##// END OF EJS Templates
improve comment
Srinivas Reddy Thatiparthy -
Show More
@@ -1,221 +1,221 b''
1 1 # encoding: utf-8
2 2 """Tests for IPython.utils.text"""
3 3
4 4 #-----------------------------------------------------------------------------
5 5 # Copyright (C) 2011 The IPython Development Team
6 6 #
7 7 # Distributed under the terms of the BSD License. The full license is in
8 8 # the file COPYING, distributed as part of this software.
9 9 #-----------------------------------------------------------------------------
10 10
11 11 #-----------------------------------------------------------------------------
12 12 # Imports
13 13 #-----------------------------------------------------------------------------
14 14
15 15 import os
16 16 import math
17 17 import random
18 18 import sys
19 19
20 20 import nose.tools as nt
21 21 try:
22 22 from pathlib import Path
23 23 except ImportError:
24 # Python 2 backport
24 # for Python 3.3
25 25 from pathlib2 import Path
26 26
27 27 from IPython.utils import text
28 28
29 29 #-----------------------------------------------------------------------------
30 30 # Globals
31 31 #-----------------------------------------------------------------------------
32 32
33 33 def test_columnize():
34 34 """Basic columnize tests."""
35 35 size = 5
36 36 items = [l*size for l in 'abcd']
37 37
38 38 out = text.columnize(items, displaywidth=80)
39 39 nt.assert_equal(out, 'aaaaa bbbbb ccccc ddddd\n')
40 40 out = text.columnize(items, displaywidth=25)
41 41 nt.assert_equal(out, 'aaaaa ccccc\nbbbbb ddddd\n')
42 42 out = text.columnize(items, displaywidth=12)
43 43 nt.assert_equal(out, 'aaaaa ccccc\nbbbbb ddddd\n')
44 44 out = text.columnize(items, displaywidth=10)
45 45 nt.assert_equal(out, 'aaaaa\nbbbbb\nccccc\nddddd\n')
46 46
47 47 out = text.columnize(items, row_first=True, displaywidth=80)
48 48 nt.assert_equal(out, 'aaaaa bbbbb ccccc ddddd\n')
49 49 out = text.columnize(items, row_first=True, displaywidth=25)
50 50 nt.assert_equal(out, 'aaaaa bbbbb\nccccc ddddd\n')
51 51 out = text.columnize(items, row_first=True, displaywidth=12)
52 52 nt.assert_equal(out, 'aaaaa bbbbb\nccccc ddddd\n')
53 53 out = text.columnize(items, row_first=True, displaywidth=10)
54 54 nt.assert_equal(out, 'aaaaa\nbbbbb\nccccc\nddddd\n')
55 55
56 56 out = text.columnize(items, displaywidth=40, spread=True)
57 57 nt.assert_equal(out, 'aaaaa bbbbb ccccc ddddd\n')
58 58 out = text.columnize(items, displaywidth=20, spread=True)
59 59 nt.assert_equal(out, 'aaaaa ccccc\nbbbbb ddddd\n')
60 60 out = text.columnize(items, displaywidth=12, spread=True)
61 61 nt.assert_equal(out, 'aaaaa ccccc\nbbbbb ddddd\n')
62 62 out = text.columnize(items, displaywidth=10, spread=True)
63 63 nt.assert_equal(out, 'aaaaa\nbbbbb\nccccc\nddddd\n')
64 64
65 65
66 66 def test_columnize_random():
67 67 """Test with random input to hopfully catch edge case """
68 68 for row_first in [True, False]:
69 69 for nitems in [random.randint(2,70) for i in range(2,20)]:
70 70 displaywidth = random.randint(20,200)
71 71 rand_len = [random.randint(2,displaywidth) for i in range(nitems)]
72 72 items = ['x'*l for l in rand_len]
73 73 out = text.columnize(items, row_first=row_first, displaywidth=displaywidth)
74 74 longer_line = max([len(x) for x in out.split('\n')])
75 75 longer_element = max(rand_len)
76 76 if longer_line > displaywidth:
77 77 print("Columnize displayed something lager than displaywidth : %s " % longer_line)
78 78 print("longer element : %s " % longer_element)
79 79 print("displaywidth : %s " % displaywidth)
80 80 print("number of element : %s " % nitems)
81 81 print("size of each element :\n %s" % rand_len)
82 82 assert False, "row_first={0}".format(row_first)
83 83
84 84 def test_columnize_medium():
85 85 """Test with inputs than shouldn't be wider than 80"""
86 86 size = 40
87 87 items = [l*size for l in 'abc']
88 88 for row_first in [True, False]:
89 89 out = text.columnize(items, row_first=row_first, displaywidth=80)
