2017-09-10 21:30:46 -03:00
|
|
|
"""Format a paragraph, comment block, or selection to a max width.
|
2013-08-10 17:56:28 -03:00
|
|
|
|
|
|
|
Does basic, standard text formatting, and also understands Python
|
|
|
|
comment blocks. Thus, for editing Python source code, this
|
|
|
|
extension is really only suitable for reformatting these comment
|
|
|
|
blocks or triple-quoted strings.
|
|
|
|
|
|
|
|
Known problems with comment reformatting:
|
|
|
|
* If there is a selection marked, and the first line of the
|
|
|
|
selection is not complete, the block will probably not be detected
|
|
|
|
as comments, and will have the normal "text formatting" rules
|
|
|
|
applied.
|
|
|
|
* If a comment block has leading whitespace that mixes tabs and
|
|
|
|
spaces, they will not be considered part of the same block.
|
|
|
|
* Fancy comments, like this bulleted list, aren't handled :-)
|
|
|
|
"""
|
2000-08-14 22:13:23 -03:00
|
|
|
import re
|
2016-08-31 01:50:55 -03:00
|
|
|
|
2016-05-28 14:22:31 -03:00
|
|
|
from idlelib.config import idleConf
|
2000-08-14 22:13:23 -03:00
|
|
|
|
2016-08-31 01:50:55 -03:00
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
class FormatParagraph:
|
|
|
|
|
|
|
|
def __init__(self, editwin):
|
|
|
|
self.editwin = editwin
|
|
|
|
|
2017-09-10 21:30:46 -03:00
|
|
|
@classmethod
|
|
|
|
def reload(cls):
|
|
|
|
cls.max_width = idleConf.GetOption('extensions', 'FormatParagraph',
|
|
|
|
'max-width', type='int', default=72)
|
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
def close(self):
|
|
|
|
self.editwin = None
|
|
|
|
|
2014-04-22 02:11:03 -03:00
|
|
|
def format_paragraph_event(self, event, limit=None):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Formats paragraph to a max width specified in idleConf.
|
|
|
|
|
|
|
|
If text is selected, format_paragraph_event will start breaking lines
|
|
|
|
at the max width, starting from the beginning selection.
|
|
|
|
|
|
|
|
If no text is selected, format_paragraph_event uses the current
|
|
|
|
cursor location to determine the paragraph (lines of text surrounded
|
|
|
|
by blank lines) and formats it.
|
2014-04-22 02:11:03 -03:00
|
|
|
|
|
|
|
The length limit parameter is for testing with a known value.
|
2013-08-10 17:56:28 -03:00
|
|
|
"""
|
2017-09-10 21:30:46 -03:00
|
|
|
limit = self.max_width if limit is None else limit
|
2000-08-14 22:13:23 -03:00
|
|
|
text = self.editwin.text
|
|
|
|
first, last = self.editwin.get_selection_indices()
|
|
|
|
if first and last:
|
|
|
|
data = text.get(first, last)
|
2013-08-10 17:56:28 -03:00
|
|
|
comment_header = get_comment_header(data)
|
2000-08-14 22:13:23 -03:00
|
|
|
else:
|
|
|
|
first, last, comment_header, data = \
|
|
|
|
find_paragraph(text, text.index("insert"))
|
|
|
|
if comment_header:
|
2014-04-22 02:11:03 -03:00
|
|
|
newdata = reformat_comment(data, limit, comment_header)
|
2000-08-14 22:13:23 -03:00
|
|
|
else:
|
2014-04-22 02:11:03 -03:00
|
|
|
newdata = reformat_paragraph(data, limit)
|
2000-08-14 22:13:23 -03:00
|
|
|
text.tag_remove("sel", "1.0", "end")
|
2013-08-10 17:56:28 -03:00
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
if newdata != data:
|
|
|
|
text.mark_set("insert", first)
|
|
|
|
text.undo_block_start()
|
|
|
|
text.delete(first, last)
|
|
|
|
text.insert(first, newdata)
|
|
|
|
text.undo_block_stop()
|
|
|
|
else:
|
|
|
|
text.mark_set("insert", last)
|
|
|
|
text.see("insert")
|
2007-12-10 21:06:40 -04:00
|
|
|
return "break"
|
2000-08-14 22:13:23 -03:00
|
|
|
|
2017-09-10 21:30:46 -03:00
|
|
|
|
|
|
|
FormatParagraph.reload()
|
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
def find_paragraph(text, mark):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Returns the start/stop indices enclosing the paragraph that mark is in.
