ENH: Add islower/isupper/istitle ufuncs for unicode/bytes dtypes

numpy · Feb 8, 2024 · fdb3062 · fdb3062
1 parent 295db0a
commit fdb3062
Show file tree

Hide file tree

Showing 6 changed files with 497 additions and 107 deletions.
diff --git a/numpy/_core/code_generators/generate_umath.py b/numpy/_core/code_generators/generate_umath.py
@@ -1170,6 +1170,21 @@ def english_upper(s):
           docstrings.get('numpy._core.umath.isspace'),
           None,
           ),
+'islower':
+    Ufunc(1, 1, False_,
+          docstrings.get('numpy._core.umath.islower'),
+          None,
+          ),
+'isupper':
+    Ufunc(1, 1, False_,
+          docstrings.get('numpy._core.umath.isupper'),
+          None,
+          ),
+'istitle':
+    Ufunc(1, 1, False_,
+          docstrings.get('numpy._core.umath.istitle'),
+          None,
+          ),
 'isdecimal':
     Ufunc(1, 1, False_,
           docstrings.get('numpy._core.umath.isdecimal'),

diff --git a/numpy/_core/code_generators/ufunc_docstrings.py b/numpy/_core/code_generators/ufunc_docstrings.py
@@ -4441,6 +4441,97 @@ def add_newdoc(place, name, doc):
 
     """)
 
+add_newdoc('numpy._core.umath', 'islower',
+    """
+    Returns true for each element if all cased characters in the
+    string are lowercase and there is at least one cased character,
+    false otherwise.
+
+    Parameters
+    ----------
+    x : array_like, with `np.bytes_` or `np.str_` dtype
+    $PARAMS
+
+    Returns
+    -------
+    out : ndarray
+        Output array of bools
+        $OUT_SCALAR_1
+
+    See Also
+    --------
+    str.islower
+
+    Examples
+    --------
+    >>> np.strings.islower("GHC")
+    array(False)
+    >>> np.strings.islower("ghc")
+    array(True)
+
+    """)
+
+add_newdoc('numpy._core.umath', 'isupper',
+    """
+    Return true for each element if all cased characters in the
+    string are uppercase and there is at least one character, false
+    otherwise.
+
+    Parameters
+    ----------
+    x : array_like, with `np.bytes_` or `np.str_` dtype
+    $PARAMS
+
+    Returns
+    -------
+    out : ndarray
+        Output array of bools
+        $OUT_SCALAR_1
+
+    See Also
+    --------
+    str.isupper
+
+    Examples
+    --------
+    >>> np.strings.isupper("GHC")
+    array(True)     
+    >>> a = np.array(["hello", "HELLO", "Hello"])
+    >>> np.strings.isupper(a)
+    array([False,  True, False]) 
+
+    """)
+
+add_newdoc('numpy._core.umath', 'istitle',
+    """
+    Returns true for each element if the element is a titlecased
+    string and there is at least one character, false otherwise.
+
+    Parameters
+    ----------
+    x : array_like, with `np.bytes_` or `np.str_` dtype
+    $PARAMS
+
+    Returns
+    -------
+    out : ndarray
+        Output array of bools
+        $OUT_SCALAR_1
+
+    See Also
+    --------
+    str.istitle
+
+    Examples
+    --------
+    >>> np.strings.istitle("Numpy Is Great")
+    array(True)
+
+    >>> np.strings.istitle("Numpy is great")
+    array(False)
+    
+    """)
+
 add_newdoc('numpy._core.umath', 'isdecimal',
     """
     For each element, return True if there are only decimal

diff --git a/numpy/_core/src/umath/string_buffer.h b/numpy/_core/src/umath/string_buffer.h
@@ -26,6 +26,9 @@ enum class IMPLEMENTED_UNARY_FUNCTIONS {
     ISDECIMAL,
     ISDIGIT,
     ISSPACE,
+    ISLOWER,
+    ISUPPER,
+    ISTITLE,
     ISNUMERIC,
     STR_LEN,
 };
@@ -136,6 +139,81 @@ codepoint_isspace<ENCODING::UTF8>(npy_ucs4 code)
     return Py_UNICODE_ISSPACE(code);
 }
 