90 90 nt.assert_equal(out, '\n'.join(items+['']), "row_first={0}".format(row_first))
91 91
92 92 def test_columnize_long():
93 93 """Test columnize with inputs longer than the display window"""
94 94 size = 11
95 95 items = [l*size for l in 'abc']
96 96 for row_first in [True, False]:
97 97 out = text.columnize(items, row_first=row_first, displaywidth=size-1)
98 98 nt.assert_equal(out, '\n'.join(items+['']), "row_first={0}".format(row_first))
99 99
100 100 def eval_formatter_check(f):
101 101 ns = dict(n=12, pi=math.pi, stuff='hello there', os=os, u=u"cafΓ©", b="cafΓ©")
102 102 s = f.format("{n} {n//4} {stuff.split()[0]}", **ns)
103 103 nt.assert_equal(s, "12 3 hello")
104 104 s = f.format(' '.join(['{n//%i}'%i for i in range(1,8)]), **ns)
105 105 nt.assert_equal(s, "12 6 4 3 2 2 1")
106 106 s = f.format('{[n//i for i in range(1,8)]}', **ns)
107 107 nt.assert_equal(s, "[12, 6, 4, 3, 2, 2, 1]")
108 108 s = f.format("{stuff!s}", **ns)
109 109 nt.assert_equal(s, ns['stuff'])
110 110 s = f.format("{stuff!r}", **ns)
111 111 nt.assert_equal(s, repr(ns['stuff']))
112 112
113 113 # Check with unicode:
114 114 s = f.format("{u}", **ns)
115 115 nt.assert_equal(s, ns['u'])
116 116 # This decodes in a platform dependent manner, but it shouldn't error out
117 117 s = f.format("{b}", **ns)
118 118
119 119 nt.assert_raises(NameError, f.format, '{dne}', **ns)
120 120
121 121 def eval_formatter_slicing_check(f):
122 122 ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
123 123 s = f.format(" {stuff.split()[:]} ", **ns)
124 124 nt.assert_equal(s, " ['hello', 'there'] ")
125 125 s = f.format(" {stuff.split()[::-1]} ", **ns)
126 126 nt.assert_equal(s, " ['there', 'hello'] ")
127 127 s = f.format("{stuff[::2]}", **ns)
128 128 nt.assert_equal(s, ns['stuff'][::2])
129 129
130 130 nt.assert_raises(SyntaxError, f.format, "{n:x}", **ns)
131 131
132 132 def eval_formatter_no_slicing_check(f):
133 133 ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
134 134
135 135 s = f.format('{n:x} {pi**2:+f}', **ns)
136 136 nt.assert_equal(s, "c +9.869604")
137 137
138 138 s = f.format('{stuff[slice(1,4)]}', **ns)
139 139 nt.assert_equal(s, 'ell')
140 140
141 141 if sys.version_info >= (3, 4):
142 142 # String formatting has changed in Python 3.4, so this now works.
143 143 s = f.format("{a[:]}", a=[1, 2])
144 144 nt.assert_equal(s, "[1, 2]")
145 145 else:
146 146 nt.assert_raises(SyntaxError, f.format, "{a[:]}")
147 147
148 148 def test_eval_formatter():
149 149 f = text.EvalFormatter()
150 150 eval_formatter_check(f)
151 151 eval_formatter_no_slicing_check(f)
152 152
153 153 def test_full_eval_formatter():
154 154 f = text.FullEvalFormatter()
155 155 eval_formatter_check(f)
156 156 eval_formatter_slicing_check(f)
157 157
158 158 def test_dollar_formatter():
159 159 f = text.DollarFormatter()
160 160 eval_formatter_check(f)
161 161 eval_formatter_slicing_check(f)
162 162
163 163 ns = dict(n=12, pi=math.pi, stuff='hello there', os=os)
164 164 s = f.format("$n", **ns)
165 165 nt.assert_equal(s, "12")
166 166 s = f.format("$n.real", **ns)
167 167 nt.assert_equal(s, "12")
168 168 s = f.format("$n/{stuff[:5]}", **ns)
169 169 nt.assert_equal(s, "12/hello")
170 170 s = f.format("$n $$HOME", **ns)
171 171 nt.assert_equal(s, "12 $HOME")
172 172 s = f.format("${foo}", foo="HOME")
173 173 nt.assert_equal(s, "$HOME")
174 174
175 175
176 176 def test_long_substr():
177 177 data = ['hi']
178 178 nt.assert_equal(text.long_substr(data), 'hi')
179 179
180 180
181 181 def test_long_substr2():
182 182 data = ['abc', 'abd', 'abf', 'ab']
183 183 nt.assert_equal(text.long_substr(data), 'ab')
184 184
185 185 def test_long_substr_empty():
186 186 data = []
187 187 nt.assert_equal(text.long_substr(data), '')
188 188
189 189 def test_strip_email():
190 190 src = """\
191 191 >> >>> def f(x):
192 192 >> ... return x+1
193 193 >> ...
194 194 >> >>> zz = f(2.5)"""
195 195 cln = """\
196 196 >>> def f(x):
197 197 ... return x+1
198 198 ...
