inspectr 0.0.1__tar.gz → 0.0.3__tar.gz

{inspectr-0.0.1 → inspectr-0.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: inspectr
-Version: 0.0.1
+Version: 0.0.3
 Summary: A collection of python tools to inspect code quality. 
 Maintainer-email: Alex Mueller <amueller474@gmail.com>
 License-Expression: Apache-2.0

inspectr-0.0.3/inspectr/duplicates.py

@@ -0,0 +1,51 @@
+# mytool/duplicates.py
+import sys
+import hashlib
+from collections import defaultdict
+
+def find_duplicates(files, block_size=1, min_occur=2):
+    """
+    Find duplicate blocks of code across files.
+
+    Args:
+        files: list of file paths
+        block_size: number of consecutive lines in a block
+        min_occur: minimum number of occurrences to report
+
+    Yields:
+        (filename, line_number, count)
+    """
+    blocks = defaultdict(list)  # hash -> list of (file, line)
+
+    for fname in files:
+        try:
+            with open(fname, encoding="utf-8") as f:
+                lines = f.readlines()
+        except OSError as e:
+            print(f"Could not read {fname}: {e}", file=sys.stderr)
+            continue
+
+        for i in range(len(lines) - block_size + 1):
+            # join block of lines
+            block = "".join(lines[i:i + block_size])
+            # stable hash
+            h = hashlib.sha1(block.encode("utf-8")).hexdigest()
+            blocks[h].append((fname, i + 1))
+
+    for locs in blocks.values():
+        if len(locs) >= min_occur:
+            for fname, lnum in locs:
+                yield fname, lnum, len(locs)
+
+
+def main(args=None) -> None:
+    if args is None:
+        args = sys.argv[1:]
+
+    if not args:
+        print("Usage: inspectr duplicates file1.py [file2.py ...]")
+        return
+
+    for fname, lnum, count in find_duplicates(args, block_size=1, min_occur=2):
+        print(f"{fname}:{lnum}  (occurs {count} times)")
+

{inspectr-0.0.1 → inspectr-0.0.3}/inspectr/size_counts.py

@@ -12,8 +12,6 @@ CLASS_METHODS = 20
 
 
 def main(files: List[pathlib.Path]) -> None:
-    files = list(pathlib.Path.cwd().rglob("*.py"))
-
     # TODO: rename these
     # Counters and diagnostics
     files_over_1000 = []

{inspectr-0.0.1 → inspectr-0.0.3}/inspectr.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: inspectr
-Version: 0.0.1
+Version: 0.0.3
 Summary: A collection of python tools to inspect code quality. 
 Maintainer-email: Alex Mueller <amueller474@gmail.com>
 License-Expression: Apache-2.0

{inspectr-0.0.1 → inspectr-0.0.3}/inspectr.egg-info/SOURCES.txt

@@ -6,6 +6,7 @@ inspectr/__main__.py
 inspectr/authenticity.py
 inspectr/bare_ratio.py
 inspectr/count_exceptions.py
+inspectr/duplicates.py
 inspectr/size_counts.py
 inspectr/with_open.py
 inspectr.egg-info/PKG-INFO

{inspectr-0.0.1 → inspectr-0.0.3}/pyproject.toml

@@ -5,7 +5,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "inspectr"
-version = "0.0.1"
+version = "0.0.3"
 description = "A collection of python tools to inspect code quality. "
 maintainers = [
     {name = "Alex Mueller", email="amueller474@gmail.com"},
@@ -31,4 +31,4 @@ inspectr = "inspectr.__main__:main"
 
 [project.urls]
 Homepage = "https://github.com/ajcm474/inspectr"
-Issues = "https://github.com/ajcm474/inspectr/issues"
+Issues = "https://github.com/ajcm474/inspectr/issues"