fileset: make it robust for bad function calls

Before, it could crash or show cryptic message.

mercurial/fileset.py

@@ -1,634 +1,639 b''
 # fileset.py - file set queries for mercurial
 #
 # Copyright 2010 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import re
 
 from .i18n import _
 from . import (
     error,
     merge,
     parser,
     pycompat,
     registrar,
     scmutil,
     util,
 )
 
 elements = {
     # token-type: binding-strength, primary, prefix, infix, suffix
     "(": (20, None, ("group", 1, ")"), ("func", 1, ")"), None),
     "-": (5, None, ("negate", 19), ("minus", 5), None),
     "not": (10, None, ("not", 10), None, None),
     "!": (10, None, ("not", 10), None, None),
     "and": (5, None, None, ("and", 5), None),
     "&": (5, None, None, ("and", 5), None),
     "or": (4, None, None, ("or", 4), None),
     "|": (4, None, None, ("or", 4), None),
     "+": (4, None, None, ("or", 4), None),
     ",": (2, None, None, ("list", 2), None),
     ")": (0, None, None, None, None),
     "symbol": (0, "symbol", None, None, None),
     "string": (0, "string", None, None, None),
     "end": (0, None, None, None, None),
 }
 
 keywords = {'and', 'or', 'not'}
 
 globchars = ".*{}[]?/\\_"
 
 def tokenize(program):
     pos, l = 0, len(program)
     program = pycompat.bytestr(program)
     while pos < l:
         c = program[pos]
         if c.isspace(): # skip inter-token whitespace
             pass
         elif c in "(),-|&+!": # handle simple operators
             yield (c, None, pos)
         elif (c in '"\'' or c == 'r' and
               program[pos:pos + 2] in ("r'", 'r"')): # handle quoted strings
             if c == 'r':
                 pos += 1
                 c = program[pos]
                 decode = lambda x: x
             else:
                 decode = parser.unescapestr
             pos += 1
             s = pos
             while pos < l: # find closing quote
                 d = program[pos]
                 if d == '\\': # skip over escaped characters
                     pos += 2
                     continue
                 if d == c:
                     yield ('string', decode(program[s:pos]), s)
                     break
                 pos += 1
             else:
                 raise error.ParseError(_("unterminated string"), s)
         elif c.isalnum() or c in globchars or ord(c) > 127:
             # gather up a symbol/keyword
             s = pos
             pos += 1
             while pos < l: # find end of symbol
                 d = program[pos]
                 if not (d.isalnum() or d in globchars or ord(d) > 127):
                     break
                 pos += 1
             sym = program[s:pos]
             if sym in keywords: # operator keywords
                 yield (sym, None, s)
             else:
                 yield ('symbol', sym, s)
             pos -= 1
         else:
             raise error.ParseError(_("syntax error"), pos)
         pos += 1
     yield ('end', None, pos)
 
 def parse(expr):
     p = parser.parser(elements)
     tree, pos = p.parse(tokenize(expr))
     if pos != len(expr):
         raise error.ParseError(_("invalid token"), pos)
     return tree
 
+def getsymbol(x):
+    if x and x[0] == 'symbol':
+        return x[1]
+    raise error.ParseError(_('not a symbol'))
+
 def getstring(x, err):
     if x and (x[0] == 'string' or x[0] == 'symbol'):
         return x[1]
     raise error.ParseError(err)
 
 def getset(mctx, x):
     if not x:
         raise error.ParseError(_("missing argument"))
     return methods[x[0]](mctx, *x[1:])
 
 def stringset(mctx, x):
     m = mctx.matcher([x])
     return [f for f in mctx.subset if m(f)]
 
 def andset(mctx, x, y):
     return getset(mctx.narrow(getset(mctx, x)), y)
 
 def orset(mctx, x, y):
     # needs optimizing
     xl = getset(mctx, x)
     yl = getset(mctx, y)
     return xl + [f for f in yl if f not in xl]
 
 def notset(mctx, x):
     s = set(getset(mctx, x))
     return [r for r in mctx.subset if r not in s]
 
 def minusset(mctx, x, y):
     xl = getset(mctx, x)
     yl = set(getset(mctx, y))
     return [f for f in xl if f not in yl]
 
 def listset(mctx, a, b):
     raise error.ParseError(_("can't use a list in this context"),
                            hint=_('see hg help "filesets.x or y"'))
 
