-
Notifications
You must be signed in to change notification settings - Fork 1.1k
/
string.rs
1566 lines (1434 loc) · 58.5 KB
/
string.rs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
use std::borrow::Cow;
use bitflags::bitflags;
use ruff_formatter::{format_args, write, Printed};
use ruff_python_ast::AnyNodeRef;
use ruff_python_ast::{
self as ast, ExprBytesLiteral, ExprFString, ExprStringLiteral, ExpressionRef,
};
use ruff_source_file::Locator;
use ruff_text_size::{Ranged, TextLen, TextRange, TextSize};
use crate::comments::{leading_comments, trailing_comments};
use crate::expression::parentheses::{
in_parentheses_only_group, in_parentheses_only_soft_line_break_or_space,
};
use crate::expression::Expr;
use crate::prelude::*;
use crate::FormatModuleError;
use crate::QuoteStyle;
#[derive(Copy, Clone, Debug)]
enum Quoting {
CanChange,
Preserve,
}
#[derive(Clone, Debug)]
pub(super) enum AnyString<'a> {
String(&'a ExprStringLiteral),
Bytes(&'a ExprBytesLiteral),
FString(&'a ExprFString),
}
impl<'a> AnyString<'a> {
pub(crate) fn from_expression(expression: &'a Expr) -> Option<AnyString<'a>> {
match expression {
Expr::StringLiteral(string) => Some(AnyString::String(string)),
Expr::BytesLiteral(bytes) => Some(AnyString::Bytes(bytes)),
Expr::FString(fstring) => Some(AnyString::FString(fstring)),
_ => None,
}
}
fn quoting(&self, locator: &Locator) -> Quoting {
match self {
Self::String(_) | Self::Bytes(_) => Quoting::CanChange,
Self::FString(f_string) => {
let unprefixed = locator
.slice(f_string.range)
.trim_start_matches(|c| c != '"' && c != '\'');
let triple_quoted =
unprefixed.starts_with(r#"""""#) || unprefixed.starts_with(r"'''");
if f_string.value.elements().any(|value| match value {
Expr::FormattedValue(ast::ExprFormattedValue { range, .. }) => {
let string_content = locator.slice(*range);
if triple_quoted {
string_content.contains(r#"""""#) || string_content.contains("'''")
} else {
string_content.contains(['"', '\''])
}
}
_ => false,
}) {
Quoting::Preserve
} else {
Quoting::CanChange
}
}
}
}
/// Returns `true` if the string is implicitly concatenated.
pub(super) fn is_implicit_concatenated(&self) -> bool {
match self {
Self::String(ExprStringLiteral { value, .. }) => value.is_implicit_concatenated(),
Self::Bytes(ExprBytesLiteral { value, .. }) => value.is_implicit_concatenated(),
Self::FString(ExprFString { value, .. }) => value.is_implicit_concatenated(),
}
}
fn parts(&self) -> Vec<AnyStringPart<'a>> {
match self {
Self::String(ExprStringLiteral { value, .. }) => {
value.parts().map(AnyStringPart::String).collect()
}
Self::Bytes(ExprBytesLiteral { value, .. }) => {
value.parts().map(AnyStringPart::Bytes).collect()
}
Self::FString(ExprFString { value, .. }) => value
.parts()
.map(|f_string_part| match f_string_part {
ast::FStringPart::Literal(string_literal) => {
AnyStringPart::String(string_literal)
}
ast::FStringPart::FString(f_string) => AnyStringPart::FString(f_string),
})
.collect(),
}
}
}
impl Ranged for AnyString<'_> {
fn range(&self) -> TextRange {
match self {
Self::String(expr) => expr.range(),
Self::Bytes(expr) => expr.range(),
Self::FString(expr) => expr.range(),
}
}
}
impl<'a> From<&AnyString<'a>> for AnyNodeRef<'a> {
fn from(value: &AnyString<'a>) -> Self {
match value {
AnyString::String(expr) => AnyNodeRef::ExprStringLiteral(expr),
AnyString::Bytes(expr) => AnyNodeRef::ExprBytesLiteral(expr),
AnyString::FString(expr) => AnyNodeRef::ExprFString(expr),
}
}
}
impl<'a> From<&AnyString<'a>> for ExpressionRef<'a> {
fn from(value: &AnyString<'a>) -> Self {
match value {
AnyString::String(expr) => ExpressionRef::StringLiteral(expr),
AnyString::Bytes(expr) => ExpressionRef::BytesLiteral(expr),
AnyString::FString(expr) => ExpressionRef::FString(expr),
}
}
}