199 199 >>> zz = f(2.5)"""
200 200 nt.assert_equal(text.strip_email_quotes(src), cln)
201 201
202 202
203 203 def test_strip_email2():
204 204 src = '> > > list()'
205 205 cln = 'list()'
206 206 nt.assert_equal(text.strip_email_quotes(src), cln)
207 207
208 208 def test_LSString():
209 209 lss = text.LSString("abc\ndef")
210 210 nt.assert_equal(lss.l, ['abc', 'def'])
211 211 nt.assert_equal(lss.s, 'abc def')
212 212 lss = text.LSString(os.getcwd())
213 213 nt.assert_is_instance(lss.p[0], Path)
214 214
215 215 def test_SList():
216 216 sl = text.SList(['a 11', 'b 1', 'a 2'])
217 217 nt.assert_equal(sl.n, 'a 11\nb 1\na 2')
218 218 nt.assert_equal(sl.s, 'a 11 b 1 a 2')
219 219 nt.assert_equal(sl.grep(lambda x: x.startswith('a')), text.SList(['a 11', 'a 2']))
220 220 nt.assert_equal(sl.fields(0), text.SList(['a', 'b', 'a']))
221 221 nt.assert_equal(sl.sort(field=1, nums=True), text.SList(['b 1', 'a 2', 'a 11']))
@@ -1,776 +1,776 b''
1 1 # encoding: utf-8
2 2 """
3 3 Utilities for working with strings and text.
4 4
5 5 Inheritance diagram:
6 6
7 7 .. inheritance-diagram:: IPython.utils.text
8 8 :parts: 3
9 9 """
10 10
11 11 import os
12 12 import re
13 13 import sys
14 14 import textwrap
15 15 from string import Formatter
16 16 try:
17 17 from pathlib import Path
18 18 except ImportError:
19 # Python 2 backport
19 # for Python 3.3
20 20 from pathlib2 import Path
21 21
22 22 from IPython.utils import py3compat
23 23
24 24 # datetime.strftime date format for ipython
25 25 if sys.platform == 'win32':
26 26 date_format = "%B %d, %Y"
27 27 else:
28 28 date_format = "%B %-d, %Y"
29 29
30 30 class LSString(str):
31 31 """String derivative with a special access attributes.
32 32
33 33 These are normal strings, but with the special attributes:
34 34
35 35 .l (or .list) : value as list (split on newlines).
36 36 .n (or .nlstr): original value (the string itself).
37 37 .s (or .spstr): value as whitespace-separated string.
38 38 .p (or .paths): list of path objects (requires path.py package)
39 39
40 40 Any values which require transformations are computed only once and
41 41 cached.
42 42
43 43 Such strings are very useful to efficiently interact with the shell, which
44 44 typically only understands whitespace-separated options for commands."""
45 45
46 46 def get_list(self):
47 47 try:
48 48 return self.__list
49 49 except AttributeError:
50 50 self.__list = self.split('\n')
51 51 return self.__list
52 52
53 53 l = list = property(get_list)
54 54
55 55 def get_spstr(self):
56 56 try:
57 57 return self.__spstr
58 58 except AttributeError:
59 59 self.__spstr = self.replace('\n',' ')
60 60 return self.__spstr
61 61
62 62 s = spstr = property(get_spstr)
63 63
64 64 def get_nlstr(self):
65 65 return self
66 66
67 67 n = nlstr = property(get_nlstr)
68 68
69 69 def get_paths(self):
70 70 try:
71 71 return self.__paths
72 72 except AttributeError:
73 73 self.__paths = [Path(p) for p in self.split('\n') if os.path.exists(p)]
74 74 return self.__paths
75 75
76 76 p = paths = property(get_paths)
77 77
78 78 # FIXME: We need to reimplement type specific displayhook and then add this
79 79 # back as a custom printer. This should also be moved outside utils into the
80 80 # core.
81 81
82 82 # def print_lsstring(arg):
83 83 # """ Prettier (non-repr-like) and more informative printer for LSString """
84 84 # print "LSString (.p, .n, .l, .s available). Value:"
85 85 # print arg
86 86 #
87 87 #
88 88 # print_lsstring = result_display.when_type(LSString)(print_lsstring)
89 89
90 90
91 91 class SList(list):
92 92 """List derivative with a special access attributes.
93 93
94 94 These are normal lists, but with the special attributes:
95 95
96 96 * .l (or .list) : value as list (the list itself).
97 97 * .n (or .nlstr): value as a string, joined on newlines.
98 98 * .s (or .spstr): value as a string, joined on spaces.
99 99 * .p (or .paths): list of path objects (requires path.py package)
100 100
101 101 Any values which require transformations are computed only once and
102 102 cached."""