 # symbols are callable like:
 #  fun(mctx, x)
 # with:
 #  mctx - current matchctx instance
 #  x - argument in tree form
 symbols = {}
 
 # filesets using matchctx.status()
 _statuscallers = set()
 
 # filesets using matchctx.existing()
 _existingcallers = set()
 
 predicate = registrar.filesetpredicate()
 
 @predicate('modified()', callstatus=True)
 def modified(mctx, x):
     """File that is modified according to :hg:`status`.
     """
     # i18n: "modified" is a keyword
     getargs(x, 0, 0, _("modified takes no arguments"))
     s = set(mctx.status().modified)
     return [f for f in mctx.subset if f in s]
 
 @predicate('added()', callstatus=True)
 def added(mctx, x):
     """File that is added according to :hg:`status`.
     """
     # i18n: "added" is a keyword
     getargs(x, 0, 0, _("added takes no arguments"))
     s = set(mctx.status().added)
     return [f for f in mctx.subset if f in s]
 
 @predicate('removed()', callstatus=True)
 def removed(mctx, x):
     """File that is removed according to :hg:`status`.
     """
     # i18n: "removed" is a keyword
     getargs(x, 0, 0, _("removed takes no arguments"))
     s = set(mctx.status().removed)
     return [f for f in mctx.subset if f in s]
 
 @predicate('deleted()', callstatus=True)
 def deleted(mctx, x):
     """Alias for ``missing()``.
     """
     # i18n: "deleted" is a keyword
     getargs(x, 0, 0, _("deleted takes no arguments"))
     s = set(mctx.status().deleted)
     return [f for f in mctx.subset if f in s]
 
 @predicate('missing()', callstatus=True)
 def missing(mctx, x):
     """File that is missing according to :hg:`status`.
     """
     # i18n: "missing" is a keyword
     getargs(x, 0, 0, _("missing takes no arguments"))
     s = set(mctx.status().deleted)
     return [f for f in mctx.subset if f in s]
 
 @predicate('unknown()', callstatus=True)
 def unknown(mctx, x):
     """File that is unknown according to :hg:`status`. These files will only be
     considered if this predicate is used.
     """
     # i18n: "unknown" is a keyword
     getargs(x, 0, 0, _("unknown takes no arguments"))
     s = set(mctx.status().unknown)
     return [f for f in mctx.subset if f in s]
 
 @predicate('ignored()', callstatus=True)
 def ignored(mctx, x):
     """File that is ignored according to :hg:`status`. These files will only be
     considered if this predicate is used.
     """
     # i18n: "ignored" is a keyword
     getargs(x, 0, 0, _("ignored takes no arguments"))
     s = set(mctx.status().ignored)
     return [f for f in mctx.subset if f in s]
 
 @predicate('clean()', callstatus=True)
 def clean(mctx, x):
     """File that is clean according to :hg:`status`.
     """
     # i18n: "clean" is a keyword
     getargs(x, 0, 0, _("clean takes no arguments"))
     s = set(mctx.status().clean)
     return [f for f in mctx.subset if f in s]
 
 def func(mctx, a, b):
-    if a[0] == 'symbol' and a[1] in symbols:
-        funcname = a[1]
+    funcname = getsymbol(a)
+    if funcname in symbols:
         enabled = mctx._existingenabled
         mctx._existingenabled = funcname in _existingcallers
         try:
             return symbols[funcname](mctx, b)
         finally:
             mctx._existingenabled = enabled
 
     keep = lambda fn: getattr(fn, '__doc__', None) is not None
 
     syms = [s for (s, fn) in symbols.items() if keep(fn)]
-    raise error.UnknownIdentifier(a[1], syms)
+    raise error.UnknownIdentifier(funcname, syms)
 
 def getlist(x):
     if not x:
         return []
     if x[0] == 'list':
         return getlist(x[1]) + [x[2]]
     return [x]
 
 def getargs(x, min, max, err):
     l = getlist(x)
     if len(l) < min or len(l) > max:
         raise error.ParseError(err)
     return l
 
