Merge branch 'main' into arm-runners

Author: diegorusso

Doc/library/contextlib.rst

@@ -151,9 +151,9 @@ Functions and classes provided:
    created by :func:`asynccontextmanager` to meet the requirement that context
    managers support multiple invocations in order to be used as decorators.
 
-  .. versionchanged:: 3.10
-     Async context managers created with :func:`asynccontextmanager` can
-     be used as decorators.
+   .. versionchanged:: 3.10
+      Async context managers created with :func:`asynccontextmanager` can
+      be used as decorators.
 
 
 .. function:: closing(thing)

InternalDocs/garbage_collector.md

@@ -17,26 +17,26 @@ value returned by this function is always 1 more as the function also has a refe
 to the object when called):
 
 ```pycon
-    >>> x = object()
-    >>> sys.getrefcount(x)
-    2
-    >>> y = x
-    >>> sys.getrefcount(x)
-    3
-    >>> del y
-    >>> sys.getrefcount(x)
-    2
+>>> x = object()
+>>> sys.getrefcount(x)
+2
+>>> y = x
+>>> sys.getrefcount(x)
+3
+>>> del y
+>>> sys.getrefcount(x)
+2
 ```
 
 The main problem with the reference counting scheme is that it does not handle reference
 cycles. For instance, consider this code:
 
 ```pycon
-    >>> container = []
-    >>> container.append(container)
-    >>> sys.getrefcount(container)
-    3
-    >>> del container
+>>> container = []
+>>> container.append(container)
+>>> sys.getrefcount(container)
+3
+>>> del container
 ```
 
 In this example, `container` holds a reference to itself, so even when we remove
@@ -199,26 +199,26 @@ variable `A`, and one self-referencing object which is completely
 unreachable:
 
 ```pycon
-    >>> import gc
-
-    >>> class Link:
-    ...    def __init__(self, next_link=None):
-    ...        self.next_link = next_link
-
-    >>> link_3 = Link()
-    >>> link_2 = Link(link_3)
-    >>> link_1 = Link(link_2)
-    >>> link_3.next_link = link_1
-    >>> A = link_1
-    >>> del link_1, link_2, link_3
-
-    >>> link_4 = Link()
-    >>> link_4.next_link = link_4
-    >>> del link_4
-
-    # Collect the unreachable Link object (and its .__dict__ dict).
-    >>> gc.collect()
-    2
+>>> import gc
+>>> 
+>>> class Link:
+...    def __init__(self, next_link=None):
+...        self.next_link = next_link
+...  
+>>> link_3 = Link()
+>>> link_2 = Link(link_3)
+>>> link_1 = Link(link_2)
+>>> link_3.next_link = link_1
+>>> A = link_1
+>>> del link_1, link_2, link_3
+>>> 
+>>> link_4 = Link()
+>>> link_4.next_link = link_4
+>>> del link_4
+>>> 
+>>> # Collect the unreachable Link object (and its .__dict__ dict).
+>>> gc.collect()
+2
 ```
 
 The GC starts with a set of candidate objects it wants to scan.  In the
@@ -439,48 +439,42 @@ These thresholds can be examined using the
 function:
 
 ```pycon
-    >>> import gc
-    >>> gc.get_threshold()
-    (700, 10, 10)
+>>> import gc
+>>> gc.get_threshold()
+(700, 10, 10)
 ```
 
 The content of these generations can be examined using the
 `gc.get_objects(generation=NUM)` function and collections can be triggered
 specifically in a generation by calling `gc.collect(generation=NUM)`.
 
 ```pycon
-    >>> import gc
-    >>> class MyObj:
-    ...     pass
-    ...
-
-    # Move everything to the old generation so it's easier to inspect
-    # the young generation.
-
-    >>> gc.collect()
-    0
-
-    # Create a reference cycle.
-
-    >>> x = MyObj()
-    >>> x.self = x
-
-    # Initially the object is in the young generation.
-
-    >>> gc.get_objects(generation=0)
-    [..., <__main__.MyObj object at 0x7fbcc12a3400>, ...]
-
-    # After a collection of the youngest generation the object
-    # moves to the old generation.
-
-    >>> gc.collect(generation=0)
-    0
-    >>> gc.get_objects(generation=0)
-    []
-    >>> gc.get_objects(generation=1)
-    []
-    >>> gc.get_objects(generation=2)
-    [..., <__main__.MyObj object at 0x7fbcc12a3400>, ...]
+>>> import gc
+>>> class MyObj:
+...     pass
+...
+>>> # Move everything to the old generation so it's easier to inspect
+>>> # the young generation.
+>>> gc.collect()
+0
+>>> # Create a reference cycle.
+>>> x = MyObj()
+>>> x.self = x
+>>> 
+>>> # Initially the object is in the young generation.
+>>> gc.get_objects(generation=0)
+[..., <__main__.MyObj object at 0x7fbcc12a3400>, ...]
+>>> 
+>>> # After a collection of the youngest generation the object
+>>> # moves to the old generation.
+>>> gc.collect(generation=0)
+0
+>>> gc.get_objects(generation=0)
+[]
+>>> gc.get_objects(generation=1)
+[]
+>>> gc.get_objects(generation=2)
+[..., <__main__.MyObj object at 0x7fbcc12a3400>, ...]
 ```
 