103 103
104 104 def get_list(self):
105 105 return self
106 106
107 107 l = list = property(get_list)
108 108
109 109 def get_spstr(self):
110 110 try:
111 111 return self.__spstr
112 112 except AttributeError:
113 113 self.__spstr = ' '.join(self)
114 114 return self.__spstr
115 115
116 116 s = spstr = property(get_spstr)
117 117
118 118 def get_nlstr(self):
119 119 try:
120 120 return self.__nlstr
121 121 except AttributeError:
122 122 self.__nlstr = '\n'.join(self)
123 123 return self.__nlstr
124 124
125 125 n = nlstr = property(get_nlstr)
126 126
127 127 def get_paths(self):
128 128 try:
129 129 return self.__paths
130 130 except AttributeError:
131 131 self.__paths = [Path(p) for p in self if os.path.exists(p)]
132 132 return self.__paths
133 133
134 134 p = paths = property(get_paths)
135 135
136 136 def grep(self, pattern, prune = False, field = None):
137 137 """ Return all strings matching 'pattern' (a regex or callable)
138 138
139 139 This is case-insensitive. If prune is true, return all items
140 140 NOT matching the pattern.
141 141
142 142 If field is specified, the match must occur in the specified
143 143 whitespace-separated field.
144 144
145 145 Examples::
146 146
147 147 a.grep( lambda x: x.startswith('C') )
148 148 a.grep('Cha.*log', prune=1)
149 149 a.grep('chm', field=-1)
150 150 """
151 151
152 152 def match_target(s):
153 153 if field is None:
154 154 return s
155 155 parts = s.split()
156 156 try:
157 157 tgt = parts[field]
158 158 return tgt
159 159 except IndexError:
160 160 return ""
161 161
162 162 if isinstance(pattern, str):
163 163 pred = lambda x : re.search(pattern, x, re.IGNORECASE)
164 164 else:
165 165 pred = pattern
166 166 if not prune:
167 167 return SList([el for el in self if pred(match_target(el))])
168 168 else:
169 169 return SList([el for el in self if not pred(match_target(el))])
170 170
171 171 def fields(self, *fields):
172 172 """ Collect whitespace-separated fields from string list
173 173
174 174 Allows quick awk-like usage of string lists.
175 175
176 176 Example data (in var a, created by 'a = !ls -l')::
177 177
178 178 -rwxrwxrwx 1 ville None 18 Dec 14 2006 ChangeLog
179 179 drwxrwxrwx+ 6 ville None 0 Oct 24 18:05 IPython
180 180
181 181 * ``a.fields(0)`` is ``['-rwxrwxrwx', 'drwxrwxrwx+']``
182 182 * ``a.fields(1,0)`` is ``['1 -rwxrwxrwx', '6 drwxrwxrwx+']``
183 183 (note the joining by space).
184 184 * ``a.fields(-1)`` is ``['ChangeLog', 'IPython']``
185 185
186 186 IndexErrors are ignored.
187 187
188 188 Without args, fields() just split()'s the strings.
189 189 """
190 190 if len(fields) == 0:
191 191 return [el.split() for el in self]
192 192
193 193 res = SList()
194 194 for el in [f.split() for f in self]:
195 195 lineparts = []
196 196
197 197 for fd in fields:
198 198 try:
199 199 lineparts.append(el[fd])
200 200 except IndexError:
201 201 pass
202 202 if lineparts:
203 203 res.append(" ".join(lineparts))
204 204
205 205 return res
206 206
207 207 def sort(self,field= None, nums = False):
208 208 """ sort by specified fields (see fields())
209 209
210 210 Example::
211 211
212 212 a.sort(1, nums = True)
213 213
214 214 Sorts a by second field, in numerical order (so that 21 > 3)
215 215
216 216 """
217 217
218 218 #decorate, sort, undecorate
219 219 if field is not None:
220 220 dsu = [[SList([line]).fields(field), line] for line in self]
221 221 else:
222 222 dsu = [[line, line] for line in self]
223 223 if nums:
224 224 for i in range(len(dsu)):
225 225 numstr = "".join([ch for ch in dsu[i][0] if ch.isdigit()])
226 226 try:
227 227 n = int(numstr)
228 228 except ValueError:
229 229 n = 0
230 230 dsu[i][0] = n
231 231
232 232
233 233 dsu.sort()
234 234 return SList([t[1] for t in dsu])
235 235
236 236
237 237 # FIXME: We need to reimplement type specific displayhook and then add this
238 238 # back as a custom printer. This should also be moved outside utils into the
239 239 # core.
240 240
241 241 # def print_slist(arg):
242 242 # """ Prettier (non-repr-like) and more informative printer for SList """
243 243 # print "SList (.p, .n, .l, .s, .grep(), .fields(), sort() available):"
244 244 # if hasattr(arg, 'hideonce') and arg.hideonce:
245 245 # arg.hideonce = False
246 246 # return
247 247 #
248 248 # nlprint(arg) # This was a nested list printer, now removed.
249 249 #
250 250 # print_slist = result_display.when_type(SList)(print_slist)
251 251
252 252
253 253 def indent(instr,nspaces=4, ntabs=0, flatten=False):
254 254 """Indent a string a given number of spaces or tabstops.