 @predicate('binary()', callexisting=True)
 def binary(mctx, x):
     """File that appears to be binary (contains NUL bytes).
     """
     # i18n: "binary" is a keyword
     getargs(x, 0, 0, _("binary takes no arguments"))
     return [f for f in mctx.existing() if mctx.ctx[f].isbinary()]
 
 @predicate('exec()', callexisting=True)
 def exec_(mctx, x):
     """File that is marked as executable.
     """
     # i18n: "exec" is a keyword
     getargs(x, 0, 0, _("exec takes no arguments"))
     return [f for f in mctx.existing() if mctx.ctx.flags(f) == 'x']
 
 @predicate('symlink()', callexisting=True)
 def symlink(mctx, x):
     """File that is marked as a symlink.
     """
     # i18n: "symlink" is a keyword
     getargs(x, 0, 0, _("symlink takes no arguments"))
     return [f for f in mctx.existing() if mctx.ctx.flags(f) == 'l']
 
 @predicate('resolved()')
 def resolved(mctx, x):
     """File that is marked resolved according to :hg:`resolve -l`.
     """
     # i18n: "resolved" is a keyword
     getargs(x, 0, 0, _("resolved takes no arguments"))
     if mctx.ctx.rev() is not None:
         return []
     ms = merge.mergestate.read(mctx.ctx.repo())
     return [f for f in mctx.subset if f in ms and ms[f] == 'r']
 
 @predicate('unresolved()')
 def unresolved(mctx, x):
     """File that is marked unresolved according to :hg:`resolve -l`.
     """
     # i18n: "unresolved" is a keyword
     getargs(x, 0, 0, _("unresolved takes no arguments"))
     if mctx.ctx.rev() is not None:
         return []
     ms = merge.mergestate.read(mctx.ctx.repo())
     return [f for f in mctx.subset if f in ms and ms[f] == 'u']
 
 @predicate('hgignore()')
 def hgignore(mctx, x):
     """File that matches the active .hgignore pattern.
     """
     # i18n: "hgignore" is a keyword
     getargs(x, 0, 0, _("hgignore takes no arguments"))
     ignore = mctx.ctx.repo().dirstate._ignore
     return [f for f in mctx.subset if ignore(f)]
 
 @predicate('portable()')
 def portable(mctx, x):
     """File that has a portable name. (This doesn't include filenames with case
     collisions.)
     """
     # i18n: "portable" is a keyword
     getargs(x, 0, 0, _("portable takes no arguments"))
     checkwinfilename = util.checkwinfilename
     return [f for f in mctx.subset if checkwinfilename(f) is None]
 
 @predicate('grep(regex)', callexisting=True)
 def grep(mctx, x):
     """File contains the given regular expression.
     """
     try:
         # i18n: "grep" is a keyword
         r = re.compile(getstring(x, _("grep requires a pattern")))
     except re.error as e:
         raise error.ParseError(_('invalid match pattern: %s') % e)
     return [f for f in mctx.existing() if r.search(mctx.ctx[f].data())]
 
 def _sizetomax(s):
     try:
         s = s.strip().lower()
         for k, v in util._sizeunits:
             if s.endswith(k):
                 # max(4k) = 5k - 1, max(4.5k) = 4.6k - 1
                 n = s[:-len(k)]
                 inc = 1.0
                 if "." in n:
                     inc /= 10 ** len(n.split(".")[1])
                 return int((float(n) + inc) * v) - 1
         # no extension, this is a precise value
         return int(s)
     except ValueError:
         raise error.ParseError(_("couldn't parse size: %s") % s)
 
 def sizematcher(x):
     """Return a function(size) -> bool from the ``size()`` expression"""
 
