Lib/textwrap.py

Source:

cpython 3.14 @ ab2d84fe1023/Lib/textwrap.py ↗

Lib/textwrap.py provides TextWrapper, wrap, fill, dedent, indent, and shorten. TextWrapper handles word-wrapping at a maximum line width while respecting break_on_hyphens, break_long_words, initial_indent, subsequent_indent, and expand_tabs.

Map

Lines	Symbol	Role
1-50	imports, _whitespace_only_re	Regex constants
51-300	TextWrapper	Main class; wrap, fill, _split, _wrap_chunks
301-380	wrap, fill, shorten	Module-level convenience wrappers
381-450	dedent, indent	Whitespace normalization functions

Reading

TextWrapper._split: tokenizing text

_split breaks the input text into a list of chunks using a regex that splits on whitespace boundaries and optional hyphenation points. The resulting chunks alternate between word tokens and whitespace tokens.

# CPython: Lib/textwrap.py:51 TextWrapper._split
def _split(self, text):
    chunks = self.wordsep_re.split(text)
    chunks = [c for c in chunks if c]
    return chunks

_wrap_chunks: line assembly

_wrap_chunks greedily accumulates chunks onto the current line until adding the next chunk would exceed width. When the line is full it is flushed to the result list and a new line starts with subsequent_indent.

# CPython: Lib/textwrap.py:200 TextWrapper._wrap_chunks
def _wrap_chunks(self, chunks):
    lines = []
    cur_line = []
    cur_len = 0
    width = self.width - len(self.subsequent_indent)
    while chunks:
        l = self.colwidth(chunks[-1])
        if cur_len + l <= width:
            cur_line.append(chunks.pop())
            cur_len += l
        else:
            if cur_line:
                lines.append(self.subsequent_indent + ''.join(cur_line[::-1]))
            cur_line = []
            cur_len = 0
    if cur_line:
        lines.append(self.subsequent_indent + ''.join(cur_line[::-1]))
    return lines

dedent: stripping common indentation

dedent finds the longest common leading whitespace prefix across all non-empty lines and strips it. This is useful for unindenting docstrings or heredoc-style multi-line strings in source code.

# CPython: Lib/textwrap.py:381 dedent
def dedent(text):
    lines = text.splitlines(keepends=True)
    margin = None
    for line in lines:
        content = line.lstrip()
        if content:
            indent = line[:len(line) - len(content)]
            margin = indent if margin is None else os.path.commonprefix([margin, indent])
    return re.sub(r'(?m)^' + margin, '', text) if margin else text

gopy notes

Not yet ported. The planned package path is module/textwrap/. Go's strings package does not include a word-wrap equivalent; a faithful port would implement TextWrapper as a Go struct with exported fields matching the Python attributes.