refactor cdef-s in .pyx files

Author: noamher

pandas/_libs/hashtable.pyx

@@ -52,9 +52,10 @@ include "hashtable_class_helper.pxi"
 include "hashtable_func_helper.pxi"
 
 cdef class Factorizer:
-    cdef public PyObjectHashTable table
-    cdef public ObjectVector uniques
-    cdef public Py_ssize_t count
+    cdef public:
+        PyObjectHashTable table
+        ObjectVector uniques
+        Py_ssize_t count
 
     def __init__(self, size_hint):
         self.table = PyObjectHashTable(size_hint)
@@ -96,9 +97,10 @@ cdef class Factorizer:
 
 
 cdef class Int64Factorizer:
-    cdef public Int64HashTable table
-    cdef public Int64Vector uniques
-    cdef public Py_ssize_t count
+    cdef public:
+        Int64HashTable table
+        Int64Vector uniques
+        Py_ssize_t count
 
     def __init__(self, size_hint):
         self.table = Int64HashTable(size_hint)

pandas/_libs/internals.pyx

@@ -23,10 +23,11 @@ from pandas._libs.algos import ensure_int64
 
 cdef class BlockPlacement:
     # __slots__ = '_as_slice', '_as_array', '_len'
-    cdef slice _as_slice
-    cdef object _as_array
+    cdef:
+        slice _as_slice
+        cdef object _as_array
 
-    cdef bint _has_slice, _has_array, _is_known_slice_like
+        bint _has_slice, _has_array, _is_known_slice_like
 
     def __init__(self, val):
         cdef:

pandas/_libs/lib.pyx

@@ -40,11 +40,12 @@ cdef extern from "numpy/arrayobject.h":
         # Use PyDataType_* macros when possible, however there are no macros
         # for accessing some of the fields, so some are defined. Please
         # ask on cython-dev if you need more.
-        cdef int type_num
-        cdef int itemsize "elsize"
-        cdef char byteorder
-        cdef object fields
-        cdef tuple names
+        cdef:
+            int type_num
+            int itemsize "elsize"
+            char byteorder
+            object fields
+            tuple names
 
 
 cdef extern from "src/parse_helper.h":
@@ -67,17 +68,18 @@ from pandas._libs.missing cimport (
 
 # constants that will be compared to potentially arbitrarily large
 # python int
-cdef object oINT64_MAX = <int64_t>INT64_MAX
-cdef object oINT64_MIN = <int64_t>INT64_MIN
-cdef object oUINT64_MAX = <uint64_t>UINT64_MAX
+cdef:
+    object oINT64_MAX = <int64_t>INT64_MAX
+    object oINT64_MIN = <int64_t>INT64_MIN
+    object oUINT64_MAX = <uint64_t>UINT64_MAX
 
-cdef bint PY2 = sys.version_info[0] == 2
-cdef float64_t NaN = <float64_t>np.NaN
+    bint PY2 = sys.version_info[0] == 2
+    float64_t NaN = <float64_t>np.NaN
 
 
 def values_from_object(obj: object):
     """ return my values or the object if we are say an ndarray """
-    func: object
+    cdef object func
 
     func = getattr(obj, 'get_values', None)
     if func is not None:
@@ -91,9 +93,9 @@ def values_from_object(obj: object):
 def memory_usage_of_objects(arr: object[:]) -> int64_t:
     """ return the memory usage of an object array in bytes,
     does not include the actual bytes of the pointers """
-    i: Py_ssize_t
-    n: Py_ssize_t
-    size: int64_t
+    cdef:
+        Py_ssize_t i, n
+        int64_t size
 
     size = 0
     n = len(arr)

pandas/_libs/missing.pyx

@@ -16,10 +16,11 @@ from pandas._libs.tslibs.nattype cimport (
     checknull_with_nat, c_NaT as NaT, is_null_datetimelike)
 
 
-cdef float64_t INF = <float64_t>np.inf
-cdef float64_t NEGINF = -INF
+cdef:
+    float64_t INF = <float64_t>np.inf
+    float64_t NEGINF = -INF
 
-cdef int64_t NPY_NAT = util.get_nat()
+    int64_t NPY_NAT = util.get_nat()
 
 
 cpdef bint checknull(object val):

pandas/_libs/parsers.pyx

@@ -64,10 +64,11 @@ from pandas.errors import (ParserError, DtypeWarning,
 CParserError = ParserError
 
 
-cdef bint PY3 = (sys.version_info[0] >= 3)
+cdef:
+    bint PY3 = (sys.version_info[0] >= 3)
 