     # i18n: "size" is a keyword
     expr = getstring(x, _("size requires an expression")).strip()
     if '-' in expr: # do we have a range?
         a, b = expr.split('-', 1)
         a = util.sizetoint(a)
         b = util.sizetoint(b)
         return lambda x: x >= a and x <= b
     elif expr.startswith("<="):
         a = util.sizetoint(expr[2:])
         return lambda x: x <= a
     elif expr.startswith("<"):
         a = util.sizetoint(expr[1:])
         return lambda x: x < a
     elif expr.startswith(">="):
         a = util.sizetoint(expr[2:])
         return lambda x: x >= a
     elif expr.startswith(">"):
         a = util.sizetoint(expr[1:])
         return lambda x: x > a
     elif expr[0].isdigit or expr[0] == '.':
         a = util.sizetoint(expr)
         b = _sizetomax(expr)
         return lambda x: x >= a and x <= b
     raise error.ParseError(_("couldn't parse size: %s") % expr)
 
 @predicate('size(expression)', callexisting=True)
 def size(mctx, x):
     """File size matches the given expression. Examples:
 
     - size('1k') - files from 1024 to 2047 bytes
     - size('< 20k') - files less than 20480 bytes
     - size('>= .5MB') - files at least 524288 bytes
     - size('4k - 1MB') - files from 4096 bytes to 1048576 bytes
     """
     m = sizematcher(x)
     return [f for f in mctx.existing() if m(mctx.ctx[f].size())]
 
 @predicate('encoding(name)', callexisting=True)
 def encoding(mctx, x):
     """File can be successfully decoded with the given character
     encoding. May not be useful for encodings other than ASCII and
     UTF-8.
     """
 
     # i18n: "encoding" is a keyword
     enc = getstring(x, _("encoding requires an encoding name"))
 
     s = []
     for f in mctx.existing():
         d = mctx.ctx[f].data()
         try:
             d.decode(enc)
         except LookupError:
             raise error.Abort(_("unknown encoding '%s'") % enc)
         except UnicodeDecodeError:
             continue
         s.append(f)
 
     return s
 
 @predicate('eol(style)', callexisting=True)
 def eol(mctx, x):
     """File contains newlines of the given style (dos, unix, mac). Binary
     files are excluded, files with mixed line endings match multiple
     styles.
     """
 
     # i18n: "eol" is a keyword
     enc = getstring(x, _("eol requires a style name"))
 
     s = []
     for f in mctx.existing():
         d = mctx.ctx[f].data()
         if util.binary(d):
             continue
         if (enc == 'dos' or enc == 'win') and '\r\n' in d:
             s.append(f)
         elif enc == 'unix' and re.search('(?<!\r)\n', d):
             s.append(f)
         elif enc == 'mac' and re.search('\r(?!\n)', d):
             s.append(f)
     return s
 
 @predicate('copied()')
 def copied(mctx, x):
     """File that is recorded as being copied.
     """
     # i18n: "copied" is a keyword
     getargs(x, 0, 0, _("copied takes no arguments"))
     s = []
     for f in mctx.subset:
         p = mctx.ctx[f].parents()
         if p and p[0].path() != f:
             s.append(f)
     return s
 
 @predicate('revs(revs, pattern)')
 def revs(mctx, x):
     """Evaluate set in the specified revisions. If the revset match multiple
     revs, this will return file matching pattern in any of the revision.
     """
     # i18n: "revs" is a keyword
     r, x = getargs(x, 2, 2, _("revs takes two arguments"))
     # i18n: "revs" is a keyword
     revspec = getstring(r, _("first argument to revs must be a revision"))
     repo = mctx.ctx.repo()
     revs = scmutil.revrange(repo, [revspec])
 
     found = set()
     result = []
     for r in revs:
         ctx = repo[r]
         for f in getset(mctx.switch(ctx, _buildstatus(ctx, x)), x):
             if f not in found:
                 found.add(f)
                 result.append(f)
     return result
 