#[derive(Clone, Debug)]
enum AnyStringPart<'a> {
String(&'a ast::StringLiteral),
Bytes(&'a ast::BytesLiteral),
FString(&'a ast::FString),
}
impl<'a> From<&AnyStringPart<'a>> for AnyNodeRef<'a> {
fn from(value: &AnyStringPart<'a>) -> Self {
match value {
AnyStringPart::String(part) => AnyNodeRef::StringLiteral(part),
AnyStringPart::Bytes(part) => AnyNodeRef::BytesLiteral(part),
AnyStringPart::FString(part) => AnyNodeRef::FString(part),
}
}
}
impl Ranged for AnyStringPart<'_> {
fn range(&self) -> TextRange {
match self {
Self::String(part) => part.range(),
Self::Bytes(part) => part.range(),
Self::FString(part) => part.range(),
}
}
}
pub(super) struct FormatString<'a> {
string: &'a AnyString<'a>,
layout: StringLayout,
}
#[derive(Default, Copy, Clone, Debug)]
pub enum StringLayout {
#[default]
Default,
DocString,
/// An implicit concatenated string in a binary like (e.g. `a + b` or `a < b`) expression.
///
/// Formats the implicit concatenated string parts without the enclosing group because the group
/// is added by the binary like formatting.
ImplicitConcatenatedStringInBinaryLike,
}
impl<'a> FormatString<'a> {
pub(super) fn new(string: &'a AnyString<'a>) -> Self {
Self {
string,
layout: StringLayout::Default,
}
}
pub(super) fn with_layout(mut self, layout: StringLayout) -> Self {
self.layout = layout;
self
}
}
impl<'a> Format<PyFormatContext<'_>> for FormatString<'a> {
fn fmt(&self, f: &mut PyFormatter) -> FormatResult<()> {
let parent_docstring_quote_style = f.context().docstring();
let locator = f.context().locator();
let result = match self.layout {
StringLayout::Default => {
if self.string.is_implicit_concatenated() {
in_parentheses_only_group(&FormatStringContinuation::new(self.string)).fmt(f)
} else {
StringPart::from_source(self.string.range(), &locator)
.normalize(
self.string.quoting(&locator),
&locator,
f.options().quote_style(),
parent_docstring_quote_style,
)
.fmt(f)
}
}
StringLayout::DocString => {
let string_part = StringPart::from_source(self.string.range(), &locator);
let normalized = string_part.normalize(
Quoting::CanChange,
&locator,
f.options().quote_style(),
parent_docstring_quote_style,
);
format_docstring(&normalized, f)
}
StringLayout::ImplicitConcatenatedStringInBinaryLike => {
FormatStringContinuation::new(self.string).fmt(f)
}
};
// TODO(dhruvmanila): With PEP 701, comments can be inside f-strings.
// This is to mark all of those comments as formatted but we need to
// figure out how to handle them. Note that this needs to be done only
// after the f-string is formatted, so only for all the non-formatted
// comments.
if let AnyString::FString(fstring) = self.string {
let comments = f.context().comments();
fstring.value.elements().for_each(|value| {
comments.mark_verbatim_node_comments_formatted(value.into());
});
}
result
}
}
struct FormatStringContinuation<'a> {
string: &'a AnyString<'a>,
}
impl<'a> FormatStringContinuation<'a> {
fn new(string: &'a AnyString<'a>) -> Self {
Self { string }
}
}
impl Format<PyFormatContext<'_>> for FormatStringContinuation<'_> {
fn fmt(&self, f: &mut PyFormatter) -> FormatResult<()> {
let comments = f.context().comments().clone();
let locator = f.context().locator();
let in_docstring = f.context().docstring();
let quote_style = f.options().quote_style();
let mut joiner = f.join_with(in_parentheses_only_soft_line_break_or_space());
for part in self.string.parts() {
let normalized = StringPart::from_source(part.range(), &locator).normalize(
self.string.quoting(&locator),
&locator,
quote_style,
in_docstring,
);
joiner.entry(&format_args![
line_suffix_boundary(),
leading_comments(comments.leading(&part)),
normalized,
trailing_comments(comments.trailing(&part))
]);
}
joiner.finish()
}
}
#[derive(Debug)]
struct StringPart {
/// The prefix.
prefix: StringPrefix,
/// The actual quotes of the string in the source
quotes: StringQuotes,
/// The range of the string's content (full range minus quotes and prefix)
content_range: TextRange,
}
impl StringPart {
fn from_source(range: TextRange, locator: &Locator) -> Self {
let string_content = locator.slice(range);
let prefix = StringPrefix::parse(string_content);
let after_prefix = &string_content[usize::from(prefix.text_len())..];
let quotes =
StringQuotes::parse(after_prefix).expect("Didn't find string quotes after prefix");
let relative_raw_content_range = TextRange::new(
prefix.text_len() + quotes.text_len(),
string_content.text_len() - quotes.text_len(),
);
let raw_content_range = relative_raw_content_range + range.start();
Self {
prefix,
content_range: raw_content_range,
quotes,
}
}
/// Computes the strings preferred quotes and normalizes its content.
///
/// The parent docstring quote style should be set when formatting a code
/// snippet within the docstring. The quote style should correspond to the
/// style of quotes used by said docstring. Normalization will ensure the
/// quoting styles don't conflict.
fn normalize<'a>(
self,
quoting: Quoting,
locator: &'a Locator,
configured_style: QuoteStyle,
parent_docstring_quote_style: Option<QuoteStyle>,
) -> NormalizedString<'a> {
// Per PEP 8 and PEP 257, always prefer double quotes for docstrings
// and triple-quoted strings. (We assume docstrings are always
// triple-quoted.)
let preferred_style = if self.quotes.triple {
// ... unless we're formatting a code snippet inside a docstring,
// then we specifically want to invert our quote style to avoid
// writing out invalid Python.
//
// It's worth pointing out that we can actually wind up being
// somewhat out of sync with PEP8 in this case. Consider this
// example:
//
// def foo():
// '''
// Something.
//
// >>> """tricksy"""
// '''
// pass
//
// Ideally, this would be reformatted as:
//
// def foo():
// """
// Something.
//
// >>> '''tricksy'''
// """
// pass
//
// But the logic here results in the original quoting being
// preserved. This is because the quoting style of the outer
// docstring is determined, in part, by looking at its contents. In
// this case, it notices that it contains a `"""` and thus infers
// that using `'''` would overall read better because it avoids
// the need to escape the interior `"""`. Except... in this case,
// the `"""` is actually part of a code snippet that could get
// reformatted to using a different quoting style itself.
//
// Fixing this would, I believe, require some fairly seismic
// changes to how formatting strings works. Namely, we would need
// to look for code snippets before normalizing the docstring, and
// then figure out the quoting style more holistically by looking
// at the various kinds of quotes used in the code snippets and
// what reformatting them might look like.
//
// Overall this is a bit of a corner case and just inverting the
// style from what the parent ultimately decided upon works, even
// if it doesn't have perfect alignment with PEP8.
if let Some(style) = parent_docstring_quote_style {
style.invert()
} else {
QuoteStyle::Double
}
} else {
configured_style
};
let raw_content = locator.slice(self.content_range);
let quotes = match quoting {
Quoting::Preserve => self.quotes,
Quoting::CanChange => {
if self.prefix.is_raw_string() {
choose_quotes_raw(raw_content, self.quotes, preferred_style)
} else {
choose_quotes(raw_content, self.quotes, preferred_style)
}
}
};
let normalized = normalize_string(locator.slice(self.content_range), quotes, self.prefix);
NormalizedString {
prefix: self.prefix,
content_range: self.content_range,
text: normalized,
quotes,
}
}
}
#[derive(Debug)]
struct NormalizedString<'a> {
prefix: StringPrefix,
/// The quotes of the normalized string (preferred quotes)
quotes: StringQuotes,
/// The range of the string's content in the source (minus prefix and quotes).
content_range: TextRange,
/// The normalized text
text: Cow<'a, str>,
}
impl Ranged for NormalizedString<'_> {
fn range(&self) -> TextRange {
self.content_range
}
}
impl Format<PyFormatContext<'_>> for NormalizedString<'_> {
fn fmt(&self, f: &mut Formatter<PyFormatContext<'_>>) -> FormatResult<()> {
write!(f, [self.prefix, self.quotes])?;
match &self.text {
Cow::Borrowed(_) => {
source_text_slice(self.range()).fmt(f)?;
}
Cow::Owned(normalized) => {
text(normalized, Some(self.start())).fmt(f)?;
}
}
self.quotes.fmt(f)
}
}
bitflags! {
#[derive(Copy, Clone, Debug, PartialEq, Eq)]
pub(super) struct StringPrefix: u8 {
const UNICODE = 0b0000_0001;
/// `r"test"`
const RAW = 0b0000_0010;
/// `R"test"
const RAW_UPPER = 0b0000_0100;
const BYTE = 0b0000_1000;
const F_STRING = 0b0001_0000;
}
}
impl StringPrefix {
pub(super) fn parse(input: &str) -> StringPrefix {
let chars = input.chars();
let mut prefix = StringPrefix::empty();
for c in chars {
let flag = match c {
'u' | 'U' => StringPrefix::UNICODE,
'f' | 'F' => StringPrefix::F_STRING,
'b' | 'B' => StringPrefix::BYTE,
'r' => StringPrefix::RAW,
'R' => StringPrefix::RAW_UPPER,
'\'' | '"' => break,
c => {
unreachable!(
"Unexpected character '{c}' terminating the prefix of a string literal"
);
}
};
prefix |= flag;
}
prefix
}
pub(super) const fn text_len(self) -> TextSize {
TextSize::new(self.bits().count_ones())
}
pub(super) const fn is_raw_string(self) -> bool {
self.contains(StringPrefix::RAW) || self.contains(StringPrefix::RAW_UPPER)
}
pub(super) const fn is_fstring(self) -> bool {
self.contains(StringPrefix::F_STRING)
}
}
impl Format<PyFormatContext<'_>> for StringPrefix {
fn fmt(&self, f: &mut PyFormatter) -> FormatResult<()> {
// Retain the casing for the raw prefix:
// https://black.readthedocs.io/en/stable/the_black_code_style/current_style.html#r-strings-and-r-strings
if self.contains(StringPrefix::RAW) {
token("r").fmt(f)?;
} else if self.contains(StringPrefix::RAW_UPPER) {
token("R").fmt(f)?;
}
if self.contains(StringPrefix::BYTE) {
token("b").fmt(f)?;
}
if self.contains(StringPrefix::F_STRING) {
token("f").fmt(f)?;
}
// Remove the unicode prefix `u` if any because it is meaningless in Python 3+.
Ok(())
}
}
/// Choose the appropriate quote style for a raw string.
///
/// The preferred quote style is chosen unless the string contains unescaped quotes of the
/// preferred style. For example, `r"foo"` is chosen over `r'foo'` if the preferred quote
/// style is double quotes.
fn choose_quotes_raw(
input: &str,
quotes: StringQuotes,
preferred_style: QuoteStyle,
) -> StringQuotes {
let preferred_quote_char = preferred_style.as_char();
let mut chars = input.chars().peekable();
let contains_unescaped_configured_quotes = loop {
match chars.next() {
Some('\\') => {
// Ignore escaped characters
chars.next();
}
// `"` or `'`
Some(c) if c == preferred_quote_char => {
if !quotes.triple {
break true;
}
match chars.peek() {
// We can't turn `r'''\""'''` into `r"""\"""""`, this would confuse the parser
// about where the closing triple quotes start
None => break true,
Some(next) if *next == preferred_quote_char => {
// `""` or `''`
chars.next();
// We can't turn `r'''""'''` into `r""""""""`, nor can we have
// `"""` or `'''` respectively inside the string
if chars.peek().is_none() || chars.peek() == Some(&preferred_quote_char) {
break true;
}
}
_ => {}
}
}
Some(_) => continue,
None => break false,
}
};
StringQuotes {
triple: quotes.triple,
style: if contains_unescaped_configured_quotes {
quotes.style
} else {
preferred_style
},
}
}
/// Choose the appropriate quote style for a string.
///
/// For single quoted strings, the preferred quote style is used, unless the alternative quote style
/// would require fewer escapes.
///
/// For triple quoted strings, the preferred quote style is always used, unless the string contains
/// a triplet of the quote character (e.g., if double quotes are preferred, double quotes will be
/// used unless the string contains `"""`).
fn choose_quotes(input: &str, quotes: StringQuotes, preferred_style: QuoteStyle) -> StringQuotes {
let style = if quotes.triple {
// True if the string contains a triple quote sequence of the configured quote style.
let mut uses_triple_quotes = false;
let mut chars = input.chars().peekable();
while let Some(c) = chars.next() {
let preferred_quote_char = preferred_style.as_char();
match c {
'\\' => {
if matches!(chars.peek(), Some('"' | '\\')) {
chars.next();
}
}
// `"` or `'`
c if c == preferred_quote_char => {
match chars.peek().copied() {
Some(c) if c == preferred_quote_char => {
// `""` or `''`
chars.next();
match chars.peek().copied() {
Some(c) if c == preferred_quote_char => {
// `"""` or `'''`
chars.next();
uses_triple_quotes = true;
break;
}
Some(_) => {}
None => {
// Handle `''' ""'''`. At this point we have consumed both
// double quotes, so on the next iteration the iterator is empty
// and we'd miss the string ending with a preferred quote
uses_triple_quotes = true;
break;
}
}
}
Some(_) => {
// A single quote char, this is ok
}
None => {
// Trailing quote at the end of the comment
uses_triple_quotes = true;
break;
}
}
}
_ => continue,
}
}
if uses_triple_quotes {
// String contains a triple quote sequence of the configured quote style.
// Keep the existing quote style.
quotes.style
} else {
preferred_style
}
} else {
let mut single_quotes = 0u32;
let mut double_quotes = 0u32;
for c in input.chars() {
match c {
'\'' => {
single_quotes += 1;
}
'"' => {
double_quotes += 1;
}
_ => continue,
}
}
match preferred_style {
QuoteStyle::Single => {
if single_quotes > double_quotes {
QuoteStyle::Double
} else {
QuoteStyle::Single
}
}
QuoteStyle::Double => {
if double_quotes > single_quotes {
QuoteStyle::Single
} else {
QuoteStyle::Double
}
}
}
};
StringQuotes {
triple: quotes.triple,
style,
}
}
#[derive(Copy, Clone, Debug)]
pub(super) struct StringQuotes {
triple: bool,
style: QuoteStyle,
}
impl StringQuotes {
pub(super) fn parse(input: &str) -> Option<StringQuotes> {
let mut chars = input.chars();
let quote_char = chars.next()?;
let style = QuoteStyle::try_from(quote_char).ok()?;
let triple = chars.next() == Some(quote_char) && chars.next() == Some(quote_char);
Some(Self { triple, style })
}
pub(super) const fn is_triple(self) -> bool {
self.triple
}
const fn text_len(self) -> TextSize {
if self.triple {
TextSize::new(3)
} else {
TextSize::new(1)
}
}
}
impl Format<PyFormatContext<'_>> for StringQuotes {
fn fmt(&self, f: &mut PyFormatter) -> FormatResult<()> {
let quotes = match (self.style, self.triple) {
(QuoteStyle::Single, false) => "'",
(QuoteStyle::Single, true) => "'''",
(QuoteStyle::Double, false) => "\"",
(QuoteStyle::Double, true) => "\"\"\"",
};
token(quotes).fmt(f)
}
}
/// Adds the necessary quote escapes and removes unnecessary escape sequences when quoting `input`
/// with the provided [`StringQuotes`] style.
///
/// Returns the normalized string and whether it contains new lines.
fn normalize_string(input: &str, quotes: StringQuotes, prefix: StringPrefix) -> Cow<str> {
// The normalized string if `input` is not yet normalized.
// `output` must remain empty if `input` is already normalized.
let mut output = String::new();
// Tracks the last index of `input` that has been written to `output`.
// If `last_index` is `0` at the end, then the input is already normalized and can be returned as is.
let mut last_index = 0;
let style = quotes.style;
let preferred_quote = style.as_char();
let opposite_quote = style.invert().as_char();
let mut chars = input.char_indices().peekable();
let is_raw = prefix.is_raw_string();
let is_fstring = prefix.is_fstring();
let mut formatted_value_nesting = 0u32;
while let Some((index, c)) = chars.next() {
if is_fstring && matches!(c, '{' | '}') {
if chars.peek().copied().is_some_and(|(_, next)| next == c) {
// Skip over the second character of the double braces
chars.next();
} else if c == '{' {
formatted_value_nesting += 1;
} else {
// Safe to assume that `c == '}'` here because of the matched pattern above
formatted_value_nesting = formatted_value_nesting.saturating_sub(1);
}
continue;
}
if c == '\r' {
output.push_str(&input[last_index..index]);
// Skip over the '\r' character, keep the `\n`
if chars.peek().copied().is_some_and(|(_, next)| next == '\n') {
chars.next();
}
// Replace the `\r` with a `\n`
else {
output.push('\n');
}
last_index = index + '\r'.len_utf8();
} else if !quotes.triple && !is_raw {
if c == '\\' {
if let Some((_, next)) = chars.peek().copied() {
#[allow(clippy::if_same_then_else)]
if next == opposite_quote && formatted_value_nesting == 0 {
// Remove the escape by ending before the backslash and starting again with the quote
chars.next();
output.push_str(&input[last_index..index]);
last_index = index + '\\'.len_utf8();
} else if next == preferred_quote {
// Quote is already escaped, skip over it.
chars.next();
} else if next == '\\' {
// Skip over escaped backslashes
chars.next();
}
}
} else if c == preferred_quote && formatted_value_nesting == 0 {
// Escape the quote
output.push_str(&input[last_index..index]);
output.push('\\');
output.push(c);
last_index = index + preferred_quote.len_utf8();
}
}
}
let normalized = if last_index == 0 {
Cow::Borrowed(input)
} else {
output.push_str(&input[last_index..]);
Cow::Owned(output)
};
normalized
}
/// For docstring indentation, black counts spaces as 1 and tabs by increasing the indentation up
/// to the next multiple of 8. This is effectively a port of
/// [`str.expandtabs`](https://docs.python.org/3/library/stdtypes.html#str.expandtabs),
/// which black [calls with the default tab width of 8](https://github.com/psf/black/blob/c36e468794f9256d5e922c399240d49782ba04f1/src/black/strings.py#L61).
fn indentation_length(line: &str) -> TextSize {
let mut indentation = 0u32;
for char in line.chars() {
if char == '\t' {
// Pad to the next multiple of tab_width
indentation += 8 - (indentation.rem_euclid(8));
} else if char.is_whitespace() {
indentation += u32::from(char.text_len());
} else {
break;
}
}
TextSize::new(indentation)
}
/// Format a docstring by trimming whitespace and adjusting the indentation.
///
/// Summary of changes we make:
/// * Normalize the string like all other strings
/// * Ignore docstring that have an escaped newline
/// * Trim all trailing whitespace, except for a chaperone space that avoids quotes or backslashes
/// in the last line.
/// * Trim leading whitespace on the first line, again except for a chaperone space
/// * If there is only content in the first line and after that only whitespace, collapse the
/// docstring into one line
/// * Adjust the indentation (see below)
///
/// # Docstring indentation
///
/// Unlike any other string, like black we change the indentation of docstring lines.
///
/// We want to preserve the indentation inside the docstring relative to the suite statement/block
/// indent that the docstring statement is in, but also want to apply the change of the outer
/// indentation in the docstring, e.g.
/// ```python
/// def sparkle_sky():
/// """Make a pretty sparkly sky.
/// * * ✨ *. .
/// * * ✨ .
/// . * . ✨ * . .
/// """
/// ```
/// should become
/// ```python
/// def sparkle_sky():
/// """Make a pretty sparkly sky.
/// * * ✨ *. .
/// * * ✨ .
/// . * . ✨ * . .
/// """
/// ```
/// We can't compute the full indentation here since we don't know what the block indent of
/// the doc comment will be yet and which we can only have added by formatting each line
/// separately with a hard line break. This means we need to strip shared indentation from
/// docstring while preserving the in-docstring bigger-than-suite-statement indentation. Example:
/// ```python
/// def f():
/// """first line
/// line a
/// line b
/// """
/// ```
/// The docstring indentation is 2, the block indents will change this to 4 (but we can't
/// determine this at this point). The indentation of line a is 2, so we trim ` line a`
/// to `line a`. For line b it's 5, so we trim it to `line b` and pad with 5-2=3 spaces to
/// ` line b`. The closing quotes, being on their own line, are stripped get only the
/// default indentation. Fully formatted:
/// ```python
/// def f():
/// """first line
/// line a
/// line b
/// """
/// ```
///
/// Tabs are counted by padding them to the next multiple of 8 according to
/// [`str.expandtabs`](https://docs.python.org/3/library/stdtypes.html#str.expandtabs). When
/// we see indentation that contains a tab or any other none ascii-space whitespace we rewrite the
/// string.
///
/// Additionally, if any line in the docstring has less indentation than the docstring
/// (effectively a negative indentation wrt. to the current level), we pad all lines to the
/// level of the docstring with spaces.
/// ```python
/// def f():
/// """first line
/// line a
/// line b
/// line c
/// """
/// ```
/// Here line a is 3 columns negatively indented, so we pad all lines by an extra 3 spaces:
/// ```python
/// def f():
/// """first line
/// line a
/// line b
/// line c
/// """
/// ```
fn format_docstring(normalized: &NormalizedString, f: &mut PyFormatter) -> FormatResult<()> {
let docstring = &normalized.text;
// Black doesn't change the indentation of docstrings that contain an escaped newline
if docstring.contains("\\\n") {
return normalized.fmt(f);
}
// is_borrowed is unstable :/
let already_normalized = matches!(docstring, Cow::Borrowed(_));
let mut lines = docstring.lines().peekable();
// Start the string
write!(
f,
[
normalized.prefix,
normalized.quotes,
source_position(normalized.start()),
]
)?;
// We track where in the source docstring we are (in source code byte offsets)
let mut offset = normalized.start();
// The first line directly after the opening quotes has different rules than the rest, mainly
// that we remove all leading whitespace as there's no indentation
let first = lines.next().unwrap_or_default();
// Black trims whitespace using [`str.strip()`](https://docs.python.org/3/library/stdtypes.html#str.strip)
// https://github.com/psf/black/blob/b4dca26c7d93f930bbd5a7b552807370b60d4298/src/black/strings.py#L77-L85
// So we use the unicode whitespace definition through `trim_{start,end}` instead of the python
// tokenizer whitespace definition in `trim_whitespace_{start,end}`.
let trim_end = first.trim_end();
let trim_both = trim_end.trim_start();
// Edge case: The first line is `""" "content`, so we need to insert chaperone space that keep
// inner quotes and closing quotes from getting to close to avoid `""""content`
if trim_both.starts_with(normalized.quotes.style.as_char()) {
space().fmt(f)?;
}
if !trim_end.is_empty() {
// For the first line of the docstring we strip the leading and trailing whitespace, e.g.
// `""" content ` to `"""content`
let leading_whitespace = trim_end.text_len() - trim_both.text_len();
let trimmed_line_range =
TextRange::at(offset, trim_end.text_len()).add_start(leading_whitespace);
if already_normalized {
source_text_slice(trimmed_line_range).fmt(f)?;
} else {
text(trim_both, Some(trimmed_line_range.start())).fmt(f)?;
}
}
offset += first.text_len();
// Check if we have a single line (or empty) docstring
if docstring[first.len()..].trim().is_empty() {
// For `"""\n"""` or other whitespace between the quotes, black keeps a single whitespace,
// but `""""""` doesn't get one inserted.
if needs_chaperone_space(normalized, trim_end)
|| (trim_end.is_empty() && !docstring.is_empty())
{
space().fmt(f)?;
}
normalized.quotes.fmt(f)?;
return Ok(());
}
hard_line_break().fmt(f)?;
// We know that the normalized string has \n line endings
offset += "\n".text_len();
// If some line of the docstring is less indented than the function body, we pad all lines to
// align it with the docstring statement. Conversely, if all lines are over-indented, we strip
// the extra indentation. We call this stripped indentation since it's relative to the block
// indent printer-made indentation.
let stripped_indentation_length = lines
.clone()
// We don't want to count whitespace-only lines as miss-indented
.filter(|line| !line.trim().is_empty())
.map(indentation_length)
.min()
.unwrap_or_default();
DocstringLinePrinter {