 
@@ -563,18 +557,18 @@ the current tracking status of the object. Subsequent garbage collections may ch
 tracking status of the object.
 
 ```pycon
-      >>> gc.is_tracked(0)
-      False
-      >>> gc.is_tracked("a")
-      False
-      >>> gc.is_tracked([])
-      True
-      >>> gc.is_tracked(())
-      False
-      >>> gc.is_tracked({})
-      True
-      >>> gc.is_tracked({"a": 1})
-      True
+>>> gc.is_tracked(0)
+False
+>>> gc.is_tracked("a")
+False
+>>> gc.is_tracked([])
+True
+>>> gc.is_tracked(())
+False
+>>> gc.is_tracked({})
+True
+>>> gc.is_tracked({"a": 1})
+True
 ```
 
 Differences between GC implementations

Lib/fnmatch.py

@@ -77,24 +77,30 @@ def translate(pat):
     There is no way to quote meta-characters.
     """
 
-    STAR = object()
-    parts = _translate(pat, STAR, '.')
-    return _join_translated_parts(parts, STAR)
+    parts, star_indices = _translate(pat, '*', '.')
+    return _join_translated_parts(parts, star_indices)
 
+_re_setops_sub = re.compile(r'([&~|])').sub
+_re_escape = functools.lru_cache(maxsize=512)(re.escape)
 
-def _translate(pat, STAR, QUESTION_MARK):
+def _translate(pat, star, question_mark):
     res = []
     add = res.append
+    star_indices = []
+
     i, n = 0, len(pat)
     while i < n:
         c = pat[i]
         i = i+1
         if c == '*':
+            # store the position of the wildcard
+            star_indices.append(len(res))
+            add(star)
             # compress consecutive `*` into one
-            if (not res) or res[-1] is not STAR:
-                add(STAR)
+            while i < n and pat[i] == '*':
+                i += 1
         elif c == '?':
-            add(QUESTION_MARK)
+            add(question_mark)
         elif c == '[':
             j = i
             if j < n and pat[j] == '!':
@@ -133,8 +139,6 @@ def _translate(pat, STAR, QUESTION_MARK):
                     # Hyphens that create ranges shouldn't be escaped.
                     stuff = '-'.join(s.replace('\\', r'\\').replace('-', r'\-')
                                      for s in chunks)
-                # Escape set operations (&&, ~~ and ||).
-                stuff = re.sub(r'([&~|])', r'\\\1', stuff)
                 i = j+1
                 if not stuff:
                     # Empty range: never match.
@@ -143,50 +147,40 @@ def _translate(pat, STAR, QUESTION_MARK):
                     # Negated empty range: match any character.
                     add('.')
                 else:
+                    # Escape set operations (&&, ~~ and ||).
+                    stuff = _re_setops_sub(r'\\\1', stuff)
                     if stuff[0] == '!':
                         stuff = '^' + stuff[1:]
                     elif stuff[0] in ('^', '['):
                         stuff = '\\' + stuff
                     add(f'[{stuff}]')
         else:
-            add(re.escape(c))
-    assert i == n
-    return res
-
-
-def _join_translated_parts(inp, STAR):
-    # Deal with STARs.
-    res = []
-    add = res.append
-    i, n = 0, len(inp)
-    # Fixed pieces at the start?
-    while i < n and inp[i] is not STAR:
-        add(inp[i])
-        i += 1
-    # Now deal with STAR fixed STAR fixed ...
-    # For an interior `STAR fixed` pairing, we want to do a minimal
-    # .*? match followed by `fixed`, with no possibility of backtracking.
-    # Atomic groups ("(?>...)") allow us to spell that directly.
-    # Note: people rely on the undocumented ability to join multiple
-    # translate() results together via "|" to build large regexps matching
-    # "one of many" shell patterns.
-    while i < n:
-        assert inp[i] is STAR
-        i += 1
-        if i == n:
-            add(".*")
-            break
-        assert inp[i] is not STAR
-        fixed = []
-        while i < n and inp[i] is not STAR:
-            fixed.append(inp[i])
-            i += 1
-        fixed = "".join(fixed)
-        if i == n:
-            add(".*")
-            add(fixed)
-        else:
-            add(f"(?>.*?{fixed})")
+            add(_re_escape(c))
     assert i == n
-    res = "".join(res)
+    return res, star_indices
+
+
+def _join_translated_parts(parts, star_indices):
+    if not star_indices:
+        return fr'(?s:{"".join(parts)})\Z'
+    iter_star_indices = iter(star_indices)
+    j = next(iter_star_indices)
+    buffer = parts[:j]  # fixed pieces at the start
+    append, extend = buffer.append, buffer.extend
+    i = j + 1
+    for j in iter_star_indices:
+        # Now deal with STAR fixed STAR fixed ...
+        # For an interior `STAR fixed` pairing, we want to do a minimal
+        # .*? match followed by `fixed`, with no possibility of backtracking.
+        # Atomic groups ("(?>...)") allow us to spell that directly.
+        # Note: people rely on the undocumented ability to join multiple
+        # translate() results together via "|" to build large regexps matching
+        # "one of many" shell patterns.
+        append('(?>.*?')
+        extend(parts[i:j])
+        append(')')
+        i = j + 1
+    append('.*')
+    extend(parts[i:])
+    res = ''.join(buffer)
     return fr'(?s:{res})\Z'

Lib/glob.py

@@ -312,7 +312,7 @@ def translate(pat, *, recursive=False, include_hidden=False, seps=None):
             if part:
                 if not include_hidden and part[0] in '*?':
                     results.append(r'(?!\.)')
-                results.extend(fnmatch._translate(part, f'{not_sep}*', not_sep))
+                results.extend(fnmatch._translate(part, f'{not_sep}*', not_sep)[0])
             if idx < last_part_idx:
                 results.append(any_sep)
     res = ''.join(results)