 @predicate('status(base, rev, pattern)')
 def status(mctx, x):
     """Evaluate predicate using status change between ``base`` and
     ``rev``. Examples:
 
     - ``status(3, 7, added())`` - matches files added from "3" to "7"
     """
     repo = mctx.ctx.repo()
     # i18n: "status" is a keyword
     b, r, x = getargs(x, 3, 3, _("status takes three arguments"))
     # i18n: "status" is a keyword
     baseerr = _("first argument to status must be a revision")
     baserevspec = getstring(b, baseerr)
     if not baserevspec:
         raise error.ParseError(baseerr)
     reverr = _("second argument to status must be a revision")
     revspec = getstring(r, reverr)
     if not revspec:
         raise error.ParseError(reverr)
     basenode, node = scmutil.revpair(repo, [baserevspec, revspec])
     basectx = repo[basenode]
     ctx = repo[node]
     return getset(mctx.switch(ctx, _buildstatus(ctx, x, basectx=basectx)), x)
 
 @predicate('subrepo([pattern])')
 def subrepo(mctx, x):
     """Subrepositories whose paths match the given pattern.
     """
     # i18n: "subrepo" is a keyword
     getargs(x, 0, 1, _("subrepo takes at most one argument"))
     ctx = mctx.ctx
     sstate = sorted(ctx.substate)
     if x:
         # i18n: "subrepo" is a keyword
         pat = getstring(x, _("subrepo requires a pattern or no arguments"))
 
         from . import match as matchmod # avoid circular import issues
         fast = not matchmod.patkind(pat)
         if fast:
             def m(s):
                 return (s == pat)
         else:
             m = matchmod.match(ctx.repo().root, '', [pat], ctx=ctx)
         return [sub for sub in sstate if m(sub)]
     else:
         return [sub for sub in sstate]
 
 methods = {
     'string': stringset,
     'symbol': stringset,
     'and': andset,
     'or': orset,
     'minus': minusset,
     'list': listset,
     'group': getset,
     'not': notset,
     'func': func,
 }
 
 class matchctx(object):
     def __init__(self, ctx, subset, status=None):
         self.ctx = ctx
         self.subset = subset
         self._status = status
         self._existingenabled = False
     def status(self):
         return self._status
     def matcher(self, patterns):
         return self.ctx.match(patterns)
     def filter(self, files):
         return [f for f in files if f in self.subset]
     def existing(self):
         assert self._existingenabled, 'unexpected existing() invocation'
         if self._status is not None:
             removed = set(self._status[3])
             unknown = set(self._status[4] + self._status[5])
         else:
             removed = set()
             unknown = set()
         return (f for f in self.subset
                 if (f in self.ctx and f not in removed) or f in unknown)
     def narrow(self, files):
         return matchctx(self.ctx, self.filter(files), self._status)
     def switch(self, ctx, status=None):
         subset = self.filter(_buildsubset(ctx, status))
         return matchctx(ctx, subset, status)
 
 class fullmatchctx(matchctx):
     """A match context where any files in any revisions should be valid"""
 
     def __init__(self, ctx, status=None):
         subset = _buildsubset(ctx, status)
         super(fullmatchctx, self).__init__(ctx, subset, status)
     def switch(self, ctx, status=None):
         return fullmatchctx(ctx, status)
 
 # filesets using matchctx.switch()
 _switchcallers = [
     'revs',
     'status',
 ]
 
 def _intree(funcs, tree):
     if isinstance(tree, tuple):
         if tree[0] == 'func' and tree[1][0] == 'symbol':
             if tree[1][1] in funcs:
                 return True
             if tree[1][1] in _switchcallers:
                 # arguments won't be evaluated in the current context
                 return False
         for s in tree[1:]:
             if _intree(funcs, s):
                 return True
     return False
 
 def _buildsubset(ctx, status):
     if status:
         subset = []
         for c in status:
             subset.extend(c)
         return subset
     else:
         return list(ctx.walk(ctx.match([])))
 
 def getfileset(ctx, expr):
     tree = parse(expr)
     return getset(fullmatchctx(ctx, _buildstatus(ctx, tree)), tree)
 
 def _buildstatus(ctx, tree, basectx=None):
     # do we need status info?
 