255 255
256 256 indent(str,nspaces=4,ntabs=0) -> indent str by ntabs+nspaces.
257 257
258 258 Parameters
259 259 ----------
260 260
261 261 instr : basestring
262 262 The string to be indented.
263 263 nspaces : int (default: 4)
264 264 The number of spaces to be indented.
265 265 ntabs : int (default: 0)
266 266 The number of tabs to be indented.
267 267 flatten : bool (default: False)
268 268 Whether to scrub existing indentation. If True, all lines will be
269 269 aligned to the same indentation. If False, existing indentation will
270 270 be strictly increased.
271 271
272 272 Returns
273 273 -------
274 274
275 275 str|unicode : string indented by ntabs and nspaces.
276 276
277 277 """
278 278 if instr is None:
279 279 return
280 280 ind = '\t'*ntabs+' '*nspaces
281 281 if flatten:
282 282 pat = re.compile(r'^\s*', re.MULTILINE)
283 283 else:
284 284 pat = re.compile(r'^', re.MULTILINE)
285 285 outstr = re.sub(pat, ind, instr)
286 286 if outstr.endswith(os.linesep+ind):
287 287 return outstr[:-len(ind)]
288 288 else:
289 289 return outstr
290 290
291 291
292 292 def list_strings(arg):
293 293 """Always return a list of strings, given a string or list of strings
294 294 as input.
295 295
296 296 Examples
297 297 --------
298 298 ::
299 299
300 300 In [7]: list_strings('A single string')
301 301 Out[7]: ['A single string']
302 302
303 303 In [8]: list_strings(['A single string in a list'])
304 304 Out[8]: ['A single string in a list']
305 305
306 306 In [9]: list_strings(['A','list','of','strings'])
307 307 Out[9]: ['A', 'list', 'of', 'strings']
308 308 """
309 309
310 310 if isinstance(arg, str):
311 311 return [arg]
312 312 else:
313 313 return arg
314 314
315 315
316 316 def marquee(txt='',width=78,mark='*'):
317 317 """Return the input string centered in a 'marquee'.
318 318
319 319 Examples
320 320 --------
321 321 ::
322 322
323 323 In [16]: marquee('A test',40)
324 324 Out[16]: '**************** A test ****************'
325 325
326 326 In [17]: marquee('A test',40,'-')
327 327 Out[17]: '---------------- A test ----------------'
328 328
329 329 In [18]: marquee('A test',40,' ')
330 330 Out[18]: ' A test '
331 331
332 332 """
333 333 if not txt:
334 334 return (mark*width)[:width]
335 335 nmark = (width-len(txt)-2)//len(mark)//2
336 336 if nmark < 0: nmark =0
337 337 marks = mark*nmark
338 338 return '%s %s %s' % (marks,txt,marks)
339 339
340 340
341 341 ini_spaces_re = re.compile(r'^(\s+)')
342 342
343 343 def num_ini_spaces(strng):
344 344 """Return the number of initial spaces in a string"""
345 345
346 346 ini_spaces = ini_spaces_re.match(strng)
347 347 if ini_spaces:
348 348 return ini_spaces.end()
349 349 else:
350 350 return 0
351 351
352 352
353 353 def format_screen(strng):
354 354 """Format a string for screen printing.
355 355
356 356 This removes some latex-type format codes."""
357 357 # Paragraph continue
358 358 par_re = re.compile(r'\\$',re.MULTILINE)
359 359 strng = par_re.sub('',strng)
360 360 return strng
361 361
362 362
363 363 def dedent(text):
364 364 """Equivalent of textwrap.dedent that ignores unindented first line.
365 365
366 366 This means it will still dedent strings like:
367 367 '''foo
368 368 is a bar
369 369 '''
370 370
371 371 For use in wrap_paragraphs.
372 372 """
373 373
374 374 if text.startswith('\n'):
375 375 # text starts with blank line, don't ignore the first line
376 376 return textwrap.dedent(text)
377 377
378 378 # split first line
379 379 splits = text.split('\n',1)
380 380 if len(splits) == 1:
381 381 # only one line
382 382 return textwrap.dedent(text)
383 383
384 384 first, rest = splits
385 385 # dedent everything but the first line
386 386 rest = textwrap.dedent(rest)
387 387 return '\n'.join([first, rest])
388 388
389 389
390 390 def wrap_paragraphs(text, ncols=80):
391 391 """Wrap multiple paragraphs to fit a specified width.
392 392
393 393 This is equivalent to textwrap.wrap, but with support for multiple
394 394 paragraphs, as separated by empty lines.
395 395
396 396 Returns
397 397 -------
398 398
399 399 list of complete paragraphs, wrapped to fill `ncols` columns.