-cdef float64_t INF = <float64_t>np.inf
-cdef float64_t NEGINF = -INF
+    float64_t INF = <float64_t>np.inf
+    float64_t NEGINF = -INF
 
 
 cdef extern from "errno.h":
@@ -735,7 +736,7 @@ cdef class TextReader:
             int status
             int64_t hr, data_line
             char *errors = "strict"
-            cdef StringPath path = _string_path(self.c_encoding)
+            StringPath path = _string_path(self.c_encoding)
 
         header = []
         unnamed_cols = set()
@@ -1389,8 +1390,9 @@ cdef class TextReader:
                 return None
 
 
-cdef object _true_values = [b'True', b'TRUE', b'true']
-cdef object _false_values = [b'False', b'FALSE', b'false']
+cdef:
+    object _true_values = [b'True', b'TRUE', b'true']
+    object _false_values = [b'False', b'FALSE', b'false']
 
 
 def _ensure_encoded(list lst):
@@ -1637,7 +1639,7 @@ cdef _categorical_convert(parser_t *parser, int64_t col,
         int64_t current_category = 0
 
         char *errors = "strict"
-        cdef StringPath path = _string_path(encoding)
+        StringPath path = _string_path(encoding)
 
         int ret = 0
         kh_str_t *table
@@ -1727,9 +1729,10 @@ cdef inline void _to_fw_string_nogil(parser_t *parser, int64_t col,
         data += width
 
 
-cdef char* cinf = b'inf'
-cdef char* cposinf = b'+inf'
-cdef char* cneginf = b'-inf'
+cdef:
+    char* cinf = b'inf'
+    char* cposinf = b'+inf'
+    char* cneginf = b'-inf'
 
 
 cdef _try_double(parser_t *parser, int64_t col,

pandas/_libs/skiplist.pyx

@@ -21,11 +21,6 @@ cdef double Log2(double x):
 # TODO: optimize this, make less messy
 
 cdef class Node:
-    # cdef public:
-    #    double value
-    #    list next
-    #    list width
-
     def __init__(self, double value, list next, list width):
         self.value = value
         self.next = next
@@ -41,9 +36,6 @@ cdef class IndexableSkiplist:
     Sorted collection supporting O(lg n) insertion, removal, and
     lookup by rank.
     """
-    # cdef:
-    #    Py_ssize_t size, maxlevels
-    #    Node head
 
     def __init__(self, expected_size=100):
         self.size = 0
@@ -57,8 +49,9 @@ cdef class IndexableSkiplist:
         return self.get(i)
 
     cpdef get(self, Py_ssize_t i):
-        cdef Py_ssize_t level
-        cdef Node node
+        cdef:
+            Py_ssize_t level
+            Node node
 
         node = self.head
         i += 1
@@ -71,9 +64,10 @@ cdef class IndexableSkiplist:
         return node.value
 
     cpdef insert(self, double value):
-        cdef Py_ssize_t level, steps, d
-        cdef Node node, prevnode, newnode, next_at_level, tmp
-        cdef list chain, steps_at_level
+        cdef:
+            Py_ssize_t level, steps, d
+            Node node, prevnode, newnode, next_at_level, tmp
+            list chain, steps_at_level
 
         # find first node on each level where node.next[levels].value > value
         chain = [None] * self.maxlevels
@@ -110,9 +104,10 @@ cdef class IndexableSkiplist:
         self.size += 1
 
     cpdef remove(self, double value):
-        cdef Py_ssize_t level, d
-        cdef Node node, prevnode, tmpnode, next_at_level
-        cdef list chain
+        cdef:
+            Py_ssize_t level, d
+            Node node, prevnode, tmpnode, next_at_level
+            list chain
 
         # find first node on each level where node.next[levels].value >= value
         chain = [None] * self.maxlevels

pandas/_libs/tslibs/nattype.pyx

@@ -90,10 +90,6 @@ def __nat_unpickle(*args):
 
 
 cdef class _NaT(datetime):
-    # cdef readonly:
-    #    int64_t value
-    #    object freq
-
     def __hash__(_NaT self):
         # py3k needs this defined here
         return hash(self.value)

pandas/_libs/tslibs/parsing.pyx

@@ -44,9 +44,10 @@ class DateParseError(ValueError):
 _DEFAULT_DATETIME = datetime(1, 1, 1).replace(hour=0, minute=0,
                                               second=0, microsecond=0)
 
