Objects/unicodeobject.c (part 6)

Source:

cpython 3.14 @ ab2d84fe1023/Objects/unicodeobject.c ↗

This annotation covers the string splitting and stripping methods. See previous objects_unicodeobject* annotations for construction, encoding, and str.format.

Map

Lines	Symbol	Role
1-100	unicode_split	str.split(sep, maxsplit)
101-200	unicode_rsplit	str.rsplit(sep, maxsplit) — split from right
201-350	unicode_splitlines	Split on line endings (\n, \r, \r\n, \v, \f, etc.)
351-500	unicode_partition	str.partition(sep) — (before, sep, after)
501-650	unicode_rpartition	str.rpartition(sep) — last occurrence
651-900	unicode_strip	str.strip(chars) — strip leading and trailing
901-1200	_Py_unicode_Whitespace	Unicode whitespace character set for whitespace-split

Reading

str.split

// CPython: Objects/unicodeobject.c:10280 unicode_split
static PyObject *
unicode_split(PyObject *self, PyObject *args, PyObject *kwds)
{
    PyObject *substring = Py_None;
    Py_ssize_t maxcount = -1;
    if (substring == Py_None)
        /* Whitespace split */
        return split_whitespace(self, maxcount);
    return split(self, substring, maxcount);
}

'a b c'.split() collapses multiple whitespace characters. 'a::b::c'.split('::', 1) uses the explicit separator with a max count.

str.splitlines

// CPython: Objects/unicodeobject.c:10380 unicode_splitlines
static PyObject *
unicode_splitlines(PyObject *self, int keepends)
{
    /* Line endings: \n, \r, \r\n, \v, \f, \x1c, \x1d, \x1e, \x85, 
, 
 */
    return PyUnicode_Splitlines(self, keepends);
}

keepends=True includes the line ending in each element. str.splitlines() splits on all Unicode line boundaries (not just \n).

str.partition

// CPython: Objects/unicodeobject.c:10480 unicode_partition
static PyObject *
unicode_partition(PyObject *self, PyObject *sep_obj)
{
    Py_ssize_t pos = PyUnicode_Find(self, sep_obj, 0, PY_SSIZE_T_MAX, 1);
    if (pos < 0) {
        /* sep not found: return (self, '', '') */
        return PyTuple_Pack(3, self, unicode_empty, unicode_empty);
    }
    Py_ssize_t sep_len = PyUnicode_GET_LENGTH(sep_obj);
    return PyTuple_Pack(3,
        PyUnicode_Substring(self, 0, pos),
        sep_obj,
        PyUnicode_Substring(self, pos + sep_len, PY_SSIZE_T_MAX));
}

str.strip

// CPython: Objects/unicodeobject.c:10680 unicode_strip
static PyObject *
unicode_strip(PyObject *self, PyObject *args)
{
    PyObject *chars = Py_None;
    if (chars == Py_None)
        return _PyUnicode_XStrip(self, 3, Py_None);  /* strip whitespace */
    return _PyUnicode_XStrip(self, 3, chars);  /* strip given chars */
}

_PyUnicode_XStrip with mode 1 = lstrip, 2 = rstrip, 3 = both. For character sets, it builds a fast lookup set from the chars argument.

gopy notes

str.split is objects.UnicodeSplit in objects/unicode_split.go. str.splitlines uses the same Unicode line boundary table as CPython. str.partition is objects.UnicodePartition. str.strip uses objects.UnicodeStrip with a UnicodeSet for fast character membership.