400 400 """
401 401 paragraph_re = re.compile(r'\n(\s*\n)+', re.MULTILINE)
402 402 text = dedent(text).strip()
403 403 paragraphs = paragraph_re.split(text)[::2] # every other entry is space
404 404 out_ps = []
405 405 indent_re = re.compile(r'\n\s+', re.MULTILINE)
406 406 for p in paragraphs:
407 407 # presume indentation that survives dedent is meaningful formatting,
408 408 # so don't fill unless text is flush.
409 409 if indent_re.search(p) is None:
410 410 # wrap paragraph
411 411 p = textwrap.fill(p, ncols)
412 412 out_ps.append(p)
413 413 return out_ps
414 414
415 415
416 416 def long_substr(data):
417 417 """Return the longest common substring in a list of strings.
418 418
419 419 Credit: http://stackoverflow.com/questions/2892931/longest-common-substring-from-more-than-two-strings-python
420 420 """
421 421 substr = ''
422 422 if len(data) > 1 and len(data[0]) > 0:
423 423 for i in range(len(data[0])):
424 424 for j in range(len(data[0])-i+1):
425 425 if j > len(substr) and all(data[0][i:i+j] in x for x in data):
426 426 substr = data[0][i:i+j]
427 427 elif len(data) == 1:
428 428 substr = data[0]
429 429 return substr
430 430
431 431
432 432 def strip_email_quotes(text):
433 433 """Strip leading email quotation characters ('>').
434 434
435 435 Removes any combination of leading '>' interspersed with whitespace that
436 436 appears *identically* in all lines of the input text.
437 437
438 438 Parameters
439 439 ----------
440 440 text : str
441 441
442 442 Examples
443 443 --------
444 444
445 445 Simple uses::
446 446
447 447 In [2]: strip_email_quotes('> > text')
448 448 Out[2]: 'text'
449 449
450 450 In [3]: strip_email_quotes('> > text\\n> > more')
451 451 Out[3]: 'text\\nmore'
452 452
453 453 Note how only the common prefix that appears in all lines is stripped::
454 454
455 455 In [4]: strip_email_quotes('> > text\\n> > more\\n> more...')
456 456 Out[4]: '> text\\n> more\\nmore...'
457 457
458 458 So if any line has no quote marks ('>') , then none are stripped from any
459 459 of them ::
460 460
461 461 In [5]: strip_email_quotes('> > text\\n> > more\\nlast different')
462 462 Out[5]: '> > text\\n> > more\\nlast different'
463 463 """
464 464 lines = text.splitlines()
465 465 matches = set()
466 466 for line in lines:
467 467 prefix = re.match(r'^(\s*>[ >]*)', line)
468 468 if prefix:
469 469 matches.add(prefix.group(1))
470 470 else:
471 471 break
472 472 else:
473 473 prefix = long_substr(list(matches))
474 474 if prefix:
475 475 strip = len(prefix)
476 476 text = '\n'.join([ ln[strip:] for ln in lines])
477 477 return text
478 478
479 479 def strip_ansi(source):
480 480 """
481 481 Remove ansi escape codes from text.
482 482
483 483 Parameters
484 484 ----------
485 485 source : str
486 486 Source to remove the ansi from
487 487 """
488 488 return re.sub(r'\033\[(\d|;)+?m', '', source)
489 489
490 490
491 491 class EvalFormatter(Formatter):
492 492 """A String Formatter that allows evaluation of simple expressions.
493 493
494 494 Note that this version interprets a : as specifying a format string (as per
495 495 standard string formatting), so if slicing is required, you must explicitly
496 496 create a slice.
497 497
498 498 This is to be used in templating cases, such as the parallel batch
499 499 script templates, where simple arithmetic on arguments is useful.
500 500
501 501 Examples
502 502 --------
503 503 ::
504 504
505 505 In [1]: f = EvalFormatter()
506 506 In [2]: f.format('{n//4}', n=8)
507 507 Out[2]: '2'
508 508
509 509 In [3]: f.format("{greeting[slice(2,4)]}", greeting="Hello")
510 510 Out[3]: 'll'
511 511 """
512 512 def get_field(self, name, args, kwargs):
513 513 v = eval(name, kwargs)
514 514 return v, name
515 515
516 516 #XXX: As of Python 3.4, the format string parsing no longer splits on a colon
517 517 # inside [], so EvalFormatter can handle slicing. Once we only support 3.4 and
518 518 # above, it should be possible to remove FullEvalFormatter.
519 519
520 520 class FullEvalFormatter(Formatter):
521 521 """A String Formatter that allows evaluation of simple expressions.
522 522
523 523 Any time a format key is not found in the kwargs,
524 524 it will be tried as an expression in the kwargs namespace.
525 525
526 526 Note that this version allows slicing using [1:2], so you cannot specify
527 527 a format string. Use :class:`EvalFormatter` to permit format strings.
