| 1 |
/*! |
| 2 |
Defines a translator that converts an `Ast` to an `Hir`. |
| 3 |
*/ |
| 4 |
|
| 5 |
use core::cell::{Cell, RefCell}; |
| 6 |
|
| 7 |
use alloc::{boxed::Box, string::ToString, vec, vec::Vec}; |
| 8 |
|
| 9 |
use crate::{ |
| 10 |
ast::{self, Ast, Span, Visitor}, |
| 11 |
either::Either, |
| 12 |
hir::{self, Error, ErrorKind, Hir, HirKind}, |
| 13 |
unicode::{self, ClassQuery}, |
| 14 |
}; |
| 15 |
|
| 16 |
type Result<T> = core::result::Result<T, Error>; |
| 17 |
|
| 18 |
/// A builder for constructing an AST->HIR translator. |
| 19 |
#[derive(Clone, Debug)] |
| 20 |
pub struct TranslatorBuilder { |
| 21 |
utf8: bool, |
| 22 |
line_terminator: u8, |
| 23 |
flags: Flags, |
| 24 |
} |
| 25 |
|
| 26 |
impl Default for TranslatorBuilder { |
| 27 |
fn default() -> TranslatorBuilder { |
| 28 |
TranslatorBuilder::new() |
| 29 |
} |
| 30 |
} |
| 31 |
|
| 32 |
impl TranslatorBuilder { |
| 33 |
/// Create a new translator builder with a default c onfiguration. |
| 34 |
pub fn new() -> TranslatorBuilder { |
| 35 |
TranslatorBuilder { |
| 36 |
utf8: true, |
| 37 |
line_terminator: b'\n', |
| 38 |
flags: Flags::default(), |
| 39 |
} |
| 40 |
} |
| 41 |
|
| 42 |
/// Build a translator using the current configuration. |
| 43 |
pub fn build(&self) -> Translator { |
| 44 |
Translator { |
| 45 |
stack: RefCell::new(vec![]), |
| 46 |
flags: Cell::new(self.flags), |
| 47 |
utf8: self.utf8, |
| 48 |
line_terminator: self.line_terminator, |
| 49 |
} |
| 50 |
} |
| 51 |
|
| 52 |
/// When disabled, translation will permit the construction of a regular |
| 53 |
/// expression that may match invalid UTF-8. |
| 54 |
/// |
| 55 |
/// When enabled (the default), the translator is guaranteed to produce an |
| 56 |
/// expression that, for non-empty matches, will only ever produce spans |
| 57 |
/// that are entirely valid UTF-8 (otherwise, the translator will return an |
| 58 |
/// error). |
| 59 |
/// |
| 60 |
/// Perhaps surprisingly, when UTF-8 is enabled, an empty regex or even |
| 61 |
/// a negated ASCII word boundary (uttered as `(?-u:\B)` in the concrete |
| 62 |
/// syntax) will be allowed even though they can produce matches that split |
| 63 |
/// a UTF-8 encoded codepoint. This only applies to zero-width or "empty" |
| 64 |
/// matches, and it is expected that the regex engine itself must handle |
| 65 |
/// these cases if necessary (perhaps by suppressing any zero-width matches |
| 66 |
/// that split a codepoint). |
| 67 |
pub fn utf8(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 68 |
self.utf8 = yes; |
| 69 |
self |
| 70 |
} |
| 71 |
|
| 72 |
/// Sets the line terminator for use with `(?u-s:.)` and `(?-us:.)`. |
| 73 |
/// |
| 74 |
/// Namely, instead of `.` (by default) matching everything except for `\n`, |
| 75 |
/// this will cause `.` to match everything except for the byte given. |
| 76 |
/// |
| 77 |
/// If `.` is used in a context where Unicode mode is enabled and this byte |
| 78 |
/// isn't ASCII, then an error will be returned. When Unicode mode is |
| 79 |
/// disabled, then any byte is permitted, but will return an error if UTF-8 |
| 80 |
/// mode is enabled and it is a non-ASCII byte. |
| 81 |
/// |
| 82 |
/// In short, any ASCII value for a line terminator is always okay. But a |
| 83 |
/// non-ASCII byte might result in an error depending on whether Unicode |
| 84 |
/// mode or UTF-8 mode are enabled. |
| 85 |
/// |
| 86 |
/// Note that if `R` mode is enabled then it always takes precedence and |
| 87 |
/// the line terminator will be treated as `\r` and `\n` simultaneously. |
| 88 |
/// |
| 89 |
/// Note also that this *doesn't* impact the look-around assertions |
| 90 |
/// `(?m:^)` and `(?m:$)`. That's usually controlled by additional |
| 91 |
/// configuration in the regex engine itself. |
| 92 |
pub fn line_terminator(&mut self, byte: u8) -> &mut TranslatorBuilder { |
| 93 |
self.line_terminator = byte; |
| 94 |
self |
| 95 |
} |
| 96 |
|
| 97 |
/// Enable or disable the case insensitive flag (`i`) by default. |
| 98 |
pub fn case_insensitive(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 99 |
self.flags.case_insensitive = if yes { Some(true) } else { None }; |
| 100 |
self |
| 101 |
} |
| 102 |
|
| 103 |
/// Enable or disable the multi-line matching flag (`m`) by default. |
| 104 |
pub fn multi_line(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 105 |
self.flags.multi_line = if yes { Some(true) } else { None }; |
| 106 |
self |
| 107 |
} |
| 108 |
|
| 109 |
/// Enable or disable the "dot matches any character" flag (`s`) by |
| 110 |
/// default. |
| 111 |
pub fn dot_matches_new_line( |
| 112 |
&mut self, |
| 113 |
yes: bool, |
| 114 |
) -> &mut TranslatorBuilder { |
| 115 |
self.flags.dot_matches_new_line = if yes { Some(true) } else { None }; |
| 116 |
self |
| 117 |
} |
| 118 |
|
| 119 |
/// Enable or disable the CRLF mode flag (`R`) by default. |
| 120 |
pub fn crlf(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 121 |
self.flags.crlf = if yes { Some(true) } else { None }; |
| 122 |
self |
| 123 |
} |
| 124 |
|
| 125 |
/// Enable or disable the "swap greed" flag (`U`) by default. |
| 126 |
pub fn swap_greed(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 127 |
self.flags.swap_greed = if yes { Some(true) } else { None }; |
| 128 |
self |
| 129 |
} |
| 130 |
|
| 131 |
/// Enable or disable the Unicode flag (`u`) by default. |
| 132 |
pub fn unicode(&mut self, yes: bool) -> &mut TranslatorBuilder { |
| 133 |
self.flags.unicode = if yes { None } else { Some(false) }; |
| 134 |
self |
| 135 |
} |
| 136 |
} |
| 137 |
|
| 138 |
/// A translator maps abstract syntax to a high level intermediate |
| 139 |
/// representation. |
| 140 |
/// |
| 141 |
/// A translator may be benefit from reuse. That is, a translator can translate |
| 142 |
/// many abstract syntax trees. |
| 143 |
/// |
| 144 |
/// A `Translator` can be configured in more detail via a |
| 145 |
/// [`TranslatorBuilder`]. |
| 146 |
#[derive(Clone, Debug)] |
| 147 |
pub struct Translator { |
| 148 |
/// Our call stack, but on the heap. |
| 149 |
stack: RefCell<Vec<HirFrame>>, |
| 150 |
/// The current flag settings. |
| 151 |
flags: Cell<Flags>, |
| 152 |
/// Whether we're allowed to produce HIR that can match arbitrary bytes. |
| 153 |
utf8: bool, |
| 154 |
/// The line terminator to use for `.`. |
| 155 |
line_terminator: u8, |
| 156 |
} |
| 157 |
|
| 158 |
impl Translator { |
| 159 |
/// Create a new translator using the default configuration. |
| 160 |
pub fn new() -> Translator { |
| 161 |
TranslatorBuilder::new().build() |
| 162 |
} |
| 163 |
|
| 164 |
/// Translate the given abstract syntax tree (AST) into a high level |
| 165 |
/// intermediate representation (HIR). |
| 166 |
/// |
| 167 |
/// If there was a problem doing the translation, then an HIR-specific |
| 168 |
/// error is returned. |
| 169 |
/// |
| 170 |
/// The original pattern string used to produce the `Ast` *must* also be |
| 171 |
/// provided. The translator does not use the pattern string during any |
| 172 |
/// correct translation, but is used for error reporting. |
| 173 |
pub fn translate(&mut self, pattern: &str, ast: &Ast) -> Result<Hir> { |
| 174 |
ast::visit(ast, TranslatorI::new(self, pattern)) |
| 175 |
} |
| 176 |
} |
| 177 |
|
| 178 |
/// An HirFrame is a single stack frame, represented explicitly, which is |
| 179 |
/// created for each item in the Ast that we traverse. |
| 180 |
/// |
| 181 |
/// Note that technically, this type doesn't represent our entire stack |
| 182 |
/// frame. In particular, the Ast visitor represents any state associated with |
| 183 |
/// traversing the Ast itself. |
| 184 |
#[derive(Clone, Debug)] |
| 185 |
enum HirFrame { |
| 186 |
/// An arbitrary HIR expression. These get pushed whenever we hit a base |
| 187 |
/// case in the Ast. They get popped after an inductive (i.e., recursive) |
| 188 |
/// step is complete. |
| 189 |
Expr(Hir), |
| 190 |
/// A literal that is being constructed, character by character, from the |
| 191 |
/// AST. We need this because the AST gives each individual character its |
| 192 |
/// own node. So as we see characters, we peek at the top-most HirFrame. |
| 193 |
/// If it's a literal, then we add to it. Otherwise, we push a new literal. |
| 194 |
/// When it comes time to pop it, we convert it to an Hir via Hir::literal. |
| 195 |
Literal(Vec<u8>), |
| 196 |
/// A Unicode character class. This frame is mutated as we descend into |
| 197 |
/// the Ast of a character class (which is itself its own mini recursive |
| 198 |
/// structure). |
| 199 |
ClassUnicode(hir::ClassUnicode), |
| 200 |
/// A byte-oriented character class. This frame is mutated as we descend |
| 201 |
/// into the Ast of a character class (which is itself its own mini |
| 202 |
/// recursive structure). |
| 203 |
/// |
| 204 |
/// Byte character classes are created when Unicode mode (`u`) is disabled. |
| 205 |
/// If `utf8` is enabled (the default), then a byte character is only |
| 206 |
/// permitted to match ASCII text. |
| 207 |
ClassBytes(hir::ClassBytes), |
| 208 |
/// This is pushed whenever a repetition is observed. After visiting every |
| 209 |
/// sub-expression in the repetition, the translator's stack is expected to |
| 210 |
/// have this sentinel at the top. |
| 211 |
/// |
| 212 |
/// This sentinel only exists to stop other things (like flattening |
| 213 |
/// literals) from reaching across repetition operators. |
| 214 |
Repetition, |
| 215 |
/// This is pushed on to the stack upon first seeing any kind of capture, |
| 216 |
/// indicated by parentheses (including non-capturing groups). It is popped |
| 217 |
/// upon leaving a group. |
| 218 |
Group { |
| 219 |
/// The old active flags when this group was opened. |
| 220 |
/// |
| 221 |
/// If this group sets flags, then the new active flags are set to the |
| 222 |
/// result of merging the old flags with the flags introduced by this |
| 223 |
/// group. If the group doesn't set any flags, then this is simply |
| 224 |
/// equivalent to whatever flags were set when the group was opened. |
| 225 |
/// |
| 226 |
/// When this group is popped, the active flags should be restored to |
| 227 |
/// the flags set here. |
| 228 |
/// |
| 229 |
/// The "active" flags correspond to whatever flags are set in the |
| 230 |
/// Translator. |
| 231 |
old_flags: Flags, |
| 232 |
}, |
| 233 |
/// This is pushed whenever a concatenation is observed. After visiting |
| 234 |
/// every sub-expression in the concatenation, the translator's stack is |
| 235 |
/// popped until it sees a Concat frame. |
| 236 |
Concat, |
| 237 |
/// This is pushed whenever an alternation is observed. After visiting |
| 238 |
/// every sub-expression in the alternation, the translator's stack is |
| 239 |
/// popped until it sees an Alternation frame. |
| 240 |
Alternation, |
| 241 |
/// This is pushed immediately before each sub-expression in an |
| 242 |
/// alternation. This separates the branches of an alternation on the |
| 243 |
/// stack and prevents literal flattening from reaching across alternation |
| 244 |
/// branches. |
| 245 |
/// |
| 246 |
/// It is popped after each expression in a branch until an 'Alternation' |
| 247 |
/// frame is observed when doing a post visit on an alternation. |
| 248 |
AlternationBranch, |
| 249 |
} |
| 250 |
|
| 251 |
impl HirFrame { |
| 252 |
/// Assert that the current stack frame is an Hir expression and return it. |
| 253 |
fn unwrap_expr(self) -> Hir { |
| 254 |
match self { |
| 255 |
HirFrame::Expr(expr) => expr, |
| 256 |
HirFrame::Literal(lit) => Hir::literal(lit), |
| 257 |
_ => panic!("tried to unwrap expr from HirFrame, got: {:?}", self), |
| 258 |
} |
| 259 |
} |
| 260 |
|
| 261 |
/// Assert that the current stack frame is a Unicode class expression and |
| 262 |
/// return it. |
| 263 |
fn unwrap_class_unicode(self) -> hir::ClassUnicode { |
| 264 |
match self { |
| 265 |
HirFrame::ClassUnicode(cls) => cls, |
| 266 |
_ => panic!( |
| 267 |
"tried to unwrap Unicode class \ |
| 268 |
from HirFrame, got: {:?}", |
| 269 |
self |
| 270 |
), |
| 271 |
} |
| 272 |
} |
| 273 |
|
| 274 |
/// Assert that the current stack frame is a byte class expression and |
| 275 |
/// return it. |
| 276 |
fn unwrap_class_bytes(self) -> hir::ClassBytes { |
| 277 |
match self { |
| 278 |
HirFrame::ClassBytes(cls) => cls, |
| 279 |
_ => panic!( |
| 280 |
"tried to unwrap byte class \ |
| 281 |
from HirFrame, got: {:?}", |
| 282 |
self |
| 283 |
), |
| 284 |
} |
| 285 |
} |
| 286 |
|
| 287 |
/// Assert that the current stack frame is a repetition sentinel. If it |
| 288 |
/// isn't, then panic. |
| 289 |
fn unwrap_repetition(self) { |
| 290 |
match self { |
| 291 |
HirFrame::Repetition => {} |
| 292 |
_ => { |
| 293 |
panic!( |
| 294 |
"tried to unwrap repetition from HirFrame, got: {:?}", |
| 295 |
self |
| 296 |
) |
| 297 |
} |
| 298 |
} |
| 299 |
} |
| 300 |
|
| 301 |
/// Assert that the current stack frame is a group indicator and return |
| 302 |
/// its corresponding flags (the flags that were active at the time the |
| 303 |
/// group was entered). |
| 304 |
fn unwrap_group(self) -> Flags { |
| 305 |
match self { |
| 306 |
HirFrame::Group { old_flags } => old_flags, |
| 307 |
_ => { |
| 308 |
panic!("tried to unwrap group from HirFrame, got: {:?}", self) |
| 309 |
} |
| 310 |
} |
| 311 |
} |
| 312 |
|
| 313 |
/// Assert that the current stack frame is an alternation pipe sentinel. If |
| 314 |
/// it isn't, then panic. |
| 315 |
fn unwrap_alternation_pipe(self) { |
| 316 |
match self { |
| 317 |
HirFrame::AlternationBranch => {} |
| 318 |
_ => { |
| 319 |
panic!( |
| 320 |
"tried to unwrap alt pipe from HirFrame, got: {:?}", |
| 321 |
self |
| 322 |
) |
| 323 |
} |
| 324 |
} |
| 325 |
} |
| 326 |
} |
| 327 |
|
| 328 |
impl<'t, 'p> Visitor for TranslatorI<'t, 'p> { |
| 329 |
type Output = Hir; |
| 330 |
type Err = Error; |
| 331 |
|
| 332 |
fn finish(self) -> Result<Hir> { |
| 333 |
// ... otherwise, we should have exactly one HIR on the stack. |
| 334 |
assert_eq!(self.trans().stack.borrow().len(), 1); |
| 335 |
Ok(self.pop().unwrap().unwrap_expr()) |
| 336 |
} |
| 337 |
|
| 338 |
fn visit_pre(&mut self, ast: &Ast) -> Result<()> { |
| 339 |
match *ast { |
| 340 |
Ast::ClassBracketed(_) => { |
| 341 |
if self.flags().unicode() { |
| 342 |
let cls = hir::ClassUnicode::empty(); |
| 343 |
self.push(HirFrame::ClassUnicode(cls)); |
| 344 |
} else { |
| 345 |
let cls = hir::ClassBytes::empty(); |
| 346 |
self.push(HirFrame::ClassBytes(cls)); |
| 347 |
} |
| 348 |
} |
| 349 |
Ast::Repetition(_) => self.push(HirFrame::Repetition), |
| 350 |
Ast::Group(ref x) => { |
| 351 |
let old_flags = x |
| 352 |
.flags() |
| 353 |
.map(|ast| self.set_flags(ast)) |
| 354 |
.unwrap_or_else(|| self.flags()); |
| 355 |
self.push(HirFrame::Group { old_flags }); |
| 356 |
} |
| 357 |
Ast::Concat(_) => { |
| 358 |
self.push(HirFrame::Concat); |
| 359 |
} |
| 360 |
Ast::Alternation(ref x) => { |
| 361 |
self.push(HirFrame::Alternation); |
| 362 |
if !x.asts.is_empty() { |
| 363 |
self.push(HirFrame::AlternationBranch); |
| 364 |
} |
| 365 |
} |
| 366 |
_ => {} |
| 367 |
} |
| 368 |
Ok(()) |
| 369 |
} |
| 370 |
|
| 371 |
fn visit_post(&mut self, ast: &Ast) -> Result<()> { |
| 372 |
match *ast { |
| 373 |
Ast::Empty(_) => { |
| 374 |
self.push(HirFrame::Expr(Hir::empty())); |
| 375 |
} |
| 376 |
Ast::Flags(ref x) => { |
| 377 |
self.set_flags(&x.flags); |
| 378 |
// Flags in the AST are generally considered directives and |
| 379 |
// not actual sub-expressions. However, they can be used in |
| 380 |
// the concrete syntax like `((?i))`, and we need some kind of |
| 381 |
// indication of an expression there, and Empty is the correct |
| 382 |
// choice. |
| 383 |
// |
| 384 |
// There can also be things like `(?i)+`, but we rule those out |
| 385 |
// in the parser. In the future, we might allow them for |
| 386 |
// consistency sake. |
| 387 |
self.push(HirFrame::Expr(Hir::empty())); |
| 388 |
} |
| 389 |
Ast::Literal(ref x) => match self.ast_literal_to_scalar(x)? { |
| 390 |
Either::Right(byte) => self.push_byte(byte), |
| 391 |
Either::Left(ch) => match self.case_fold_char(x.span, ch)? { |
| 392 |
None => self.push_char(ch), |
| 393 |
Some(expr) => self.push(HirFrame::Expr(expr)), |
| 394 |
}, |
| 395 |
}, |
| 396 |
Ast::Dot(ref span) => { |
| 397 |
self.push(HirFrame::Expr(self.hir_dot(**span)?)); |
| 398 |
} |
| 399 |
Ast::Assertion(ref x) => { |
| 400 |
self.push(HirFrame::Expr(self.hir_assertion(x)?)); |
| 401 |
} |
| 402 |
Ast::ClassPerl(ref x) => { |
| 403 |
if self.flags().unicode() { |
| 404 |
let cls = self.hir_perl_unicode_class(x)?; |
| 405 |
let hcls = hir::Class::Unicode(cls); |
| 406 |
self.push(HirFrame::Expr(Hir::class(hcls))); |
| 407 |
} else { |
| 408 |
let cls = self.hir_perl_byte_class(x)?; |
| 409 |
let hcls = hir::Class::Bytes(cls); |
| 410 |
self.push(HirFrame::Expr(Hir::class(hcls))); |
| 411 |
} |
| 412 |
} |
| 413 |
Ast::ClassUnicode(ref x) => { |
| 414 |
let cls = hir::Class::Unicode(self.hir_unicode_class(x)?); |
| 415 |
self.push(HirFrame::Expr(Hir::class(cls))); |
| 416 |
} |
| 417 |
Ast::ClassBracketed(ref ast) => { |
| 418 |
if self.flags().unicode() { |
| 419 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 420 |
self.unicode_fold_and_negate( |
| 421 |
&ast.span, |
| 422 |
ast.negated, |
| 423 |
&mut cls, |
| 424 |
)?; |
| 425 |
let expr = Hir::class(hir::Class::Unicode(cls)); |
| 426 |
self.push(HirFrame::Expr(expr)); |
| 427 |
} else { |
| 428 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 429 |
self.bytes_fold_and_negate( |
| 430 |
&ast.span, |
| 431 |
ast.negated, |
| 432 |
&mut cls, |
| 433 |
)?; |
| 434 |
let expr = Hir::class(hir::Class::Bytes(cls)); |
| 435 |
self.push(HirFrame::Expr(expr)); |
| 436 |
} |
| 437 |
} |
| 438 |
Ast::Repetition(ref x) => { |
| 439 |
let expr = self.pop().unwrap().unwrap_expr(); |
| 440 |
self.pop().unwrap().unwrap_repetition(); |
| 441 |
self.push(HirFrame::Expr(self.hir_repetition(x, expr))); |
| 442 |
} |
| 443 |
Ast::Group(ref x) => { |
| 444 |
let expr = self.pop().unwrap().unwrap_expr(); |
| 445 |
let old_flags = self.pop().unwrap().unwrap_group(); |
| 446 |
self.trans().flags.set(old_flags); |
| 447 |
self.push(HirFrame::Expr(self.hir_capture(x, expr))); |
| 448 |
} |
| 449 |
Ast::Concat(_) => { |
| 450 |
let mut exprs = vec![]; |
| 451 |
while let Some(expr) = self.pop_concat_expr() { |
| 452 |
if !matches!(*expr.kind(), HirKind::Empty) { |
| 453 |
exprs.push(expr); |
| 454 |
} |
| 455 |
} |
| 456 |
exprs.reverse(); |
| 457 |
self.push(HirFrame::Expr(Hir::concat(exprs))); |
| 458 |
} |
| 459 |
Ast::Alternation(_) => { |
| 460 |
let mut exprs = vec![]; |
| 461 |
while let Some(expr) = self.pop_alt_expr() { |
| 462 |
self.pop().unwrap().unwrap_alternation_pipe(); |
| 463 |
exprs.push(expr); |
| 464 |
} |
| 465 |
exprs.reverse(); |
| 466 |
self.push(HirFrame::Expr(Hir::alternation(exprs))); |
| 467 |
} |
| 468 |
} |
| 469 |
Ok(()) |
| 470 |
} |
| 471 |
|
| 472 |
fn visit_alternation_in(&mut self) -> Result<()> { |
| 473 |
self.push(HirFrame::AlternationBranch); |
| 474 |
Ok(()) |
| 475 |
} |
| 476 |
|
| 477 |
fn visit_class_set_item_pre( |
| 478 |
&mut self, |
| 479 |
ast: &ast::ClassSetItem, |
| 480 |
) -> Result<()> { |
| 481 |
match *ast { |
| 482 |
ast::ClassSetItem::Bracketed(_) => { |
| 483 |
if self.flags().unicode() { |
| 484 |
let cls = hir::ClassUnicode::empty(); |
| 485 |
self.push(HirFrame::ClassUnicode(cls)); |
| 486 |
} else { |
| 487 |
let cls = hir::ClassBytes::empty(); |
| 488 |
self.push(HirFrame::ClassBytes(cls)); |
| 489 |
} |
| 490 |
} |
| 491 |
// We needn't handle the Union case here since the visitor will |
| 492 |
// do it for us. |
| 493 |
_ => {} |
| 494 |
} |
| 495 |
Ok(()) |
| 496 |
} |
| 497 |
|
| 498 |
fn visit_class_set_item_post( |
| 499 |
&mut self, |
| 500 |
ast: &ast::ClassSetItem, |
| 501 |
) -> Result<()> { |
| 502 |
match *ast { |
| 503 |
ast::ClassSetItem::Empty(_) => {} |
| 504 |
ast::ClassSetItem::Literal(ref x) => { |
| 505 |
if self.flags().unicode() { |
| 506 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 507 |
cls.push(hir::ClassUnicodeRange::new(x.c, x.c)); |
| 508 |
self.push(HirFrame::ClassUnicode(cls)); |
| 509 |
} else { |
| 510 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 511 |
let byte = self.class_literal_byte(x)?; |
| 512 |
cls.push(hir::ClassBytesRange::new(byte, byte)); |
| 513 |
self.push(HirFrame::ClassBytes(cls)); |
| 514 |
} |
| 515 |
} |
| 516 |
ast::ClassSetItem::Range(ref x) => { |
| 517 |
if self.flags().unicode() { |
| 518 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 519 |
cls.push(hir::ClassUnicodeRange::new(x.start.c, x.end.c)); |
| 520 |
self.push(HirFrame::ClassUnicode(cls)); |
| 521 |
} else { |
| 522 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 523 |
let start = self.class_literal_byte(&x.start)?; |
| 524 |
let end = self.class_literal_byte(&x.end)?; |
| 525 |
cls.push(hir::ClassBytesRange::new(start, end)); |
| 526 |
self.push(HirFrame::ClassBytes(cls)); |
| 527 |
} |
| 528 |
} |
| 529 |
ast::ClassSetItem::Ascii(ref x) => { |
| 530 |
if self.flags().unicode() { |
| 531 |
let xcls = self.hir_ascii_unicode_class(x)?; |
| 532 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 533 |
cls.union(&xcls); |
| 534 |
self.push(HirFrame::ClassUnicode(cls)); |
| 535 |
} else { |
| 536 |
let xcls = self.hir_ascii_byte_class(x)?; |
| 537 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 538 |
cls.union(&xcls); |
| 539 |
self.push(HirFrame::ClassBytes(cls)); |
| 540 |
} |
| 541 |
} |
| 542 |
ast::ClassSetItem::Unicode(ref x) => { |
| 543 |
let xcls = self.hir_unicode_class(x)?; |
| 544 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 545 |
cls.union(&xcls); |
| 546 |
self.push(HirFrame::ClassUnicode(cls)); |
| 547 |
} |
| 548 |
ast::ClassSetItem::Perl(ref x) => { |
| 549 |
if self.flags().unicode() { |
| 550 |
let xcls = self.hir_perl_unicode_class(x)?; |
| 551 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 552 |
cls.union(&xcls); |
| 553 |
self.push(HirFrame::ClassUnicode(cls)); |
| 554 |
} else { |
| 555 |
let xcls = self.hir_perl_byte_class(x)?; |
| 556 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 557 |
cls.union(&xcls); |
| 558 |
self.push(HirFrame::ClassBytes(cls)); |
| 559 |
} |
| 560 |
} |
| 561 |
ast::ClassSetItem::Bracketed(ref ast) => { |
| 562 |
if self.flags().unicode() { |
| 563 |
let mut cls1 = self.pop().unwrap().unwrap_class_unicode(); |
| 564 |
self.unicode_fold_and_negate( |
| 565 |
&ast.span, |
| 566 |
ast.negated, |
| 567 |
&mut cls1, |
| 568 |
)?; |
| 569 |
|
| 570 |
let mut cls2 = self.pop().unwrap().unwrap_class_unicode(); |
| 571 |
cls2.union(&cls1); |
| 572 |
self.push(HirFrame::ClassUnicode(cls2)); |
| 573 |
} else { |
| 574 |
let mut cls1 = self.pop().unwrap().unwrap_class_bytes(); |
| 575 |
self.bytes_fold_and_negate( |
| 576 |
&ast.span, |
| 577 |
ast.negated, |
| 578 |
&mut cls1, |
| 579 |
)?; |
| 580 |
|
| 581 |
let mut cls2 = self.pop().unwrap().unwrap_class_bytes(); |
| 582 |
cls2.union(&cls1); |
| 583 |
self.push(HirFrame::ClassBytes(cls2)); |
| 584 |
} |
| 585 |
} |
| 586 |
// This is handled automatically by the visitor. |
| 587 |
ast::ClassSetItem::Union(_) => {} |
| 588 |
} |
| 589 |
Ok(()) |
| 590 |
} |
| 591 |
|
| 592 |
fn visit_class_set_binary_op_pre( |
| 593 |
&mut self, |
| 594 |
_op: &ast::ClassSetBinaryOp, |
| 595 |
) -> Result<()> { |
| 596 |
if self.flags().unicode() { |
| 597 |
let cls = hir::ClassUnicode::empty(); |
| 598 |
self.push(HirFrame::ClassUnicode(cls)); |
| 599 |
} else { |
| 600 |
let cls = hir::ClassBytes::empty(); |
| 601 |
self.push(HirFrame::ClassBytes(cls)); |
| 602 |
} |
| 603 |
Ok(()) |
| 604 |
} |
| 605 |
|
| 606 |
fn visit_class_set_binary_op_in( |
| 607 |
&mut self, |
| 608 |
_op: &ast::ClassSetBinaryOp, |
| 609 |
) -> Result<()> { |
| 610 |
if self.flags().unicode() { |
| 611 |
let cls = hir::ClassUnicode::empty(); |
| 612 |
self.push(HirFrame::ClassUnicode(cls)); |
| 613 |
} else { |
| 614 |
let cls = hir::ClassBytes::empty(); |
| 615 |
self.push(HirFrame::ClassBytes(cls)); |
| 616 |
} |
| 617 |
Ok(()) |
| 618 |
} |
| 619 |
|
| 620 |
fn visit_class_set_binary_op_post( |
| 621 |
&mut self, |
| 622 |
op: &ast::ClassSetBinaryOp, |
| 623 |
) -> Result<()> { |
| 624 |
use crate::ast::ClassSetBinaryOpKind::*; |
| 625 |
|
| 626 |
if self.flags().unicode() { |
| 627 |
let mut rhs = self.pop().unwrap().unwrap_class_unicode(); |
| 628 |
let mut lhs = self.pop().unwrap().unwrap_class_unicode(); |
| 629 |
let mut cls = self.pop().unwrap().unwrap_class_unicode(); |
| 630 |
if self.flags().case_insensitive() { |
| 631 |
rhs.try_case_fold_simple().map_err(|_| { |
| 632 |
self.error( |
| 633 |
op.rhs.span().clone(), |
| 634 |
ErrorKind::UnicodeCaseUnavailable, |
| 635 |
) |
| 636 |
})?; |
| 637 |
lhs.try_case_fold_simple().map_err(|_| { |
| 638 |
self.error( |
| 639 |
op.lhs.span().clone(), |
| 640 |
ErrorKind::UnicodeCaseUnavailable, |
| 641 |
) |
| 642 |
})?; |
| 643 |
} |
| 644 |
match op.kind { |
| 645 |
Intersection => lhs.intersect(&rhs), |
| 646 |
Difference => lhs.difference(&rhs), |
| 647 |
SymmetricDifference => lhs.symmetric_difference(&rhs), |
| 648 |
} |
| 649 |
cls.union(&lhs); |
| 650 |
self.push(HirFrame::ClassUnicode(cls)); |
| 651 |
} else { |
| 652 |
let mut rhs = self.pop().unwrap().unwrap_class_bytes(); |
| 653 |
let mut lhs = self.pop().unwrap().unwrap_class_bytes(); |
| 654 |
let mut cls = self.pop().unwrap().unwrap_class_bytes(); |
| 655 |
if self.flags().case_insensitive() { |
| 656 |
rhs.case_fold_simple(); |
| 657 |
lhs.case_fold_simple(); |
| 658 |
} |
| 659 |
match op.kind { |
| 660 |
Intersection => lhs.intersect(&rhs), |
| 661 |
Difference => lhs.difference(&rhs), |
| 662 |
SymmetricDifference => lhs.symmetric_difference(&rhs), |
| 663 |
} |
| 664 |
cls.union(&lhs); |
| 665 |
self.push(HirFrame::ClassBytes(cls)); |
| 666 |
} |
| 667 |
Ok(()) |
| 668 |
} |
| 669 |
} |
| 670 |
|
| 671 |
/// The internal implementation of a translator. |
| 672 |
/// |
| 673 |
/// This type is responsible for carrying around the original pattern string, |
| 674 |
/// which is not tied to the internal state of a translator. |
| 675 |
/// |
| 676 |
/// A TranslatorI exists for the time it takes to translate a single Ast. |
| 677 |
#[derive(Clone, Debug)] |
| 678 |
struct TranslatorI<'t, 'p> { |
| 679 |
trans: &'t Translator, |
| 680 |
pattern: &'p str, |
| 681 |
} |
| 682 |
|
| 683 |
impl<'t, 'p> TranslatorI<'t, 'p> { |
| 684 |
/// Build a new internal translator. |
| 685 |
fn new(trans: &'t Translator, pattern: &'p str) -> TranslatorI<'t, 'p> { |
| 686 |
TranslatorI { trans, pattern } |
| 687 |
} |
| 688 |
|
| 689 |
/// Return a reference to the underlying translator. |
| 690 |
fn trans(&self) -> &Translator { |
| 691 |
&self.trans |
| 692 |
} |
| 693 |
|
| 694 |
/// Push the given frame on to the call stack. |
| 695 |
fn push(&self, frame: HirFrame) { |
| 696 |
self.trans().stack.borrow_mut().push(frame); |
| 697 |
} |
| 698 |
|
| 699 |
/// Push the given literal char on to the call stack. |
| 700 |
/// |
| 701 |
/// If the top-most element of the stack is a literal, then the char |
| 702 |
/// is appended to the end of that literal. Otherwise, a new literal |
| 703 |
/// containing just the given char is pushed to the top of the stack. |
| 704 |
fn push_char(&self, ch: char) { |
| 705 |
let mut buf = [0; 4]; |
| 706 |
let bytes = ch.encode_utf8(&mut buf).as_bytes(); |
| 707 |
let mut stack = self.trans().stack.borrow_mut(); |
| 708 |
if let Some(HirFrame::Literal(ref mut literal)) = stack.last_mut() { |
| 709 |
literal.extend_from_slice(bytes); |
| 710 |
} else { |
| 711 |
stack.push(HirFrame::Literal(bytes.to_vec())); |
| 712 |
} |
| 713 |
} |
| 714 |
|
| 715 |
/// Push the given literal byte on to the call stack. |
| 716 |
/// |
| 717 |
/// If the top-most element of the stack is a literal, then the byte |
| 718 |
/// is appended to the end of that literal. Otherwise, a new literal |
| 719 |
/// containing just the given byte is pushed to the top of the stack. |
| 720 |
fn push_byte(&self, byte: u8) { |
| 721 |
let mut stack = self.trans().stack.borrow_mut(); |
| 722 |
if let Some(HirFrame::Literal(ref mut literal)) = stack.last_mut() { |
| 723 |
literal.push(byte); |
| 724 |
} else { |
| 725 |
stack.push(HirFrame::Literal(vec![byte])); |
| 726 |
} |
| 727 |
} |
| 728 |
|
| 729 |
/// Pop the top of the call stack. If the call stack is empty, return None. |
| 730 |
fn pop(&self) -> Option<HirFrame> { |
| 731 |
self.trans().stack.borrow_mut().pop() |
| 732 |
} |
| 733 |
|
| 734 |
/// Pop an HIR expression from the top of the stack for a concatenation. |
| 735 |
/// |
| 736 |
/// This returns None if the stack is empty or when a concat frame is seen. |
| 737 |
/// Otherwise, it panics if it could not find an HIR expression. |
| 738 |
fn pop_concat_expr(&self) -> Option<Hir> { |
| 739 |
let frame = self.pop()?; |
| 740 |
match frame { |
| 741 |
HirFrame::Concat => None, |
| 742 |
HirFrame::Expr(expr) => Some(expr), |
| 743 |
HirFrame::Literal(lit) => Some(Hir::literal(lit)), |
| 744 |
HirFrame::ClassUnicode(_) => { |
| 745 |
unreachable!("expected expr or concat, got Unicode class") |
| 746 |
} |
| 747 |
HirFrame::ClassBytes(_) => { |
| 748 |
unreachable!("expected expr or concat, got byte class") |
| 749 |
} |
| 750 |
HirFrame::Repetition => { |
| 751 |
unreachable!("expected expr or concat, got repetition") |
| 752 |
} |
| 753 |
HirFrame::Group { .. } => { |
| 754 |
unreachable!("expected expr or concat, got group") |
| 755 |
} |
| 756 |
HirFrame::Alternation => { |
| 757 |
unreachable!("expected expr or concat, got alt marker") |
| 758 |
} |
| 759 |
HirFrame::AlternationBranch => { |
| 760 |
unreachable!("expected expr or concat, got alt branch marker") |
| 761 |
} |
| 762 |
} |
| 763 |
} |
| 764 |
|
| 765 |
/// Pop an HIR expression from the top of the stack for an alternation. |
| 766 |
/// |
| 767 |
/// This returns None if the stack is empty or when an alternation frame is |
| 768 |
/// seen. Otherwise, it panics if it could not find an HIR expression. |
| 769 |
fn pop_alt_expr(&self) -> Option<Hir> { |
| 770 |
let frame = self.pop()?; |
| 771 |
match frame { |
| 772 |
HirFrame::Alternation => None, |
| 773 |
HirFrame::Expr(expr) => Some(expr), |
| 774 |
HirFrame::Literal(lit) => Some(Hir::literal(lit)), |
| 775 |
HirFrame::ClassUnicode(_) => { |
| 776 |
unreachable!("expected expr or alt, got Unicode class") |
| 777 |
} |
| 778 |
HirFrame::ClassBytes(_) => { |
| 779 |
unreachable!("expected expr or alt, got byte class") |
| 780 |
} |
| 781 |
HirFrame::Repetition => { |
| 782 |
unreachable!("expected expr or alt, got repetition") |
| 783 |
} |
| 784 |
HirFrame::Group { .. } => { |
| 785 |
unreachable!("expected expr or alt, got group") |
| 786 |
} |
| 787 |
HirFrame::Concat => { |
| 788 |
unreachable!("expected expr or alt, got concat marker") |
| 789 |
} |
| 790 |
HirFrame::AlternationBranch => { |
| 791 |
unreachable!("expected expr or alt, got alt branch marker") |
| 792 |
} |
| 793 |
} |
| 794 |
} |
| 795 |
|
| 796 |
/// Create a new error with the given span and error type. |
| 797 |
fn error(&self, span: Span, kind: ErrorKind) -> Error { |
| 798 |
Error { kind, pattern: self.pattern.to_string(), span } |
| 799 |
} |
| 800 |
|
| 801 |
/// Return a copy of the active flags. |
| 802 |
fn flags(&self) -> Flags { |
| 803 |
self.trans().flags.get() |
| 804 |
} |
| 805 |
|
| 806 |
/// Set the flags of this translator from the flags set in the given AST. |
| 807 |
/// Then, return the old flags. |
| 808 |
fn set_flags(&self, ast_flags: &ast::Flags) -> Flags { |
| 809 |
let old_flags = self.flags(); |
| 810 |
let mut new_flags = Flags::from_ast(ast_flags); |
| 811 |
new_flags.merge(&old_flags); |
| 812 |
self.trans().flags.set(new_flags); |
| 813 |
old_flags |
| 814 |
} |
| 815 |
|
| 816 |
/// Convert an Ast literal to its scalar representation. |
| 817 |
/// |
| 818 |
/// When Unicode mode is enabled, then this always succeeds and returns a |
| 819 |
/// `char` (Unicode scalar value). |
| 820 |
/// |
| 821 |
/// When Unicode mode is disabled, then a `char` will still be returned |
| 822 |
/// whenever possible. A byte is returned only when invalid UTF-8 is |
| 823 |
/// allowed and when the byte is not ASCII. Otherwise, a non-ASCII byte |
| 824 |
/// will result in an error when invalid UTF-8 is not allowed. |
| 825 |
fn ast_literal_to_scalar( |
| 826 |
&self, |
| 827 |
lit: &ast::Literal, |
| 828 |
) -> Result<Either<char, u8>> { |
| 829 |
if self.flags().unicode() { |
| 830 |
return Ok(Either::Left(lit.c)); |
| 831 |
} |
| 832 |
let byte = match lit.byte() { |
| 833 |
None => return Ok(Either::Left(lit.c)), |
| 834 |
Some(byte) => byte, |
| 835 |
}; |
| 836 |
if byte <= 0x7F { |
| 837 |
return Ok(Either::Left(char::try_from(byte).unwrap())); |
| 838 |
} |
| 839 |
if self.trans().utf8 { |
| 840 |
return Err(self.error(lit.span, ErrorKind::InvalidUtf8)); |
| 841 |
} |
| 842 |
Ok(Either::Right(byte)) |
| 843 |
} |
| 844 |
|
| 845 |
fn case_fold_char(&self, span: Span, c: char) -> Result<Option<Hir>> { |
| 846 |
if !self.flags().case_insensitive() { |
| 847 |
return Ok(None); |
| 848 |
} |
| 849 |
if self.flags().unicode() { |
| 850 |
// If case folding won't do anything, then don't bother trying. |
| 851 |
let map = unicode::SimpleCaseFolder::new() |
| 852 |
.map(|f| f.overlaps(c, c)) |
| 853 |
.map_err(|_| { |
| 854 |
self.error(span, ErrorKind::UnicodeCaseUnavailable) |
| 855 |
})?; |
| 856 |
if !map { |
| 857 |
return Ok(None); |
| 858 |
} |
| 859 |
let mut cls = |
| 860 |
hir::ClassUnicode::new(vec![hir::ClassUnicodeRange::new( |
| 861 |
c, c, |
| 862 |
)]); |
| 863 |
cls.try_case_fold_simple().map_err(|_| { |
| 864 |
self.error(span, ErrorKind::UnicodeCaseUnavailable) |
| 865 |
})?; |
| 866 |
Ok(Some(Hir::class(hir::Class::Unicode(cls)))) |
| 867 |
} else { |
| 868 |
if !c.is_ascii() { |
| 869 |
return Ok(None); |
| 870 |
} |
| 871 |
// If case folding won't do anything, then don't bother trying. |
| 872 |
match c { |
| 873 |
'A'..='Z' | 'a'..='z' => {} |
| 874 |
_ => return Ok(None), |
| 875 |
} |
| 876 |
let mut cls = |
| 877 |
hir::ClassBytes::new(vec![hir::ClassBytesRange::new( |
| 878 |
// OK because 'c.len_utf8() == 1' which in turn implies |
| 879 |
// that 'c' is ASCII. |
| 880 |
u8::try_from(c).unwrap(), |
| 881 |
u8::try_from(c).unwrap(), |
| 882 |
)]); |
| 883 |
cls.case_fold_simple(); |
| 884 |
Ok(Some(Hir::class(hir::Class::Bytes(cls)))) |
| 885 |
} |
| 886 |
} |
| 887 |
|
| 888 |
fn hir_dot(&self, span: Span) -> Result<Hir> { |
| 889 |
let (utf8, lineterm, flags) = |
| 890 |
(self.trans().utf8, self.trans().line_terminator, self.flags()); |
| 891 |
if utf8 && (!flags.unicode() || !lineterm.is_ascii()) { |
| 892 |
return Err(self.error(span, ErrorKind::InvalidUtf8)); |
| 893 |
} |
| 894 |
let dot = if flags.dot_matches_new_line() { |
| 895 |
if flags.unicode() { |
| 896 |
hir::Dot::AnyChar |
| 897 |
} else { |
| 898 |
hir::Dot::AnyByte |
| 899 |
} |
| 900 |
} else { |
| 901 |
if flags.unicode() { |
| 902 |
if flags.crlf() { |
| 903 |
hir::Dot::AnyCharExceptCRLF |
| 904 |
} else { |
| 905 |
if !lineterm.is_ascii() { |
| 906 |
return Err( |
| 907 |
self.error(span, ErrorKind::InvalidLineTerminator) |
| 908 |
); |
| 909 |
} |
| 910 |
hir::Dot::AnyCharExcept(char::from(lineterm)) |
| 911 |
} |
| 912 |
} else { |
| 913 |
if flags.crlf() { |
| 914 |
hir::Dot::AnyByteExceptCRLF |
| 915 |
} else { |
| 916 |
hir::Dot::AnyByteExcept(lineterm) |
| 917 |
} |
| 918 |
} |
| 919 |
}; |
| 920 |
Ok(Hir::dot(dot)) |
| 921 |
} |
| 922 |
|
| 923 |
fn hir_assertion(&self, asst: &ast::Assertion) -> Result<Hir> { |
| 924 |
let unicode = self.flags().unicode(); |
| 925 |
let multi_line = self.flags().multi_line(); |
| 926 |
let crlf = self.flags().crlf(); |
| 927 |
Ok(match asst.kind { |
| 928 |
ast::AssertionKind::StartLine => Hir::look(if multi_line { |
| 929 |
if crlf { |
| 930 |
hir::Look::StartCRLF |
| 931 |
} else { |
| 932 |
hir::Look::StartLF |
| 933 |
} |
| 934 |
} else { |
| 935 |
hir::Look::Start |
| 936 |
}), |
| 937 |
ast::AssertionKind::EndLine => Hir::look(if multi_line { |
| 938 |
if crlf { |
| 939 |
hir::Look::EndCRLF |
| 940 |
} else { |
| 941 |
hir::Look::EndLF |
| 942 |
} |
| 943 |
} else { |
| 944 |
hir::Look::End |
| 945 |
}), |
| 946 |
ast::AssertionKind::StartText => Hir::look(hir::Look::Start), |
| 947 |
ast::AssertionKind::EndText => Hir::look(hir::Look::End), |
| 948 |
ast::AssertionKind::WordBoundary => Hir::look(if unicode { |
| 949 |
hir::Look::WordUnicode |
| 950 |
} else { |
| 951 |
hir::Look::WordAscii |
| 952 |
}), |
| 953 |
ast::AssertionKind::NotWordBoundary => Hir::look(if unicode { |
| 954 |
hir::Look::WordUnicodeNegate |
| 955 |
} else { |
| 956 |
hir::Look::WordAsciiNegate |
| 957 |
}), |
| 958 |
ast::AssertionKind::WordBoundaryStart |
| 959 |
| ast::AssertionKind::WordBoundaryStartAngle => { |
| 960 |
Hir::look(if unicode { |
| 961 |
hir::Look::WordStartUnicode |
| 962 |
} else { |
| 963 |
hir::Look::WordStartAscii |
| 964 |
}) |
| 965 |
} |
| 966 |
ast::AssertionKind::WordBoundaryEnd |
| 967 |
| ast::AssertionKind::WordBoundaryEndAngle => { |
| 968 |
Hir::look(if unicode { |
| 969 |
hir::Look::WordEndUnicode |
| 970 |
} else { |
| 971 |
hir::Look::WordEndAscii |
| 972 |
}) |
| 973 |
} |
| 974 |
ast::AssertionKind::WordBoundaryStartHalf => { |
| 975 |
Hir::look(if unicode { |
| 976 |
hir::Look::WordStartHalfUnicode |
| 977 |
} else { |
| 978 |
hir::Look::WordStartHalfAscii |
| 979 |
}) |
| 980 |
} |
| 981 |
ast::AssertionKind::WordBoundaryEndHalf => Hir::look(if unicode { |
| 982 |
hir::Look::WordEndHalfUnicode |
| 983 |
} else { |
| 984 |
hir::Look::WordEndHalfAscii |
| 985 |
}), |
| 986 |
}) |
| 987 |
} |
| 988 |
|
| 989 |
fn hir_capture(&self, group: &ast::Group, expr: Hir) -> Hir { |
| 990 |
let (index, name) = match group.kind { |
| 991 |
ast::GroupKind::CaptureIndex(index) => (index, None), |
| 992 |
ast::GroupKind::CaptureName { ref name, .. } => { |
| 993 |
(name.index, Some(name.name.clone().into_boxed_str())) |
| 994 |
} |
| 995 |
// The HIR doesn't need to use non-capturing groups, since the way |
| 996 |
// in which the data type is defined handles this automatically. |
| 997 |
ast::GroupKind::NonCapturing(_) => return expr, |
| 998 |
}; |
| 999 |
Hir::capture(hir::Capture { index, name, sub: Box::new(expr) }) |
| 1000 |
} |
| 1001 |
|
| 1002 |
fn hir_repetition(&self, rep: &ast::Repetition, expr: Hir) -> Hir { |
| 1003 |
let (min, max) = match rep.op.kind { |
| 1004 |
ast::RepetitionKind::ZeroOrOne => (0, Some(1)), |
| 1005 |
ast::RepetitionKind::ZeroOrMore => (0, None), |
| 1006 |
ast::RepetitionKind::OneOrMore => (1, None), |
| 1007 |
ast::RepetitionKind::Range(ast::RepetitionRange::Exactly(m)) => { |
| 1008 |
(m, Some(m)) |
| 1009 |
} |
| 1010 |
ast::RepetitionKind::Range(ast::RepetitionRange::AtLeast(m)) => { |
| 1011 |
(m, None) |
| 1012 |
} |
| 1013 |
ast::RepetitionKind::Range(ast::RepetitionRange::Bounded( |
| 1014 |
m, |
| 1015 |
n, |
| 1016 |
)) => (m, Some(n)), |
| 1017 |
}; |
| 1018 |
let greedy = |
| 1019 |
if self.flags().swap_greed() { !rep.greedy } else { rep.greedy }; |
| 1020 |
Hir::repetition(hir::Repetition { |
| 1021 |
min, |
| 1022 |
max, |
| 1023 |
greedy, |
| 1024 |
sub: Box::new(expr), |
| 1025 |
}) |
| 1026 |
} |
| 1027 |
|
| 1028 |
fn hir_unicode_class( |
| 1029 |
&self, |
| 1030 |
ast_class: &ast::ClassUnicode, |
| 1031 |
) -> Result<hir::ClassUnicode> { |
| 1032 |
use crate::ast::ClassUnicodeKind::*; |
| 1033 |
|
| 1034 |
if !self.flags().unicode() { |
| 1035 |
return Err( |
| 1036 |
self.error(ast_class.span, ErrorKind::UnicodeNotAllowed) |
| 1037 |
); |
| 1038 |
} |
| 1039 |
let query = match ast_class.kind { |
| 1040 |
OneLetter(name) => ClassQuery::OneLetter(name), |
| 1041 |
Named(ref name) => ClassQuery::Binary(name), |
| 1042 |
NamedValue { ref name, ref value, .. } => ClassQuery::ByValue { |
| 1043 |
property_name: name, |
| 1044 |
property_value: value, |
| 1045 |
}, |
| 1046 |
}; |
| 1047 |
let mut result = self.convert_unicode_class_error( |
| 1048 |
&ast_class.span, |
| 1049 |
unicode::class(query), |
| 1050 |
); |
| 1051 |
if let Ok(ref mut class) = result { |
| 1052 |
self.unicode_fold_and_negate( |
| 1053 |
&ast_class.span, |
| 1054 |
ast_class.negated, |
| 1055 |
class, |
| 1056 |
)?; |
| 1057 |
} |
| 1058 |
result |
| 1059 |
} |
| 1060 |
|
| 1061 |
fn hir_ascii_unicode_class( |
| 1062 |
&self, |
| 1063 |
ast: &ast::ClassAscii, |
| 1064 |
) -> Result<hir::ClassUnicode> { |
| 1065 |
let mut cls = hir::ClassUnicode::new( |
| 1066 |
ascii_class_as_chars(&ast.kind) |
| 1067 |
.map(|(s, e)| hir::ClassUnicodeRange::new(s, e)), |
| 1068 |
); |
| 1069 |
self.unicode_fold_and_negate(&ast.span, ast.negated, &mut cls)?; |
| 1070 |
Ok(cls) |
| 1071 |
} |
| 1072 |
|
| 1073 |
fn hir_ascii_byte_class( |
| 1074 |
&self, |
| 1075 |
ast: &ast::ClassAscii, |
| 1076 |
) -> Result<hir::ClassBytes> { |
| 1077 |
let mut cls = hir::ClassBytes::new( |
| 1078 |
ascii_class(&ast.kind) |
| 1079 |
.map(|(s, e)| hir::ClassBytesRange::new(s, e)), |
| 1080 |
); |
| 1081 |
self.bytes_fold_and_negate(&ast.span, ast.negated, &mut cls)?; |
| 1082 |
Ok(cls) |
| 1083 |
} |
| 1084 |
|
| 1085 |
fn hir_perl_unicode_class( |
| 1086 |
&self, |
| 1087 |
ast_class: &ast::ClassPerl, |
| 1088 |
) -> Result<hir::ClassUnicode> { |
| 1089 |
use crate::ast::ClassPerlKind::*; |
| 1090 |
|
| 1091 |
assert!(self.flags().unicode()); |
| 1092 |
let result = match ast_class.kind { |
| 1093 |
Digit => unicode::perl_digit(), |
| 1094 |
Space => unicode::perl_space(), |
| 1095 |
Word => unicode::perl_word(), |
| 1096 |
}; |
| 1097 |
let mut class = |
| 1098 |
self.convert_unicode_class_error(&ast_class.span, result)?; |
| 1099 |
// We needn't apply case folding here because the Perl Unicode classes |
| 1100 |
// are already closed under Unicode simple case folding. |
| 1101 |
if ast_class.negated { |
| 1102 |
class.negate(); |
| 1103 |
} |
| 1104 |
Ok(class) |
| 1105 |
} |
| 1106 |
|
| 1107 |
fn hir_perl_byte_class( |
| 1108 |
&self, |
| 1109 |
ast_class: &ast::ClassPerl, |
| 1110 |
) -> Result<hir::ClassBytes> { |
| 1111 |
use crate::ast::ClassPerlKind::*; |
| 1112 |
|
| 1113 |
assert!(!self.flags().unicode()); |
| 1114 |
let mut class = match ast_class.kind { |
| 1115 |
Digit => hir_ascii_class_bytes(&ast::ClassAsciiKind::Digit), |
| 1116 |
Space => hir_ascii_class_bytes(&ast::ClassAsciiKind::Space), |
| 1117 |
Word => hir_ascii_class_bytes(&ast::ClassAsciiKind::Word), |
| 1118 |
}; |
| 1119 |
// We needn't apply case folding here because the Perl ASCII classes |
| 1120 |
// are already closed (under ASCII case folding). |
| 1121 |
if ast_class.negated { |
| 1122 |
class.negate(); |
| 1123 |
} |
| 1124 |
// Negating a Perl byte class is likely to cause it to match invalid |
| 1125 |
// UTF-8. That's only OK if the translator is configured to allow such |
| 1126 |
// things. |
| 1127 |
if self.trans().utf8 && !class.is_ascii() { |
| 1128 |
return Err(self.error(ast_class.span, ErrorKind::InvalidUtf8)); |
| 1129 |
} |
| 1130 |
Ok(class) |
| 1131 |
} |
| 1132 |
|
| 1133 |
/// Converts the given Unicode specific error to an HIR translation error. |
| 1134 |
/// |
| 1135 |
/// The span given should approximate the position at which an error would |
| 1136 |
/// occur. |
| 1137 |
fn convert_unicode_class_error( |
| 1138 |
&self, |
| 1139 |
span: &Span, |
| 1140 |
result: core::result::Result<hir::ClassUnicode, unicode::Error>, |
| 1141 |
) -> Result<hir::ClassUnicode> { |
| 1142 |
result.map_err(|err| { |
| 1143 |
let sp = span.clone(); |
| 1144 |
match err { |
| 1145 |
unicode::Error::PropertyNotFound => { |
| 1146 |
self.error(sp, ErrorKind::UnicodePropertyNotFound) |
| 1147 |
} |
| 1148 |
unicode::Error::PropertyValueNotFound => { |
| 1149 |
self.error(sp, ErrorKind::UnicodePropertyValueNotFound) |
| 1150 |
} |
| 1151 |
unicode::Error::PerlClassNotFound => { |
| 1152 |
self.error(sp, ErrorKind::UnicodePerlClassNotFound) |
| 1153 |
} |
| 1154 |
} |
| 1155 |
}) |
| 1156 |
} |
| 1157 |
|
| 1158 |
fn unicode_fold_and_negate( |
| 1159 |
&self, |
| 1160 |
span: &Span, |
| 1161 |
negated: bool, |
| 1162 |
class: &mut hir::ClassUnicode, |
| 1163 |
) -> Result<()> { |
| 1164 |
// Note that we must apply case folding before negation! |
| 1165 |
// Consider `(?i)[^x]`. If we applied negation first, then |
| 1166 |
// the result would be the character class that matched any |
| 1167 |
// Unicode scalar value. |
| 1168 |
if self.flags().case_insensitive() { |
| 1169 |
class.try_case_fold_simple().map_err(|_| { |
| 1170 |
self.error(span.clone(), ErrorKind::UnicodeCaseUnavailable) |
| 1171 |
})?; |
| 1172 |
} |
| 1173 |
if negated { |
| 1174 |
class.negate(); |
| 1175 |
} |
| 1176 |
Ok(()) |
| 1177 |
} |
| 1178 |
|
| 1179 |
fn bytes_fold_and_negate( |
| 1180 |
&self, |
| 1181 |
span: &Span, |
| 1182 |
negated: bool, |
| 1183 |
class: &mut hir::ClassBytes, |
| 1184 |
) -> Result<()> { |
| 1185 |
// Note that we must apply case folding before negation! |
| 1186 |
// Consider `(?i)[^x]`. If we applied negation first, then |
| 1187 |
// the result would be the character class that matched any |
| 1188 |
// Unicode scalar value. |
| 1189 |
if self.flags().case_insensitive() { |
| 1190 |
class.case_fold_simple(); |
| 1191 |
} |
| 1192 |
if negated { |
| 1193 |
class.negate(); |
| 1194 |
} |
| 1195 |
if self.trans().utf8 && !class.is_ascii() { |
| 1196 |
return Err(self.error(span.clone(), ErrorKind::InvalidUtf8)); |
| 1197 |
} |
| 1198 |
Ok(()) |
| 1199 |
} |
| 1200 |
|
| 1201 |
/// Return a scalar byte value suitable for use as a literal in a byte |
| 1202 |
/// character class. |
| 1203 |
fn class_literal_byte(&self, ast: &ast::Literal) -> Result<u8> { |
| 1204 |
match self.ast_literal_to_scalar(ast)? { |
| 1205 |
Either::Right(byte) => Ok(byte), |
| 1206 |
Either::Left(ch) => { |
| 1207 |
if ch.is_ascii() { |
| 1208 |
Ok(u8::try_from(ch).unwrap()) |
| 1209 |
} else { |
| 1210 |
// We can't feasibly support Unicode in |
| 1211 |
// byte oriented classes. Byte classes don't |
| 1212 |
// do Unicode case folding. |
| 1213 |
Err(self.error(ast.span, ErrorKind::UnicodeNotAllowed)) |
| 1214 |
} |
| 1215 |
} |
| 1216 |
} |
| 1217 |
} |
| 1218 |
} |
| 1219 |
|
| 1220 |
/// A translator's representation of a regular expression's flags at any given |
| 1221 |
/// moment in time. |
| 1222 |
/// |
| 1223 |
/// Each flag can be in one of three states: absent, present but disabled or |
| 1224 |
/// present but enabled. |
| 1225 |
#[derive(Clone, Copy, Debug, Default)] |
| 1226 |
struct Flags { |
| 1227 |
case_insensitive: Option<bool>, |
| 1228 |
multi_line: Option<bool>, |
| 1229 |
dot_matches_new_line: Option<bool>, |
| 1230 |
swap_greed: Option<bool>, |
| 1231 |
unicode: Option<bool>, |
| 1232 |
crlf: Option<bool>, |
| 1233 |
// Note that `ignore_whitespace` is omitted here because it is handled |
| 1234 |
// entirely in the parser. |
| 1235 |
} |
| 1236 |
|
| 1237 |
impl Flags { |
| 1238 |
fn from_ast(ast: &ast::Flags) -> Flags { |
| 1239 |
let mut flags = Flags::default(); |
| 1240 |
let mut enable = true; |
| 1241 |
for item in &ast.items { |
| 1242 |
match item.kind { |
| 1243 |
ast::FlagsItemKind::Negation => { |
| 1244 |
enable = false; |
| 1245 |
} |
| 1246 |
ast::FlagsItemKind::Flag(ast::Flag::CaseInsensitive) => { |
| 1247 |
flags.case_insensitive = Some(enable); |
| 1248 |
} |
| 1249 |
ast::FlagsItemKind::Flag(ast::Flag::MultiLine) => { |
| 1250 |
flags.multi_line = Some(enable); |
| 1251 |
} |
| 1252 |
ast::FlagsItemKind::Flag(ast::Flag::DotMatchesNewLine) => { |
| 1253 |
flags.dot_matches_new_line = Some(enable); |
| 1254 |
} |
| 1255 |
ast::FlagsItemKind::Flag(ast::Flag::SwapGreed) => { |
| 1256 |
flags.swap_greed = Some(enable); |
| 1257 |
} |
| 1258 |
ast::FlagsItemKind::Flag(ast::Flag::Unicode) => { |
| 1259 |
flags.unicode = Some(enable); |
| 1260 |
} |
| 1261 |
ast::FlagsItemKind::Flag(ast::Flag::CRLF) => { |
| 1262 |
flags.crlf = Some(enable); |
| 1263 |
} |
| 1264 |
ast::FlagsItemKind::Flag(ast::Flag::IgnoreWhitespace) => {} |
| 1265 |
} |
| 1266 |
} |
| 1267 |
flags |
| 1268 |
} |
| 1269 |
|
| 1270 |
fn merge(&mut self, previous: &Flags) { |
| 1271 |
if self.case_insensitive.is_none() { |
| 1272 |
self.case_insensitive = previous.case_insensitive; |
| 1273 |
} |
| 1274 |
if self.multi_line.is_none() { |
| 1275 |
self.multi_line = previous.multi_line; |
| 1276 |
} |
| 1277 |
if self.dot_matches_new_line.is_none() { |
| 1278 |
self.dot_matches_new_line = previous.dot_matches_new_line; |
| 1279 |
} |
| 1280 |
if self.swap_greed.is_none() { |
| 1281 |
self.swap_greed = previous.swap_greed; |
| 1282 |
} |
| 1283 |
if self.unicode.is_none() { |
| 1284 |
self.unicode = previous.unicode; |
| 1285 |
} |
| 1286 |
if self.crlf.is_none() { |
| 1287 |
self.crlf = previous.crlf; |
| 1288 |
} |
| 1289 |
} |
| 1290 |
|
| 1291 |
fn case_insensitive(&self) -> bool { |
| 1292 |
self.case_insensitive.unwrap_or(false) |
| 1293 |
} |
| 1294 |
|
| 1295 |
fn multi_line(&self) -> bool { |
| 1296 |
self.multi_line.unwrap_or(false) |
| 1297 |
} |
| 1298 |
|
| 1299 |
fn dot_matches_new_line(&self) -> bool { |
| 1300 |
self.dot_matches_new_line.unwrap_or(false) |
| 1301 |
} |
| 1302 |
|
| 1303 |
fn swap_greed(&self) -> bool { |
| 1304 |
self.swap_greed.unwrap_or(false) |
| 1305 |
} |
| 1306 |
|
| 1307 |
fn unicode(&self) -> bool { |
| 1308 |
self.unicode.unwrap_or(true) |
| 1309 |
} |
| 1310 |
|
| 1311 |
fn crlf(&self) -> bool { |
| 1312 |
self.crlf.unwrap_or(false) |
| 1313 |
} |
| 1314 |
} |
| 1315 |
|
| 1316 |
fn hir_ascii_class_bytes(kind: &ast::ClassAsciiKind) -> hir::ClassBytes { |
| 1317 |
let ranges: Vec<_> = ascii_class(kind) |
| 1318 |
.map(|(s, e)| hir::ClassBytesRange::new(s, e)) |
| 1319 |
.collect(); |
| 1320 |
hir::ClassBytes::new(ranges) |
| 1321 |
} |
| 1322 |
|
| 1323 |
fn ascii_class(kind: &ast::ClassAsciiKind) -> impl Iterator<Item = (u8, u8)> { |
| 1324 |
use crate::ast::ClassAsciiKind::*; |
| 1325 |
|
| 1326 |
let slice: &'static [(u8, u8)] = match *kind { |
| 1327 |
Alnum => &[(b'0', b'9'), (b'A', b'Z'), (b'a', b'z')], |
| 1328 |
Alpha => &[(b'A', b'Z'), (b'a', b'z')], |
| 1329 |
Ascii => &[(b'\x00', b'\x7F')], |
| 1330 |
Blank => &[(b'\t', b'\t'), (b' ', b' ')], |
| 1331 |
Cntrl => &[(b'\x00', b'\x1F'), (b'\x7F', b'\x7F')], |
| 1332 |
Digit => &[(b'0', b'9')], |
| 1333 |
Graph => &[(b'!', b'~')], |
| 1334 |
Lower => &[(b'a', b'z')], |
| 1335 |
Print => &[(b' ', b'~')], |
| 1336 |
Punct => &[(b'!', b'/'), (b':', b'@'), (b'[', b'`'), (b'{', b'~')], |
| 1337 |
Space => &[ |
| 1338 |
(b'\t', b'\t'), |
| 1339 |
(b'\n', b'\n'), |
| 1340 |
(b'\x0B', b'\x0B'), |
| 1341 |
(b'\x0C', b'\x0C'), |
| 1342 |
(b'\r', b'\r'), |
| 1343 |
(b' ', b' '), |
| 1344 |
], |
| 1345 |
Upper => &[(b'A', b'Z')], |
| 1346 |
Word => &[(b'0', b'9'), (b'A', b'Z'), (b'_', b'_'), (b'a', b'z')], |
| 1347 |
Xdigit => &[(b'0', b'9'), (b'A', b'F'), (b'a', b'f')], |
| 1348 |
}; |
| 1349 |
slice.iter().copied() |
| 1350 |
} |
| 1351 |
|
| 1352 |
fn ascii_class_as_chars( |
| 1353 |
kind: &ast::ClassAsciiKind, |
| 1354 |
) -> impl Iterator<Item = (char, char)> { |
| 1355 |
ascii_class(kind).map(|(s, e)| (char::from(s), char::from(e))) |
| 1356 |
} |
| 1357 |
|
| 1358 |
#[cfg(test)] |
| 1359 |
mod tests { |
| 1360 |
use crate::{ |
| 1361 |
ast::{parse::ParserBuilder, Position}, |
| 1362 |
hir::{Look, Properties}, |
| 1363 |
}; |
| 1364 |
|
| 1365 |
use super::*; |
| 1366 |
|
| 1367 |
// We create these errors to compare with real hir::Errors in the tests. |
| 1368 |
// We define equality between TestError and hir::Error to disregard the |
| 1369 |
// pattern string in hir::Error, which is annoying to provide in tests. |
| 1370 |
#[derive(Clone, Debug)] |
| 1371 |
struct TestError { |
| 1372 |
span: Span, |
| 1373 |
kind: hir::ErrorKind, |
| 1374 |
} |
| 1375 |
|
| 1376 |
impl PartialEq<hir::Error> for TestError { |
| 1377 |
fn eq(&self, other: &hir::Error) -> bool { |
| 1378 |
self.span == other.span && self.kind == other.kind |
| 1379 |
} |
| 1380 |
} |
| 1381 |
|
| 1382 |
impl PartialEq<TestError> for hir::Error { |
| 1383 |
fn eq(&self, other: &TestError) -> bool { |
| 1384 |
self.span == other.span && self.kind == other.kind |
| 1385 |
} |
| 1386 |
} |
| 1387 |
|
| 1388 |
fn parse(pattern: &str) -> Ast { |
| 1389 |
ParserBuilder::new().octal(true).build().parse(pattern).unwrap() |
| 1390 |
} |
| 1391 |
|
| 1392 |
fn t(pattern: &str) -> Hir { |
| 1393 |
TranslatorBuilder::new() |
| 1394 |
.utf8(true) |
| 1395 |
.build() |
| 1396 |
.translate(pattern, &parse(pattern)) |
| 1397 |
.unwrap() |
| 1398 |
} |
| 1399 |
|
| 1400 |
fn t_err(pattern: &str) -> hir::Error { |
| 1401 |
TranslatorBuilder::new() |
| 1402 |
.utf8(true) |
| 1403 |
.build() |
| 1404 |
.translate(pattern, &parse(pattern)) |
| 1405 |
.unwrap_err() |
| 1406 |
} |
| 1407 |
|
| 1408 |
fn t_bytes(pattern: &str) -> Hir { |
| 1409 |
TranslatorBuilder::new() |
| 1410 |
.utf8(false) |
| 1411 |
.build() |
| 1412 |
.translate(pattern, &parse(pattern)) |
| 1413 |
.unwrap() |
| 1414 |
} |
| 1415 |
|
| 1416 |
fn props(pattern: &str) -> Properties { |
| 1417 |
t(pattern).properties().clone() |
| 1418 |
} |
| 1419 |
|
| 1420 |
fn props_bytes(pattern: &str) -> Properties { |
| 1421 |
t_bytes(pattern).properties().clone() |
| 1422 |
} |
| 1423 |
|
| 1424 |
fn hir_lit(s: &str) -> Hir { |
| 1425 |
hir_blit(s.as_bytes()) |
| 1426 |
} |
| 1427 |
|
| 1428 |
fn hir_blit(s: &[u8]) -> Hir { |
| 1429 |
Hir::literal(s) |
| 1430 |
} |
| 1431 |
|
| 1432 |
fn hir_capture(index: u32, expr: Hir) -> Hir { |
| 1433 |
Hir::capture(hir::Capture { index, name: None, sub: Box::new(expr) }) |
| 1434 |
} |
| 1435 |
|
| 1436 |
fn hir_capture_name(index: u32, name: &str, expr: Hir) -> Hir { |
| 1437 |
Hir::capture(hir::Capture { |
| 1438 |
index, |
| 1439 |
name: Some(name.into()), |
| 1440 |
sub: Box::new(expr), |
| 1441 |
}) |
| 1442 |
} |
| 1443 |
|
| 1444 |
fn hir_quest(greedy: bool, expr: Hir) -> Hir { |
| 1445 |
Hir::repetition(hir::Repetition { |
| 1446 |
min: 0, |
| 1447 |
max: Some(1), |
| 1448 |
greedy, |
| 1449 |
sub: Box::new(expr), |
| 1450 |
}) |
| 1451 |
} |
| 1452 |
|
| 1453 |
fn hir_star(greedy: bool, expr: Hir) -> Hir { |
| 1454 |
Hir::repetition(hir::Repetition { |
| 1455 |
min: 0, |
| 1456 |
max: None, |
| 1457 |
greedy, |
| 1458 |
sub: Box::new(expr), |
| 1459 |
}) |
| 1460 |
} |
| 1461 |
|
| 1462 |
fn hir_plus(greedy: bool, expr: Hir) -> Hir { |
| 1463 |
Hir::repetition(hir::Repetition { |
| 1464 |
min: 1, |
| 1465 |
max: None, |
| 1466 |
greedy, |
| 1467 |
sub: Box::new(expr), |
| 1468 |
}) |
| 1469 |
} |
| 1470 |
|
| 1471 |
fn hir_range(greedy: bool, min: u32, max: Option<u32>, expr: Hir) -> Hir { |
| 1472 |
Hir::repetition(hir::Repetition { |
| 1473 |
min, |
| 1474 |
max, |
| 1475 |
greedy, |
| 1476 |
sub: Box::new(expr), |
| 1477 |
}) |
| 1478 |
} |
| 1479 |
|
| 1480 |
fn hir_alt(alts: Vec<Hir>) -> Hir { |
| 1481 |
Hir::alternation(alts) |
| 1482 |
} |
| 1483 |
|
| 1484 |
fn hir_cat(exprs: Vec<Hir>) -> Hir { |
| 1485 |
Hir::concat(exprs) |
| 1486 |
} |
| 1487 |
|
| 1488 |
#[allow(dead_code)] |
| 1489 |
fn hir_uclass_query(query: ClassQuery<'_>) -> Hir { |
| 1490 |
Hir::class(hir::Class::Unicode(unicode::class(query).unwrap())) |
| 1491 |
} |
| 1492 |
|
| 1493 |
#[allow(dead_code)] |
| 1494 |
fn hir_uclass_perl_word() -> Hir { |
| 1495 |
Hir::class(hir::Class::Unicode(unicode::perl_word().unwrap())) |
| 1496 |
} |
| 1497 |
|
| 1498 |
fn hir_ascii_uclass(kind: &ast::ClassAsciiKind) -> Hir { |
| 1499 |
Hir::class(hir::Class::Unicode(hir::ClassUnicode::new( |
| 1500 |
ascii_class_as_chars(kind) |
| 1501 |
.map(|(s, e)| hir::ClassUnicodeRange::new(s, e)), |
| 1502 |
))) |
| 1503 |
} |
| 1504 |
|
| 1505 |
fn hir_ascii_bclass(kind: &ast::ClassAsciiKind) -> Hir { |
| 1506 |
Hir::class(hir::Class::Bytes(hir::ClassBytes::new( |
| 1507 |
ascii_class(kind).map(|(s, e)| hir::ClassBytesRange::new(s, e)), |
| 1508 |
))) |
| 1509 |
} |
| 1510 |
|
| 1511 |
fn hir_uclass(ranges: &[(char, char)]) -> Hir { |
| 1512 |
Hir::class(uclass(ranges)) |
| 1513 |
} |
| 1514 |
|
| 1515 |
fn hir_bclass(ranges: &[(u8, u8)]) -> Hir { |
| 1516 |
Hir::class(bclass(ranges)) |
| 1517 |
} |
| 1518 |
|
| 1519 |
fn hir_case_fold(expr: Hir) -> Hir { |
| 1520 |
match expr.into_kind() { |
| 1521 |
HirKind::Class(mut cls) => { |
| 1522 |
cls.case_fold_simple(); |
| 1523 |
Hir::class(cls) |
| 1524 |
} |
| 1525 |
_ => panic!("cannot case fold non-class Hir expr"), |
| 1526 |
} |
| 1527 |
} |
| 1528 |
|
| 1529 |
fn hir_negate(expr: Hir) -> Hir { |
| 1530 |
match expr.into_kind() { |
| 1531 |
HirKind::Class(mut cls) => { |
| 1532 |
cls.negate(); |
| 1533 |
Hir::class(cls) |
| 1534 |
} |
| 1535 |
_ => panic!("cannot negate non-class Hir expr"), |
| 1536 |
} |
| 1537 |
} |
| 1538 |
|
| 1539 |
fn uclass(ranges: &[(char, char)]) -> hir::Class { |
| 1540 |
let ranges: Vec<hir::ClassUnicodeRange> = ranges |
| 1541 |
.iter() |
| 1542 |
.map(|&(s, e)| hir::ClassUnicodeRange::new(s, e)) |
| 1543 |
.collect(); |
| 1544 |
hir::Class::Unicode(hir::ClassUnicode::new(ranges)) |
| 1545 |
} |
| 1546 |
|
| 1547 |
fn bclass(ranges: &[(u8, u8)]) -> hir::Class { |
| 1548 |
let ranges: Vec<hir::ClassBytesRange> = ranges |
| 1549 |
.iter() |
| 1550 |
.map(|&(s, e)| hir::ClassBytesRange::new(s, e)) |
| 1551 |
.collect(); |
| 1552 |
hir::Class::Bytes(hir::ClassBytes::new(ranges)) |
| 1553 |
} |
| 1554 |
|
| 1555 |
#[cfg(feature = "unicode-case")] |
| 1556 |
fn class_case_fold(mut cls: hir::Class) -> Hir { |
| 1557 |
cls.case_fold_simple(); |
| 1558 |
Hir::class(cls) |
| 1559 |
} |
| 1560 |
|
| 1561 |
fn class_negate(mut cls: hir::Class) -> Hir { |
| 1562 |
cls.negate(); |
| 1563 |
Hir::class(cls) |
| 1564 |
} |
| 1565 |
|
| 1566 |
#[allow(dead_code)] |
| 1567 |
fn hir_union(expr1: Hir, expr2: Hir) -> Hir { |
| 1568 |
use crate::hir::Class::{Bytes, Unicode}; |
| 1569 |
|
| 1570 |
match (expr1.into_kind(), expr2.into_kind()) { |
| 1571 |
(HirKind::Class(Unicode(mut c1)), HirKind::Class(Unicode(c2))) => { |
| 1572 |
c1.union(&c2); |
| 1573 |
Hir::class(hir::Class::Unicode(c1)) |
| 1574 |
} |
| 1575 |
(HirKind::Class(Bytes(mut c1)), HirKind::Class(Bytes(c2))) => { |
| 1576 |
c1.union(&c2); |
| 1577 |
Hir::class(hir::Class::Bytes(c1)) |
| 1578 |
} |
| 1579 |
_ => panic!("cannot union non-class Hir exprs"), |
| 1580 |
} |
| 1581 |
} |
| 1582 |
|
| 1583 |
#[allow(dead_code)] |
| 1584 |
fn hir_difference(expr1: Hir, expr2: Hir) -> Hir { |
| 1585 |
use crate::hir::Class::{Bytes, Unicode}; |
| 1586 |
|
| 1587 |
match (expr1.into_kind(), expr2.into_kind()) { |
| 1588 |
(HirKind::Class(Unicode(mut c1)), HirKind::Class(Unicode(c2))) => { |
| 1589 |
c1.difference(&c2); |
| 1590 |
Hir::class(hir::Class::Unicode(c1)) |
| 1591 |
} |
| 1592 |
(HirKind::Class(Bytes(mut c1)), HirKind::Class(Bytes(c2))) => { |
| 1593 |
c1.difference(&c2); |
| 1594 |
Hir::class(hir::Class::Bytes(c1)) |
| 1595 |
} |
| 1596 |
_ => panic!("cannot difference non-class Hir exprs"), |
| 1597 |
} |
| 1598 |
} |
| 1599 |
|
| 1600 |
fn hir_look(look: hir::Look) -> Hir { |
| 1601 |
Hir::look(look) |
| 1602 |
} |
| 1603 |
|
| 1604 |
#[test] |
| 1605 |
fn empty() { |
| 1606 |
assert_eq!(t(""), Hir::empty()); |
| 1607 |
assert_eq!(t("(?i)"), Hir::empty()); |
| 1608 |
assert_eq!(t("()"), hir_capture(1, Hir::empty())); |
| 1609 |
assert_eq!(t("(?:)"), Hir::empty()); |
| 1610 |
assert_eq!(t("(?P<wat>)"), hir_capture_name(1, "wat", Hir::empty())); |
| 1611 |
assert_eq!(t("|"), hir_alt(vec![Hir::empty(), Hir::empty()])); |
| 1612 |
assert_eq!( |
| 1613 |
t("()|()"), |
| 1614 |
hir_alt(vec![ |
| 1615 |
hir_capture(1, Hir::empty()), |
| 1616 |
hir_capture(2, Hir::empty()), |
| 1617 |
]) |
| 1618 |
); |
| 1619 |
assert_eq!( |
| 1620 |
t("(|b)"), |
| 1621 |
hir_capture(1, hir_alt(vec![Hir::empty(), hir_lit("b"),])) |
| 1622 |
); |
| 1623 |
assert_eq!( |
| 1624 |
t("(a|)"), |
| 1625 |
hir_capture(1, hir_alt(vec![hir_lit("a"), Hir::empty(),])) |
| 1626 |
); |
| 1627 |
assert_eq!( |
| 1628 |
t("(a||c)"), |
| 1629 |
hir_capture( |
| 1630 |
1, |
| 1631 |
hir_alt(vec![hir_lit("a"), Hir::empty(), hir_lit("c"),]) |
| 1632 |
) |
| 1633 |
); |
| 1634 |
assert_eq!( |
| 1635 |
t("(||)"), |
| 1636 |
hir_capture( |
| 1637 |
1, |
| 1638 |
hir_alt(vec![Hir::empty(), Hir::empty(), Hir::empty(),]) |
| 1639 |
) |
| 1640 |
); |
| 1641 |
} |
| 1642 |
|
| 1643 |
#[test] |
| 1644 |
fn literal() { |
| 1645 |
assert_eq!(t("a"), hir_lit("a")); |
| 1646 |
assert_eq!(t("(?-u)a"), hir_lit("a")); |
| 1647 |
assert_eq!(t("☃"), hir_lit("☃")); |
| 1648 |
assert_eq!(t("abcd"), hir_lit("abcd")); |
| 1649 |
|
| 1650 |
assert_eq!(t_bytes("(?-u)a"), hir_lit("a")); |
| 1651 |
assert_eq!(t_bytes("(?-u)\x61"), hir_lit("a")); |
| 1652 |
assert_eq!(t_bytes(r"(?-u)\x61"), hir_lit("a")); |
| 1653 |
assert_eq!(t_bytes(r"(?-u)\xFF"), hir_blit(b"\xFF")); |
| 1654 |
|
| 1655 |
assert_eq!(t("(?-u)☃"), hir_lit("☃")); |
| 1656 |
assert_eq!( |
| 1657 |
t_err(r"(?-u)\xFF"), |
| 1658 |
TestError { |
| 1659 |
kind: hir::ErrorKind::InvalidUtf8, |
| 1660 |
span: Span::new( |
| 1661 |
Position::new(5, 1, 6), |
| 1662 |
Position::new(9, 1, 10) |
| 1663 |
), |
| 1664 |
} |
| 1665 |
); |
| 1666 |
} |
| 1667 |
|
| 1668 |
#[test] |
| 1669 |
fn literal_case_insensitive() { |
| 1670 |
#[cfg(feature = "unicode-case")] |
| 1671 |
assert_eq!(t("(?i)a"), hir_uclass(&[('A', 'A'), ('a', 'a'),])); |
| 1672 |
#[cfg(feature = "unicode-case")] |
| 1673 |
assert_eq!(t("(?i:a)"), hir_uclass(&[('A', 'A'), ('a', 'a')])); |
| 1674 |
#[cfg(feature = "unicode-case")] |
| 1675 |
assert_eq!( |
| 1676 |
t("a(?i)a(?-i)a"), |
| 1677 |
hir_cat(vec![ |
| 1678 |
hir_lit("a"), |
| 1679 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1680 |
hir_lit("a"), |
| 1681 |
]) |
| 1682 |
); |
| 1683 |
#[cfg(feature = "unicode-case")] |
| 1684 |
assert_eq!( |
| 1685 |
t("(?i)ab@c"), |
| 1686 |
hir_cat(vec![ |
| 1687 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1688 |
hir_uclass(&[('B', 'B'), ('b', 'b')]), |
| 1689 |
hir_lit("@"), |
| 1690 |
hir_uclass(&[('C', 'C'), ('c', 'c')]), |
| 1691 |
]) |
| 1692 |
); |
| 1693 |
#[cfg(feature = "unicode-case")] |
| 1694 |
assert_eq!( |
| 1695 |
t("(?i)β"), |
| 1696 |
hir_uclass(&[('Β', 'Β'), ('β', 'β'), ('ϐ', 'ϐ'),]) |
| 1697 |
); |
| 1698 |
|
| 1699 |
assert_eq!(t("(?i-u)a"), hir_bclass(&[(b'A', b'A'), (b'a', b'a'),])); |
| 1700 |
#[cfg(feature = "unicode-case")] |
| 1701 |
assert_eq!( |
| 1702 |
t("(?-u)a(?i)a(?-i)a"), |
| 1703 |
hir_cat(vec![ |
| 1704 |
hir_lit("a"), |
| 1705 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a')]), |
| 1706 |
hir_lit("a"), |
| 1707 |
]) |
| 1708 |
); |
| 1709 |
assert_eq!( |
| 1710 |
t("(?i-u)ab@c"), |
| 1711 |
hir_cat(vec![ |
| 1712 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a')]), |
| 1713 |
hir_bclass(&[(b'B', b'B'), (b'b', b'b')]), |
| 1714 |
hir_lit("@"), |
| 1715 |
hir_bclass(&[(b'C', b'C'), (b'c', b'c')]), |
| 1716 |
]) |
| 1717 |
); |
| 1718 |
|
| 1719 |
assert_eq!( |
| 1720 |
t_bytes("(?i-u)a"), |
| 1721 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a'),]) |
| 1722 |
); |
| 1723 |
assert_eq!( |
| 1724 |
t_bytes("(?i-u)\x61"), |
| 1725 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a'),]) |
| 1726 |
); |
| 1727 |
assert_eq!( |
| 1728 |
t_bytes(r"(?i-u)\x61"), |
| 1729 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a'),]) |
| 1730 |
); |
| 1731 |
assert_eq!(t_bytes(r"(?i-u)\xFF"), hir_blit(b"\xFF")); |
| 1732 |
|
| 1733 |
assert_eq!(t("(?i-u)β"), hir_lit("β"),); |
| 1734 |
} |
| 1735 |
|
| 1736 |
#[test] |
| 1737 |
fn dot() { |
| 1738 |
assert_eq!( |
| 1739 |
t("."), |
| 1740 |
hir_uclass(&[('\0', '\t'), ('\x0B', '\u{10FFFF}')]) |
| 1741 |
); |
| 1742 |
assert_eq!( |
| 1743 |
t("(?R)."), |
| 1744 |
hir_uclass(&[ |
| 1745 |
('\0', '\t'), |
| 1746 |
('\x0B', '\x0C'), |
| 1747 |
('\x0E', '\u{10FFFF}'), |
| 1748 |
]) |
| 1749 |
); |
| 1750 |
assert_eq!(t("(?s)."), hir_uclass(&[('\0', '\u{10FFFF}')])); |
| 1751 |
assert_eq!(t("(?Rs)."), hir_uclass(&[('\0', '\u{10FFFF}')])); |
| 1752 |
assert_eq!( |
| 1753 |
t_bytes("(?-u)."), |
| 1754 |
hir_bclass(&[(b'\0', b'\t'), (b'\x0B', b'\xFF')]) |
| 1755 |
); |
| 1756 |
assert_eq!( |
| 1757 |
t_bytes("(?R-u)."), |
| 1758 |
hir_bclass(&[ |
| 1759 |
(b'\0', b'\t'), |
| 1760 |
(b'\x0B', b'\x0C'), |
| 1761 |
(b'\x0E', b'\xFF'), |
| 1762 |
]) |
| 1763 |
); |
| 1764 |
assert_eq!(t_bytes("(?s-u)."), hir_bclass(&[(b'\0', b'\xFF'),])); |
| 1765 |
assert_eq!(t_bytes("(?Rs-u)."), hir_bclass(&[(b'\0', b'\xFF'),])); |
| 1766 |
|
| 1767 |
// If invalid UTF-8 isn't allowed, then non-Unicode `.` isn't allowed. |
| 1768 |
assert_eq!( |
| 1769 |
t_err("(?-u)."), |
| 1770 |
TestError { |
| 1771 |
kind: hir::ErrorKind::InvalidUtf8, |
| 1772 |
span: Span::new( |
| 1773 |
Position::new(5, 1, 6), |
| 1774 |
Position::new(6, 1, 7) |
| 1775 |
), |
| 1776 |
} |
| 1777 |
); |
| 1778 |
assert_eq!( |
| 1779 |
t_err("(?R-u)."), |
| 1780 |
TestError { |
| 1781 |
kind: hir::ErrorKind::InvalidUtf8, |
| 1782 |
span: Span::new( |
| 1783 |
Position::new(6, 1, 7), |
| 1784 |
Position::new(7, 1, 8) |
| 1785 |
), |
| 1786 |
} |
| 1787 |
); |
| 1788 |
assert_eq!( |
| 1789 |
t_err("(?s-u)."), |
| 1790 |
TestError { |
| 1791 |
kind: hir::ErrorKind::InvalidUtf8, |
| 1792 |
span: Span::new( |
| 1793 |
Position::new(6, 1, 7), |
| 1794 |
Position::new(7, 1, 8) |
| 1795 |
), |
| 1796 |
} |
| 1797 |
); |
| 1798 |
assert_eq!( |
| 1799 |
t_err("(?Rs-u)."), |
| 1800 |
TestError { |
| 1801 |
kind: hir::ErrorKind::InvalidUtf8, |
| 1802 |
span: Span::new( |
| 1803 |
Position::new(7, 1, 8), |
| 1804 |
Position::new(8, 1, 9) |
| 1805 |
), |
| 1806 |
} |
| 1807 |
); |
| 1808 |
} |
| 1809 |
|
| 1810 |
#[test] |
| 1811 |
fn assertions() { |
| 1812 |
assert_eq!(t("^"), hir_look(hir::Look::Start)); |
| 1813 |
assert_eq!(t("$"), hir_look(hir::Look::End)); |
| 1814 |
assert_eq!(t(r"\A"), hir_look(hir::Look::Start)); |
| 1815 |
assert_eq!(t(r"\z"), hir_look(hir::Look::End)); |
| 1816 |
assert_eq!(t("(?m)^"), hir_look(hir::Look::StartLF)); |
| 1817 |
assert_eq!(t("(?m)$"), hir_look(hir::Look::EndLF)); |
| 1818 |
assert_eq!(t(r"(?m)\A"), hir_look(hir::Look::Start)); |
| 1819 |
assert_eq!(t(r"(?m)\z"), hir_look(hir::Look::End)); |
| 1820 |
|
| 1821 |
assert_eq!(t(r"\b"), hir_look(hir::Look::WordUnicode)); |
| 1822 |
assert_eq!(t(r"\B"), hir_look(hir::Look::WordUnicodeNegate)); |
| 1823 |
assert_eq!(t(r"(?-u)\b"), hir_look(hir::Look::WordAscii)); |
| 1824 |
assert_eq!(t(r"(?-u)\B"), hir_look(hir::Look::WordAsciiNegate)); |
| 1825 |
} |
| 1826 |
|
| 1827 |
#[test] |
| 1828 |
fn group() { |
| 1829 |
assert_eq!(t("(a)"), hir_capture(1, hir_lit("a"))); |
| 1830 |
assert_eq!( |
| 1831 |
t("(a)(b)"), |
| 1832 |
hir_cat(vec![ |
| 1833 |
hir_capture(1, hir_lit("a")), |
| 1834 |
hir_capture(2, hir_lit("b")), |
| 1835 |
]) |
| 1836 |
); |
| 1837 |
assert_eq!( |
| 1838 |
t("(a)|(b)"), |
| 1839 |
hir_alt(vec![ |
| 1840 |
hir_capture(1, hir_lit("a")), |
| 1841 |
hir_capture(2, hir_lit("b")), |
| 1842 |
]) |
| 1843 |
); |
| 1844 |
assert_eq!(t("(?P<foo>)"), hir_capture_name(1, "foo", Hir::empty())); |
| 1845 |
assert_eq!(t("(?P<foo>a)"), hir_capture_name(1, "foo", hir_lit("a"))); |
| 1846 |
assert_eq!( |
| 1847 |
t("(?P<foo>a)(?P<bar>b)"), |
| 1848 |
hir_cat(vec![ |
| 1849 |
hir_capture_name(1, "foo", hir_lit("a")), |
| 1850 |
hir_capture_name(2, "bar", hir_lit("b")), |
| 1851 |
]) |
| 1852 |
); |
| 1853 |
assert_eq!(t("(?:)"), Hir::empty()); |
| 1854 |
assert_eq!(t("(?:a)"), hir_lit("a")); |
| 1855 |
assert_eq!( |
| 1856 |
t("(?:a)(b)"), |
| 1857 |
hir_cat(vec![hir_lit("a"), hir_capture(1, hir_lit("b")),]) |
| 1858 |
); |
| 1859 |
assert_eq!( |
| 1860 |
t("(a)(?:b)(c)"), |
| 1861 |
hir_cat(vec![ |
| 1862 |
hir_capture(1, hir_lit("a")), |
| 1863 |
hir_lit("b"), |
| 1864 |
hir_capture(2, hir_lit("c")), |
| 1865 |
]) |
| 1866 |
); |
| 1867 |
assert_eq!( |
| 1868 |
t("(a)(?P<foo>b)(c)"), |
| 1869 |
hir_cat(vec![ |
| 1870 |
hir_capture(1, hir_lit("a")), |
| 1871 |
hir_capture_name(2, "foo", hir_lit("b")), |
| 1872 |
hir_capture(3, hir_lit("c")), |
| 1873 |
]) |
| 1874 |
); |
| 1875 |
assert_eq!(t("()"), hir_capture(1, Hir::empty())); |
| 1876 |
assert_eq!(t("((?i))"), hir_capture(1, Hir::empty())); |
| 1877 |
assert_eq!(t("((?x))"), hir_capture(1, Hir::empty())); |
| 1878 |
assert_eq!( |
| 1879 |
t("(((?x)))"), |
| 1880 |
hir_capture(1, hir_capture(2, Hir::empty())) |
| 1881 |
); |
| 1882 |
} |
| 1883 |
|
| 1884 |
#[test] |
| 1885 |
fn line_anchors() { |
| 1886 |
assert_eq!(t("^"), hir_look(hir::Look::Start)); |
| 1887 |
assert_eq!(t("$"), hir_look(hir::Look::End)); |
| 1888 |
assert_eq!(t(r"\A"), hir_look(hir::Look::Start)); |
| 1889 |
assert_eq!(t(r"\z"), hir_look(hir::Look::End)); |
| 1890 |
|
| 1891 |
assert_eq!(t(r"(?m)\A"), hir_look(hir::Look::Start)); |
| 1892 |
assert_eq!(t(r"(?m)\z"), hir_look(hir::Look::End)); |
| 1893 |
assert_eq!(t("(?m)^"), hir_look(hir::Look::StartLF)); |
| 1894 |
assert_eq!(t("(?m)$"), hir_look(hir::Look::EndLF)); |
| 1895 |
|
| 1896 |
assert_eq!(t(r"(?R)\A"), hir_look(hir::Look::Start)); |
| 1897 |
assert_eq!(t(r"(?R)\z"), hir_look(hir::Look::End)); |
| 1898 |
assert_eq!(t("(?R)^"), hir_look(hir::Look::Start)); |
| 1899 |
assert_eq!(t("(?R)$"), hir_look(hir::Look::End)); |
| 1900 |
|
| 1901 |
assert_eq!(t(r"(?Rm)\A"), hir_look(hir::Look::Start)); |
| 1902 |
assert_eq!(t(r"(?Rm)\z"), hir_look(hir::Look::End)); |
| 1903 |
assert_eq!(t("(?Rm)^"), hir_look(hir::Look::StartCRLF)); |
| 1904 |
assert_eq!(t("(?Rm)$"), hir_look(hir::Look::EndCRLF)); |
| 1905 |
} |
| 1906 |
|
| 1907 |
#[test] |
| 1908 |
fn flags() { |
| 1909 |
#[cfg(feature = "unicode-case")] |
| 1910 |
assert_eq!( |
| 1911 |
t("(?i:a)a"), |
| 1912 |
hir_cat( |
| 1913 |
vec![hir_uclass(&[('A', 'A'), ('a', 'a')]), hir_lit("a"),] |
| 1914 |
) |
| 1915 |
); |
| 1916 |
assert_eq!( |
| 1917 |
t("(?i-u:a)β"), |
| 1918 |
hir_cat(vec![ |
| 1919 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a')]), |
| 1920 |
hir_lit("β"), |
| 1921 |
]) |
| 1922 |
); |
| 1923 |
assert_eq!( |
| 1924 |
t("(?:(?i-u)a)b"), |
| 1925 |
hir_cat(vec![ |
| 1926 |
hir_bclass(&[(b'A', b'A'), (b'a', b'a')]), |
| 1927 |
hir_lit("b"), |
| 1928 |
]) |
| 1929 |
); |
| 1930 |
assert_eq!( |
| 1931 |
t("((?i-u)a)b"), |
| 1932 |
hir_cat(vec![ |
| 1933 |
hir_capture(1, hir_bclass(&[(b'A', b'A'), (b'a', b'a')])), |
| 1934 |
hir_lit("b"), |
| 1935 |
]) |
| 1936 |
); |
| 1937 |
#[cfg(feature = "unicode-case")] |
| 1938 |
assert_eq!( |
| 1939 |
t("(?i)(?-i:a)a"), |
| 1940 |
hir_cat( |
| 1941 |
vec![hir_lit("a"), hir_uclass(&[('A', 'A'), ('a', 'a')]),] |
| 1942 |
) |
| 1943 |
); |
| 1944 |
#[cfg(feature = "unicode-case")] |
| 1945 |
assert_eq!( |
| 1946 |
t("(?im)a^"), |
| 1947 |
hir_cat(vec![ |
| 1948 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1949 |
hir_look(hir::Look::StartLF), |
| 1950 |
]) |
| 1951 |
); |
| 1952 |
#[cfg(feature = "unicode-case")] |
| 1953 |
assert_eq!( |
| 1954 |
t("(?im)a^(?i-m)a^"), |
| 1955 |
hir_cat(vec![ |
| 1956 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1957 |
hir_look(hir::Look::StartLF), |
| 1958 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1959 |
hir_look(hir::Look::Start), |
| 1960 |
]) |
| 1961 |
); |
| 1962 |
assert_eq!( |
| 1963 |
t("(?U)a*a*?(?-U)a*a*?"), |
| 1964 |
hir_cat(vec![ |
| 1965 |
hir_star(false, hir_lit("a")), |
| 1966 |
hir_star(true, hir_lit("a")), |
| 1967 |
hir_star(true, hir_lit("a")), |
| 1968 |
hir_star(false, hir_lit("a")), |
| 1969 |
]) |
| 1970 |
); |
| 1971 |
#[cfg(feature = "unicode-case")] |
| 1972 |
assert_eq!( |
| 1973 |
t("(?:a(?i)a)a"), |
| 1974 |
hir_cat(vec![ |
| 1975 |
hir_cat(vec![ |
| 1976 |
hir_lit("a"), |
| 1977 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1978 |
]), |
| 1979 |
hir_lit("a"), |
| 1980 |
]) |
| 1981 |
); |
| 1982 |
#[cfg(feature = "unicode-case")] |
| 1983 |
assert_eq!( |
| 1984 |
t("(?i)(?:a(?-i)a)a"), |
| 1985 |
hir_cat(vec![ |
| 1986 |
hir_cat(vec![ |
| 1987 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1988 |
hir_lit("a"), |
| 1989 |
]), |
| 1990 |
hir_uclass(&[('A', 'A'), ('a', 'a')]), |
| 1991 |
]) |
| 1992 |
); |
| 1993 |
} |
| 1994 |
|
| 1995 |
#[test] |
| 1996 |
fn escape() { |
| 1997 |
assert_eq!( |
| 1998 |
t(r"\\\.\+\*\?\(\)\|\[\]\{\}\^\$\#"), |
| 1999 |
hir_lit(r"\.+*?()|[]{}^$#") |
| 2000 |
); |
| 2001 |
} |
| 2002 |
|
| 2003 |
#[test] |
| 2004 |
fn repetition() { |
| 2005 |
assert_eq!(t("a?"), hir_quest(true, hir_lit("a"))); |
| 2006 |
assert_eq!(t("a*"), hir_star(true, hir_lit("a"))); |
| 2007 |
assert_eq!(t("a+"), hir_plus(true, hir_lit("a"))); |
| 2008 |
assert_eq!(t("a??"), hir_quest(false, hir_lit("a"))); |
| 2009 |
assert_eq!(t("a*?"), hir_star(false, hir_lit("a"))); |
| 2010 |
assert_eq!(t("a+?"), hir_plus(false, hir_lit("a"))); |
| 2011 |
|
| 2012 |
assert_eq!(t("a{1}"), hir_range(true, 1, Some(1), hir_lit("a"),)); |
| 2013 |
assert_eq!(t("a{1,}"), hir_range(true, 1, None, hir_lit("a"),)); |
| 2014 |
assert_eq!(t("a{1,2}"), hir_range(true, 1, Some(2), hir_lit("a"),)); |
| 2015 |
assert_eq!(t("a{1}?"), hir_range(false, 1, Some(1), hir_lit("a"),)); |
| 2016 |
assert_eq!(t("a{1,}?"), hir_range(false, 1, None, hir_lit("a"),)); |
| 2017 |
assert_eq!(t("a{1,2}?"), hir_range(false, 1, Some(2), hir_lit("a"),)); |
| 2018 |
|
| 2019 |
assert_eq!( |
| 2020 |
t("ab?"), |
| 2021 |
hir_cat(vec![hir_lit("a"), hir_quest(true, hir_lit("b")),]) |
| 2022 |
); |
| 2023 |
assert_eq!(t("(ab)?"), hir_quest(true, hir_capture(1, hir_lit("ab")))); |
| 2024 |
assert_eq!( |
| 2025 |
t("a|b?"), |
| 2026 |
hir_alt(vec![hir_lit("a"), hir_quest(true, hir_lit("b")),]) |
| 2027 |
); |
| 2028 |
} |
| 2029 |
|
| 2030 |
#[test] |
| 2031 |
fn cat_alt() { |
| 2032 |
let a = || hir_look(hir::Look::Start); |
| 2033 |
let b = || hir_look(hir::Look::End); |
| 2034 |
let c = || hir_look(hir::Look::WordUnicode); |
| 2035 |
let d = || hir_look(hir::Look::WordUnicodeNegate); |
| 2036 |
|
| 2037 |
assert_eq!(t("(^$)"), hir_capture(1, hir_cat(vec![a(), b()]))); |
| 2038 |
assert_eq!(t("^|$"), hir_alt(vec![a(), b()])); |
| 2039 |
assert_eq!(t(r"^|$|\b"), hir_alt(vec![a(), b(), c()])); |
| 2040 |
assert_eq!( |
| 2041 |
t(r"^$|$\b|\b\B"), |
| 2042 |
hir_alt(vec![ |
| 2043 |
hir_cat(vec![a(), b()]), |
| 2044 |
hir_cat(vec![b(), c()]), |
| 2045 |
hir_cat(vec![c(), d()]), |
| 2046 |
]) |
| 2047 |
); |
| 2048 |
assert_eq!(t("(^|$)"), hir_capture(1, hir_alt(vec![a(), b()]))); |
| 2049 |
assert_eq!( |
| 2050 |
t(r"(^|$|\b)"), |
| 2051 |
hir_capture(1, hir_alt(vec![a(), b(), c()])) |
| 2052 |
); |
| 2053 |
assert_eq!( |
| 2054 |
t(r"(^$|$\b|\b\B)"), |
| 2055 |
hir_capture( |
| 2056 |
1, |
| 2057 |
hir_alt(vec![ |
| 2058 |
hir_cat(vec![a(), b()]), |
| 2059 |
hir_cat(vec![b(), c()]), |
| 2060 |
hir_cat(vec![c(), d()]), |
| 2061 |
]) |
| 2062 |
) |
| 2063 |
); |
| 2064 |
assert_eq!( |
| 2065 |
t(r"(^$|($\b|(\b\B)))"), |
| 2066 |
hir_capture( |
| 2067 |
1, |
| 2068 |
hir_alt(vec![ |
| 2069 |
hir_cat(vec![a(), b()]), |
| 2070 |
hir_capture( |
| 2071 |
2, |
| 2072 |
hir_alt(vec![ |
| 2073 |
hir_cat(vec![b(), c()]), |
| 2074 |
hir_capture(3, hir_cat(vec![c(), d()])), |
| 2075 |
]) |
| 2076 |
), |
| 2077 |
]) |
| 2078 |
) |
| 2079 |
); |
| 2080 |
} |
| 2081 |
|
| 2082 |
// Tests the HIR transformation of things like '[a-z]|[A-Z]' into |
| 2083 |
// '[A-Za-z]'. In other words, an alternation of just classes is always |
| 2084 |
// equivalent to a single class corresponding to the union of the branches |
| 2085 |
// in that class. (Unless some branches match invalid UTF-8 and others |
| 2086 |
// match non-ASCII Unicode.) |
| 2087 |
#[test] |
| 2088 |
fn cat_class_flattened() { |
| 2089 |
assert_eq!(t(r"[a-z]|[A-Z]"), hir_uclass(&[('A', 'Z'), ('a', 'z')])); |
| 2090 |
// Combining all of the letter properties should give us the one giant |
| 2091 |
// letter property. |
| 2092 |
#[cfg(feature = "unicode-gencat")] |
| 2093 |
assert_eq!( |
| 2094 |
t(r"(?x) |
| 2095 |
\p{Lowercase_Letter} |
| 2096 |
|\p{Uppercase_Letter} |
| 2097 |
|\p{Titlecase_Letter} |
| 2098 |
|\p{Modifier_Letter} |
| 2099 |
|\p{Other_Letter} |
| 2100 |
"), |
| 2101 |
hir_uclass_query(ClassQuery::Binary("letter")) |
| 2102 |
); |
| 2103 |
// Byte classes that can truly match invalid UTF-8 cannot be combined |
| 2104 |
// with Unicode classes. |
| 2105 |
assert_eq!( |
| 2106 |
t_bytes(r"[Δδ]|(?-u:[\x90-\xFF])|[Λλ]"), |
| 2107 |
hir_alt(vec![ |
| 2108 |
hir_uclass(&[('Δ', 'Δ'), ('δ', 'δ')]), |
| 2109 |
hir_bclass(&[(b'\x90', b'\xFF')]), |
| 2110 |
hir_uclass(&[('Λ', 'Λ'), ('λ', 'λ')]), |
| 2111 |
]) |
| 2112 |
); |
| 2113 |
// Byte classes on their own can be combined, even if some are ASCII |
| 2114 |
// and others are invalid UTF-8. |
| 2115 |
assert_eq!( |
| 2116 |
t_bytes(r"[a-z]|(?-u:[\x90-\xFF])|[A-Z]"), |
| 2117 |
hir_bclass(&[(b'A', b'Z'), (b'a', b'z'), (b'\x90', b'\xFF')]), |
| 2118 |
); |
| 2119 |
} |
| 2120 |
|
| 2121 |
#[test] |
| 2122 |
fn class_ascii() { |
| 2123 |
assert_eq!( |
| 2124 |
t("[[:alnum:]]"), |
| 2125 |
hir_ascii_uclass(&ast::ClassAsciiKind::Alnum) |
| 2126 |
); |
| 2127 |
assert_eq!( |
| 2128 |
t("[[:alpha:]]"), |
| 2129 |
hir_ascii_uclass(&ast::ClassAsciiKind::Alpha) |
| 2130 |
); |
| 2131 |
assert_eq!( |
| 2132 |
t("[[:ascii:]]"), |
| 2133 |
hir_ascii_uclass(&ast::ClassAsciiKind::Ascii) |
| 2134 |
); |
| 2135 |
assert_eq!( |
| 2136 |
t("[[:blank:]]"), |
| 2137 |
hir_ascii_uclass(&ast::ClassAsciiKind::Blank) |
| 2138 |
); |
| 2139 |
assert_eq!( |
| 2140 |
t("[[:cntrl:]]"), |
| 2141 |
hir_ascii_uclass(&ast::ClassAsciiKind::Cntrl) |
| 2142 |
); |
| 2143 |
assert_eq!( |
| 2144 |
t("[[:digit:]]"), |
| 2145 |
hir_ascii_uclass(&ast::ClassAsciiKind::Digit) |
| 2146 |
); |
| 2147 |
assert_eq!( |
| 2148 |
t("[[:graph:]]"), |
| 2149 |
hir_ascii_uclass(&ast::ClassAsciiKind::Graph) |
| 2150 |
); |
| 2151 |
assert_eq!( |
| 2152 |
t("[[:lower:]]"), |
| 2153 |
hir_ascii_uclass(&ast::ClassAsciiKind::Lower) |
| 2154 |
); |
| 2155 |
assert_eq!( |
| 2156 |
t("[[:print:]]"), |
| 2157 |
hir_ascii_uclass(&ast::ClassAsciiKind::Print) |
| 2158 |
); |
| 2159 |
assert_eq!( |
| 2160 |
t("[[:punct:]]"), |
| 2161 |
hir_ascii_uclass(&ast::ClassAsciiKind::Punct) |
| 2162 |
); |
| 2163 |
assert_eq!( |
| 2164 |
t("[[:space:]]"), |
| 2165 |
hir_ascii_uclass(&ast::ClassAsciiKind::Space) |
| 2166 |
); |
| 2167 |
assert_eq!( |
| 2168 |
t("[[:upper:]]"), |
| 2169 |
hir_ascii_uclass(&ast::ClassAsciiKind::Upper) |
| 2170 |
); |
| 2171 |
assert_eq!( |
| 2172 |
t("[[:word:]]"), |
| 2173 |
hir_ascii_uclass(&ast::ClassAsciiKind::Word) |
| 2174 |
); |
| 2175 |
assert_eq!( |
| 2176 |
t("[[:xdigit:]]"), |
| 2177 |
hir_ascii_uclass(&ast::ClassAsciiKind::Xdigit) |
| 2178 |
); |
| 2179 |
|
| 2180 |
assert_eq!( |
| 2181 |
t("[[:^lower:]]"), |
| 2182 |
hir_negate(hir_ascii_uclass(&ast::ClassAsciiKind::Lower)) |
| 2183 |
); |
| 2184 |
#[cfg(feature = "unicode-case")] |
| 2185 |
assert_eq!( |
| 2186 |
t("(?i)[[:lower:]]"), |
| 2187 |
hir_uclass(&[ |
| 2188 |
('A', 'Z'), |
| 2189 |
('a', 'z'), |
| 2190 |
('\u{17F}', '\u{17F}'), |
| 2191 |
('\u{212A}', '\u{212A}'), |
| 2192 |
]) |
| 2193 |
); |
| 2194 |
|
| 2195 |
assert_eq!( |
| 2196 |
t("(?-u)[[:lower:]]"), |
| 2197 |
hir_ascii_bclass(&ast::ClassAsciiKind::Lower) |
| 2198 |
); |
| 2199 |
assert_eq!( |
| 2200 |
t("(?i-u)[[:lower:]]"), |
| 2201 |
hir_case_fold(hir_ascii_bclass(&ast::ClassAsciiKind::Lower)) |
| 2202 |
); |
| 2203 |
|
| 2204 |
assert_eq!( |
| 2205 |
t_err("(?-u)[[:^lower:]]"), |
| 2206 |
TestError { |
| 2207 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2208 |
span: Span::new( |
| 2209 |
Position::new(6, 1, 7), |
| 2210 |
Position::new(16, 1, 17) |
| 2211 |
), |
| 2212 |
} |
| 2213 |
); |
| 2214 |
assert_eq!( |
| 2215 |
t_err("(?i-u)[[:^lower:]]"), |
| 2216 |
TestError { |
| 2217 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2218 |
span: Span::new( |
| 2219 |
Position::new(7, 1, 8), |
| 2220 |
Position::new(17, 1, 18) |
| 2221 |
), |
| 2222 |
} |
| 2223 |
); |
| 2224 |
} |
| 2225 |
|
| 2226 |
#[test] |
| 2227 |
fn class_ascii_multiple() { |
| 2228 |
// See: https://github.com/rust-lang/regex/issues/680 |
| 2229 |
assert_eq!( |
| 2230 |
t("[[:alnum:][:^ascii:]]"), |
| 2231 |
hir_union( |
| 2232 |
hir_ascii_uclass(&ast::ClassAsciiKind::Alnum), |
| 2233 |
hir_uclass(&[('\u{80}', '\u{10FFFF}')]), |
| 2234 |
), |
| 2235 |
); |
| 2236 |
assert_eq!( |
| 2237 |
t_bytes("(?-u)[[:alnum:][:^ascii:]]"), |
| 2238 |
hir_union( |
| 2239 |
hir_ascii_bclass(&ast::ClassAsciiKind::Alnum), |
| 2240 |
hir_bclass(&[(0x80, 0xFF)]), |
| 2241 |
), |
| 2242 |
); |
| 2243 |
} |
| 2244 |
|
| 2245 |
#[test] |
| 2246 |
#[cfg(feature = "unicode-perl")] |
| 2247 |
fn class_perl_unicode() { |
| 2248 |
// Unicode |
| 2249 |
assert_eq!(t(r"\d"), hir_uclass_query(ClassQuery::Binary("digit"))); |
| 2250 |
assert_eq!(t(r"\s"), hir_uclass_query(ClassQuery::Binary("space"))); |
| 2251 |
assert_eq!(t(r"\w"), hir_uclass_perl_word()); |
| 2252 |
#[cfg(feature = "unicode-case")] |
| 2253 |
assert_eq!( |
| 2254 |
t(r"(?i)\d"), |
| 2255 |
hir_uclass_query(ClassQuery::Binary("digit")) |
| 2256 |
); |
| 2257 |
#[cfg(feature = "unicode-case")] |
| 2258 |
assert_eq!( |
| 2259 |
t(r"(?i)\s"), |
| 2260 |
hir_uclass_query(ClassQuery::Binary("space")) |
| 2261 |
); |
| 2262 |
#[cfg(feature = "unicode-case")] |
| 2263 |
assert_eq!(t(r"(?i)\w"), hir_uclass_perl_word()); |
| 2264 |
|
| 2265 |
// Unicode, negated |
| 2266 |
assert_eq!( |
| 2267 |
t(r"\D"), |
| 2268 |
hir_negate(hir_uclass_query(ClassQuery::Binary("digit"))) |
| 2269 |
); |
| 2270 |
assert_eq!( |
| 2271 |
t(r"\S"), |
| 2272 |
hir_negate(hir_uclass_query(ClassQuery::Binary("space"))) |
| 2273 |
); |
| 2274 |
assert_eq!(t(r"\W"), hir_negate(hir_uclass_perl_word())); |
| 2275 |
#[cfg(feature = "unicode-case")] |
| 2276 |
assert_eq!( |
| 2277 |
t(r"(?i)\D"), |
| 2278 |
hir_negate(hir_uclass_query(ClassQuery::Binary("digit"))) |
| 2279 |
); |
| 2280 |
#[cfg(feature = "unicode-case")] |
| 2281 |
assert_eq!( |
| 2282 |
t(r"(?i)\S"), |
| 2283 |
hir_negate(hir_uclass_query(ClassQuery::Binary("space"))) |
| 2284 |
); |
| 2285 |
#[cfg(feature = "unicode-case")] |
| 2286 |
assert_eq!(t(r"(?i)\W"), hir_negate(hir_uclass_perl_word())); |
| 2287 |
} |
| 2288 |
|
| 2289 |
#[test] |
| 2290 |
fn class_perl_ascii() { |
| 2291 |
// ASCII only |
| 2292 |
assert_eq!( |
| 2293 |
t(r"(?-u)\d"), |
| 2294 |
hir_ascii_bclass(&ast::ClassAsciiKind::Digit) |
| 2295 |
); |
| 2296 |
assert_eq!( |
| 2297 |
t(r"(?-u)\s"), |
| 2298 |
hir_ascii_bclass(&ast::ClassAsciiKind::Space) |
| 2299 |
); |
| 2300 |
assert_eq!( |
| 2301 |
t(r"(?-u)\w"), |
| 2302 |
hir_ascii_bclass(&ast::ClassAsciiKind::Word) |
| 2303 |
); |
| 2304 |
assert_eq!( |
| 2305 |
t(r"(?i-u)\d"), |
| 2306 |
hir_ascii_bclass(&ast::ClassAsciiKind::Digit) |
| 2307 |
); |
| 2308 |
assert_eq!( |
| 2309 |
t(r"(?i-u)\s"), |
| 2310 |
hir_ascii_bclass(&ast::ClassAsciiKind::Space) |
| 2311 |
); |
| 2312 |
assert_eq!( |
| 2313 |
t(r"(?i-u)\w"), |
| 2314 |
hir_ascii_bclass(&ast::ClassAsciiKind::Word) |
| 2315 |
); |
| 2316 |
|
| 2317 |
// ASCII only, negated |
| 2318 |
assert_eq!( |
| 2319 |
t_bytes(r"(?-u)\D"), |
| 2320 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Digit)) |
| 2321 |
); |
| 2322 |
assert_eq!( |
| 2323 |
t_bytes(r"(?-u)\S"), |
| 2324 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Space)) |
| 2325 |
); |
| 2326 |
assert_eq!( |
| 2327 |
t_bytes(r"(?-u)\W"), |
| 2328 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Word)) |
| 2329 |
); |
| 2330 |
assert_eq!( |
| 2331 |
t_bytes(r"(?i-u)\D"), |
| 2332 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Digit)) |
| 2333 |
); |
| 2334 |
assert_eq!( |
| 2335 |
t_bytes(r"(?i-u)\S"), |
| 2336 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Space)) |
| 2337 |
); |
| 2338 |
assert_eq!( |
| 2339 |
t_bytes(r"(?i-u)\W"), |
| 2340 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Word)) |
| 2341 |
); |
| 2342 |
|
| 2343 |
// ASCII only, negated, with UTF-8 mode enabled. |
| 2344 |
// In this case, negating any Perl class results in an error because |
| 2345 |
// all such classes can match invalid UTF-8. |
| 2346 |
assert_eq!( |
| 2347 |
t_err(r"(?-u)\D"), |
| 2348 |
TestError { |
| 2349 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2350 |
span: Span::new( |
| 2351 |
Position::new(5, 1, 6), |
| 2352 |
Position::new(7, 1, 8), |
| 2353 |
), |
| 2354 |
}, |
| 2355 |
); |
| 2356 |
assert_eq!( |
| 2357 |
t_err(r"(?-u)\S"), |
| 2358 |
TestError { |
| 2359 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2360 |
span: Span::new( |
| 2361 |
Position::new(5, 1, 6), |
| 2362 |
Position::new(7, 1, 8), |
| 2363 |
), |
| 2364 |
}, |
| 2365 |
); |
| 2366 |
assert_eq!( |
| 2367 |
t_err(r"(?-u)\W"), |
| 2368 |
TestError { |
| 2369 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2370 |
span: Span::new( |
| 2371 |
Position::new(5, 1, 6), |
| 2372 |
Position::new(7, 1, 8), |
| 2373 |
), |
| 2374 |
}, |
| 2375 |
); |
| 2376 |
assert_eq!( |
| 2377 |
t_err(r"(?i-u)\D"), |
| 2378 |
TestError { |
| 2379 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2380 |
span: Span::new( |
| 2381 |
Position::new(6, 1, 7), |
| 2382 |
Position::new(8, 1, 9), |
| 2383 |
), |
| 2384 |
}, |
| 2385 |
); |
| 2386 |
assert_eq!( |
| 2387 |
t_err(r"(?i-u)\S"), |
| 2388 |
TestError { |
| 2389 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2390 |
span: Span::new( |
| 2391 |
Position::new(6, 1, 7), |
| 2392 |
Position::new(8, 1, 9), |
| 2393 |
), |
| 2394 |
}, |
| 2395 |
); |
| 2396 |
assert_eq!( |
| 2397 |
t_err(r"(?i-u)\W"), |
| 2398 |
TestError { |
| 2399 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2400 |
span: Span::new( |
| 2401 |
Position::new(6, 1, 7), |
| 2402 |
Position::new(8, 1, 9), |
| 2403 |
), |
| 2404 |
}, |
| 2405 |
); |
| 2406 |
} |
| 2407 |
|
| 2408 |
#[test] |
| 2409 |
#[cfg(not(feature = "unicode-perl"))] |
| 2410 |
fn class_perl_word_disabled() { |
| 2411 |
assert_eq!( |
| 2412 |
t_err(r"\w"), |
| 2413 |
TestError { |
| 2414 |
kind: hir::ErrorKind::UnicodePerlClassNotFound, |
| 2415 |
span: Span::new( |
| 2416 |
Position::new(0, 1, 1), |
| 2417 |
Position::new(2, 1, 3) |
| 2418 |
), |
| 2419 |
} |
| 2420 |
); |
| 2421 |
} |
| 2422 |
|
| 2423 |
#[test] |
| 2424 |
#[cfg(all(not(feature = "unicode-perl"), not(feature = "unicode-bool")))] |
| 2425 |
fn class_perl_space_disabled() { |
| 2426 |
assert_eq!( |
| 2427 |
t_err(r"\s"), |
| 2428 |
TestError { |
| 2429 |
kind: hir::ErrorKind::UnicodePerlClassNotFound, |
| 2430 |
span: Span::new( |
| 2431 |
Position::new(0, 1, 1), |
| 2432 |
Position::new(2, 1, 3) |
| 2433 |
), |
| 2434 |
} |
| 2435 |
); |
| 2436 |
} |
| 2437 |
|
| 2438 |
#[test] |
| 2439 |
#[cfg(all( |
| 2440 |
not(feature = "unicode-perl"), |
| 2441 |
not(feature = "unicode-gencat") |
| 2442 |
))] |
| 2443 |
fn class_perl_digit_disabled() { |
| 2444 |
assert_eq!( |
| 2445 |
t_err(r"\d"), |
| 2446 |
TestError { |
| 2447 |
kind: hir::ErrorKind::UnicodePerlClassNotFound, |
| 2448 |
span: Span::new( |
| 2449 |
Position::new(0, 1, 1), |
| 2450 |
Position::new(2, 1, 3) |
| 2451 |
), |
| 2452 |
} |
| 2453 |
); |
| 2454 |
} |
| 2455 |
|
| 2456 |
#[test] |
| 2457 |
#[cfg(feature = "unicode-gencat")] |
| 2458 |
fn class_unicode_gencat() { |
| 2459 |
assert_eq!(t(r"\pZ"), hir_uclass_query(ClassQuery::Binary("Z"))); |
| 2460 |
assert_eq!(t(r"\pz"), hir_uclass_query(ClassQuery::Binary("Z"))); |
| 2461 |
assert_eq!( |
| 2462 |
t(r"\p{Separator}"), |
| 2463 |
hir_uclass_query(ClassQuery::Binary("Z")) |
| 2464 |
); |
| 2465 |
assert_eq!( |
| 2466 |
t(r"\p{se PaRa ToR}"), |
| 2467 |
hir_uclass_query(ClassQuery::Binary("Z")) |
| 2468 |
); |
| 2469 |
assert_eq!( |
| 2470 |
t(r"\p{gc:Separator}"), |
| 2471 |
hir_uclass_query(ClassQuery::Binary("Z")) |
| 2472 |
); |
| 2473 |
assert_eq!( |
| 2474 |
t(r"\p{gc=Separator}"), |
| 2475 |
hir_uclass_query(ClassQuery::Binary("Z")) |
| 2476 |
); |
| 2477 |
assert_eq!( |
| 2478 |
t(r"\p{Other}"), |
| 2479 |
hir_uclass_query(ClassQuery::Binary("Other")) |
| 2480 |
); |
| 2481 |
assert_eq!(t(r"\pC"), hir_uclass_query(ClassQuery::Binary("Other"))); |
| 2482 |
|
| 2483 |
assert_eq!( |
| 2484 |
t(r"\PZ"), |
| 2485 |
hir_negate(hir_uclass_query(ClassQuery::Binary("Z"))) |
| 2486 |
); |
| 2487 |
assert_eq!( |
| 2488 |
t(r"\P{separator}"), |
| 2489 |
hir_negate(hir_uclass_query(ClassQuery::Binary("Z"))) |
| 2490 |
); |
| 2491 |
assert_eq!( |
| 2492 |
t(r"\P{gc!=separator}"), |
| 2493 |
hir_negate(hir_uclass_query(ClassQuery::Binary("Z"))) |
| 2494 |
); |
| 2495 |
|
| 2496 |
assert_eq!(t(r"\p{any}"), hir_uclass_query(ClassQuery::Binary("Any"))); |
| 2497 |
assert_eq!( |
| 2498 |
t(r"\p{assigned}"), |
| 2499 |
hir_uclass_query(ClassQuery::Binary("Assigned")) |
| 2500 |
); |
| 2501 |
assert_eq!( |
| 2502 |
t(r"\p{ascii}"), |
| 2503 |
hir_uclass_query(ClassQuery::Binary("ASCII")) |
| 2504 |
); |
| 2505 |
assert_eq!( |
| 2506 |
t(r"\p{gc:any}"), |
| 2507 |
hir_uclass_query(ClassQuery::Binary("Any")) |
| 2508 |
); |
| 2509 |
assert_eq!( |
| 2510 |
t(r"\p{gc:assigned}"), |
| 2511 |
hir_uclass_query(ClassQuery::Binary("Assigned")) |
| 2512 |
); |
| 2513 |
assert_eq!( |
| 2514 |
t(r"\p{gc:ascii}"), |
| 2515 |
hir_uclass_query(ClassQuery::Binary("ASCII")) |
| 2516 |
); |
| 2517 |
|
| 2518 |
assert_eq!( |
| 2519 |
t_err(r"(?-u)\pZ"), |
| 2520 |
TestError { |
| 2521 |
kind: hir::ErrorKind::UnicodeNotAllowed, |
| 2522 |
span: Span::new( |
| 2523 |
Position::new(5, 1, 6), |
| 2524 |
Position::new(8, 1, 9) |
| 2525 |
), |
| 2526 |
} |
| 2527 |
); |
| 2528 |
assert_eq!( |
| 2529 |
t_err(r"(?-u)\p{Separator}"), |
| 2530 |
TestError { |
| 2531 |
kind: hir::ErrorKind::UnicodeNotAllowed, |
| 2532 |
span: Span::new( |
| 2533 |
Position::new(5, 1, 6), |
| 2534 |
Position::new(18, 1, 19) |
| 2535 |
), |
| 2536 |
} |
| 2537 |
); |
| 2538 |
assert_eq!( |
| 2539 |
t_err(r"\pE"), |
| 2540 |
TestError { |
| 2541 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2542 |
span: Span::new( |
| 2543 |
Position::new(0, 1, 1), |
| 2544 |
Position::new(3, 1, 4) |
| 2545 |
), |
| 2546 |
} |
| 2547 |
); |
| 2548 |
assert_eq!( |
| 2549 |
t_err(r"\p{Foo}"), |
| 2550 |
TestError { |
| 2551 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2552 |
span: Span::new( |
| 2553 |
Position::new(0, 1, 1), |
| 2554 |
Position::new(7, 1, 8) |
| 2555 |
), |
| 2556 |
} |
| 2557 |
); |
| 2558 |
assert_eq!( |
| 2559 |
t_err(r"\p{gc:Foo}"), |
| 2560 |
TestError { |
| 2561 |
kind: hir::ErrorKind::UnicodePropertyValueNotFound, |
| 2562 |
span: Span::new( |
| 2563 |
Position::new(0, 1, 1), |
| 2564 |
Position::new(10, 1, 11) |
| 2565 |
), |
| 2566 |
} |
| 2567 |
); |
| 2568 |
} |
| 2569 |
|
| 2570 |
#[test] |
| 2571 |
#[cfg(not(feature = "unicode-gencat"))] |
| 2572 |
fn class_unicode_gencat_disabled() { |
| 2573 |
assert_eq!( |
| 2574 |
t_err(r"\p{Separator}"), |
| 2575 |
TestError { |
| 2576 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2577 |
span: Span::new( |
| 2578 |
Position::new(0, 1, 1), |
| 2579 |
Position::new(13, 1, 14) |
| 2580 |
), |
| 2581 |
} |
| 2582 |
); |
| 2583 |
|
| 2584 |
assert_eq!( |
| 2585 |
t_err(r"\p{Any}"), |
| 2586 |
TestError { |
| 2587 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2588 |
span: Span::new( |
| 2589 |
Position::new(0, 1, 1), |
| 2590 |
Position::new(7, 1, 8) |
| 2591 |
), |
| 2592 |
} |
| 2593 |
); |
| 2594 |
} |
| 2595 |
|
| 2596 |
#[test] |
| 2597 |
#[cfg(feature = "unicode-script")] |
| 2598 |
fn class_unicode_script() { |
| 2599 |
assert_eq!( |
| 2600 |
t(r"\p{Greek}"), |
| 2601 |
hir_uclass_query(ClassQuery::Binary("Greek")) |
| 2602 |
); |
| 2603 |
#[cfg(feature = "unicode-case")] |
| 2604 |
assert_eq!( |
| 2605 |
t(r"(?i)\p{Greek}"), |
| 2606 |
hir_case_fold(hir_uclass_query(ClassQuery::Binary("Greek"))) |
| 2607 |
); |
| 2608 |
#[cfg(feature = "unicode-case")] |
| 2609 |
assert_eq!( |
| 2610 |
t(r"(?i)\P{Greek}"), |
| 2611 |
hir_negate(hir_case_fold(hir_uclass_query(ClassQuery::Binary( |
| 2612 |
"Greek" |
| 2613 |
)))) |
| 2614 |
); |
| 2615 |
|
| 2616 |
assert_eq!( |
| 2617 |
t_err(r"\p{sc:Foo}"), |
| 2618 |
TestError { |
| 2619 |
kind: hir::ErrorKind::UnicodePropertyValueNotFound, |
| 2620 |
span: Span::new( |
| 2621 |
Position::new(0, 1, 1), |
| 2622 |
Position::new(10, 1, 11) |
| 2623 |
), |
| 2624 |
} |
| 2625 |
); |
| 2626 |
assert_eq!( |
| 2627 |
t_err(r"\p{scx:Foo}"), |
| 2628 |
TestError { |
| 2629 |
kind: hir::ErrorKind::UnicodePropertyValueNotFound, |
| 2630 |
span: Span::new( |
| 2631 |
Position::new(0, 1, 1), |
| 2632 |
Position::new(11, 1, 12) |
| 2633 |
), |
| 2634 |
} |
| 2635 |
); |
| 2636 |
} |
| 2637 |
|
| 2638 |
#[test] |
| 2639 |
#[cfg(not(feature = "unicode-script"))] |
| 2640 |
fn class_unicode_script_disabled() { |
| 2641 |
assert_eq!( |
| 2642 |
t_err(r"\p{Greek}"), |
| 2643 |
TestError { |
| 2644 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2645 |
span: Span::new( |
| 2646 |
Position::new(0, 1, 1), |
| 2647 |
Position::new(9, 1, 10) |
| 2648 |
), |
| 2649 |
} |
| 2650 |
); |
| 2651 |
|
| 2652 |
assert_eq!( |
| 2653 |
t_err(r"\p{scx:Greek}"), |
| 2654 |
TestError { |
| 2655 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2656 |
span: Span::new( |
| 2657 |
Position::new(0, 1, 1), |
| 2658 |
Position::new(13, 1, 14) |
| 2659 |
), |
| 2660 |
} |
| 2661 |
); |
| 2662 |
} |
| 2663 |
|
| 2664 |
#[test] |
| 2665 |
#[cfg(feature = "unicode-age")] |
| 2666 |
fn class_unicode_age() { |
| 2667 |
assert_eq!( |
| 2668 |
t_err(r"\p{age:Foo}"), |
| 2669 |
TestError { |
| 2670 |
kind: hir::ErrorKind::UnicodePropertyValueNotFound, |
| 2671 |
span: Span::new( |
| 2672 |
Position::new(0, 1, 1), |
| 2673 |
Position::new(11, 1, 12) |
| 2674 |
), |
| 2675 |
} |
| 2676 |
); |
| 2677 |
} |
| 2678 |
|
| 2679 |
#[test] |
| 2680 |
#[cfg(feature = "unicode-gencat")] |
| 2681 |
fn class_unicode_any_empty() { |
| 2682 |
assert_eq!(t(r"\P{any}"), hir_uclass(&[]),); |
| 2683 |
} |
| 2684 |
|
| 2685 |
#[test] |
| 2686 |
#[cfg(not(feature = "unicode-age"))] |
| 2687 |
fn class_unicode_age_disabled() { |
| 2688 |
assert_eq!( |
| 2689 |
t_err(r"\p{age:3.0}"), |
| 2690 |
TestError { |
| 2691 |
kind: hir::ErrorKind::UnicodePropertyNotFound, |
| 2692 |
span: Span::new( |
| 2693 |
Position::new(0, 1, 1), |
| 2694 |
Position::new(11, 1, 12) |
| 2695 |
), |
| 2696 |
} |
| 2697 |
); |
| 2698 |
} |
| 2699 |
|
| 2700 |
#[test] |
| 2701 |
fn class_bracketed() { |
| 2702 |
assert_eq!(t("[a]"), hir_lit("a")); |
| 2703 |
assert_eq!(t("[ab]"), hir_uclass(&[('a', 'b')])); |
| 2704 |
assert_eq!(t("[^[a]]"), class_negate(uclass(&[('a', 'a')]))); |
| 2705 |
assert_eq!(t("[a-z]"), hir_uclass(&[('a', 'z')])); |
| 2706 |
assert_eq!(t("[a-fd-h]"), hir_uclass(&[('a', 'h')])); |
| 2707 |
assert_eq!(t("[a-fg-m]"), hir_uclass(&[('a', 'm')])); |
| 2708 |
assert_eq!(t(r"[\x00]"), hir_uclass(&[('\0', '\0')])); |
| 2709 |
assert_eq!(t(r"[\n]"), hir_uclass(&[('\n', '\n')])); |
| 2710 |
assert_eq!(t("[\n]"), hir_uclass(&[('\n', '\n')])); |
| 2711 |
#[cfg(any(feature = "unicode-perl", feature = "unicode-gencat"))] |
| 2712 |
assert_eq!(t(r"[\d]"), hir_uclass_query(ClassQuery::Binary("digit"))); |
| 2713 |
#[cfg(feature = "unicode-gencat")] |
| 2714 |
assert_eq!( |
| 2715 |
t(r"[\pZ]"), |
| 2716 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2717 |
); |
| 2718 |
#[cfg(feature = "unicode-gencat")] |
| 2719 |
assert_eq!( |
| 2720 |
t(r"[\p{separator}]"), |
| 2721 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2722 |
); |
| 2723 |
#[cfg(any(feature = "unicode-perl", feature = "unicode-gencat"))] |
| 2724 |
assert_eq!(t(r"[^\D]"), hir_uclass_query(ClassQuery::Binary("digit"))); |
| 2725 |
#[cfg(feature = "unicode-gencat")] |
| 2726 |
assert_eq!( |
| 2727 |
t(r"[^\PZ]"), |
| 2728 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2729 |
); |
| 2730 |
#[cfg(feature = "unicode-gencat")] |
| 2731 |
assert_eq!( |
| 2732 |
t(r"[^\P{separator}]"), |
| 2733 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2734 |
); |
| 2735 |
#[cfg(all( |
| 2736 |
feature = "unicode-case", |
| 2737 |
any(feature = "unicode-perl", feature = "unicode-gencat") |
| 2738 |
))] |
| 2739 |
assert_eq!( |
| 2740 |
t(r"(?i)[^\D]"), |
| 2741 |
hir_uclass_query(ClassQuery::Binary("digit")) |
| 2742 |
); |
| 2743 |
#[cfg(all(feature = "unicode-case", feature = "unicode-script"))] |
| 2744 |
assert_eq!( |
| 2745 |
t(r"(?i)[^\P{greek}]"), |
| 2746 |
hir_case_fold(hir_uclass_query(ClassQuery::Binary("greek"))) |
| 2747 |
); |
| 2748 |
|
| 2749 |
assert_eq!(t("(?-u)[a]"), hir_bclass(&[(b'a', b'a')])); |
| 2750 |
assert_eq!(t(r"(?-u)[\x00]"), hir_bclass(&[(b'\0', b'\0')])); |
| 2751 |
assert_eq!(t_bytes(r"(?-u)[\xFF]"), hir_bclass(&[(b'\xFF', b'\xFF')])); |
| 2752 |
|
| 2753 |
#[cfg(feature = "unicode-case")] |
| 2754 |
assert_eq!(t("(?i)[a]"), hir_uclass(&[('A', 'A'), ('a', 'a')])); |
| 2755 |
#[cfg(feature = "unicode-case")] |
| 2756 |
assert_eq!( |
| 2757 |
t("(?i)[k]"), |
| 2758 |
hir_uclass(&[('K', 'K'), ('k', 'k'), ('\u{212A}', '\u{212A}'),]) |
| 2759 |
); |
| 2760 |
#[cfg(feature = "unicode-case")] |
| 2761 |
assert_eq!( |
| 2762 |
t("(?i)[β]"), |
| 2763 |
hir_uclass(&[('Β', 'Β'), ('β', 'β'), ('ϐ', 'ϐ'),]) |
| 2764 |
); |
| 2765 |
assert_eq!(t("(?i-u)[k]"), hir_bclass(&[(b'K', b'K'), (b'k', b'k'),])); |
| 2766 |
|
| 2767 |
assert_eq!(t("[^a]"), class_negate(uclass(&[('a', 'a')]))); |
| 2768 |
assert_eq!(t(r"[^\x00]"), class_negate(uclass(&[('\0', '\0')]))); |
| 2769 |
assert_eq!( |
| 2770 |
t_bytes("(?-u)[^a]"), |
| 2771 |
class_negate(bclass(&[(b'a', b'a')])) |
| 2772 |
); |
| 2773 |
#[cfg(any(feature = "unicode-perl", feature = "unicode-gencat"))] |
| 2774 |
assert_eq!( |
| 2775 |
t(r"[^\d]"), |
| 2776 |
hir_negate(hir_uclass_query(ClassQuery::Binary("digit"))) |
| 2777 |
); |
| 2778 |
#[cfg(feature = "unicode-gencat")] |
| 2779 |
assert_eq!( |
| 2780 |
t(r"[^\pZ]"), |
| 2781 |
hir_negate(hir_uclass_query(ClassQuery::Binary("separator"))) |
| 2782 |
); |
| 2783 |
#[cfg(feature = "unicode-gencat")] |
| 2784 |
assert_eq!( |
| 2785 |
t(r"[^\p{separator}]"), |
| 2786 |
hir_negate(hir_uclass_query(ClassQuery::Binary("separator"))) |
| 2787 |
); |
| 2788 |
#[cfg(all(feature = "unicode-case", feature = "unicode-script"))] |
| 2789 |
assert_eq!( |
| 2790 |
t(r"(?i)[^\p{greek}]"), |
| 2791 |
hir_negate(hir_case_fold(hir_uclass_query(ClassQuery::Binary( |
| 2792 |
"greek" |
| 2793 |
)))) |
| 2794 |
); |
| 2795 |
#[cfg(all(feature = "unicode-case", feature = "unicode-script"))] |
| 2796 |
assert_eq!( |
| 2797 |
t(r"(?i)[\P{greek}]"), |
| 2798 |
hir_negate(hir_case_fold(hir_uclass_query(ClassQuery::Binary( |
| 2799 |
"greek" |
| 2800 |
)))) |
| 2801 |
); |
| 2802 |
|
| 2803 |
// Test some weird cases. |
| 2804 |
assert_eq!(t(r"[\[]"), hir_uclass(&[('[', '[')])); |
| 2805 |
|
| 2806 |
assert_eq!(t(r"[&]"), hir_uclass(&[('&', '&')])); |
| 2807 |
assert_eq!(t(r"[\&]"), hir_uclass(&[('&', '&')])); |
| 2808 |
assert_eq!(t(r"[\&\&]"), hir_uclass(&[('&', '&')])); |
| 2809 |
assert_eq!(t(r"[\x00-&]"), hir_uclass(&[('\0', '&')])); |
| 2810 |
assert_eq!(t(r"[&-\xFF]"), hir_uclass(&[('&', '\u{FF}')])); |
| 2811 |
|
| 2812 |
assert_eq!(t(r"[~]"), hir_uclass(&[('~', '~')])); |
| 2813 |
assert_eq!(t(r"[\~]"), hir_uclass(&[('~', '~')])); |
| 2814 |
assert_eq!(t(r"[\~\~]"), hir_uclass(&[('~', '~')])); |
| 2815 |
assert_eq!(t(r"[\x00-~]"), hir_uclass(&[('\0', '~')])); |
| 2816 |
assert_eq!(t(r"[~-\xFF]"), hir_uclass(&[('~', '\u{FF}')])); |
| 2817 |
|
| 2818 |
assert_eq!(t(r"[-]"), hir_uclass(&[('-', '-')])); |
| 2819 |
assert_eq!(t(r"[\-]"), hir_uclass(&[('-', '-')])); |
| 2820 |
assert_eq!(t(r"[\-\-]"), hir_uclass(&[('-', '-')])); |
| 2821 |
assert_eq!(t(r"[\x00-\-]"), hir_uclass(&[('\0', '-')])); |
| 2822 |
assert_eq!(t(r"[\--\xFF]"), hir_uclass(&[('-', '\u{FF}')])); |
| 2823 |
|
| 2824 |
assert_eq!( |
| 2825 |
t_err("(?-u)[^a]"), |
| 2826 |
TestError { |
| 2827 |
kind: hir::ErrorKind::InvalidUtf8, |
| 2828 |
span: Span::new( |
| 2829 |
Position::new(5, 1, 6), |
| 2830 |
Position::new(9, 1, 10) |
| 2831 |
), |
| 2832 |
} |
| 2833 |
); |
| 2834 |
#[cfg(any(feature = "unicode-perl", feature = "unicode-bool"))] |
| 2835 |
assert_eq!(t(r"[^\s\S]"), hir_uclass(&[]),); |
| 2836 |
#[cfg(any(feature = "unicode-perl", feature = "unicode-bool"))] |
| 2837 |
assert_eq!(t_bytes(r"(?-u)[^\s\S]"), hir_bclass(&[]),); |
| 2838 |
} |
| 2839 |
|
| 2840 |
#[test] |
| 2841 |
fn class_bracketed_union() { |
| 2842 |
assert_eq!(t("[a-zA-Z]"), hir_uclass(&[('A', 'Z'), ('a', 'z')])); |
| 2843 |
#[cfg(feature = "unicode-gencat")] |
| 2844 |
assert_eq!( |
| 2845 |
t(r"[a\pZb]"), |
| 2846 |
hir_union( |
| 2847 |
hir_uclass(&[('a', 'b')]), |
| 2848 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2849 |
) |
| 2850 |
); |
| 2851 |
#[cfg(all(feature = "unicode-gencat", feature = "unicode-script"))] |
| 2852 |
assert_eq!( |
| 2853 |
t(r"[\pZ\p{Greek}]"), |
| 2854 |
hir_union( |
| 2855 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2856 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2857 |
) |
| 2858 |
); |
| 2859 |
#[cfg(all( |
| 2860 |
feature = "unicode-age", |
| 2861 |
feature = "unicode-gencat", |
| 2862 |
feature = "unicode-script" |
| 2863 |
))] |
| 2864 |
assert_eq!( |
| 2865 |
t(r"[\p{age:3.0}\pZ\p{Greek}]"), |
| 2866 |
hir_union( |
| 2867 |
hir_uclass_query(ClassQuery::ByValue { |
| 2868 |
property_name: "age", |
| 2869 |
property_value: "3.0", |
| 2870 |
}), |
| 2871 |
hir_union( |
| 2872 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2873 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2874 |
) |
| 2875 |
) |
| 2876 |
); |
| 2877 |
#[cfg(all( |
| 2878 |
feature = "unicode-age", |
| 2879 |
feature = "unicode-gencat", |
| 2880 |
feature = "unicode-script" |
| 2881 |
))] |
| 2882 |
assert_eq!( |
| 2883 |
t(r"[[[\p{age:3.0}\pZ]\p{Greek}][\p{Cyrillic}]]"), |
| 2884 |
hir_union( |
| 2885 |
hir_uclass_query(ClassQuery::ByValue { |
| 2886 |
property_name: "age", |
| 2887 |
property_value: "3.0", |
| 2888 |
}), |
| 2889 |
hir_union( |
| 2890 |
hir_uclass_query(ClassQuery::Binary("cyrillic")), |
| 2891 |
hir_union( |
| 2892 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2893 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2894 |
) |
| 2895 |
) |
| 2896 |
) |
| 2897 |
); |
| 2898 |
|
| 2899 |
#[cfg(all( |
| 2900 |
feature = "unicode-age", |
| 2901 |
feature = "unicode-case", |
| 2902 |
feature = "unicode-gencat", |
| 2903 |
feature = "unicode-script" |
| 2904 |
))] |
| 2905 |
assert_eq!( |
| 2906 |
t(r"(?i)[\p{age:3.0}\pZ\p{Greek}]"), |
| 2907 |
hir_case_fold(hir_union( |
| 2908 |
hir_uclass_query(ClassQuery::ByValue { |
| 2909 |
property_name: "age", |
| 2910 |
property_value: "3.0", |
| 2911 |
}), |
| 2912 |
hir_union( |
| 2913 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2914 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2915 |
) |
| 2916 |
)) |
| 2917 |
); |
| 2918 |
#[cfg(all( |
| 2919 |
feature = "unicode-age", |
| 2920 |
feature = "unicode-gencat", |
| 2921 |
feature = "unicode-script" |
| 2922 |
))] |
| 2923 |
assert_eq!( |
| 2924 |
t(r"[^\p{age:3.0}\pZ\p{Greek}]"), |
| 2925 |
hir_negate(hir_union( |
| 2926 |
hir_uclass_query(ClassQuery::ByValue { |
| 2927 |
property_name: "age", |
| 2928 |
property_value: "3.0", |
| 2929 |
}), |
| 2930 |
hir_union( |
| 2931 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2932 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2933 |
) |
| 2934 |
)) |
| 2935 |
); |
| 2936 |
#[cfg(all( |
| 2937 |
feature = "unicode-age", |
| 2938 |
feature = "unicode-case", |
| 2939 |
feature = "unicode-gencat", |
| 2940 |
feature = "unicode-script" |
| 2941 |
))] |
| 2942 |
assert_eq!( |
| 2943 |
t(r"(?i)[^\p{age:3.0}\pZ\p{Greek}]"), |
| 2944 |
hir_negate(hir_case_fold(hir_union( |
| 2945 |
hir_uclass_query(ClassQuery::ByValue { |
| 2946 |
property_name: "age", |
| 2947 |
property_value: "3.0", |
| 2948 |
}), |
| 2949 |
hir_union( |
| 2950 |
hir_uclass_query(ClassQuery::Binary("greek")), |
| 2951 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 2952 |
) |
| 2953 |
))) |
| 2954 |
); |
| 2955 |
} |
| 2956 |
|
| 2957 |
#[test] |
| 2958 |
fn class_bracketed_nested() { |
| 2959 |
assert_eq!(t(r"[a[^c]]"), class_negate(uclass(&[('c', 'c')]))); |
| 2960 |
assert_eq!(t(r"[a-b[^c]]"), class_negate(uclass(&[('c', 'c')]))); |
| 2961 |
assert_eq!(t(r"[a-c[^c]]"), class_negate(uclass(&[]))); |
| 2962 |
|
| 2963 |
assert_eq!(t(r"[^a[^c]]"), hir_uclass(&[('c', 'c')])); |
| 2964 |
assert_eq!(t(r"[^a-b[^c]]"), hir_uclass(&[('c', 'c')])); |
| 2965 |
|
| 2966 |
#[cfg(feature = "unicode-case")] |
| 2967 |
assert_eq!( |
| 2968 |
t(r"(?i)[a[^c]]"), |
| 2969 |
hir_negate(class_case_fold(uclass(&[('c', 'c')]))) |
| 2970 |
); |
| 2971 |
#[cfg(feature = "unicode-case")] |
| 2972 |
assert_eq!( |
| 2973 |
t(r"(?i)[a-b[^c]]"), |
| 2974 |
hir_negate(class_case_fold(uclass(&[('c', 'c')]))) |
| 2975 |
); |
| 2976 |
|
| 2977 |
#[cfg(feature = "unicode-case")] |
| 2978 |
assert_eq!(t(r"(?i)[^a[^c]]"), hir_uclass(&[('C', 'C'), ('c', 'c')])); |
| 2979 |
#[cfg(feature = "unicode-case")] |
| 2980 |
assert_eq!( |
| 2981 |
t(r"(?i)[^a-b[^c]]"), |
| 2982 |
hir_uclass(&[('C', 'C'), ('c', 'c')]) |
| 2983 |
); |
| 2984 |
|
| 2985 |
assert_eq!(t(r"[^a-c[^c]]"), hir_uclass(&[]),); |
| 2986 |
#[cfg(feature = "unicode-case")] |
| 2987 |
assert_eq!(t(r"(?i)[^a-c[^c]]"), hir_uclass(&[]),); |
| 2988 |
} |
| 2989 |
|
| 2990 |
#[test] |
| 2991 |
fn class_bracketed_intersect() { |
| 2992 |
assert_eq!(t("[abc&&b-c]"), hir_uclass(&[('b', 'c')])); |
| 2993 |
assert_eq!(t("[abc&&[b-c]]"), hir_uclass(&[('b', 'c')])); |
| 2994 |
assert_eq!(t("[[abc]&&[b-c]]"), hir_uclass(&[('b', 'c')])); |
| 2995 |
assert_eq!(t("[a-z&&b-y&&c-x]"), hir_uclass(&[('c', 'x')])); |
| 2996 |
assert_eq!(t("[c-da-b&&a-d]"), hir_uclass(&[('a', 'd')])); |
| 2997 |
assert_eq!(t("[a-d&&c-da-b]"), hir_uclass(&[('a', 'd')])); |
| 2998 |
assert_eq!(t(r"[a-z&&a-c]"), hir_uclass(&[('a', 'c')])); |
| 2999 |
assert_eq!(t(r"[[a-z&&a-c]]"), hir_uclass(&[('a', 'c')])); |
| 3000 |
assert_eq!(t(r"[^[a-z&&a-c]]"), hir_negate(hir_uclass(&[('a', 'c')]))); |
| 3001 |
|
| 3002 |
assert_eq!(t("(?-u)[abc&&b-c]"), hir_bclass(&[(b'b', b'c')])); |
| 3003 |
assert_eq!(t("(?-u)[abc&&[b-c]]"), hir_bclass(&[(b'b', b'c')])); |
| 3004 |
assert_eq!(t("(?-u)[[abc]&&[b-c]]"), hir_bclass(&[(b'b', b'c')])); |
| 3005 |
assert_eq!(t("(?-u)[a-z&&b-y&&c-x]"), hir_bclass(&[(b'c', b'x')])); |
| 3006 |
assert_eq!(t("(?-u)[c-da-b&&a-d]"), hir_bclass(&[(b'a', b'd')])); |
| 3007 |
assert_eq!(t("(?-u)[a-d&&c-da-b]"), hir_bclass(&[(b'a', b'd')])); |
| 3008 |
|
| 3009 |
#[cfg(feature = "unicode-case")] |
| 3010 |
assert_eq!( |
| 3011 |
t("(?i)[abc&&b-c]"), |
| 3012 |
hir_case_fold(hir_uclass(&[('b', 'c')])) |
| 3013 |
); |
| 3014 |
#[cfg(feature = "unicode-case")] |
| 3015 |
assert_eq!( |
| 3016 |
t("(?i)[abc&&[b-c]]"), |
| 3017 |
hir_case_fold(hir_uclass(&[('b', 'c')])) |
| 3018 |
); |
| 3019 |
#[cfg(feature = "unicode-case")] |
| 3020 |
assert_eq!( |
| 3021 |
t("(?i)[[abc]&&[b-c]]"), |
| 3022 |
hir_case_fold(hir_uclass(&[('b', 'c')])) |
| 3023 |
); |
| 3024 |
#[cfg(feature = "unicode-case")] |
| 3025 |
assert_eq!( |
| 3026 |
t("(?i)[a-z&&b-y&&c-x]"), |
| 3027 |
hir_case_fold(hir_uclass(&[('c', 'x')])) |
| 3028 |
); |
| 3029 |
#[cfg(feature = "unicode-case")] |
| 3030 |
assert_eq!( |
| 3031 |
t("(?i)[c-da-b&&a-d]"), |
| 3032 |
hir_case_fold(hir_uclass(&[('a', 'd')])) |
| 3033 |
); |
| 3034 |
#[cfg(feature = "unicode-case")] |
| 3035 |
assert_eq!( |
| 3036 |
t("(?i)[a-d&&c-da-b]"), |
| 3037 |
hir_case_fold(hir_uclass(&[('a', 'd')])) |
| 3038 |
); |
| 3039 |
|
| 3040 |
assert_eq!( |
| 3041 |
t("(?i-u)[abc&&b-c]"), |
| 3042 |
hir_case_fold(hir_bclass(&[(b'b', b'c')])) |
| 3043 |
); |
| 3044 |
assert_eq!( |
| 3045 |
t("(?i-u)[abc&&[b-c]]"), |
| 3046 |
hir_case_fold(hir_bclass(&[(b'b', b'c')])) |
| 3047 |
); |
| 3048 |
assert_eq!( |
| 3049 |
t("(?i-u)[[abc]&&[b-c]]"), |
| 3050 |
hir_case_fold(hir_bclass(&[(b'b', b'c')])) |
| 3051 |
); |
| 3052 |
assert_eq!( |
| 3053 |
t("(?i-u)[a-z&&b-y&&c-x]"), |
| 3054 |
hir_case_fold(hir_bclass(&[(b'c', b'x')])) |
| 3055 |
); |
| 3056 |
assert_eq!( |
| 3057 |
t("(?i-u)[c-da-b&&a-d]"), |
| 3058 |
hir_case_fold(hir_bclass(&[(b'a', b'd')])) |
| 3059 |
); |
| 3060 |
assert_eq!( |
| 3061 |
t("(?i-u)[a-d&&c-da-b]"), |
| 3062 |
hir_case_fold(hir_bclass(&[(b'a', b'd')])) |
| 3063 |
); |
| 3064 |
|
| 3065 |
// In `[a^]`, `^` does not need to be escaped, so it makes sense that |
| 3066 |
// `^` is also allowed to be unescaped after `&&`. |
| 3067 |
assert_eq!(t(r"[\^&&^]"), hir_uclass(&[('^', '^')])); |
| 3068 |
// `]` needs to be escaped after `&&` since it's not at start of class. |
| 3069 |
assert_eq!(t(r"[]&&\]]"), hir_uclass(&[(']', ']')])); |
| 3070 |
assert_eq!(t(r"[-&&-]"), hir_uclass(&[('-', '-')])); |
| 3071 |
assert_eq!(t(r"[\&&&&]"), hir_uclass(&[('&', '&')])); |
| 3072 |
assert_eq!(t(r"[\&&&\&]"), hir_uclass(&[('&', '&')])); |
| 3073 |
// Test precedence. |
| 3074 |
assert_eq!( |
| 3075 |
t(r"[a-w&&[^c-g]z]"), |
| 3076 |
hir_uclass(&[('a', 'b'), ('h', 'w')]) |
| 3077 |
); |
| 3078 |
} |
| 3079 |
|
| 3080 |
#[test] |
| 3081 |
fn class_bracketed_intersect_negate() { |
| 3082 |
#[cfg(feature = "unicode-perl")] |
| 3083 |
assert_eq!( |
| 3084 |
t(r"[^\w&&\d]"), |
| 3085 |
hir_negate(hir_uclass_query(ClassQuery::Binary("digit"))) |
| 3086 |
); |
| 3087 |
assert_eq!(t(r"[^[a-z&&a-c]]"), hir_negate(hir_uclass(&[('a', 'c')]))); |
| 3088 |
#[cfg(feature = "unicode-perl")] |
| 3089 |
assert_eq!( |
| 3090 |
t(r"[^[\w&&\d]]"), |
| 3091 |
hir_negate(hir_uclass_query(ClassQuery::Binary("digit"))) |
| 3092 |
); |
| 3093 |
#[cfg(feature = "unicode-perl")] |
| 3094 |
assert_eq!( |
| 3095 |
t(r"[^[^\w&&\d]]"), |
| 3096 |
hir_uclass_query(ClassQuery::Binary("digit")) |
| 3097 |
); |
| 3098 |
#[cfg(feature = "unicode-perl")] |
| 3099 |
assert_eq!(t(r"[[[^\w]&&[^\d]]]"), hir_negate(hir_uclass_perl_word())); |
| 3100 |
|
| 3101 |
#[cfg(feature = "unicode-perl")] |
| 3102 |
assert_eq!( |
| 3103 |
t_bytes(r"(?-u)[^\w&&\d]"), |
| 3104 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Digit)) |
| 3105 |
); |
| 3106 |
assert_eq!( |
| 3107 |
t_bytes(r"(?-u)[^[a-z&&a-c]]"), |
| 3108 |
hir_negate(hir_bclass(&[(b'a', b'c')])) |
| 3109 |
); |
| 3110 |
assert_eq!( |
| 3111 |
t_bytes(r"(?-u)[^[\w&&\d]]"), |
| 3112 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Digit)) |
| 3113 |
); |
| 3114 |
assert_eq!( |
| 3115 |
t_bytes(r"(?-u)[^[^\w&&\d]]"), |
| 3116 |
hir_ascii_bclass(&ast::ClassAsciiKind::Digit) |
| 3117 |
); |
| 3118 |
assert_eq!( |
| 3119 |
t_bytes(r"(?-u)[[[^\w]&&[^\d]]]"), |
| 3120 |
hir_negate(hir_ascii_bclass(&ast::ClassAsciiKind::Word)) |
| 3121 |
); |
| 3122 |
} |
| 3123 |
|
| 3124 |
#[test] |
| 3125 |
fn class_bracketed_difference() { |
| 3126 |
#[cfg(feature = "unicode-gencat")] |
| 3127 |
assert_eq!( |
| 3128 |
t(r"[\pL--[:ascii:]]"), |
| 3129 |
hir_difference( |
| 3130 |
hir_uclass_query(ClassQuery::Binary("letter")), |
| 3131 |
hir_uclass(&[('\0', '\x7F')]) |
| 3132 |
) |
| 3133 |
); |
| 3134 |
|
| 3135 |
assert_eq!( |
| 3136 |
t(r"(?-u)[[:alpha:]--[:lower:]]"), |
| 3137 |
hir_bclass(&[(b'A', b'Z')]) |
| 3138 |
); |
| 3139 |
} |
| 3140 |
|
| 3141 |
#[test] |
| 3142 |
fn class_bracketed_symmetric_difference() { |
| 3143 |
#[cfg(feature = "unicode-script")] |
| 3144 |
assert_eq!( |
| 3145 |
t(r"[\p{sc:Greek}~~\p{scx:Greek}]"), |
| 3146 |
hir_uclass(&[ |
| 3147 |
('\u{0342}', '\u{0342}'), |
| 3148 |
('\u{0345}', '\u{0345}'), |
| 3149 |
('\u{1DC0}', '\u{1DC1}'), |
| 3150 |
]) |
| 3151 |
); |
| 3152 |
assert_eq!(t(r"[a-g~~c-j]"), hir_uclass(&[('a', 'b'), ('h', 'j')])); |
| 3153 |
|
| 3154 |
assert_eq!( |
| 3155 |
t(r"(?-u)[a-g~~c-j]"), |
| 3156 |
hir_bclass(&[(b'a', b'b'), (b'h', b'j')]) |
| 3157 |
); |
| 3158 |
} |
| 3159 |
|
| 3160 |
#[test] |
| 3161 |
fn ignore_whitespace() { |
| 3162 |
assert_eq!(t(r"(?x)\12 3"), hir_lit("\n3")); |
| 3163 |
assert_eq!(t(r"(?x)\x { 53 }"), hir_lit("S")); |
| 3164 |
assert_eq!( |
| 3165 |
t(r"(?x)\x # comment |
| 3166 |
{ # comment |
| 3167 |
53 # comment |
| 3168 |
} #comment"), |
| 3169 |
hir_lit("S") |
| 3170 |
); |
| 3171 |
|
| 3172 |
assert_eq!(t(r"(?x)\x 53"), hir_lit("S")); |
| 3173 |
assert_eq!( |
| 3174 |
t(r"(?x)\x # comment |
| 3175 |
53 # comment"), |
| 3176 |
hir_lit("S") |
| 3177 |
); |
| 3178 |
assert_eq!(t(r"(?x)\x5 3"), hir_lit("S")); |
| 3179 |
|
| 3180 |
#[cfg(feature = "unicode-gencat")] |
| 3181 |
assert_eq!( |
| 3182 |
t(r"(?x)\p # comment |
| 3183 |
{ # comment |
| 3184 |
Separator # comment |
| 3185 |
} # comment"), |
| 3186 |
hir_uclass_query(ClassQuery::Binary("separator")) |
| 3187 |
); |
| 3188 |
|
| 3189 |
assert_eq!( |
| 3190 |
t(r"(?x)a # comment |
| 3191 |
{ # comment |
| 3192 |
5 # comment |
| 3193 |
, # comment |
| 3194 |
10 # comment |
| 3195 |
} # comment"), |
| 3196 |
hir_range(true, 5, Some(10), hir_lit("a")) |
| 3197 |
); |
| 3198 |
|
| 3199 |
assert_eq!(t(r"(?x)a\ # hi there"), hir_lit("a ")); |
| 3200 |
} |
| 3201 |
|
| 3202 |
#[test] |
| 3203 |
fn analysis_is_utf8() { |
| 3204 |
// Positive examples. |
| 3205 |
assert!(props_bytes(r"a").is_utf8()); |
| 3206 |
assert!(props_bytes(r"ab").is_utf8()); |
| 3207 |
assert!(props_bytes(r"(?-u)a").is_utf8()); |
| 3208 |
assert!(props_bytes(r"(?-u)ab").is_utf8()); |
| 3209 |
assert!(props_bytes(r"\xFF").is_utf8()); |
| 3210 |
assert!(props_bytes(r"\xFF\xFF").is_utf8()); |
| 3211 |
assert!(props_bytes(r"[^a]").is_utf8()); |
| 3212 |
assert!(props_bytes(r"[^a][^a]").is_utf8()); |
| 3213 |
assert!(props_bytes(r"\b").is_utf8()); |
| 3214 |
assert!(props_bytes(r"\B").is_utf8()); |
| 3215 |
assert!(props_bytes(r"(?-u)\b").is_utf8()); |
| 3216 |
assert!(props_bytes(r"(?-u)\B").is_utf8()); |
| 3217 |
|
| 3218 |
// Negative examples. |
| 3219 |
assert!(!props_bytes(r"(?-u)\xFF").is_utf8()); |
| 3220 |
assert!(!props_bytes(r"(?-u)\xFF\xFF").is_utf8()); |
| 3221 |
assert!(!props_bytes(r"(?-u)[^a]").is_utf8()); |
| 3222 |
assert!(!props_bytes(r"(?-u)[^a][^a]").is_utf8()); |
| 3223 |
} |
| 3224 |
|
| 3225 |
#[test] |
| 3226 |
fn analysis_captures_len() { |
| 3227 |
assert_eq!(0, props(r"a").explicit_captures_len()); |
| 3228 |
assert_eq!(0, props(r"(?:a)").explicit_captures_len()); |
| 3229 |
assert_eq!(0, props(r"(?i-u:a)").explicit_captures_len()); |
| 3230 |
assert_eq!(0, props(r"(?i-u)a").explicit_captures_len()); |
| 3231 |
assert_eq!(1, props(r"(a)").explicit_captures_len()); |
| 3232 |
assert_eq!(1, props(r"(?P<foo>a)").explicit_captures_len()); |
| 3233 |
assert_eq!(1, props(r"()").explicit_captures_len()); |
| 3234 |
assert_eq!(1, props(r"()a").explicit_captures_len()); |
| 3235 |
assert_eq!(1, props(r"(a)+").explicit_captures_len()); |
| 3236 |
assert_eq!(2, props(r"(a)(b)").explicit_captures_len()); |
| 3237 |
assert_eq!(2, props(r"(a)|(b)").explicit_captures_len()); |
| 3238 |
assert_eq!(2, props(r"((a))").explicit_captures_len()); |
| 3239 |
assert_eq!(1, props(r"([a&&b])").explicit_captures_len()); |
| 3240 |
} |
| 3241 |
|
| 3242 |
#[test] |
| 3243 |
fn analysis_static_captures_len() { |
| 3244 |
let len = |pattern| props(pattern).static_explicit_captures_len(); |
| 3245 |
assert_eq!(Some(0), len(r"")); |
| 3246 |
assert_eq!(Some(0), len(r"foo|bar")); |
| 3247 |
assert_eq!(None, len(r"(foo)|bar")); |
| 3248 |
assert_eq!(None, len(r"foo|(bar)")); |
| 3249 |
assert_eq!(Some(1), len(r"(foo|bar)")); |
| 3250 |
assert_eq!(Some(1), len(r"(a|b|c|d|e|f)")); |
| 3251 |
assert_eq!(Some(1), len(r"(a)|(b)|(c)|(d)|(e)|(f)")); |
| 3252 |
assert_eq!(Some(2), len(r"(a)(b)|(c)(d)|(e)(f)")); |
| 3253 |
assert_eq!(Some(6), len(r"(a)(b)(c)(d)(e)(f)")); |
| 3254 |
assert_eq!(Some(3), len(r"(a)(b)(extra)|(a)(b)()")); |
| 3255 |
assert_eq!(Some(3), len(r"(a)(b)((?:extra)?)")); |
| 3256 |
assert_eq!(None, len(r"(a)(b)(extra)?")); |
| 3257 |
assert_eq!(Some(1), len(r"(foo)|(bar)")); |
| 3258 |
assert_eq!(Some(2), len(r"(foo)(bar)")); |
| 3259 |
assert_eq!(Some(2), len(r"(foo)+(bar)")); |
| 3260 |
assert_eq!(None, len(r"(foo)*(bar)")); |
| 3261 |
assert_eq!(Some(0), len(r"(foo)?{0}")); |
| 3262 |
assert_eq!(None, len(r"(foo)?{1}")); |
| 3263 |
assert_eq!(Some(1), len(r"(foo){1}")); |
| 3264 |
assert_eq!(Some(1), len(r"(foo){1,}")); |
| 3265 |
assert_eq!(Some(1), len(r"(foo){1,}?")); |
| 3266 |
assert_eq!(None, len(r"(foo){1,}??")); |
| 3267 |
assert_eq!(None, len(r"(foo){0,}")); |
| 3268 |
assert_eq!(Some(1), len(r"(foo)(?:bar)")); |
| 3269 |
assert_eq!(Some(2), len(r"(foo(?:bar)+)(?:baz(boo))")); |
| 3270 |
assert_eq!(Some(2), len(r"(?P<bar>foo)(?:bar)(bal|loon)")); |
| 3271 |
assert_eq!( |
| 3272 |
Some(2), |
| 3273 |
len(r#"<(a)[^>]+href="([^"]+)"|<(img)[^>]+src="([^"]+)""#) |
| 3274 |
); |
| 3275 |
} |
| 3276 |
|
| 3277 |
#[test] |
| 3278 |
fn analysis_is_all_assertions() { |
| 3279 |
// Positive examples. |
| 3280 |
let p = props(r"\b"); |
| 3281 |
assert!(!p.look_set().is_empty()); |
| 3282 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3283 |
|
| 3284 |
let p = props(r"\B"); |
| 3285 |
assert!(!p.look_set().is_empty()); |
| 3286 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3287 |
|
| 3288 |
let p = props(r"^"); |
| 3289 |
assert!(!p.look_set().is_empty()); |
| 3290 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3291 |
|
| 3292 |
let p = props(r"$"); |
| 3293 |
assert!(!p.look_set().is_empty()); |
| 3294 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3295 |
|
| 3296 |
let p = props(r"\A"); |
| 3297 |
assert!(!p.look_set().is_empty()); |
| 3298 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3299 |
|
| 3300 |
let p = props(r"\z"); |
| 3301 |
assert!(!p.look_set().is_empty()); |
| 3302 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3303 |
|
| 3304 |
let p = props(r"$^\z\A\b\B"); |
| 3305 |
assert!(!p.look_set().is_empty()); |
| 3306 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3307 |
|
| 3308 |
let p = props(r"$|^|\z|\A|\b|\B"); |
| 3309 |
assert!(!p.look_set().is_empty()); |
| 3310 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3311 |
|
| 3312 |
let p = props(r"^$|$^"); |
| 3313 |
assert!(!p.look_set().is_empty()); |
| 3314 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3315 |
|
| 3316 |
let p = props(r"((\b)+())*^"); |
| 3317 |
assert!(!p.look_set().is_empty()); |
| 3318 |
assert_eq!(p.minimum_len(), Some(0)); |
| 3319 |
|
| 3320 |
// Negative examples. |
| 3321 |
let p = props(r"^a"); |
| 3322 |
assert!(!p.look_set().is_empty()); |
| 3323 |
assert_eq!(p.minimum_len(), Some(1)); |
| 3324 |
} |
| 3325 |
|
| 3326 |
#[test] |
| 3327 |
fn analysis_look_set_prefix_any() { |
| 3328 |
let p = props(r"(?-u)(?i:(?:\b|_)win(?:32|64|dows)?(?:\b|_))"); |
| 3329 |
assert!(p.look_set_prefix_any().contains(Look::WordAscii)); |
| 3330 |
} |
| 3331 |
|
| 3332 |
#[test] |
| 3333 |
fn analysis_is_anchored() { |
| 3334 |
let is_start = |p| props(p).look_set_prefix().contains(Look::Start); |
| 3335 |
let is_end = |p| props(p).look_set_suffix().contains(Look::End); |
| 3336 |
|
| 3337 |
// Positive examples. |
| 3338 |
assert!(is_start(r"^")); |
| 3339 |
assert!(is_end(r"$")); |
| 3340 |
|
| 3341 |
assert!(is_start(r"^^")); |
| 3342 |
assert!(props(r"$$").look_set_suffix().contains(Look::End)); |
| 3343 |
|
| 3344 |
assert!(is_start(r"^$")); |
| 3345 |
assert!(is_end(r"^$")); |
| 3346 |
|
| 3347 |
assert!(is_start(r"^foo")); |
| 3348 |
assert!(is_end(r"foo$")); |
| 3349 |
|
| 3350 |
assert!(is_start(r"^foo|^bar")); |
| 3351 |
assert!(is_end(r"foo$|bar$")); |
| 3352 |
|
| 3353 |
assert!(is_start(r"^(foo|bar)")); |
| 3354 |
assert!(is_end(r"(foo|bar)$")); |
| 3355 |
|
| 3356 |
assert!(is_start(r"^+")); |
| 3357 |
assert!(is_end(r"$+")); |
| 3358 |
assert!(is_start(r"^++")); |
| 3359 |
assert!(is_end(r"$++")); |
| 3360 |
assert!(is_start(r"(^)+")); |
| 3361 |
assert!(is_end(r"($)+")); |
| 3362 |
|
| 3363 |
assert!(is_start(r"$^")); |
| 3364 |
assert!(is_start(r"$^")); |
| 3365 |
assert!(is_start(r"$^|^$")); |
| 3366 |
assert!(is_end(r"$^|^$")); |
| 3367 |
|
| 3368 |
assert!(is_start(r"\b^")); |
| 3369 |
assert!(is_end(r"$\b")); |
| 3370 |
assert!(is_start(r"^(?m:^)")); |
| 3371 |
assert!(is_end(r"(?m:$)$")); |
| 3372 |
assert!(is_start(r"(?m:^)^")); |
| 3373 |
assert!(is_end(r"$(?m:$)")); |
| 3374 |
|
| 3375 |
// Negative examples. |
| 3376 |
assert!(!is_start(r"(?m)^")); |
| 3377 |
assert!(!is_end(r"(?m)$")); |
| 3378 |
assert!(!is_start(r"(?m:^$)|$^")); |
| 3379 |
assert!(!is_end(r"(?m:^$)|$^")); |
| 3380 |
assert!(!is_start(r"$^|(?m:^$)")); |
| 3381 |
assert!(!is_end(r"$^|(?m:^$)")); |
| 3382 |
|
| 3383 |
assert!(!is_start(r"a^")); |
| 3384 |
assert!(!is_start(r"$a")); |
| 3385 |
|
| 3386 |
assert!(!is_end(r"a^")); |
| 3387 |
assert!(!is_end(r"$a")); |
| 3388 |
|
| 3389 |
assert!(!is_start(r"^foo|bar")); |
| 3390 |
assert!(!is_end(r"foo|bar$")); |
| 3391 |
|
| 3392 |
assert!(!is_start(r"^*")); |
| 3393 |
assert!(!is_end(r"$*")); |
| 3394 |
assert!(!is_start(r"^*+")); |
| 3395 |
assert!(!is_end(r"$*+")); |
| 3396 |
assert!(!is_start(r"^+*")); |
| 3397 |
assert!(!is_end(r"$+*")); |
| 3398 |
assert!(!is_start(r"(^)*")); |
| 3399 |
assert!(!is_end(r"($)*")); |
| 3400 |
} |
| 3401 |
|
| 3402 |
#[test] |
| 3403 |
fn analysis_is_any_anchored() { |
| 3404 |
let is_start = |p| props(p).look_set().contains(Look::Start); |
| 3405 |
let is_end = |p| props(p).look_set().contains(Look::End); |
| 3406 |
|
| 3407 |
// Positive examples. |
| 3408 |
assert!(is_start(r"^")); |
| 3409 |
assert!(is_end(r"$")); |
| 3410 |
assert!(is_start(r"\A")); |
| 3411 |
assert!(is_end(r"\z")); |
| 3412 |
|
| 3413 |
// Negative examples. |
| 3414 |
assert!(!is_start(r"(?m)^")); |
| 3415 |
assert!(!is_end(r"(?m)$")); |
| 3416 |
assert!(!is_start(r"$")); |
| 3417 |
assert!(!is_end(r"^")); |
| 3418 |
} |
| 3419 |
|
| 3420 |
#[test] |
| 3421 |
fn analysis_can_empty() { |
| 3422 |
// Positive examples. |
| 3423 |
let assert_empty = |
| 3424 |
|p| assert_eq!(Some(0), props_bytes(p).minimum_len()); |
| 3425 |
assert_empty(r""); |
| 3426 |
assert_empty(r"()"); |
| 3427 |
assert_empty(r"()*"); |
| 3428 |
assert_empty(r"()+"); |
| 3429 |
assert_empty(r"()?"); |
| 3430 |
assert_empty(r"a*"); |
| 3431 |
assert_empty(r"a?"); |
| 3432 |
assert_empty(r"a{0}"); |
| 3433 |
assert_empty(r"a{0,}"); |
| 3434 |
assert_empty(r"a{0,1}"); |
| 3435 |
assert_empty(r"a{0,10}"); |
| 3436 |
#[cfg(feature = "unicode-gencat")] |
| 3437 |
assert_empty(r"\pL*"); |
| 3438 |
assert_empty(r"a*|b"); |
| 3439 |
assert_empty(r"b|a*"); |
| 3440 |
assert_empty(r"a|"); |
| 3441 |
assert_empty(r"|a"); |
| 3442 |
assert_empty(r"a||b"); |
| 3443 |
assert_empty(r"a*a?(abcd)*"); |
| 3444 |
assert_empty(r"^"); |
| 3445 |
assert_empty(r"$"); |
| 3446 |
assert_empty(r"(?m)^"); |
| 3447 |
assert_empty(r"(?m)$"); |
| 3448 |
assert_empty(r"\A"); |
| 3449 |
assert_empty(r"\z"); |
| 3450 |
assert_empty(r"\B"); |
| 3451 |
assert_empty(r"(?-u)\B"); |
| 3452 |
assert_empty(r"\b"); |
| 3453 |
assert_empty(r"(?-u)\b"); |
| 3454 |
|
| 3455 |
// Negative examples. |
| 3456 |
let assert_non_empty = |
| 3457 |
|p| assert_ne!(Some(0), props_bytes(p).minimum_len()); |
| 3458 |
assert_non_empty(r"a+"); |
| 3459 |
assert_non_empty(r"a{1}"); |
| 3460 |
assert_non_empty(r"a{1,}"); |
| 3461 |
assert_non_empty(r"a{1,2}"); |
| 3462 |
assert_non_empty(r"a{1,10}"); |
| 3463 |
assert_non_empty(r"b|a"); |
| 3464 |
assert_non_empty(r"a*a+(abcd)*"); |
| 3465 |
#[cfg(feature = "unicode-gencat")] |
| 3466 |
assert_non_empty(r"\P{any}"); |
| 3467 |
assert_non_empty(r"[a--a]"); |
| 3468 |
assert_non_empty(r"[a&&b]"); |
| 3469 |
} |
| 3470 |
|
| 3471 |
#[test] |
| 3472 |
fn analysis_is_literal() { |
| 3473 |
// Positive examples. |
| 3474 |
assert!(props(r"a").is_literal()); |
| 3475 |
assert!(props(r"ab").is_literal()); |
| 3476 |
assert!(props(r"abc").is_literal()); |
| 3477 |
assert!(props(r"(?m)abc").is_literal()); |
| 3478 |
assert!(props(r"(?:a)").is_literal()); |
| 3479 |
assert!(props(r"foo(?:a)").is_literal()); |
| 3480 |
assert!(props(r"(?:a)foo").is_literal()); |
| 3481 |
assert!(props(r"[a]").is_literal()); |
| 3482 |
|
| 3483 |
// Negative examples. |
| 3484 |
assert!(!props(r"").is_literal()); |
| 3485 |
assert!(!props(r"^").is_literal()); |
| 3486 |
assert!(!props(r"a|b").is_literal()); |
| 3487 |
assert!(!props(r"(a)").is_literal()); |
| 3488 |
assert!(!props(r"a+").is_literal()); |
| 3489 |
assert!(!props(r"foo(a)").is_literal()); |
| 3490 |
assert!(!props(r"(a)foo").is_literal()); |
| 3491 |
assert!(!props(r"[ab]").is_literal()); |
| 3492 |
} |
| 3493 |
|
| 3494 |
#[test] |
| 3495 |
fn analysis_is_alternation_literal() { |
| 3496 |
// Positive examples. |
| 3497 |
assert!(props(r"a").is_alternation_literal()); |
| 3498 |
assert!(props(r"ab").is_alternation_literal()); |
| 3499 |
assert!(props(r"abc").is_alternation_literal()); |
| 3500 |
assert!(props(r"(?m)abc").is_alternation_literal()); |
| 3501 |
assert!(props(r"foo|bar").is_alternation_literal()); |
| 3502 |
assert!(props(r"foo|bar|baz").is_alternation_literal()); |
| 3503 |
assert!(props(r"[a]").is_alternation_literal()); |
| 3504 |
assert!(props(r"(?:ab)|cd").is_alternation_literal()); |
| 3505 |
assert!(props(r"ab|(?:cd)").is_alternation_literal()); |
| 3506 |
|
| 3507 |
// Negative examples. |
| 3508 |
assert!(!props(r"").is_alternation_literal()); |
| 3509 |
assert!(!props(r"^").is_alternation_literal()); |
| 3510 |
assert!(!props(r"(a)").is_alternation_literal()); |
| 3511 |
assert!(!props(r"a+").is_alternation_literal()); |
| 3512 |
assert!(!props(r"foo(a)").is_alternation_literal()); |
| 3513 |
assert!(!props(r"(a)foo").is_alternation_literal()); |
| 3514 |
assert!(!props(r"[ab]").is_alternation_literal()); |
| 3515 |
assert!(!props(r"[ab]|b").is_alternation_literal()); |
| 3516 |
assert!(!props(r"a|[ab]").is_alternation_literal()); |
| 3517 |
assert!(!props(r"(a)|b").is_alternation_literal()); |
| 3518 |
assert!(!props(r"a|(b)").is_alternation_literal()); |
| 3519 |
assert!(!props(r"a|b").is_alternation_literal()); |
| 3520 |
assert!(!props(r"a|b|c").is_alternation_literal()); |
| 3521 |
assert!(!props(r"[a]|b").is_alternation_literal()); |
| 3522 |
assert!(!props(r"a|[b]").is_alternation_literal()); |
| 3523 |
assert!(!props(r"(?:a)|b").is_alternation_literal()); |
| 3524 |
assert!(!props(r"a|(?:b)").is_alternation_literal()); |
| 3525 |
assert!(!props(r"(?:z|xx)@|xx").is_alternation_literal()); |
| 3526 |
} |
| 3527 |
|
| 3528 |
// This tests that the smart Hir::repetition constructors does some basic |
| 3529 |
// simplifications. |
| 3530 |
#[test] |
| 3531 |
fn smart_repetition() { |
| 3532 |
assert_eq!(t(r"a{0}"), Hir::empty()); |
| 3533 |
assert_eq!(t(r"a{1}"), hir_lit("a")); |
| 3534 |
assert_eq!(t(r"\B{32111}"), hir_look(hir::Look::WordUnicodeNegate)); |
| 3535 |
} |
| 3536 |
|
| 3537 |
// This tests that the smart Hir::concat constructor simplifies the given |
| 3538 |
// exprs in a way we expect. |
| 3539 |
#[test] |
| 3540 |
fn smart_concat() { |
| 3541 |
assert_eq!(t(""), Hir::empty()); |
| 3542 |
assert_eq!(t("(?:)"), Hir::empty()); |
| 3543 |
assert_eq!(t("abc"), hir_lit("abc")); |
| 3544 |
assert_eq!(t("(?:foo)(?:bar)"), hir_lit("foobar")); |
| 3545 |
assert_eq!(t("quux(?:foo)(?:bar)baz"), hir_lit("quuxfoobarbaz")); |
| 3546 |
assert_eq!( |
| 3547 |
t("foo(?:bar^baz)quux"), |
| 3548 |
hir_cat(vec![ |
| 3549 |
hir_lit("foobar"), |
| 3550 |
hir_look(hir::Look::Start), |
| 3551 |
hir_lit("bazquux"), |
| 3552 |
]) |
| 3553 |
); |
| 3554 |
assert_eq!( |
| 3555 |
t("foo(?:ba(?:r^b)az)quux"), |
| 3556 |
hir_cat(vec![ |
| 3557 |
hir_lit("foobar"), |
| 3558 |
hir_look(hir::Look::Start), |
| 3559 |
hir_lit("bazquux"), |
| 3560 |
]) |
| 3561 |
); |
| 3562 |
} |
| 3563 |
|
| 3564 |
// This tests that the smart Hir::alternation constructor simplifies the |
| 3565 |
// given exprs in a way we expect. |
| 3566 |
#[test] |
| 3567 |
fn smart_alternation() { |
| 3568 |
assert_eq!( |
| 3569 |
t("(?:foo)|(?:bar)"), |
| 3570 |
hir_alt(vec![hir_lit("foo"), hir_lit("bar")]) |
| 3571 |
); |
| 3572 |
assert_eq!( |
| 3573 |
t("quux|(?:abc|def|xyz)|baz"), |
| 3574 |
hir_alt(vec![ |
| 3575 |
hir_lit("quux"), |
| 3576 |
hir_lit("abc"), |
| 3577 |
hir_lit("def"), |
| 3578 |
hir_lit("xyz"), |
| 3579 |
hir_lit("baz"), |
| 3580 |
]) |
| 3581 |
); |
| 3582 |
assert_eq!( |
| 3583 |
t("quux|(?:abc|(?:def|mno)|xyz)|baz"), |
| 3584 |
hir_alt(vec![ |
| 3585 |
hir_lit("quux"), |
| 3586 |
hir_lit("abc"), |
| 3587 |
hir_lit("def"), |
| 3588 |
hir_lit("mno"), |
| 3589 |
hir_lit("xyz"), |
| 3590 |
hir_lit("baz"), |
| 3591 |
]) |
| 3592 |
); |
| 3593 |
assert_eq!( |
| 3594 |
t("a|b|c|d|e|f|x|y|z"), |
| 3595 |
hir_uclass(&[('a', 'f'), ('x', 'z')]), |
| 3596 |
); |
| 3597 |
// Tests that we lift common prefixes out of an alternation. |
| 3598 |
assert_eq!( |
| 3599 |
t("[A-Z]foo|[A-Z]quux"), |
| 3600 |
hir_cat(vec![ |
| 3601 |
hir_uclass(&[('A', 'Z')]), |
| 3602 |
hir_alt(vec![hir_lit("foo"), hir_lit("quux")]), |
| 3603 |
]), |
| 3604 |
); |
| 3605 |
assert_eq!( |
| 3606 |
t("[A-Z][A-Z]|[A-Z]quux"), |
| 3607 |
hir_cat(vec![ |
| 3608 |
hir_uclass(&[('A', 'Z')]), |
| 3609 |
hir_alt(vec![hir_uclass(&[('A', 'Z')]), hir_lit("quux")]), |
| 3610 |
]), |
| 3611 |
); |
| 3612 |
assert_eq!( |
| 3613 |
t("[A-Z][A-Z]|[A-Z][A-Z]quux"), |
| 3614 |
hir_cat(vec![ |
| 3615 |
hir_uclass(&[('A', 'Z')]), |
| 3616 |
hir_uclass(&[('A', 'Z')]), |
| 3617 |
hir_alt(vec![Hir::empty(), hir_lit("quux")]), |
| 3618 |
]), |
| 3619 |
); |
| 3620 |
assert_eq!( |
| 3621 |
t("[A-Z]foo|[A-Z]foobar"), |
| 3622 |
hir_cat(vec![ |
| 3623 |
hir_uclass(&[('A', 'Z')]), |
| 3624 |
hir_alt(vec![hir_lit("foo"), hir_lit("foobar")]), |
| 3625 |
]), |
| 3626 |
); |
| 3627 |
} |
| 3628 |
|
| 3629 |
#[test] |
| 3630 |
fn regression_alt_empty_concat() { |
| 3631 |
use crate::ast::{self, Ast}; |
| 3632 |
|
| 3633 |
let span = Span::splat(Position::new(0, 0, 0)); |
| 3634 |
let ast = Ast::alternation(ast::Alternation { |
| 3635 |
span, |
| 3636 |
asts: vec![Ast::concat(ast::Concat { span, asts: vec![] })], |
| 3637 |
}); |
| 3638 |
|
| 3639 |
let mut t = Translator::new(); |
| 3640 |
assert_eq!(Ok(Hir::empty()), t.translate("", &ast)); |
| 3641 |
} |
| 3642 |
|
| 3643 |
#[test] |
| 3644 |
fn regression_empty_alt() { |
| 3645 |
use crate::ast::{self, Ast}; |
| 3646 |
|
| 3647 |
let span = Span::splat(Position::new(0, 0, 0)); |
| 3648 |
let ast = Ast::concat(ast::Concat { |
| 3649 |
span, |
| 3650 |
asts: vec![Ast::alternation(ast::Alternation { |
| 3651 |
span, |
| 3652 |
asts: vec![], |
| 3653 |
})], |
| 3654 |
}); |
| 3655 |
|
| 3656 |
let mut t = Translator::new(); |
| 3657 |
assert_eq!(Ok(Hir::fail()), t.translate("", &ast)); |
| 3658 |
} |
| 3659 |
|
| 3660 |
#[test] |
| 3661 |
fn regression_singleton_alt() { |
| 3662 |
use crate::{ |
| 3663 |
ast::{self, Ast}, |
| 3664 |
hir::Dot, |
| 3665 |
}; |
| 3666 |
|
| 3667 |
let span = Span::splat(Position::new(0, 0, 0)); |
| 3668 |
let ast = Ast::concat(ast::Concat { |
| 3669 |
span, |
| 3670 |
asts: vec![Ast::alternation(ast::Alternation { |
| 3671 |
span, |
| 3672 |
asts: vec![Ast::dot(span)], |
| 3673 |
})], |
| 3674 |
}); |
| 3675 |
|
| 3676 |
let mut t = Translator::new(); |
| 3677 |
assert_eq!(Ok(Hir::dot(Dot::AnyCharExceptLF)), t.translate("", &ast)); |
| 3678 |
} |
| 3679 |
|
| 3680 |
// See: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=63168 |
| 3681 |
#[test] |
| 3682 |
fn regression_fuzz_match() { |
| 3683 |
let pat = "[(\u{6} \0-\u{afdf5}] \0 "; |
| 3684 |
let ast = ParserBuilder::new() |
| 3685 |
.octal(false) |
| 3686 |
.ignore_whitespace(true) |
| 3687 |
.build() |
| 3688 |
.parse(pat) |
| 3689 |
.unwrap(); |
| 3690 |
let hir = TranslatorBuilder::new() |
| 3691 |
.utf8(true) |
| 3692 |
.case_insensitive(false) |
| 3693 |
.multi_line(false) |
| 3694 |
.dot_matches_new_line(false) |
| 3695 |
.swap_greed(true) |
| 3696 |
.unicode(true) |
| 3697 |
.build() |
| 3698 |
.translate(pat, &ast) |
| 3699 |
.unwrap(); |
| 3700 |
assert_eq!( |
| 3701 |
hir, |
| 3702 |
Hir::concat(vec![ |
| 3703 |
hir_uclass(&[('\0', '\u{afdf5}')]), |
| 3704 |
hir_lit("\0"), |
| 3705 |
]) |
| 3706 |
); |
| 3707 |
} |
| 3708 |
|
| 3709 |
// See: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=63155 |
| 3710 |
#[cfg(feature = "unicode")] |
| 3711 |
#[test] |
| 3712 |
fn regression_fuzz_difference1() { |
| 3713 |
let pat = r"\W\W|\W[^\v--\W\W\P{Script_Extensions:Pau_Cin_Hau}\u10A1A1-\U{3E3E3}--~~~~--~~~~~~~~------~~~~~~--~~~~~~]*"; |
| 3714 |
let _ = t(pat); // shouldn't panic |
| 3715 |
} |
| 3716 |
|
| 3717 |
// See: https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=63153 |
| 3718 |
#[test] |
| 3719 |
fn regression_fuzz_char_decrement1() { |
| 3720 |
let pat = "w[w[^w?\rw\rw[^w?\rw[^w?\rw[^w?\rw[^w?\rw[^w?\rw[^w?\r\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0w?\rw[^w?\rw[^w?\rw[^w\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\u{1}\0]\0\0\0\0\0\0\0\0\0*\0\0\u{1}\0]\0\0-*\0][^w?\rw[^w?\rw[^w?\rw[^w?\rw[^w?\rw[^w?\rw[^w\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\u{1}\0]\0\0\0\0\0\0\0\0\0x\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\0\0\0\0*??\0\u{7f}{2}\u{10}??\0\0\0\0\0\0\0\0\0\u{3}\0\0\0}\0-*\0]\0\0\0\0\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\u{1}\0]\0\0-*\0]\0\0\0\0\0\0\0\u{1}\0]\0\u{1}\u{1}H-i]-]\0\0\0\0\u{1}\0]\0\0\0\u{1}\0]\0\0-*\0\0\0\0\u{1}9-\u{7f}]\0'|-\u{7f}]\0'|(?i-ux)[-\u{7f}]\0'\u{3}\0\0\0}\0-*\0]<D\0\0\0\0\0\0\u{1}]\0\0\0\0]\0\0-*\0]\0\0 "; |
| 3721 |
let _ = t(pat); // shouldn't panic |
| 3722 |
} |
| 3723 |
} |