inspectr 0.0.2__py3-none-any.whl → 0.0.4__py3-none-any.whl

inspectr/duplicates.py

@@ -0,0 +1,50 @@
+import sys
+import hashlib
+from collections import defaultdict
+
+def find_duplicates(files, block_size=10, min_occur=3):
+    """
+    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=10, min_occur=3):
+        print(f"{fname}:{lnum}  (occurs {count} times)")
+

{inspectr-0.0.2.dist-info → inspectr-0.0.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: inspectr
-Version: 0.0.2
+Version: 0.0.4
 Summary: A collection of python tools to inspect code quality. 
 Maintainer-email: Alex Mueller <amueller474@gmail.com>
 License-Expression: Apache-2.0

{inspectr-0.0.2.dist-info → inspectr-0.0.4.dist-info}/RECORD

@@ -3,11 +3,12 @@ inspectr/__main__.py,sha256=Rojob_KTdzcHjItkGgA3TPLqPd223pIfeajpCWDP3Gc,652
 inspectr/authenticity.py,sha256=fgKJSE_1qYE1NgLdXmOjPQjo-dwD42sEhqpcsazIj3U,1703
 inspectr/bare_ratio.py,sha256=oQfag96zjGnmFHofZ5qabBGGVoKUnmRNzGq_WX_8ZZg,1997
 inspectr/count_exceptions.py,sha256=UhUc3ZZY8eDxLtlLX1USKzP_JEPY3clx8R8TPvtQ5LE,1588
+inspectr/duplicates.py,sha256=xAFNO90JVEhzZqmyHg8gXAo28e_DIrB3SYh7tONAzbU,1455
 inspectr/size_counts.py,sha256=Yw-k5v7lPylzfh-9SUs1PtlV3Bhvdy9X7w2Pi93xL9g,2920
 inspectr/with_open.py,sha256=qbCAb14cZ15yX7SbLo2C3nE1sDEKLiZ48bvFQfTAKJk,606
-inspectr-0.0.2.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-inspectr-0.0.2.dist-info/METADATA,sha256=4L9CFN6aRG8bB1Z8rkw5exLeRU6FvApMu6U0vcGpKiE,1573
-inspectr-0.0.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-inspectr-0.0.2.dist-info/entry_points.txt,sha256=IrOM4SpCRfFZzBWngg3ezXuX31ZqFBR-flSU5c8tbTo,52
-inspectr-0.0.2.dist-info/top_level.txt,sha256=NlTFBMaWgYmxFzjQtp4Y6itVQQV8sBPtAAZvFnviT-A,9
-inspectr-0.0.2.dist-info/RECORD,,
+inspectr-0.0.4.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+inspectr-0.0.4.dist-info/METADATA,sha256=2YAiL5P6MPWIgR54rlkZXzt3fXxaW6CsT1vT1krn_uo,1573
+inspectr-0.0.4.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+inspectr-0.0.4.dist-info/entry_points.txt,sha256=IrOM4SpCRfFZzBWngg3ezXuX31ZqFBR-flSU5c8tbTo,52
+inspectr-0.0.4.dist-info/top_level.txt,sha256=NlTFBMaWgYmxFzjQtp4Y6itVQQV8sBPtAAZvFnviT-A,9
+inspectr-0.0.4.dist-info/RECORD,,