Lib/idlelib/FormatParagraph.py

   1 # Extension to format a paragraph
   2
   3 # Does basic, standard text formatting, and also understands Python
   4 # comment blocks.  Thus, for editing Python source code, this
   5 # extension is really only suitable for reformatting these comment
   6 # blocks or triple-quoted strings.
   7
   8 # Known problems with comment reformatting:
   9 # * If there is a selection marked, and the first line of the
  10 #   selection is not complete, the block will probably not be detected
  11 #   as comments, and will have the normal "text formatting" rules
  12 #   applied.
  13 # * If a comment block has leading whitespace that mixes tabs and
  14 #   spaces, they will not be considered part of the same block.
  15 # * Fancy comments, like this bulleted list, arent handled :-)
  16
  17 import string
  18 import re
  19
  20 class FormatParagraph:
  21
  22     menudefs = [
  23         ('format', [   # /s/edit/format   dscherer@cmu.edu
  24             ('Format Paragraph', '<<format-paragraph>>'),
  25          ])
  26     ]
  27
  28     def __init__(self, editwin):
  29         self.editwin = editwin
  30
  31     def close(self):
  32         self.editwin = None
  33
  34     def format_paragraph_event(self, event):
  35         text = self.editwin.text
  36         first, last = self.editwin.get_selection_indices()
  37         if first and last:
  38             data = text.get(first, last)
  39             comment_header = ''
  40         else:
  41             first, last, comment_header, data = \
  42                     find_paragraph(text, text.index("insert"))
  43         if comment_header:
  44             # Reformat the comment lines - convert to text sans header.
  45             lines = string.split(data, "\n")
  46             lines = map(lambda st, l=len(comment_header): st[l:], lines)
  47             data = string.join(lines, "\n")
  48             # Reformat to 70 chars or a 20 char width, whichever is greater.
  49             format_width = max(70-len(comment_header), 20)
  50             newdata = reformat_paragraph(data, format_width)
  51             # re-split and re-insert the comment header.
  52             newdata = string.split(newdata, "\n")
  53             # If the block ends in a \n, we dont want the comment
  54             # prefix inserted after it. (Im not sure it makes sense to
  55             # reformat a comment block that isnt made of complete
  56             # lines, but whatever!)  Can't think of a clean soltution,
  57             # so we hack away
  58             block_suffix = ""
  59             if not newdata[-1]:
  60                 block_suffix = "\n"
  61                 newdata = newdata[:-1]
  62             builder = lambda item, prefix=comment_header: prefix+item
  63             newdata = string.join(map(builder, newdata), '\n') + block_suffix
  64         else:
  65             # Just a normal text format
  66             newdata = reformat_paragraph(data)
  67         text.tag_remove("sel", "1.0", "end")
  68         if newdata != data:
  69             text.mark_set("insert", first)
  70             text.undo_block_start()
  71             text.delete(first, last)
  72             text.insert(first, newdata)
  73             text.undo_block_stop()
  74         else:
  75             text.mark_set("insert", last)
  76         text.see("insert")
  77
  78 def find_paragraph(text, mark):
  79     lineno, col = map(int, string.split(mark, "."))
  80     line = text.get("%d.0" % lineno, "%d.0 lineend" % lineno)
  81     while text.compare("%d.0" % lineno, "<", "end") and is_all_white(line):
  82         lineno = lineno + 1
  83         line = text.get("%d.0" % lineno, "%d.0 lineend" % lineno)
  84     first_lineno = lineno
  85     comment_header = get_comment_header(line)
  86     comment_header_len = len(comment_header)
  87     while get_comment_header(line)==comment_header and \
  88               not is_all_white(line[comment_header_len:]):
  89         lineno = lineno + 1
  90         line = text.get("%d.0" % lineno, "%d.0 lineend" % lineno)
  91     last = "%d.0" % lineno
  92     # Search back to beginning of paragraph
  93     lineno = first_lineno - 1
  94     line = text.get("%d.0" % lineno, "%d.0 lineend" % lineno)
  95     while lineno > 0 and \
  96               get_comment_header(line)==comment_header and \
  97               not is_all_white(line[comment_header_len:]):
  98         lineno = lineno - 1
  99         line = text.get("%d.0" % lineno, "%d.0 lineend" % lineno)
 100     first = "%d.0" % (lineno+1)
 101     return first, last, comment_header, text.get(first, last)
 102
 103 def reformat_paragraph(data, limit=70):
 104     lines = string.split(data, "\n")
 105     i = 0
 106     n = len(lines)
 107     while i < n and is_all_white(lines[i]):
 108         i = i+1
 109     if i >= n:
 110         return data
 111     indent1 = get_indent(lines[i])
 112     if i+1 < n and not is_all_white(lines[i+1]):
 113         indent2 = get_indent(lines[i+1])
 114     else:
 115         indent2 = indent1
 116     new = lines[:i]
 117     partial = indent1
 118     while i < n and not is_all_white(lines[i]):
 119         # XXX Should take double space after period (etc.) into account
 120         words = re.split("(\s+)", lines[i])
 121         for j in range(0, len(words), 2):
 122             word = words[j]
 123             if not word:
 124                 continue # Can happen when line ends in whitespace
 125             if len(string.expandtabs(partial + word)) > limit and \
 126                partial != indent1:
 127                 new.append(string.rstrip(partial))
 128                 partial = indent2
 129             partial = partial + word + " "
 130             if j+1 < len(words) and words[j+1] != " ":
 131                 partial = partial + " "
 132         i = i+1
 133     new.append(string.rstrip(partial))
 134     # XXX Should reformat remaining paragraphs as well
 135     new.extend(lines[i:])
 136     return string.join(new, "\n")
 137
 138 def is_all_white(line):
 139     return re.match(r"^\s*$", line) is not None
 140
 141 def get_indent(line):
 142     return re.match(r"^(\s*)", line).group()
 143
 144 def get_comment_header(line):
 145     m = re.match(r"^(\s*#*)", line)
 146     if m is None: return ""
 147     return m.group(1)