|
|
|
|
|
|
|
|
Also returns the comment format string, if any, and paragraph of text
|
|
|
|
between the start/stop indices.
|
|
|
|
"""
|
2002-09-15 23:22:19 -03:00
|
|
|
lineno, col = map(int, mark.split("."))
|
2013-08-10 17:56:28 -03:00
|
|
|
line = text.get("%d.0" % lineno, "%d.end" % lineno)
|
|
|
|
|
|
|
|
# Look for start of next paragraph if the index passed in is a blank line
|
2000-08-14 22:13:23 -03:00
|
|
|
while text.compare("%d.0" % lineno, "<", "end") and is_all_white(line):
|
|
|
|
lineno = lineno + 1
|
2013-08-10 17:56:28 -03:00
|
|
|
line = text.get("%d.0" % lineno, "%d.end" % lineno)
|
2000-08-14 22:13:23 -03:00
|
|
|
first_lineno = lineno
|
|
|
|
comment_header = get_comment_header(line)
|
|
|
|
comment_header_len = len(comment_header)
|
2013-08-10 17:56:28 -03:00
|
|
|
|
|
|
|
# Once start line found, search for end of paragraph (a blank line)
|
2000-08-14 22:13:23 -03:00
|
|
|
while get_comment_header(line)==comment_header and \
|
|
|
|
not is_all_white(line[comment_header_len:]):
|
|
|
|
lineno = lineno + 1
|
2013-08-10 17:56:28 -03:00
|
|
|
line = text.get("%d.0" % lineno, "%d.end" % lineno)
|
2000-08-14 22:13:23 -03:00
|
|
|
last = "%d.0" % lineno
|
2013-08-10 17:56:28 -03:00
|
|
|
|
|
|
|
# Search back to beginning of paragraph (first blank line before)
|
2000-08-14 22:13:23 -03:00
|
|
|
lineno = first_lineno - 1
|
2013-08-10 17:56:28 -03:00
|
|
|
line = text.get("%d.0" % lineno, "%d.end" % lineno)
|
2000-08-14 22:13:23 -03:00
|
|
|
while lineno > 0 and \
|
|
|
|
get_comment_header(line)==comment_header and \
|
|
|
|
not is_all_white(line[comment_header_len:]):
|
|
|
|
lineno = lineno - 1
|
2013-08-10 17:56:28 -03:00
|
|
|
line = text.get("%d.0" % lineno, "%d.end" % lineno)
|
2000-08-14 22:13:23 -03:00
|
|
|
first = "%d.0" % (lineno+1)
|
2013-08-10 17:56:28 -03:00
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
return first, last, comment_header, text.get(first, last)
|
|
|
|
|
2013-08-10 17:56:28 -03:00
|
|
|
# This should perhaps be replaced with textwrap.wrap
|
2004-06-04 03:31:08 -03:00
|
|
|
def reformat_paragraph(data, limit):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Return data reformatted to specified width (limit)."""