-cdef object _TIMEPAT = re.compile(r'^([01]?[0-9]|2[0-3]):([0-5][0-9])')
+cdef:
+    object _TIMEPAT = re.compile(r'^([01]?[0-9]|2[0-3]):([0-5][0-9])')
 
-cdef set _not_datelike_strings = {'a', 'A', 'm', 'M', 'p', 'P', 't', 'T'}
+    set _not_datelike_strings = {'a', 'A', 'm', 'M', 'p', 'P', 't', 'T'}
 
 # ----------------------------------------------------------------------
 

pandas/_libs/tslibs/period.pyx

@@ -52,9 +52,10 @@ from pandas._libs.tslibs.nattype cimport (
 from pandas._libs.tslibs.offsets cimport to_offset
 from pandas._libs.tslibs.offsets import _Tick
 
-cdef bint PY2 = str == bytes
-cdef enum:
-    INT32_MIN = -2147483648
+cdef:
+    bint PY2 = str == bytes
+    enum:
+        INT32_MIN = -2147483648
 
 
 ctypedef struct asfreq_info:

pandas/_libs/tslibs/resolution.pyx

@@ -16,15 +16,16 @@ from pandas._libs.tslibs.ccalendar cimport get_days_in_month
 # ----------------------------------------------------------------------
 # Constants
 
-cdef int64_t NPY_NAT = get_nat()
-
-cdef int RESO_NS = 0
-cdef int RESO_US = 1
-cdef int RESO_MS = 2
-cdef int RESO_SEC = 3
-cdef int RESO_MIN = 4
-cdef int RESO_HR = 5
-cdef int RESO_DAY = 6
+cdef:
+    int64_t NPY_NAT = get_nat()
+
+    int RESO_NS = 0
+    int RESO_US = 1
+    int RESO_MS = 2
+    int RESO_SEC = 3
+    int RESO_MIN = 4
+    int RESO_HR = 5
+    int RESO_DAY = 6
 
 # ----------------------------------------------------------------------
 

pandas/_libs/window.pyx

@@ -26,16 +26,17 @@ from pandas._libs.skiplist cimport (
     skiplist_t, skiplist_init, skiplist_destroy, skiplist_get, skiplist_insert,
     skiplist_remove)
 
-cdef float32_t MINfloat32 = np.NINF
-cdef float64_t MINfloat64 = np.NINF
+cdef:
+    float32_t MINfloat32 = np.NINF
+    float64_t MINfloat64 = np.NINF
 
-cdef float32_t MAXfloat32 = np.inf
-cdef float64_t MAXfloat64 = np.inf
+    float32_t MAXfloat32 = np.inf
+    float64_t MAXfloat64 = np.inf
 
-cdef float64_t NaN = <float64_t>np.NaN
+    float64_t NaN = <float64_t>np.NaN
 
-cdef inline int int_max(int a, int b): return a if a >= b else b
-cdef inline int int_min(int a, int b): return a if a <= b else b
+    inline int int_max(int a, int b): return a if a >= b else b
+    inline int int_min(int a, int b): return a if a <= b else b
 
 
 # Cython implementations of rolling sum, mean, variance, skewness,

pandas/io/msgpack/_packer.pyx

@@ -74,14 +74,15 @@ cdef class Packer(object):
         Use bin type introduced in msgpack spec 2.0 for bytes.
         It also enable str8 type for unicode.
     """
-    cdef msgpack_packer pk
-    cdef object _default
-    cdef object _bencoding
-    cdef object _berrors
-    cdef char *encoding
-    cdef char *unicode_errors
-    cdef bint use_float
-    cdef bint autoreset
+    cdef:
+        msgpack_packer pk
+        object _default
+        object _bencoding
+        object _berrors
+        char *encoding
+        char *unicode_errors
+        cdef bint use_float
+        cdef bint autoreset
 
     def __cinit__(self):
         cdef int buf_size = 1024 * 1024
@@ -123,16 +124,17 @@ cdef class Packer(object):
 
     cdef int _pack(self, object o,
                    int nest_limit=DEFAULT_RECURSE_LIMIT) except -1:
-        cdef long long llval
-        cdef unsigned long long ullval
-        cdef long longval
-        cdef float fval
-        cdef double dval
-        cdef char* rawval
-        cdef int ret
-        cdef dict d
-        cdef size_t L
-        cdef int default_used = 0
+        cdef:
+            long long llval
+            unsigned long long ullval
+            long longval
+            float fval
+            double dval
+            char* rawval
+            int ret
+            dict d
+            size_t L
+            int default_used = 0
 
         if nest_limit < 0:
             raise PackValueError("recursion limit exceeded.")