     # temporaty boolean to simplify the next conditional
     purewdir = ctx.rev() is None and basectx is None
 
     if (_intree(_statuscallers, tree) or
         # Using matchctx.existing() on a workingctx requires us to check
         # for deleted files.
         (purewdir and _intree(_existingcallers, tree))):
         unknown = _intree(['unknown'], tree)
         ignored = _intree(['ignored'], tree)
 
         r = ctx.repo()
         if basectx is None:
             basectx = ctx.p1()
         return r.status(basectx, ctx,
                         unknown=unknown, ignored=ignored, clean=True)
     else:
         return None
 
 def prettyformat(tree):
     return parser.prettyformat(tree, ('string', 'symbol'))
 
 def loadpredicate(ui, extname, registrarobj):
     """Load fileset predicates from specified registrarobj
     """
     for name, func in registrarobj._table.iteritems():
         symbols[name] = func
         if func._callstatus:
             _statuscallers.add(name)
         if func._callexisting:
             _existingcallers.add(name)
 
 # load built-in predicates explicitly to setup _statuscallers/_existingcallers
 loadpredicate(None, None, predicate)
 
 # tell hggettext to extract docstrings from these functions:
 i18nfunctions = symbols.values()

mercurial/minifileset.py

@@ -1,91 +1,90 b''
 # minifileset.py - a simple language to select files
 #
 # Copyright 2017 Facebook, Inc.
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 from .i18n import _
 from . import (
     error,
     fileset,
 )
 
 def _compile(tree):
     if not tree:
         raise error.ParseError(_("missing argument"))
     op = tree[0]
     if op == 'symbol':
         name = fileset.getstring(tree, _('invalid file pattern'))
         if name.startswith('**'): # file extension test, ex. "**.tar.gz"
             ext = name[2:]
             for c in ext:
                 if c in '*{}[]?/\\':
                     raise error.ParseError(_('reserved character: %s') % c)
             return lambda n, s: n.endswith(ext)
         raise error.ParseError(_('invalid symbol: %s') % name)
     elif op == 'string':
         # TODO: teach fileset about 'path:', so that this can be a symbol and
         # not require quoting.
         name = fileset.getstring(tree, _('invalid path literal'))
         if name.startswith('path:'): # directory or full path test
             p = name[5:] # prefix
             pl = len(p)
             f = lambda n, s: n.startswith(p) and (len(n) == pl or n[pl] == '/')
             return f
         raise error.ParseError(_("invalid string"),
                                hint=_('paths must be prefixed with "path:"'))
     elif op == 'or':
         func1 = _compile(tree[1])
         func2 = _compile(tree[2])
         return lambda n, s: func1(n, s) or func2(n, s)
     elif op == 'and':
         func1 = _compile(tree[1])
         func2 = _compile(tree[2])
         return lambda n, s: func1(n, s) and func2(n, s)
     elif op == 'not':
         return lambda n, s: not _compile(tree[1])(n, s)
     elif op == 'group':
         return _compile(tree[1])
     elif op == 'func':
         symbols = {
             'all': lambda n, s: True,
             'none': lambda n, s: False,
             'size': lambda n, s: fileset.sizematcher(tree[2])(s),
         }
 
