added python2 support

removed support of unicode strings in python2

added Cython cast from py_string -> const char*

Author: anmyachev

pandas/_libs/lib.pyx

@@ -59,7 +59,7 @@ from pandas._libs.tslibs.nattype import NaT
 from pandas._libs.tslibs.conversion cimport convert_to_tsobject
 from pandas._libs.tslibs.timedeltas cimport convert_to_timedelta64
 from pandas._libs.tslibs.timezones cimport get_timezone, tz_compare
-
+from pandas.compat import string_types
 from pandas._libs.missing cimport (
     checknull, isnaobj, is_null_datetime64, is_null_timedelta64, is_null_period
 )

pandas/_libs/src/parser/tokenizer.c

@@ -1692,7 +1692,6 @@ double precise_xstrtod(const char *str, char **endptr, char decimal,
     int max_digits = 17;
     int n;
 
-    if (maybe_int != NULL) *maybe_int = 1;
     // Cache powers of 10 in memory.
     static double e[] = {
         1.,    1e1,   1e2,   1e3,   1e4,   1e5,   1e6,   1e7,   1e8,   1e9,
@@ -1727,6 +1726,7 @@ double precise_xstrtod(const char *str, char **endptr, char decimal,
         1e290, 1e291, 1e292, 1e293, 1e294, 1e295, 1e296, 1e297, 1e298, 1e299,
         1e300, 1e301, 1e302, 1e303, 1e304, 1e305, 1e306, 1e307, 1e308};
 
+    if (maybe_int != NULL) *maybe_int = 1;
     // Skip leading whitespace.
     while (isspace_ascii(*p)) p++;
 

pandas/_libs/tslibs/util.pxd

@@ -1,5 +1,6 @@
 
 from cpython cimport PyTypeObject
+from cpython.version cimport PY_MAJOR_VERSION
 
 cdef extern from *:
     """
@@ -14,6 +15,29 @@ cdef extern from *:
     object char_to_string(const char* data)
 
 
+cdef extern from *:
+    """
+    #if PY_VERSION_HEX >= 0x03000000
+    #define PyBytes_AsStringAndSize(py_string, buffer, length) \
+        PyBytes_AsStringAndSize(py_string, buffer, length)
+
+    #define PyUnicode_AsUTF8AndSize(py_string, buffer, length) \
+        buffer = PyUnicode_AsUTF8AndSize(py_string, length)
+
+    #else
+    #define PyBytes_AsStringAndSize(py_string, buffer, length) \
+        PyString_AsStringAndSize(py_string, buffer, length)
+
+    #define PyUnicode_AsUTF8AndSize(py_string, buffer, length) \
+        do {                                                   \
+            buffer = PyUnicode_AS_DATA(py_string);             \
+            *length = PyUnicode_GET_SIZE(py_string);           \
+        } while(0)
+    #endif
+    """
+    void PyUnicode_AsUTF8AndSize(object py_string, const char* buffer, Py_ssize_t* length)
+    void PyBytes_AsStringAndSize(object py_string, char** buffer, Py_ssize_t* length)
+
 cdef extern from "Python.h":
     # Note: importing extern-style allows us to declare these as nogil
     # functions, whereas `from cpython cimport` does not.
@@ -24,15 +48,6 @@ cdef extern from "Python.h":
     bint PyComplex_Check(object obj) nogil
     bint PyObject_TypeCheck(object obj, PyTypeObject* type) nogil
 
-    # Note that following functions can potentially raise an exception,
-    # thus they cannot be declared 'nogil'. Also PyUnicode_AsUTF8AndSize() can
-    # potentially allocate memory inside in unlikely case of when underlying
-    # unicode object was stored as non-utf8 and utf8 wasn't requested before.
-    bint PyBytes_AsStringAndSize(object obj, char** buf,
-                                 Py_ssize_t* length) except -1
-    const char* PyUnicode_AsUTF8AndSize(object obj,
-                                        Py_ssize_t* length) except NULL
-
 from numpy cimport int64_t, float64_t
 
 cdef extern from "numpy/arrayobject.h":
@@ -243,7 +258,7 @@ cdef inline bint is_nan(object val):
 
 
 cdef inline const char* get_c_string_buf_and_size(object py_string,
-                                                  Py_ssize_t *length):
+                                                  Py_ssize_t *length) except NULL:
     """
     Extract internal char* buffer of unicode or bytes object `py_string` with
     getting length of this internal buffer saved in `length`.
@@ -263,10 +278,13 @@ cdef inline const char* get_c_string_buf_and_size(object py_string,
     buf : const char*
     """
     cdef:
-        const char *buf
+        const char *buf = NULL
 
     if PyUnicode_Check(py_string):
-        buf = PyUnicode_AsUTF8AndSize(py_string, length)
+        if PY_MAJOR_VERSION > 2:
+            PyUnicode_AsUTF8AndSize(py_string, buf, length)
+        else:
+            buf = py_string
     else:
         PyBytes_AsStringAndSize(py_string, <char**>&buf, length)
     return buf

pandas/tests/io/parser/test_parse_dates.py

@@ -184,7 +184,6 @@ def test_multiple_date_cols_int_cast(all_parsers):
             "KORD,19990127, 23:00:00, 22:56:00, -0.5900")
     parse_dates = {"actual": [1, 2], "nominal": [1, 3]}
     parser = all_parsers
-
     result = parser.read_csv(StringIO(data), header=None,
                              date_parser=conv.parse_date_time,
                              parse_dates=parse_dates, prefix="X")