references(
 
     @classmethod
     def _parse_article_from_str(cls, article_as_str: str) -> Article:
-        article_data = collections.defaultdict(list)
+        article_data: Dict[str, List[str]] = collections.defaultdict(list)
         article_data.setdefault("CR", [])
         field = None
         for line in article_as_str.split("\n"):

src/bibx/algorithms/sap.py

@@ -95,7 +95,7 @@ def clean_graph(g: nx.DiGraph) -> nx.DiGraph:
         """
         # Extract the giant component of the graph
         giant_component_nodes = max(nx.weakly_connected_components(g), key=len)
-        giant: nx.DiGraph = g.subgraph(giant_component_nodes).copy()
+        giant: nx.DiGraph = cast(nx.DiGraph, g.subgraph(giant_component_nodes).copy())
 
         # Remove nodes that cite one element and are never cited themselves
         giant.remove_nodes_from(
@@ -312,4 +312,4 @@ def _clear(graph: nx.DiGraph) -> nx.DiGraph:
             and graph.nodes[n][TRUNK] > 0
             and graph.nodes[n][LEAF] > 0
         ]
-        return graph.subgraph(nodes)
+        return cast(nx.DiGraph, graph.subgraph(nodes))

stubs/bibtexparser/__init__.pyi

@@ -0,0 +1,3 @@
+from _typeshed import Incomplete
+
+def load(bibtex_file, parser: Incomplete | None = None): ...

stubs/networkx/__init__.pyi

@@ -0,0 +1,31 @@
+from typing import Any, Dict, Iterable, Iterator, List
+
+from _typeshed import Incomplete, Self
+
+__version__: str
+
+class Graph:
+    def subgraph(self: Self, nodes: Iterable) -> Self: ...
+    def copy(self: Self) -> Self: ...
+    @property
+    def nodes(self) -> Dict[str, Incomplete]: ...
+    def add_node(self, node: str, **kwargs) -> None: ...
+    def add_edge(self, u: str, v: str, **kwargs) -> None: ...
+    def add_edges_from(self, edges: Iterable) -> None: ...
+    def remove_nodes_from(self, edges: Iterable) -> None: ...
+    def remove_edges_from(self, edges: Iterable) -> None: ...
+    def __iter__(self) -> Iterator[str]: ...
+    def in_degree(self, node: str) -> int: ...
+    def out_degree(self, node: str) -> int: ...
+
+class DiGraph(Graph):
+    def to_undirected(self) -> Graph: ...
+    def successors(self, node: str) -> Iterable[str]: ...
+    def predecessors(self, node: str) -> Iterable[str]: ...
+
+# NOTE: The attr parameter of this function is quite more complicated than this
+def set_node_attributes(g: DiGraph, attr: Any, name: str) -> None: ...
+def topological_sort(g: DiGraph) -> Iterable[str]: ...
+def selfloop_edges(g: DiGraph) -> Iterable[str]: ...
+def weakly_connected_components(g: DiGraph) -> Iterable[List[str]]: ...
+def strongly_connected_components(g: DiGraph) -> Iterable[List[str]]: ...

stubs/networkx/algorithms/community/louvain.pyi

@@ -0,0 +1,3 @@
+from networkx import Graph
+
+def louvain_communities(g: Graph): ...

tests/algorithms/test_sap.py

@@ -1,5 +1,4 @@
 import networkx as nx
-
 from bibx.algorithms.sap import Sap
 
 

tests/entities/test_collection.py

@@ -13,7 +13,7 @@
         times_cited=0,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -28,7 +28,7 @@
         times_cited=2,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -43,7 +43,7 @@
         times_cited=12,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -58,7 +58,7 @@
         times_cited=2,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -73,7 +73,7 @@
         times_cited=0,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -88,7 +88,7 @@
         times_cited=None,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -103,7 +103,7 @@
         times_cited=1,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -118,7 +118,7 @@
         times_cited=20,
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -132,7 +132,7 @@
         doi="19",
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -146,7 +146,7 @@
         doi="19",
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),
@@ -159,7 +159,7 @@
         doi="19",
         references=[],
         keywords=[],
-        sources=[],
+        sources=set(),
         extra={},
         _label=None,
     ),