+template<ENCODING enc>
+inline bool
+codepoint_islower(npy_ucs4 code);
+
+template<>
+inline bool
+codepoint_islower<ENCODING::ASCII>(npy_ucs4 code)
+{
+    return Py_ISLOWER((char) code);
+}
+
+template<>
+inline bool
+codepoint_islower<ENCODING::UTF32>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISLOWER(code);
+}
+
+template<>
+inline bool
+codepoint_islower<ENCODING::UTF8>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISLOWER(code);
+}
+
+template<ENCODING enc>
+inline bool
+codepoint_isupper(npy_ucs4 code);
+
+template<>
+inline bool
+codepoint_isupper<ENCODING::ASCII>(npy_ucs4 code)
+{
+    return Py_ISUPPER((char) code);
+}
+
+template<>
+inline bool
+codepoint_isupper<ENCODING::UTF32>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISUPPER(code);
+}
+
+template<>
+inline bool
+codepoint_isupper<ENCODING::UTF8>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISUPPER(code);
+}
+
+template<ENCODING enc>
+inline bool
+codepoint_istitle(npy_ucs4);
+
+template<>
+inline bool
+codepoint_istitle<ENCODING::ASCII>(npy_ucs4 code)
+{
+    return false;
+}
+
+template<>
+inline bool
+codepoint_istitle<ENCODING::UTF32>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISTITLE(code);
+}
+
+template<>
+inline bool
+codepoint_istitle<ENCODING::UTF8>(npy_ucs4 code)
+{
+    return Py_UNICODE_ISTITLE(code);
+}
+
 inline bool
 codepoint_isnumeric(npy_ucs4 code)
 {
@@ -389,6 +467,84 @@ struct Buffer {
         return unary_loop<IMPLEMENTED_UNARY_FUNCTIONS::ISDIGIT>();
     }
 
+    inline bool
+    islower()
+    {
+        size_t len = num_codepoints();
+        if (len == 0) {
+            return false;
+        }
+
+        Buffer<enc> tmp = *this;
+        bool cased = 0;
+        for (size_t i = 0; i < len; i++) {
+            if (codepoint_isupper<enc>(*tmp) || codepoint_istitle<enc>(*tmp)) {
+                return false;
+            }
+            else if (!cased && codepoint_islower<enc>(*tmp)) {
+                cased = true;
+            }
+            tmp++;
+        }
+        return cased;
+    }
+
+    inline bool
+    isupper()
+    {
+        size_t len = num_codepoints();
+        if (len == 0) {
+            return false;
+        }
+
+        Buffer<enc> tmp = *this;
+        bool cased = 0;
+        for (size_t i = 0; i < len; i++) {
+            if (codepoint_islower<enc>(*tmp) || codepoint_istitle<enc>(*tmp)) {
+                return false;
+            }
+            else if (!cased && codepoint_isupper<enc>(*tmp)) {
+                cased = true;
+            }
+            tmp++;
+        }
+        return cased;
+    }
+
+    inline bool
+    istitle()
+    {
+        size_t len = num_codepoints();
+        if (len == 0) {
+            return false;
+        }
+
+        Buffer<enc> tmp = *this;
+        bool cased = false;
+        bool previous_is_cased = false;
+        for (size_t i = 0; i < len; i++) {
+            if (codepoint_isupper<enc>(*tmp) || codepoint_istitle<enc>(*tmp)) {
+                if (previous_is_cased) {
+                    return false;
+                }
+                previous_is_cased = true;
+                cased = true;
+            }
+            else if (codepoint_islower<enc>(*tmp)) {
+                if (!previous_is_cased) {
+                    return false;
+                }
+                previous_is_cased = true;
+                cased = true;
+            }
+            else {
+                previous_is_cased = false;
+            }
+            tmp++;
+        }
+        return cased;
+    }
+
     inline bool
     isnumeric()
     {
@@ -466,6 +622,12 @@ struct call_buffer_member_function {
                 return codepoint_isspace<enc>(*buf);
             case IMPLEMENTED_UNARY_FUNCTIONS::STR_LEN:
                 return (T)buf.num_codepoints();
+            case IMPLEMENTED_UNARY_FUNCTIONS::ISLOWER:
+                return (T)buf.islower();
+            case IMPLEMENTED_UNARY_FUNCTIONS::ISUPPER:
+                return (T)buf.isupper();
+            case IMPLEMENTED_UNARY_FUNCTIONS::ISTITLE:
+                return (T)buf.istitle();
             case IMPLEMENTED_UNARY_FUNCTIONS::ISNUMERIC:
                 return codepoint_isnumeric(*buf);
             case IMPLEMENTED_UNARY_FUNCTIONS::ISDECIMAL: