PageRenderTime 83ms CodeModel.GetById 18ms app.highlight 46ms RepoModel.GetById 1ms app.codeStats 0ms

/Doc/howto/regex.rst

http://unladen-swallow.googlecode.com/
ReStructuredText | 1372 lines | 1080 code | 292 blank | 0 comment | 0 complexity | 79cb0b616de7bd5d3ee085f71b56d1bc MD5 | raw file

Large files files are truncated, but you can click here to view the full file

   1.. _regex-howto:
   2
   3****************************
   4  Regular Expression HOWTO
   5****************************
   6
   7:Author: A.M. Kuchling <amk@amk.ca>
   8:Release: 0.05
   9
  10.. TODO:
  11   Document lookbehind assertions
  12   Better way of displaying a RE, a string, and what it matches
  13   Mention optional argument to match.groups()
  14   Unicode (at least a reference)
  15
  16
  17.. topic:: Abstract
  18
  19   This document is an introductory tutorial to using regular expressions in Python
  20   with the :mod:`re` module.  It provides a gentler introduction than the
  21   corresponding section in the Library Reference.
  22
  23
  24Introduction
  25============
  26
  27The :mod:`re` module was added in Python 1.5, and provides Perl-style regular
  28expression patterns.  Earlier versions of Python came with the :mod:`regex`
  29module, which provided Emacs-style patterns.  The :mod:`regex` module was
  30removed completely in Python 2.5.
  31
  32Regular expressions (called REs, or regexes, or regex patterns) are essentially
  33a tiny, highly specialized programming language embedded inside Python and made
  34available through the :mod:`re` module. Using this little language, you specify
  35the rules for the set of possible strings that you want to match; this set might
  36contain English sentences, or e-mail addresses, or TeX commands, or anything you
  37like.  You can then ask questions such as "Does this string match the pattern?",
  38or "Is there a match for the pattern anywhere in this string?".  You can also
  39use REs to modify a string or to split it apart in various ways.
  40
  41Regular expression patterns are compiled into a series of bytecodes which are
  42then executed by a matching engine written in C.  For advanced use, it may be
  43necessary to pay careful attention to how the engine will execute a given RE,
  44and write the RE in a certain way in order to produce bytecode that runs faster.
  45Optimization isn't covered in this document, because it requires that you have a
  46good understanding of the matching engine's internals.
  47
  48The regular expression language is relatively small and restricted, so not all
  49possible string processing tasks can be done using regular expressions.  There
  50are also tasks that *can* be done with regular expressions, but the expressions
  51turn out to be very complicated.  In these cases, you may be better off writing
  52Python code to do the processing; while Python code will be slower than an
  53elaborate regular expression, it will also probably be more understandable.
  54
  55
  56Simple Patterns
  57===============
  58
  59We'll start by learning about the simplest possible regular expressions.  Since
  60regular expressions are used to operate on strings, we'll begin with the most
  61common task: matching characters.
  62
  63For a detailed explanation of the computer science underlying regular
  64expressions (deterministic and non-deterministic finite automata), you can refer
  65to almost any textbook on writing compilers.
  66
  67
  68Matching Characters
  69-------------------
  70
  71Most letters and characters will simply match themselves.  For example, the
  72regular expression ``test`` will match the string ``test`` exactly.  (You can
  73enable a case-insensitive mode that would let this RE match ``Test`` or ``TEST``
  74as well; more about this later.)
  75
  76There are exceptions to this rule; some characters are special
  77:dfn:`metacharacters`, and don't match themselves.  Instead, they signal that
  78some out-of-the-ordinary thing should be matched, or they affect other portions
  79of the RE by repeating them or changing their meaning.  Much of this document is
  80devoted to discussing various metacharacters and what they do.
  81
  82Here's a complete list of the metacharacters; their meanings will be discussed
  83in the rest of this HOWTO. ::
  84
  85   . ^ $ * + ? { [ ] \ | ( )
  86
  87The first metacharacters we'll look at are ``[`` and ``]``. They're used for
  88specifying a character class, which is a set of characters that you wish to
  89match.  Characters can be listed individually, or a range of characters can be
  90indicated by giving two characters and separating them by a ``'-'``.  For
  91example, ``[abc]`` will match any of the characters ``a``, ``b``, or ``c``; this
  92is the same as ``[a-c]``, which uses a range to express the same set of
  93characters.  If you wanted to match only lowercase letters, your RE would be
  94``[a-z]``.
  95
  96Metacharacters are not active inside classes.  For example, ``[akm$]`` will
  97match any of the characters ``'a'``, ``'k'``, ``'m'``, or ``'$'``; ``'$'`` is
  98usually a metacharacter, but inside a character class it's stripped of its
  99special nature.
 100
 101You can match the characters not listed within the class by :dfn:`complementing`
 102the set.  This is indicated by including a ``'^'`` as the first character of the
 103class; ``'^'`` outside a character class will simply match the ``'^'``
 104character.  For example, ``[^5]`` will match any character except ``'5'``.
 105
 106Perhaps the most important metacharacter is the backslash, ``\``.   As in Python
 107string literals, the backslash can be followed by various characters to signal
 108various special sequences.  It's also used to escape all the metacharacters so
 109you can still match them in patterns; for example, if you need to match a ``[``
 110or  ``\``, you can precede them with a backslash to remove their special
 111meaning: ``\[`` or ``\\``.
 112
 113Some of the special sequences beginning with ``'\'`` represent predefined sets
 114of characters that are often useful, such as the set of digits, the set of
 115letters, or the set of anything that isn't whitespace.  The following predefined
 116special sequences are available:
 117
 118``\d``
 119   Matches any decimal digit; this is equivalent to the class ``[0-9]``.
 120
 121``\D``
 122   Matches any non-digit character; this is equivalent to the class ``[^0-9]``.
 123
 124``\s``
 125   Matches any whitespace character; this is equivalent to the class ``[
 126   \t\n\r\f\v]``.
 127
 128``\S``
 129   Matches any non-whitespace character; this is equivalent to the class ``[^
 130   \t\n\r\f\v]``.
 131
 132``\w``
 133   Matches any alphanumeric character; this is equivalent to the class
 134   ``[a-zA-Z0-9_]``.
 135
 136``\W``
 137   Matches any non-alphanumeric character; this is equivalent to the class
 138   ``[^a-zA-Z0-9_]``.
 139
 140These sequences can be included inside a character class.  For example,
 141``[\s,.]`` is a character class that will match any whitespace character, or
 142``','`` or ``'.'``.
 143
 144The final metacharacter in this section is ``.``.  It matches anything except a
 145newline character, and there's an alternate mode (``re.DOTALL``) where it will
 146match even a newline.  ``'.'`` is often used where you want to match "any
 147character".
 148
 149
 150Repeating Things
 151----------------
 152
 153Being able to match varying sets of characters is the first thing regular
 154expressions can do that isn't already possible with the methods available on
 155strings.  However, if that was the only additional capability of regexes, they
 156wouldn't be much of an advance. Another capability is that you can specify that
 157portions of the RE must be repeated a certain number of times.
 158
 159The first metacharacter for repeating things that we'll look at is ``*``.  ``*``
 160doesn't match the literal character ``*``; instead, it specifies that the
 161previous character can be matched zero or more times, instead of exactly once.
 162
 163For example, ``ca*t`` will match ``ct`` (0 ``a`` characters), ``cat`` (1 ``a``),
 164``caaat`` (3 ``a`` characters), and so forth.  The RE engine has various
 165internal limitations stemming from the size of C's ``int`` type that will
 166prevent it from matching over 2 billion ``a`` characters; you probably don't
 167have enough memory to construct a string that large, so you shouldn't run into
 168that limit.
 169
 170Repetitions such as ``*`` are :dfn:`greedy`; when repeating a RE, the matching
 171engine will try to repeat it as many times as possible. If later portions of the
 172pattern don't match, the matching engine will then back up and try again with
 173few repetitions.
 174
 175A step-by-step example will make this more obvious.  Let's consider the
 176expression ``a[bcd]*b``.  This matches the letter ``'a'``, zero or more letters
 177from the class ``[bcd]``, and finally ends with a ``'b'``.  Now imagine matching
 178this RE against the string ``abcbd``.
 179
 180+------+-----------+---------------------------------+
 181| Step | Matched   | Explanation                     |
 182+======+===========+=================================+
 183| 1    | ``a``     | The ``a`` in the RE matches.    |
 184+------+-----------+---------------------------------+
 185| 2    | ``abcbd`` | The engine matches ``[bcd]*``,  |
 186|      |           | going as far as it can, which   |
 187|      |           | is to the end of the string.    |
 188+------+-----------+---------------------------------+
 189| 3    | *Failure* | The engine tries to match       |
 190|      |           | ``b``, but the current position |
 191|      |           | is at the end of the string, so |
 192|      |           | it fails.                       |
 193+------+-----------+---------------------------------+
 194| 4    | ``abcb``  | Back up, so that  ``[bcd]*``    |
 195|      |           | matches one less character.     |
 196+------+-----------+---------------------------------+
 197| 5    | *Failure* | Try ``b`` again, but the        |
 198|      |           | current position is at the last |
 199|      |           | character, which is a ``'d'``.  |
 200+------+-----------+---------------------------------+
 201| 6    | ``abc``   | Back up again, so that          |
 202|      |           | ``[bcd]*`` is only matching     |
 203|      |           | ``bc``.                         |
 204+------+-----------+---------------------------------+
 205| 6    | ``abcb``  | Try ``b`` again.  This time     |
 206|      |           | the character at the            |
 207|      |           | current position is ``'b'``, so |
 208|      |           | it succeeds.                    |
 209+------+-----------+---------------------------------+
 210
 211The end of the RE has now been reached, and it has matched ``abcb``.  This
 212demonstrates how the matching engine goes as far as it can at first, and if no
 213match is found it will then progressively back up and retry the rest of the RE
 214again and again.  It will back up until it has tried zero matches for
 215``[bcd]*``, and if that subsequently fails, the engine will conclude that the
 216string doesn't match the RE at all.
 217
 218Another repeating metacharacter is ``+``, which matches one or more times.  Pay
 219careful attention to the difference between ``*`` and ``+``; ``*`` matches
 220*zero* or more times, so whatever's being repeated may not be present at all,
 221while ``+`` requires at least *one* occurrence.  To use a similar example,
 222``ca+t`` will match ``cat`` (1 ``a``), ``caaat`` (3 ``a``'s), but won't match
 223``ct``.
 224
 225There are two more repeating qualifiers.  The question mark character, ``?``,
 226matches either once or zero times; you can think of it as marking something as
 227being optional.  For example, ``home-?brew`` matches either ``homebrew`` or
 228``home-brew``.
 229
 230The most complicated repeated qualifier is ``{m,n}``, where *m* and *n* are
 231decimal integers.  This qualifier means there must be at least *m* repetitions,
 232and at most *n*.  For example, ``a/{1,3}b`` will match ``a/b``, ``a//b``, and
 233``a///b``.  It won't match ``ab``, which has no slashes, or ``a////b``, which
 234has four.
 235
 236You can omit either *m* or *n*; in that case, a reasonable value is assumed for
 237the missing value.  Omitting *m* is interpreted as a lower limit of 0, while
 238omitting *n* results in an upper bound of infinity --- actually, the upper bound
 239is the 2-billion limit mentioned earlier, but that might as well be infinity.
 240
 241Readers of a reductionist bent may notice that the three other qualifiers can
 242all be expressed using this notation.  ``{0,}`` is the same as ``*``, ``{1,}``
 243is equivalent to ``+``, and ``{0,1}`` is the same as ``?``.  It's better to use
 244``*``, ``+``, or ``?`` when you can, simply because they're shorter and easier
 245to read.
 246
 247
 248Using Regular Expressions
 249=========================
 250
 251Now that we've looked at some simple regular expressions, how do we actually use
 252them in Python?  The :mod:`re` module provides an interface to the regular
 253expression engine, allowing you to compile REs into objects and then perform
 254matches with them.
 255
 256
 257Compiling Regular Expressions
 258-----------------------------
 259
 260Regular expressions are compiled into :class:`RegexObject` instances, which have
 261methods for various operations such as searching for pattern matches or
 262performing string substitutions. ::
 263
 264   >>> import re
 265   >>> p = re.compile('ab*')
 266   >>> print p
 267   <_sre.SRE_Pattern object at 80b4150>
 268
 269:func:`re.compile` also accepts an optional *flags* argument, used to enable
 270various special features and syntax variations.  We'll go over the available
 271settings later, but for now a single example will do::
 272
 273   >>> p = re.compile('ab*', re.IGNORECASE)
 274
 275The RE is passed to :func:`re.compile` as a string.  REs are handled as strings
 276because regular expressions aren't part of the core Python language, and no
 277special syntax was created for expressing them.  (There are applications that
 278don't need REs at all, so there's no need to bloat the language specification by
 279including them.) Instead, the :mod:`re` module is simply a C extension module
 280included with Python, just like the :mod:`socket` or :mod:`zlib` modules.
 281
 282Putting REs in strings keeps the Python language simpler, but has one
 283disadvantage which is the topic of the next section.
 284
 285
 286The Backslash Plague
 287--------------------
 288
 289As stated earlier, regular expressions use the backslash character (``'\'``) to
 290indicate special forms or to allow special characters to be used without
 291invoking their special meaning. This conflicts with Python's usage of the same
 292character for the same purpose in string literals.
 293
 294Let's say you want to write a RE that matches the string ``\section``, which
 295might be found in a LaTeX file.  To figure out what to write in the program
 296code, start with the desired string to be matched.  Next, you must escape any
 297backslashes and other metacharacters by preceding them with a backslash,
 298resulting in the string ``\\section``.  The resulting string that must be passed
 299to :func:`re.compile` must be ``\\section``.  However, to express this as a
 300Python string literal, both backslashes must be escaped *again*.
 301
 302+-------------------+------------------------------------------+
 303| Characters        | Stage                                    |
 304+===================+==========================================+
 305| ``\section``      | Text string to be matched                |
 306+-------------------+------------------------------------------+
 307| ``\\section``     | Escaped backslash for :func:`re.compile` |
 308+-------------------+------------------------------------------+
 309| ``"\\\\section"`` | Escaped backslashes for a string literal |
 310+-------------------+------------------------------------------+
 311
 312In short, to match a literal backslash, one has to write ``'\\\\'`` as the RE
 313string, because the regular expression must be ``\\``, and each backslash must
 314be expressed as ``\\`` inside a regular Python string literal.  In REs that
 315feature backslashes repeatedly, this leads to lots of repeated backslashes and
 316makes the resulting strings difficult to understand.
 317
 318The solution is to use Python's raw string notation for regular expressions;
 319backslashes are not handled in any special way in a string literal prefixed with
 320``'r'``, so ``r"\n"`` is a two-character string containing ``'\'`` and ``'n'``,
 321while ``"\n"`` is a one-character string containing a newline. Regular
 322expressions will often be written in Python code using this raw string notation.
 323
 324+-------------------+------------------+
 325| Regular String    | Raw string       |
 326+===================+==================+
 327| ``"ab*"``         | ``r"ab*"``       |
 328+-------------------+------------------+
 329| ``"\\\\section"`` | ``r"\\section"`` |
 330+-------------------+------------------+
 331| ``"\\w+\\s+\\1"`` | ``r"\w+\s+\1"``  |
 332+-------------------+------------------+
 333
 334
 335Performing Matches
 336------------------
 337
 338Once you have an object representing a compiled regular expression, what do you
 339do with it?  :class:`RegexObject` instances have several methods and attributes.
 340Only the most significant ones will be covered here; consult the :mod:`re` docs
 341for a complete listing.
 342
 343+------------------+-----------------------------------------------+
 344| Method/Attribute | Purpose                                       |
 345+==================+===============================================+
 346| ``match()``      | Determine if the RE matches at the beginning  |
 347|                  | of the string.                                |
 348+------------------+-----------------------------------------------+
 349| ``search()``     | Scan through a string, looking for any        |
 350|                  | location where this RE matches.               |
 351+------------------+-----------------------------------------------+
 352| ``findall()``    | Find all substrings where the RE matches, and |
 353|                  | returns them as a list.                       |
 354+------------------+-----------------------------------------------+
 355| ``finditer()``   | Find all substrings where the RE matches, and |
 356|                  | returns them as an :term:`iterator`.          |
 357+------------------+-----------------------------------------------+
 358
 359:meth:`match` and :meth:`search` return ``None`` if no match can be found.  If
 360they're successful, a ``MatchObject`` instance is returned, containing
 361information about the match: where it starts and ends, the substring it matched,
 362and more.
 363
 364You can learn about this by interactively experimenting with the :mod:`re`
 365module.  If you have Tkinter available, you may also want to look at
 366:file:`Tools/scripts/redemo.py`, a demonstration program included with the
 367Python distribution.  It allows you to enter REs and strings, and displays
 368whether the RE matches or fails. :file:`redemo.py` can be quite useful when
 369trying to debug a complicated RE.  Phil Schwartz's `Kodos
 370<http://kodos.sourceforge.net/>`_ is also an interactive tool for developing and
 371testing RE patterns.
 372
 373This HOWTO uses the standard Python interpreter for its examples. First, run the
 374Python interpreter, import the :mod:`re` module, and compile a RE::
 375
 376   Python 2.2.2 (#1, Feb 10 2003, 12:57:01)
 377   >>> import re
 378   >>> p = re.compile('[a-z]+')
 379   >>> p
 380   <_sre.SRE_Pattern object at 80c3c28>
 381
 382Now, you can try matching various strings against the RE ``[a-z]+``.  An empty
 383string shouldn't match at all, since ``+`` means 'one or more repetitions'.
 384:meth:`match` should return ``None`` in this case, which will cause the
 385interpreter to print no output.  You can explicitly print the result of
 386:meth:`match` to make this clear. ::
 387
 388   >>> p.match("")
 389   >>> print p.match("")
 390   None
 391
 392Now, let's try it on a string that it should match, such as ``tempo``.  In this
 393case, :meth:`match` will return a :class:`MatchObject`, so you should store the
 394result in a variable for later use. ::
 395
 396   >>> m = p.match('tempo')
 397   >>> print m
 398   <_sre.SRE_Match object at 80c4f68>
 399
 400Now you can query the :class:`MatchObject` for information about the matching
 401string.   :class:`MatchObject` instances also have several methods and
 402attributes; the most important ones are:
 403
 404+------------------+--------------------------------------------+
 405| Method/Attribute | Purpose                                    |
 406+==================+============================================+
 407| ``group()``      | Return the string matched by the RE        |
 408+------------------+--------------------------------------------+
 409| ``start()``      | Return the starting position of the match  |
 410+------------------+--------------------------------------------+
 411| ``end()``        | Return the ending position of the match    |
 412+------------------+--------------------------------------------+
 413| ``span()``       | Return a tuple containing the (start, end) |
 414|                  | positions  of the match                    |
 415+------------------+--------------------------------------------+
 416
 417Trying these methods will soon clarify their meaning::
 418
 419   >>> m.group()
 420   'tempo'
 421   >>> m.start(), m.end()
 422   (0, 5)
 423   >>> m.span()
 424   (0, 5)
 425
 426:meth:`group` returns the substring that was matched by the RE.  :meth:`start`
 427and :meth:`end` return the starting and ending index of the match. :meth:`span`
 428returns both start and end indexes in a single tuple.  Since the :meth:`match`
 429method only checks if the RE matches at the start of a string, :meth:`start`
 430will always be zero.  However, the :meth:`search` method of :class:`RegexObject`
 431instances scans through the string, so  the match may not start at zero in that
 432case. ::
 433
 434   >>> print p.match('::: message')
 435   None
 436   >>> m = p.search('::: message') ; print m
 437   <re.MatchObject instance at 80c9650>
 438   >>> m.group()
 439   'message'
 440   >>> m.span()
 441   (4, 11)
 442
 443In actual programs, the most common style is to store the :class:`MatchObject`
 444in a variable, and then check if it was ``None``.  This usually looks like::
 445
 446   p = re.compile( ... )
 447   m = p.match( 'string goes here' )
 448   if m:
 449       print 'Match found: ', m.group()
 450   else:
 451       print 'No match'
 452
 453Two :class:`RegexObject` methods return all of the matches for a pattern.
 454:meth:`findall` returns a list of matching strings::
 455
 456   >>> p = re.compile('\d+')
 457   >>> p.findall('12 drummers drumming, 11 pipers piping, 10 lords a-leaping')
 458   ['12', '11', '10']
 459
 460:meth:`findall` has to create the entire list before it can be returned as the
 461result.  The :meth:`finditer` method returns a sequence of :class:`MatchObject`
 462instances as an :term:`iterator`. [#]_ ::
 463
 464   >>> iterator = p.finditer('12 drummers drumming, 11 ... 10 ...')
 465   >>> iterator
 466   <callable-iterator object at 0x401833ac>
 467   >>> for match in iterator:
 468   ...     print match.span()
 469   ...
 470   (0, 2)
 471   (22, 24)
 472   (29, 31)
 473
 474
 475Module-Level Functions
 476----------------------
 477
 478You don't have to create a :class:`RegexObject` and call its methods; the
 479:mod:`re` module also provides top-level functions called :func:`match`,
 480:func:`search`, :func:`findall`, :func:`sub`, and so forth.  These functions
 481take the same arguments as the corresponding :class:`RegexObject` method, with
 482the RE string added as the first argument, and still return either ``None`` or a
 483:class:`MatchObject` instance. ::
 484
 485   >>> print re.match(r'From\s+', 'Fromage amk')
 486   None
 487   >>> re.match(r'From\s+', 'From amk Thu May 14 19:12:10 1998')
 488   <re.MatchObject instance at 80c5978>
 489
 490Under the hood, these functions simply produce a :class:`RegexObject` for you
 491and call the appropriate method on it.  They also store the compiled object in a
 492cache, so future calls using the same RE are faster.
 493
 494Should you use these module-level functions, or should you get the
 495:class:`RegexObject` and call its methods yourself?  That choice depends on how
 496frequently the RE will be used, and on your personal coding style.  If the RE is
 497being used at only one point in the code, then the module functions are probably
 498more convenient.  If a program contains a lot of regular expressions, or re-uses
 499the same ones in several locations, then it might be worthwhile to collect all
 500the definitions in one place, in a section of code that compiles all the REs
 501ahead of time.  To take an example from the standard library, here's an extract
 502from :file:`xmllib.py`::
 503
 504   ref = re.compile( ... )
 505   entityref = re.compile( ... )
 506   charref = re.compile( ... )
 507   starttagopen = re.compile( ... )
 508
 509I generally prefer to work with the compiled object, even for one-time uses, but
 510few people will be as much of a purist about this as I am.
 511
 512
 513Compilation Flags
 514-----------------
 515
 516Compilation flags let you modify some aspects of how regular expressions work.
 517Flags are available in the :mod:`re` module under two names, a long name such as
 518:const:`IGNORECASE` and a short, one-letter form such as :const:`I`.  (If you're
 519familiar with Perl's pattern modifiers, the one-letter forms use the same
 520letters; the short form of :const:`re.VERBOSE` is :const:`re.X`, for example.)
 521Multiple flags can be specified by bitwise OR-ing them; ``re.I | re.M`` sets
 522both the :const:`I` and :const:`M` flags, for example.
 523
 524Here's a table of the available flags, followed by a more detailed explanation
 525of each one.
 526
 527+---------------------------------+--------------------------------------------+
 528| Flag                            | Meaning                                    |
 529+=================================+============================================+
 530| :const:`DOTALL`, :const:`S`     | Make ``.`` match any character, including  |
 531|                                 | newlines                                   |
 532+---------------------------------+--------------------------------------------+
 533| :const:`IGNORECASE`, :const:`I` | Do case-insensitive matches                |
 534+---------------------------------+--------------------------------------------+
 535| :const:`LOCALE`, :const:`L`     | Do a locale-aware match                    |
 536+---------------------------------+--------------------------------------------+
 537| :const:`MULTILINE`, :const:`M`  | Multi-line matching, affecting ``^`` and   |
 538|                                 | ``$``                                      |
 539+---------------------------------+--------------------------------------------+
 540| :const:`VERBOSE`, :const:`X`    | Enable verbose REs, which can be organized |
 541|                                 | more cleanly and understandably.           |
 542+---------------------------------+--------------------------------------------+
 543| :const:`UNICODE`, :const:`U`    | Makes several escapes like ``\w``, ``\b``, |
 544|                                 | ``\s`` and ``\d`` dependent on the Unicode |
 545|                                 | character database.                        |
 546+---------------------------------+--------------------------------------------+
 547
 548
 549.. data:: I
 550          IGNORECASE
 551   :noindex:
 552
 553   Perform case-insensitive matching; character class and literal strings will
 554   match letters by ignoring case.  For example, ``[A-Z]`` will match lowercase
 555   letters, too, and ``Spam`` will match ``Spam``, ``spam``, or ``spAM``. This
 556   lowercasing doesn't take the current locale into account; it will if you also
 557   set the :const:`LOCALE` flag.
 558
 559
 560.. data:: L
 561          LOCALE
 562   :noindex:
 563
 564   Make ``\w``, ``\W``, ``\b``, and ``\B``, dependent on the current locale.
 565
 566   Locales are a feature of the C library intended to help in writing programs that
 567   take account of language differences.  For example, if you're processing French
 568   text, you'd want to be able to write ``\w+`` to match words, but ``\w`` only
 569   matches the character class ``[A-Za-z]``; it won't match ``'é'`` or ``'ç'``.  If
 570   your system is configured properly and a French locale is selected, certain C
 571   functions will tell the program that ``'é'`` should also be considered a letter.
 572   Setting the :const:`LOCALE` flag when compiling a regular expression will cause
 573   the resulting compiled object to use these C functions for ``\w``; this is
 574   slower, but also enables ``\w+`` to match French words as you'd expect.
 575
 576
 577.. data:: M
 578          MULTILINE
 579   :noindex:
 580
 581   (``^`` and ``$`` haven't been explained yet;  they'll be introduced in section
 582   :ref:`more-metacharacters`.)
 583
 584   Usually ``^`` matches only at the beginning of the string, and ``$`` matches
 585   only at the end of the string and immediately before the newline (if any) at the
 586   end of the string. When this flag is specified, ``^`` matches at the beginning
 587   of the string and at the beginning of each line within the string, immediately
 588   following each newline.  Similarly, the ``$`` metacharacter matches either at
 589   the end of the string and at the end of each line (immediately preceding each
 590   newline).
 591
 592
 593.. data:: S
 594          DOTALL
 595   :noindex:
 596
 597   Makes the ``'.'`` special character match any character at all, including a
 598   newline; without this flag, ``'.'`` will match anything *except* a newline.
 599
 600
 601.. data:: U
 602          UNICODE
 603   :noindex:
 604
 605   Make ``\w``, ``\W``, ``\b``, ``\B``, ``\d``, ``\D``, ``\s`` and ``\S``
 606   dependent on the Unicode character properties database.
 607
 608
 609.. data:: X
 610          VERBOSE
 611   :noindex:
 612
 613   This flag allows you to write regular expressions that are more readable by
 614   granting you more flexibility in how you can format them.  When this flag has
 615   been specified, whitespace within the RE string is ignored, except when the
 616   whitespace is in a character class or preceded by an unescaped backslash; this
 617   lets you organize and indent the RE more clearly.  This flag also lets you put
 618   comments within a RE that will be ignored by the engine; comments are marked by
 619   a ``'#'`` that's neither in a character class or preceded by an unescaped
 620   backslash.
 621
 622   For example, here's a RE that uses :const:`re.VERBOSE`; see how much easier it
 623   is to read? ::
 624
 625      charref = re.compile(r"""
 626       &[#]                # Start of a numeric entity reference
 627       (
 628           0[0-7]+         # Octal form
 629         | [0-9]+          # Decimal form
 630         | x[0-9a-fA-F]+   # Hexadecimal form
 631       )
 632       ;                   # Trailing semicolon
 633      """, re.VERBOSE)
 634
 635   Without the verbose setting, the RE would look like this::
 636
 637      charref = re.compile("&#(0[0-7]+"
 638                           "|[0-9]+"
 639                           "|x[0-9a-fA-F]+);")
 640
 641   In the above example, Python's automatic concatenation of string literals has
 642   been used to break up the RE into smaller pieces, but it's still more difficult
 643   to understand than the version using :const:`re.VERBOSE`.
 644
 645
 646More Pattern Power
 647==================
 648
 649So far we've only covered a part of the features of regular expressions.  In
 650this section, we'll cover some new metacharacters, and how to use groups to
 651retrieve portions of the text that was matched.
 652
 653
 654.. _more-metacharacters:
 655
 656More Metacharacters
 657-------------------
 658
 659There are some metacharacters that we haven't covered yet.  Most of them will be
 660covered in this section.
 661
 662Some of the remaining metacharacters to be discussed are :dfn:`zero-width
 663assertions`.  They don't cause the engine to advance through the string;
 664instead, they consume no characters at all, and simply succeed or fail.  For
 665example, ``\b`` is an assertion that the current position is located at a word
 666boundary; the position isn't changed by the ``\b`` at all.  This means that
 667zero-width assertions should never be repeated, because if they match once at a
 668given location, they can obviously be matched an infinite number of times.
 669
 670``|``
 671   Alternation, or the "or" operator.   If A and B are regular expressions,
 672   ``A|B`` will match any string that matches either ``A`` or ``B``. ``|`` has very
 673   low precedence in order to make it work reasonably when you're alternating
 674   multi-character strings. ``Crow|Servo`` will match either ``Crow`` or ``Servo``,
 675   not ``Cro``, a ``'w'`` or an ``'S'``, and ``ervo``.
 676
 677   To match a literal ``'|'``, use ``\|``, or enclose it inside a character class,
 678   as in ``[|]``.
 679
 680``^``
 681   Matches at the beginning of lines.  Unless the :const:`MULTILINE` flag has been
 682   set, this will only match at the beginning of the string.  In :const:`MULTILINE`
 683   mode, this also matches immediately after each newline within the string.
 684
 685   For example, if you wish to match the word ``From`` only at the beginning of a
 686   line, the RE to use is ``^From``. ::
 687
 688      >>> print re.search('^From', 'From Here to Eternity')
 689      <re.MatchObject instance at 80c1520>
 690      >>> print re.search('^From', 'Reciting From Memory')
 691      None
 692
 693   .. To match a literal \character{\^}, use \regexp{\e\^} or enclose it
 694   .. inside a character class, as in \regexp{[{\e}\^]}.
 695
 696``$``
 697   Matches at the end of a line, which is defined as either the end of the string,
 698   or any location followed by a newline character.     ::
 699
 700      >>> print re.search('}$', '{block}')
 701      <re.MatchObject instance at 80adfa8>
 702      >>> print re.search('}$', '{block} ')
 703      None
 704      >>> print re.search('}$', '{block}\n')
 705      <re.MatchObject instance at 80adfa8>
 706
 707   To match a literal ``'$'``, use ``\$`` or enclose it inside a character class,
 708   as in  ``[$]``.
 709
 710``\A``
 711   Matches only at the start of the string.  When not in :const:`MULTILINE` mode,
 712   ``\A`` and ``^`` are effectively the same.  In :const:`MULTILINE` mode, they're
 713   different: ``\A`` still matches only at the beginning of the string, but ``^``
 714   may match at any location inside the string that follows a newline character.
 715
 716``\Z``
 717   Matches only at the end of the string.
 718
 719``\b``
 720   Word boundary.  This is a zero-width assertion that matches only at the
 721   beginning or end of a word.  A word is defined as a sequence of alphanumeric
 722   characters, so the end of a word is indicated by whitespace or a
 723   non-alphanumeric character.
 724
 725   The following example matches ``class`` only when it's a complete word; it won't
 726   match when it's contained inside another word. ::
 727
 728      >>> p = re.compile(r'\bclass\b')
 729      >>> print p.search('no class at all')
 730      <re.MatchObject instance at 80c8f28>
 731      >>> print p.search('the declassified algorithm')
 732      None
 733      >>> print p.search('one subclass is')
 734      None
 735
 736   There are two subtleties you should remember when using this special sequence.
 737   First, this is the worst collision between Python's string literals and regular
 738   expression sequences.  In Python's string literals, ``\b`` is the backspace
 739   character, ASCII value 8.  If you're not using raw strings, then Python will
 740   convert the ``\b`` to a backspace, and your RE won't match as you expect it to.
 741   The following example looks the same as our previous RE, but omits the ``'r'``
 742   in front of the RE string. ::
 743
 744      >>> p = re.compile('\bclass\b')
 745      >>> print p.search('no class at all')
 746      None
 747      >>> print p.search('\b' + 'class' + '\b')
 748      <re.MatchObject instance at 80c3ee0>
 749
 750   Second, inside a character class, where there's no use for this assertion,
 751   ``\b`` represents the backspace character, for compatibility with Python's
 752   string literals.
 753
 754``\B``
 755   Another zero-width assertion, this is the opposite of ``\b``, only matching when
 756   the current position is not at a word boundary.
 757
 758
 759Grouping
 760--------
 761
 762Frequently you need to obtain more information than just whether the RE matched
 763or not.  Regular expressions are often used to dissect strings by writing a RE
 764divided into several subgroups which match different components of interest.
 765For example, an RFC-822 header line is divided into a header name and a value,
 766separated by a ``':'``, like this::
 767
 768   From: author@example.com
 769   User-Agent: Thunderbird 1.5.0.9 (X11/20061227)
 770   MIME-Version: 1.0
 771   To: editor@example.com
 772
 773This can be handled by writing a regular expression which matches an entire
 774header line, and has one group which matches the header name, and another group
 775which matches the header's value.
 776
 777Groups are marked by the ``'('``, ``')'`` metacharacters. ``'('`` and ``')'``
 778have much the same meaning as they do in mathematical expressions; they group
 779together the expressions contained inside them, and you can repeat the contents
 780of a group with a repeating qualifier, such as ``*``, ``+``, ``?``, or
 781``{m,n}``.  For example, ``(ab)*`` will match zero or more repetitions of
 782``ab``. ::
 783
 784   >>> p = re.compile('(ab)*')
 785   >>> print p.match('ababababab').span()
 786   (0, 10)
 787
 788Groups indicated with ``'('``, ``')'`` also capture the starting and ending
 789index of the text that they match; this can be retrieved by passing an argument
 790to :meth:`group`, :meth:`start`, :meth:`end`, and :meth:`span`.  Groups are
 791numbered starting with 0.  Group 0 is always present; it's the whole RE, so
 792:class:`MatchObject` methods all have group 0 as their default argument.  Later
 793we'll see how to express groups that don't capture the span of text that they
 794match. ::
 795
 796   >>> p = re.compile('(a)b')
 797   >>> m = p.match('ab')
 798   >>> m.group()
 799   'ab'
 800   >>> m.group(0)
 801   'ab'
 802
 803Subgroups are numbered from left to right, from 1 upward.  Groups can be nested;
 804to determine the number, just count the opening parenthesis characters, going
 805from left to right. ::
 806
 807   >>> p = re.compile('(a(b)c)d')
 808   >>> m = p.match('abcd')
 809   >>> m.group(0)
 810   'abcd'
 811   >>> m.group(1)
 812   'abc'
 813   >>> m.group(2)
 814   'b'
 815
 816:meth:`group` can be passed multiple group numbers at a time, in which case it
 817will return a tuple containing the corresponding values for those groups. ::
 818
 819   >>> m.group(2,1,2)
 820   ('b', 'abc', 'b')
 821
 822The :meth:`groups` method returns a tuple containing the strings for all the
 823subgroups, from 1 up to however many there are. ::
 824
 825   >>> m.groups()
 826   ('abc', 'b')
 827
 828Backreferences in a pattern allow you to specify that the contents of an earlier
 829capturing group must also be found at the current location in the string.  For
 830example, ``\1`` will succeed if the exact contents of group 1 can be found at
 831the current position, and fails otherwise.  Remember that Python's string
 832literals also use a backslash followed by numbers to allow including arbitrary
 833characters in a string, so be sure to use a raw string when incorporating
 834backreferences in a RE.
 835
 836For example, the following RE detects doubled words in a string. ::
 837
 838   >>> p = re.compile(r'(\b\w+)\s+\1')
 839   >>> p.search('Paris in the the spring').group()
 840   'the the'
 841
 842Backreferences like this aren't often useful for just searching through a string
 843--- there are few text formats which repeat data in this way --- but you'll soon
 844find out that they're *very* useful when performing string substitutions.
 845
 846
 847Non-capturing and Named Groups
 848------------------------------
 849
 850Elaborate REs may use many groups, both to capture substrings of interest, and
 851to group and structure the RE itself.  In complex REs, it becomes difficult to
 852keep track of the group numbers.  There are two features which help with this
 853problem.  Both of them use a common syntax for regular expression extensions, so
 854we'll look at that first.
 855
 856Perl 5 added several additional features to standard regular expressions, and
 857the Python :mod:`re` module supports most of them.   It would have been
 858difficult to choose new single-keystroke metacharacters or new special sequences
 859beginning with ``\`` to represent the new features without making Perl's regular
 860expressions confusingly different from standard REs.  If you chose ``&`` as a
 861new metacharacter, for example, old expressions would be assuming that ``&`` was
 862a regular character and wouldn't have escaped it by writing ``\&`` or ``[&]``.
 863
 864The solution chosen by the Perl developers was to use ``(?...)`` as the
 865extension syntax.  ``?`` immediately after a parenthesis was a syntax error
 866because the ``?`` would have nothing to repeat, so this didn't introduce any
 867compatibility problems.  The characters immediately after the ``?``  indicate
 868what extension is being used, so ``(?=foo)`` is one thing (a positive lookahead
 869assertion) and ``(?:foo)`` is something else (a non-capturing group containing
 870the subexpression ``foo``).
 871
 872Python adds an extension syntax to Perl's extension syntax.  If the first
 873character after the question mark is a ``P``, you know that it's an extension
 874that's specific to Python.  Currently there are two such extensions:
 875``(?P<name>...)`` defines a named group, and ``(?P=name)`` is a backreference to
 876a named group.  If future versions of Perl 5 add similar features using a
 877different syntax, the :mod:`re` module will be changed to support the new
 878syntax, while preserving the Python-specific syntax for compatibility's sake.
 879
 880Now that we've looked at the general extension syntax, we can return to the
 881features that simplify working with groups in complex REs. Since groups are
 882numbered from left to right and a complex expression may use many groups, it can
 883become difficult to keep track of the correct numbering.  Modifying such a
 884complex RE is annoying, too: insert a new group near the beginning and you
 885change the numbers of everything that follows it.
 886
 887Sometimes you'll want to use a group to collect a part of a regular expression,
 888but aren't interested in retrieving the group's contents. You can make this fact
 889explicit by using a non-capturing group: ``(?:...)``, where you can replace the
 890``...`` with any other regular expression. ::
 891
 892   >>> m = re.match("([abc])+", "abc")
 893   >>> m.groups()
 894   ('c',)
 895   >>> m = re.match("(?:[abc])+", "abc")
 896   >>> m.groups()
 897   ()
 898
 899Except for the fact that you can't retrieve the contents of what the group
 900matched, a non-capturing group behaves exactly the same as a capturing group;
 901you can put anything inside it, repeat it with a repetition metacharacter such
 902as ``*``, and nest it within other groups (capturing or non-capturing).
 903``(?:...)`` is particularly useful when modifying an existing pattern, since you
 904can add new groups without changing how all the other groups are numbered.  It
 905should be mentioned that there's no performance difference in searching between
 906capturing and non-capturing groups; neither form is any faster than the other.
 907
 908A more significant feature is named groups: instead of referring to them by
 909numbers, groups can be referenced by a name.
 910
 911The syntax for a named group is one of the Python-specific extensions:
 912``(?P<name>...)``.  *name* is, obviously, the name of the group.  Named groups
 913also behave exactly like capturing groups, and additionally associate a name
 914with a group.  The :class:`MatchObject` methods that deal with capturing groups
 915all accept either integers that refer to the group by number or strings that
 916contain the desired group's name.  Named groups are still given numbers, so you
 917can retrieve information about a group in two ways::
 918
 919   >>> p = re.compile(r'(?P<word>\b\w+\b)')
 920   >>> m = p.search( '(((( Lots of punctuation )))' )
 921   >>> m.group('word')
 922   'Lots'
 923   >>> m.group(1)
 924   'Lots'
 925
 926Named groups are handy because they let you use easily-remembered names, instead
 927of having to remember numbers.  Here's an example RE from the :mod:`imaplib`
 928module::
 929
 930   InternalDate = re.compile(r'INTERNALDATE "'
 931           r'(?P<day>[ 123][0-9])-(?P<mon>[A-Z][a-z][a-z])-'
 932           r'(?P<year>[0-9][0-9][0-9][0-9])'
 933           r' (?P<hour>[0-9][0-9]):(?P<min>[0-9][0-9]):(?P<sec>[0-9][0-9])'
 934           r' (?P<zonen>[-+])(?P<zoneh>[0-9][0-9])(?P<zonem>[0-9][0-9])'
 935           r'"')
 936
 937It's obviously much easier to retrieve ``m.group('zonem')``, instead of having
 938to remember to retrieve group 9.
 939
 940The syntax for backreferences in an expression such as ``(...)\1`` refers to the
 941number of the group.  There's naturally a variant that uses the group name
 942instead of the number. This is another Python extension: ``(?P=name)`` indicates
 943that the contents of the group called *name* should again be matched at the
 944current point.  The regular expression for finding doubled words,
 945``(\b\w+)\s+\1`` can also be written as ``(?P<word>\b\w+)\s+(?P=word)``::
 946
 947   >>> p = re.compile(r'(?P<word>\b\w+)\s+(?P=word)')
 948   >>> p.search('Paris in the the spring').group()
 949   'the the'
 950
 951
 952Lookahead Assertions
 953--------------------
 954
 955Another zero-width assertion is the lookahead assertion.  Lookahead assertions
 956are available in both positive and negative form, and  look like this:
 957
 958``(?=...)``
 959   Positive lookahead assertion.  This succeeds if the contained regular
 960   expression, represented here by ``...``, successfully matches at the current
 961   location, and fails otherwise. But, once the contained expression has been
 962   tried, the matching engine doesn't advance at all; the rest of the pattern is
 963   tried right where the assertion started.
 964
 965``(?!...)``
 966   Negative lookahead assertion.  This is the opposite of the positive assertion;
 967   it succeeds if the contained expression *doesn't* match at the current position
 968   in the string.
 969
 970To make this concrete, let's look at a case where a lookahead is useful.
 971Consider a simple pattern to match a filename and split it apart into a base
 972name and an extension, separated by a ``.``.  For example, in ``news.rc``,
 973``news`` is the base name, and ``rc`` is the filename's extension.
 974
 975The pattern to match this is quite simple:
 976
 977``.*[.].*$``
 978
 979Notice that the ``.`` needs to be treated specially because it's a
 980metacharacter; I've put it inside a character class.  Also notice the trailing
 981``$``; this is added to ensure that all the rest of the string must be included
 982in the extension.  This regular expression matches ``foo.bar`` and
 983``autoexec.bat`` and ``sendmail.cf`` and ``printers.conf``.
 984
 985Now, consider complicating the problem a bit; what if you want to match
 986filenames where the extension is not ``bat``? Some incorrect attempts:
 987
 988``.*[.][^b].*$``  The first attempt above tries to exclude ``bat`` by requiring
 989that the first character of the extension is not a ``b``.  This is wrong,
 990because the pattern also doesn't match ``foo.bar``.
 991
 992``.*[.]([^b]..|.[^a].|..[^t])$``
 993
 994The expression gets messier when you try to patch up the first solution by
 995requiring one of the following cases to match: the first character of the
 996extension isn't ``b``; the second character isn't ``a``; or the third character
 997isn't ``t``.  This accepts ``foo.bar`` and rejects ``autoexec.bat``, but it
 998requires a three-letter extension and won't accept a filename with a two-letter
 999extension such as ``sendmail.cf``.  We'll complicate the pattern again in an
1000effort to fix it.
1001
1002``.*[.]([^b].?.?|.[^a]?.?|..?[^t]?)$``
1003
1004In the third attempt, the second and third letters are all made optional in
1005order to allow matching extensions shorter than three characters, such as
1006``sendmail.cf``.
1007
1008The pattern's getting really complicated now, which makes it hard to read and
1009understand.  Worse, if the problem changes and you want to exclude both ``bat``
1010and ``exe`` as extensions, the pattern would get even more complicated and
1011confusing.
1012
1013A negative lookahead cuts through all this confusion:
1014
1015``.*[.](?!bat$).*$``  The negative lookahead means: if the expression ``bat``
1016doesn't match at this point, try the rest of the pattern; if ``bat$`` does
1017match, the whole pattern will fail.  The trailing ``$`` is required to ensure
1018that something like ``sample.batch``, where the extension only starts with
1019``bat``, will be allowed.
1020
1021Excluding another filename extension is now easy; simply add it as an
1022alternative inside the assertion.  The following pattern excludes filenames that
1023end in either ``bat`` or ``exe``:
1024
1025``.*[.](?!bat$|exe$).*$``
1026
1027
1028Modifying Strings
1029=================
1030
1031Up to this point, we've simply performed searches against a static string.
1032Regular expressions are also commonly used to modify strings in various ways,
1033using the following :class:`RegexObject` methods:
1034
1035+------------------+-----------------------------------------------+
1036| Method/Attribute | Purpose                                       |
1037+==================+===============================================+
1038| ``split()``      | Split the string into a list, splitting it    |
1039|                  | wherever the RE matches                       |
1040+------------------+-----------------------------------------------+
1041| ``sub()``        | Find all substrings where the RE matches, and |
1042|                  | replace them with a different string          |
1043+------------------+-----------------------------------------------+
1044| ``subn()``       | Does the same thing as :meth:`sub`,  but      |
1045|                  | returns the new string and the number of      |
1046|                  | replacements                                  |
1047+------------------+-----------------------------------------------+
1048
1049
1050Splitting Strings
1051-----------------
1052
1053The :meth:`split` method of a :class:`RegexObject` splits a string apart
1054wherever the RE matches, returning a list of the pieces. It's similar to the
1055:meth:`split` method of strings but provides much more generality in the
1056delimiters that you can split by; :meth:`split` only supports splitting by
1057whitespace or by a fixed string.  As you'd expect, there's a module-level
1058:func:`re.split` function, too.
1059
1060
1061.. method:: .split(string [, maxsplit=0])
1062   :noindex:
1063
1064   Split *string* by the matches of the regular expression.  If capturing
1065   parentheses are used in the RE, then their contents will also be returned as
1066   part of the resulting list.  If *maxsplit* is nonzero, at most *maxsplit* splits
1067   are performed.
1068
1069You can limit the number of splits made, by passing a value for *maxsplit*.
1070When *maxsplit* is nonzero, at most *maxsplit* splits will be made, and the
1071remainder of the string is returned as the final element of the list.  In the
1072following example, the delimiter is any sequence of non-alphanumeric characters.
1073::
1074
1075   >>> p = re.compile(r'\W+')
1076   >>> p.split('This is a test, short and sweet, of split().')
1077   ['This', 'is', 'a', 'test', 'short', 'and', 'sweet', 'of', 'split', '']
1078   >>> p.split('This is a test, short and sweet, of split().', 3)
1079   ['This', 'is', 'a', 'test, short and sweet, of split().']
1080
1081Sometimes you're not only interested in what the text between delimiters is, but
1082also need to know what the delimiter was.  If capturing parentheses are used in
1083the RE, then their values are also returned as part of the list.  Compare the
1084following calls::
1085
1086   >>> p = re.compile(r'\W+')
1087   >>> p2 = re.compile(r'(\W+)')
1088   >>> p.split('This... is a test.')
1089   ['This', 'is', 'a', 'test', '']
1090   >>> p2.split('This... is a test.')
1091   ['This', '... ', 'is', ' ', 'a', ' ', 'test', '.', '']
1092
1093The module-level function :func:`re.split` adds the RE to be used as the first
1094argument, but is otherwise the same.   ::
1095
1096   >>> re.split('[\W]+', 'Words, words, words.')
1097   ['Words', 'words', 'words', '']
1098   >>> re.split('([\W]+)', 'Words, words, words.')
1099   ['Words', ', ', 'words', ', ', 'words', '.', '']
1100   >>> re.split('[\W]+', 'Words, words, words.', 1)
1101   ['Words', 'words, words.']
1102
1103
1104Search and Replace
1105------------------
1106
1107Another common task is to find all the matches for a pattern, and replace them
1108with a different string.  The :meth:`sub` method takes a replacement value,
1109which can be either a string or a function, and the string to be processed.
1110
1111
1112.. method:: .sub(replacement, string[, count=0])
1113   :noindex:
1114
1115   Retu…

Large files files are truncated, but you can click here to view the full file