528 528
529 529 Examples
530 530 --------
531 531 ::
532 532
533 533 In [1]: f = FullEvalFormatter()
534 534 In [2]: f.format('{n//4}', n=8)
535 535 Out[2]: '2'
536 536
537 537 In [3]: f.format('{list(range(5))[2:4]}')
538 538 Out[3]: '[2, 3]'
539 539
540 540 In [4]: f.format('{3*2}')
541 541 Out[4]: '6'
542 542 """
543 543 # copied from Formatter._vformat with minor changes to allow eval
544 544 # and replace the format_spec code with slicing
545 545 def vformat(self, format_string, args, kwargs):
546 546 result = []
547 547 for literal_text, field_name, format_spec, conversion in \
548 548 self.parse(format_string):
549 549
550 550 # output the literal text
551 551 if literal_text:
552 552 result.append(literal_text)
553 553
554 554 # if there's a field, output it
555 555 if field_name is not None:
556 556 # this is some markup, find the object and do
557 557 # the formatting
558 558
559 559 if format_spec:
560 560 # override format spec, to allow slicing:
561 561 field_name = ':'.join([field_name, format_spec])
562 562
563 563 # eval the contents of the field for the object
564 564 # to be formatted
565 565 obj = eval(field_name, kwargs)
566 566
567 567 # do any conversion on the resulting object
568 568 obj = self.convert_field(obj, conversion)
569 569
570 570 # format the object and append to the result
571 571 result.append(self.format_field(obj, ''))
572 572
573 573 return u''.join(py3compat.cast_unicode(s) for s in result)
574 574
575 575
576 576 class DollarFormatter(FullEvalFormatter):
577 577 """Formatter allowing Itpl style $foo replacement, for names and attribute
578 578 access only. Standard {foo} replacement also works, and allows full
579 579 evaluation of its arguments.
580 580
581 581 Examples
582 582 --------
583 583 ::
584 584
585 585 In [1]: f = DollarFormatter()
586 586 In [2]: f.format('{n//4}', n=8)
587 587 Out[2]: '2'
588 588
589 589 In [3]: f.format('23 * 76 is $result', result=23*76)
590 590 Out[3]: '23 * 76 is 1748'
591 591
592 592 In [4]: f.format('$a or {b}', a=1, b=2)
593 593 Out[4]: '1 or 2'
594 594 """
595 595 _dollar_pattern = re.compile("(.*?)\$(\$?[\w\.]+)")
596 596 def parse(self, fmt_string):
597 597 for literal_txt, field_name, format_spec, conversion \
598 598 in Formatter.parse(self, fmt_string):
599 599
600 600 # Find $foo patterns in the literal text.
601 601 continue_from = 0
602 602 txt = ""
603 603 for m in self._dollar_pattern.finditer(literal_txt):
604 604 new_txt, new_field = m.group(1,2)
605 605 # $$foo --> $foo
606 606 if new_field.startswith("$"):
607 607 txt += new_txt + new_field
608 608 else:
609 609 yield (txt + new_txt, new_field, "", None)
610 610 txt = ""
611 611 continue_from = m.end()
612 612
613 613 # Re-yield the {foo} style pattern
614 614 yield (txt + literal_txt[continue_from:], field_name, format_spec, conversion)
615 615
616 616 #-----------------------------------------------------------------------------
617 617 # Utils to columnize a list of string
618 618 #-----------------------------------------------------------------------------
619 619
620 620 def _col_chunks(l, max_rows, row_first=False):
621 621 """Yield successive max_rows-sized column chunks from l."""
622 622 if row_first:
623 623 ncols = (len(l) // max_rows) + (len(l) % max_rows > 0)
624 624 for i in range(ncols):
625 625 yield [l[j] for j in range(i, len(l), ncols)]
626 626 else:
627 627 for i in range(0, len(l), max_rows):
628 628 yield l[i:(i + max_rows)]
629 629
630 630
631 631 def _find_optimal(rlist, row_first=False, separator_size=2, displaywidth=80):
632 632 """Calculate optimal info to columnize a list of string"""
633 633 for max_rows in range(1, len(rlist) + 1):
634 634 col_widths = list(map(max, _col_chunks(rlist, max_rows, row_first)))
635 635 sumlength = sum(col_widths)
636 636 ncols = len(col_widths)
637 637 if sumlength + separator_size * (ncols - 1) <= displaywidth:
638 638 break
639 639 return {'num_columns': ncols,
640 640 'optimal_separator_width': (displaywidth - sumlength) // (ncols - 1) if (ncols - 1) else 0,
641 641 'max_rows': max_rows,
642 642 'column_widths': col_widths
643 643 }
644 644
645 645
646 646 def _get_or_default(mylist, i, default=None):
647 647 """return list item number, or default if don't exist"""
648 648 if i >= len(mylist):
649 649 return default
650 650 else :
651 651 return mylist[i]
652 652
653 653
654 654 def compute_item_matrix(items, row_first=False, empty=None, *args, **kwargs) :
655 655 """Returns a nested list, and info to columnize items
656 656
657 657 Parameters
658 658 ----------
659 659
660 660 items
661 661 list of strings to columize
662 662 row_first : (default False)
663 663 Whether to compute columns for a row-first matrix instead of
664 664 column-first (default).
665 665 empty : (default None)
666 666 default value to fill list if needed
667 667 separator_size : int (default=2)
668 668 How much caracters will be used as a separation between each columns.
669 669 displaywidth : int (default=80)
670 670 The width of the area onto wich the columns should enter
671 671
672 672 Returns
673 673 -------
674 674
675 675 strings_matrix
676 676
677 677 nested list of string, the outer most list contains as many list as
678 678 rows, the innermost lists have each as many element as colums. If the
679 679 total number of elements in `items` does not equal the product of
680 680 rows*columns, the last element of some lists are filled with `None`.
681 681
682 682 dict_info
683 683 some info to make columnize easier:
684 684
685 685 num_columns
686 686 number of columns
687 687 max_rows
688 688 maximum number of rows (final number may be less)
689 689 column_widths
690 690 list of with of each columns
691 691 optimal_separator_width
692 692 best separator width between columns
693 693
694 694 Examples
695 695 --------
696 696 ::
697 697
698 698 In [1]: l = ['aaa','b','cc','d','eeeee','f','g','h','i','j','k','l']
699 699 In [2]: list, info = compute_item_matrix(l, displaywidth=12)
700 700 In [3]: list
701 701 Out[3]: [['aaa', 'f', 'k'], ['b', 'g', 'l'], ['cc', 'h', None], ['d', 'i', None], ['eeeee', 'j', None]]
702 702 In [4]: ideal = {'num_columns': 3, 'column_widths': [5, 1, 1], 'optimal_separator_width': 2, 'max_rows': 5}
703 703 In [5]: all((info[k] == ideal[k] for k in ideal.keys()))
704 704 Out[5]: True
705 705 """
706 706 info = _find_optimal(list(map(len, items)), row_first, *args, **kwargs)
707 707 nrow, ncol = info['max_rows'], info['num_columns']
708 708 if row_first:
709 709 return ([[_get_or_default(items, r * ncol + c, default=empty) for c in range(ncol)] for r in range(nrow)], info)
710 710 else:
711 711 return ([[_get_or_default(items, c * nrow + r, default=empty) for c in range(ncol)] for r in range(nrow)], info)
712 712
713 713
714 714 def columnize(items, row_first=False, separator=' ', displaywidth=80, spread=False):
715 715 """ Transform a list of strings into a single string with columns.
716 716
717 717 Parameters
718 718 ----------
719 719 items : sequence of strings
720 720 The strings to process.
721 721
722 722 row_first : (default False)
723 723 Whether to compute columns for a row-first matrix instead of
724 724 column-first (default).
725 725
726 726 separator : str, optional [default is two spaces]
727 727 The string that separates columns.
728 728
729 729 displaywidth : int, optional [default is 80]
730 730 Width of the display in number of characters.
731 731
732 732 Returns
733 733 -------
734 734 The formatted string.
735 735 """
736 736 if not items:
737 737 return '\n'
738 738 matrix, info = compute_item_matrix(items, row_first=row_first, separator_size=len(separator), displaywidth=displaywidth)
739 739 if spread:
740 740 separator = separator.ljust(int(info['optimal_separator_width']))
741 741 fmatrix = [filter(None, x) for x in matrix]
742 742 sjoin = lambda x : separator.join([ y.ljust(w, ' ') for y, w in zip(x, info['column_widths'])])
743 743 return '\n'.join(map(sjoin, fmatrix))+'\n'
744 744
745 745
746 746 def get_text_list(list_, last_sep=' and ', sep=", ", wrap_item_with=""):
747 747 """
748 748 Return a string with a natural enumeration of items
749 749
750 750 >>> get_text_list(['a', 'b', 'c', 'd'])
751 751 'a, b, c and d'
752 752 >>> get_text_list(['a', 'b', 'c'], ' or ')
753 753 'a, b or c'
754 754 >>> get_text_list(['a', 'b', 'c'], ', ')
755 755 'a, b, c'
756 756 >>> get_text_list(['a', 'b'], ' or ')
757 757 'a or b'
758 758 >>> get_text_list(['a'])
759 759 'a'
760 760 >>> get_text_list([])
761 761 ''
762 762 >>> get_text_list(['a', 'b'], wrap_item_with="`")
763 763 '`a` and `b`'
764 764 >>> get_text_list(['a', 'b', 'c', 'd'], " = ", sep=" + ")
765 765 'a + b + c = d'
766 766 """
767 767 if len(list_) == 0:
768 768 return ''
769 769 if wrap_item_with:
770 770 list_ = ['%s%s%s' % (wrap_item_with, item, wrap_item_with) for
771 771 item in list_]
772 772 if len(list_) == 1:
773 773 return list_[0]
774 774 return '%s%s%s' % (
775 775 sep.join(i for i in list_[:-1]),
776 776 last_sep, list_[-1])
General Comments 0
You need to be logged in to leave comments. Login now