Re-instate raising TypeError when trying to get string data of non-string object

Author: vnlitvinov

pandas/_libs/hashtable_class_helper.pxi.in

@@ -9,7 +9,7 @@ WARNING: DO NOT edit .pxi FILE directly, .pxi is generated from .pxi.in
 # VectorData
 # ----------------------------------------------------------------------
 
-from pandas._libs.tslibs.util cimport get_string_data
+from pandas._libs.tslibs.util cimport get_string_data, get_string_data_checked
 
 {{py:
 
@@ -597,7 +597,7 @@ cdef class StringHashTable(HashTable):
         cdef:
             khiter_t k
             const char *v
-        get_string_data(val, &v, NULL)
+        get_string_data_checked(val, &v, NULL)
 
         k = kh_get_str(self.table, v)
         if k != self.table.n_buckets:
@@ -611,7 +611,7 @@ cdef class StringHashTable(HashTable):
             int ret = 0
             const char *v
 
-        get_string_data(val, &v, NULL)
+        get_string_data_checked(val, &v, NULL)
 
         k = kh_put_str(self.table, v, &ret)
         self.table.keys[k] = key
@@ -634,7 +634,7 @@ cdef class StringHashTable(HashTable):
         vecs = <const char **>malloc(n * sizeof(char *))
         for i in range(n):
             val = values[i]
-            get_string_data(val, &v, NULL)
+            get_string_data_checked(val, &v, NULL)
             vecs[i] = v
 
         with nogil:
@@ -778,7 +778,7 @@ cdef class StringHashTable(HashTable):
                 labels[i] = na_sentinel
             else:
                 # if ignore_na is False, we also stringify NaN/None/etc.
-                get_string_data(val, &v, NULL)
+                get_string_data_checked(val, &v, NULL)
                 vecs[i] = v
 
         # compute

pandas/_libs/tslibs/np_datetime.pyx

@@ -12,7 +12,7 @@ from cpython.datetime cimport (datetime, date,
 PyDateTime_IMPORT
 
 from numpy cimport int64_t
-from pandas._libs.tslibs.util cimport get_string_data
+from pandas._libs.tslibs.util cimport get_string_data_checked
 
 cdef extern from "src/datetime/np_datetime.h":
     int cmp_npy_datetimestruct(npy_datetimestruct *a,
@@ -177,8 +177,6 @@ cdef inline int _string_to_dts(object val, npy_datetimestruct* dts,
         Py_ssize_t length
         const char* tmp
 
-    if not get_string_data(val, &tmp, &length):
-        raise ValueError('Unable to parse %s' % str(val))
+    get_string_data_checked(val, &tmp, &length)
     return parse_iso_8601_datetime(tmp, length,
                                    dts, out_local, out_tzoffset)
-

pandas/_libs/tslibs/src/datetime/np_datetime_strings.c

@@ -72,7 +72,7 @@ int parse_iso_8601_datetime(const char *str, int len,
     int year_leap = 0;
     int i, numdigits;
     const char *substr;
-    char sublen;
+    int sublen;
 
     /* If year-month-day are separated by a valid separator,
      * months/days without leading zeroes will be parsed
@@ -587,7 +587,8 @@ int get_datetime_iso_8601_strlen(int local, NPY_DATETIMEUNIT base) {
  */
 int make_iso_8601_datetime(npy_datetimestruct *dts, char *outstr, int outlen,
                            NPY_DATETIMEUNIT base) {
-    char *substr = outstr, sublen = outlen;
+    char *substr = outstr;
+    int sublen = outlen;
     int tmplen;
 
     /*

pandas/_libs/tslibs/util.pxd

@@ -1,5 +1,5 @@
 
-from cpython cimport PyTypeObject
+from cpython cimport PyTypeObject, PyErr_BadArgument
 
 cdef extern from *:
     """
@@ -24,9 +24,14 @@ cdef extern from "Python.h":
     bint PyFloat_Check(object obj) nogil
     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) nogil
-    char* PyUnicode_AsUTF8AndSize(object obj, Py_ssize_t* length) nogil
+                                 Py_ssize_t* length)
+    char* PyUnicode_AsUTF8AndSize(object obj, Py_ssize_t* length)
 
 from numpy cimport int64_t
 
@@ -237,14 +242,15 @@ cdef inline bint get_string_data(object s, const char **buf,
                                  Py_ssize_t *length):
     """
     Extract internal char * buffer of unicode or bytes object `s` to `buf` with
-    getting length of this internal buffer, that save in `length`.
-    Return `False` if it failed to extract such buffer for whatever reason
-    otherwise return `True`
+    getting length of this internal buffer saved in `length`.
+    Returns `False` if it failed to extract such buffer for whatever reason,
+    otherwise returns `True`.
 
-    Note
-    ----
-    python object owns memory, `buf` should not be freed
-    `length` can be NULL
+    Notes
+    -----
+    Python object owns memory, `buf` should not be freed.
+    `length` can be NULL if getting buffer length is not needed.
+    This function should only raise exceptions in out-of-memory cases.
 
     Parameters
     ----------
@@ -262,3 +268,12 @@ cdef inline bint get_string_data(object s, const char **buf,
     if PyBytes_Check(s):
         return PyBytes_AsStringAndSize(s, <char**>buf, length) == 0
     return False
+
+cdef inline void get_string_data_checked(object s, const char **buf,
+                                         Py_ssize_t *length):
+    """
+    This is a wrapper for get_string_data() that raises TypeError
+    when supplied with neither unicode nor bytes object
+    """
+    if not get_string_data(s, buf, length):
+        PyErr_BadArgument()