Remove more-itertools dependency (#183)

* edit dependencies

* implement own ichunked

conda_index/index/convert_cache.py

@@ -13,8 +13,7 @@
 import os.path
 import re
 import sqlite3
-
-from more_itertools import ichunked
+from itertools import chain, islice
 
 from . import common
 
@@ -196,6 +195,17 @@ def db_path(match, override_channel=None):
     return f"{match['basename']}"
 
 
+def ichunked(iterable, n):
+    """
+    Break iterable into n-item iterables.
+
+    Lazier than Python 3.12 batched().
+    """
+    source = iter(iterable)
+    for item in source:
+        yield islice(chain([item], source), n)
+
+
 def convert_cache(conn, cache_generator):
     """
     Convert old style `conda index` cache to sqlite cache.

news/126-less-itertools

@@ -0,0 +1,19 @@
+### Enhancements
+
+* <news item>
+
+### Bug fixes
+
+* <news item>
+
+### Deprecations
+
+* <news item>
+
+### Docs
+
+* <news item>
+
+### Other
+
+* Remove `more-itertools` dependency. (#127)

pyproject.toml

@@ -17,7 +17,6 @@ dependencies = [
     "conda-package-streaming >=0.7.0",
     "filelock",
     "jinja2",
-    "more-itertools",
     "ruamel.yaml",
     "zstandard",
 ]

recipe/meta.yaml

@@ -30,7 +30,6 @@ requirements:
     - conda-package-streaming
     - filelock
     - jinja2
-    - more-itertools
     - ruamel.yaml
 
 test:

tests/test_utils.py

@@ -1,6 +1,7 @@
 import pathlib
 import tempfile
 
+from conda_index.index.convert_cache import ichunked
 from conda_index.utils import file_contents_match
 
 
@@ -25,3 +26,51 @@ def test_file_contents_match():
 
         assert file_contents_match(c, c)
         assert file_contents_match(c, d)
+
+
+def test_ichunked():
+    """
+    Test laziness of our version of ichunked.
+    """
+    CHUNK_SIZE = 5  # not divisible into total
+    TOTAL = 32
+    REMAINDER = TOTAL - (TOTAL // CHUNK_SIZE) * CHUNK_SIZE
+
+    consumed = -1
+    generated = 0
+
+    def counter():
+        nonlocal consumed
+        consumed += 1
+        return consumed
+
+    def counters():
+        nonlocal generated
+        for i in range(TOTAL):
+            generated = i
+            yield i, counter
+
+    print("More lazy version")
+    for chunk in ichunked(counters(), CHUNK_SIZE):
+        print("Batch")
+        chunk_size = 0
+        for i, c in chunk:
+            chunk_size += 1
+            count = c()
+            print(i, generated, count)
+            assert i == generated == count
+        assert chunk_size == CHUNK_SIZE or chunk_size == REMAINDER
+
+    try:
+        from itertools import batched
+    except ImportError:
+        return
+
+    # demonstrate that generated is sometimes greater than i, c() in
+    # gathers-into-tuples implementation
+    print("Less lazy version")
+    consumed = -1
+    for chunk in batched(counters(), CHUNK_SIZE):
+        print("Batch")
+        for i, c in chunk:
+            print(i, generated, c())