Fix a missing import bug.

[python_utils.git] / string_utils.py

diff --git a/string_utils.py b/string_utils.py
index 9a38d25c49cccddceec4da06ee8bbfe8133749aa..097dc1b092dc51bb031f104e82e09047bef1b8ad 100644 (file)
--- a/string_utils.py
+++ b/string_utils.py
@@ -1141,6 +1141,24 @@ def valid_datetime(in_str: str) -> bool:
     return False
 
 
+def squeeze(in_str: str, character_to_squeeze: str = ' ') -> str:
+    """
+    Squeeze runs of more than one character_to_squeeze into one.
+
+    >>> squeeze(' this        is       a    test    ')
+    ' this is a test '
+
+    >>> squeeze('one|!||!|two|!||!|three', character_to_squeeze='|!|')
+    'one|!|two|!|three'
+
+    """
+    return re.sub(
+        r'(' + re.escape(character_to_squeeze) + r')+',
+        character_to_squeeze,
+        in_str
+    )
+
+
 def dedent(in_str: str) -> str:
     """
     Removes tab indentation from multi line strings (inspired by analogous Scala function).
@@ -1503,12 +1521,16 @@ def from_bitstring(bits: str, encoding='utf-8', errors='surrogatepass') -> str:
     return n.to_bytes((n.bit_length() + 7) // 8, 'big').decode(encoding, errors) or '\0'
 
 
-def ip_v4_sort_key(txt: str) -> str:
+def ip_v4_sort_key(txt: str) -> Tuple[int]:
     """Turn an IPv4 address into a tuple for sorting purposes.
 
     >>> ip_v4_sort_key('10.0.0.18')
     (10, 0, 0, 18)
 
+    >>> ips = ['10.0.0.10', '100.0.0.1', '1.2.3.4', '10.0.0.9']
+    >>> sorted(ips, key=lambda x: ip_v4_sort_key(x))
+    ['1.2.3.4', '10.0.0.9', '10.0.0.10', '100.0.0.1']
+
     """
     if not is_ip_v4(txt):
         print(f"not IP: {txt}")
@@ -1516,6 +1538,21 @@ def ip_v4_sort_key(txt: str) -> str:
     return tuple([int(x) for x in txt.split('.')])
 
 
+def path_ancestors_before_descendants_sort_key(volume: str) -> Tuple[str]:
+    """Chunk up a file path so that parent/ancestor paths sort before
+    children/descendant paths.
+
+    >>> path_ancestors_before_descendants_sort_key('/usr/local/bin')
+    ('usr', 'local', 'bin')
+
+    >>> paths = ['/usr/local', '/usr/local/bin', '/usr']
+    >>> sorted(paths, key=lambda x: path_ancestors_before_descendants_sort_key(x))
+    ['/usr', '/usr/local', '/usr/local/bin']
+
+    """
+    return tuple([x for x in volume.split('/') if len(x) > 0])
+
+
 if __name__ == '__main__':
     import doctest
     doctest.testmod()