|
2002-09-15 23:22:19 -03:00
|
|
|
lines = data.split("\n")
|
2000-08-14 22:13:23 -03:00
|
|
|
i = 0
|
|
|
|
n = len(lines)
|
|
|
|
while i < n and is_all_white(lines[i]):
|
|
|
|
i = i+1
|
|
|
|
if i >= n:
|
|
|
|
return data
|
|
|
|
indent1 = get_indent(lines[i])
|
|
|
|
if i+1 < n and not is_all_white(lines[i+1]):
|
|
|
|
indent2 = get_indent(lines[i+1])
|
|
|
|
else:
|
|
|
|
indent2 = indent1
|
|
|
|
new = lines[:i]
|
|
|
|
partial = indent1
|
|
|
|
while i < n and not is_all_white(lines[i]):
|
|
|
|
# XXX Should take double space after period (etc.) into account
|
2016-09-08 14:59:53 -03:00
|
|
|
words = re.split(r"(\s+)", lines[i])
|
2000-08-14 22:13:23 -03:00
|
|
|
for j in range(0, len(words), 2):
|
|
|
|
word = words[j]
|
|
|
|
if not word:
|
|
|
|
continue # Can happen when line ends in whitespace
|
2002-09-15 23:22:19 -03:00
|
|
|
if len((partial + word).expandtabs()) > limit and \
|
2013-08-10 17:56:28 -03:00
|
|
|
partial != indent1:
|
2002-09-15 23:22:19 -03:00
|
|
|
new.append(partial.rstrip())
|
2000-08-14 22:13:23 -03:00
|
|
|
partial = indent2
|
|
|
|
partial = partial + word + " "
|
|
|
|
if j+1 < len(words) and words[j+1] != " ":
|
|
|
|
partial = partial + " "
|
|
|
|
i = i+1
|
2002-09-15 23:22:19 -03:00
|
|
|
new.append(partial.rstrip())
|
2000-08-14 22:13:23 -03:00
|
|
|
# XXX Should reformat remaining paragraphs as well
|
|
|
|
new.extend(lines[i:])
|
2002-09-15 23:22:19 -03:00
|
|
|
return "\n".join(new)
|
2000-08-14 22:13:23 -03:00
|
|
|
|
2013-08-10 17:56:28 -03:00
|
|
|
def reformat_comment(data, limit, comment_header):
|
|
|
|
"""Return data reformatted to specified width with comment header."""
|
|
|
|
|
|
|
|
# Remove header from the comment lines
|
|
|
|
lc = len(comment_header)
|
|
|
|
data = "\n".join(line[lc:] for line in data.split("\n"))
|
|
|
|
# Reformat to maxformatwidth chars or a 20 char width,
|
|
|
|
# whichever is greater.
|
|
|
|
format_width = max(limit - len(comment_header), 20)
|
|
|
|
newdata = reformat_paragraph(data, format_width)
|
|
|
|
# re-split and re-insert the comment header.
|
|
|
|
newdata = newdata.split("\n")
|
2017-11-05 21:30:06 -04:00
|
|
|
# If the block ends in a \n, we don't want the comment prefix
|
2013-08-10 17:56:28 -03:00
|
|
|
# inserted after it. (Im not sure it makes sense to reformat a
|
|
|
|
# comment block that is not made of complete lines, but whatever!)
|
|
|
|
# Can't think of a clean solution, so we hack away
|
|
|
|
block_suffix = ""
|
|
|
|
if not newdata[-1]:
|
|
|
|
block_suffix = "\n"
|
|
|
|
newdata = newdata[:-1]
|
|
|
|
return '\n'.join(comment_header+line for line in newdata) + block_suffix
|
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
def is_all_white(line):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Return True if line is empty or all whitespace."""
|
|
|
|
|
2000-08-14 22:13:23 -03:00
|
|
|
return re.match(r"^\s*$", line) is not None
|
|
|
|
|
|
|
|
def get_indent(line):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Return the initial space or tab indent of line."""
|
|
|
|
return re.match(r"^([ \t]*)", line).group()
|
2000-08-14 22:13:23 -03:00
|
|
|
|
|
|
|
def get_comment_header(line):
|
2013-08-10 17:56:28 -03:00
|
|
|
"""Return string with leading whitespace and '#' from line or ''.
|
|
|
|
|
|
|
|
A null return indicates that the line is not a comment line. A non-
|
|
|
|
null return, such as ' #', will be used to find the other lines of
|
|
|
|
a comment block with the same indent.
|
|
|
|
"""
|
|
|
|
m = re.match(r"^([ \t]*#*)", line)
|
2000-08-14 22:13:23 -03:00
|
|
|
if m is None: return ""
|
|
|
|
return m.group(1)
|
2013-08-10 17:56:28 -03:00
|
|
|
|
2016-08-31 01:50:55 -03:00
|
|
|
|
2013-08-10 17:56:28 -03:00
|
|
|
if __name__ == "__main__":
|
|
|
|
import unittest
|
2016-05-28 14:22:31 -03:00
|
|
|
unittest.main('idlelib.idle_test.test_paragraph',
|
2013-08-10 17:56:28 -03:00
|
|
|
verbosity=2, exit=False)
|