-        x = tree[1]
-        name = x[1]
-        if x[0] == 'symbol' and name in symbols:
+        name = fileset.getsymbol(tree[1])
+        if name in symbols:
             return symbols[name]
 
         raise error.UnknownIdentifier(name, symbols.keys())
     elif op == 'minus':     # equivalent to 'x and not y'
         func1 = _compile(tree[1])
         func2 = _compile(tree[2])
         return lambda n, s: func1(n, s) and not func2(n, s)
     elif op == 'negate':
         raise error.ParseError(_("can't use negate operator in this context"))
     elif op == 'list':
         raise error.ParseError(_("can't use a list in this context"),
                                hint=_('see hg help "filesets.x or y"'))
     raise error.ProgrammingError('illegal tree: %r' % (tree,))
 
 def compile(text):
     """generate a function (path, size) -> bool from filter specification.
 
     "text" could contain the operators defined by the fileset language for
     common logic operations, and parenthesis for grouping.  The supported path
     tests are '**.extname' for file extension test, and '"path:dir/subdir"'
     for prefix test.  The ``size()`` predicate is borrowed from filesets to test
     file size.  The predicates ``all()`` and ``none()`` are also supported.
 
     '(**.php & size(">10MB")) | **.zip | ("path:bin" & !"path:bin/README")' for
     example, will catch all php files whose size is greater than 10 MB, all
     files whose name ends with ".zip", and all files under "bin" in the repo
     root except for "bin/README".
     """
     tree = fileset.parse(text)
     return _compile(tree)

tests/test-fileset.t

@@ -1,606 +1,622 b''
   $ fileset() {
   >   hg debugfileset "$@"
   > }
 
   $ hg init repo
   $ cd repo
   $ echo a > a1
   $ echo a > a2
   $ echo b > b1
   $ echo b > b2
   $ hg ci -Am addfiles
   adding a1
   adding a2
   adding b1
   adding b2
 
 Test operators and basic patterns
 
   $ fileset -v a1
   (symbol 'a1')
   a1
   $ fileset -v 'a*'
   (symbol 'a*')
   a1
   a2
   $ fileset -v '"re:a\d"'
   (string 're:a\\d')
   a1
   a2
   $ fileset -v 'a1 or a2'
   (or
     (symbol 'a1')
     (symbol 'a2'))
   a1
   a2
   $ fileset 'a1 | a2'
   a1
   a2
   $ fileset 'a* and "*1"'
   a1
   $ fileset 'a* & "*1"'
   a1
   $ fileset 'not (r"a*")'
   b1
   b2
   $ fileset '! ("a*")'
   b1
   b2
   $ fileset 'a* - a1'
   a2
   $ fileset 'a_b'
   $ fileset '"\xy"'
   hg: parse error: invalid \x escape
   [255]
 
+Test invalid syntax
+
+  $ fileset -v '"added"()'
+  (func
+    (string 'added')
+    None)
+  hg: parse error: not a symbol
+  [255]
+  $ fileset -v '()()'
+  (func
+    (group
+      None)
+    None)
+  hg: parse error: not a symbol
+  [255]
+
 Test files status
 
   $ rm a1
   $ hg rm a2
   $ echo b >> b2
   $ hg cp b1 c1
   $ echo c > c2
   $ echo c > c3
   $ cat > .hgignore <<EOF
   > \.hgignore
   > 2$
   > EOF
   $ fileset 'modified()'
   b2
   $ fileset 'added()'
   c1
   $ fileset 'removed()'
   a2
   $ fileset 'deleted()'
   a1
   $ fileset 'missing()'
   a1
   $ fileset 'unknown()'
   c3
   $ fileset 'ignored()'
   .hgignore
   c2
   $ fileset 'hgignore()'
   a2
   b2
   $ fileset 'clean()'
   b1
   $ fileset 'copied()'
   c1
 
 Test files status in different revisions
 
   $ hg status -m
   M b2
   $ fileset -r0 'revs("wdir()", modified())' --traceback
   b2
   $ hg status -a
   A c1
   $ fileset -r0 'revs("wdir()", added())'
   c1
   $ hg status --change 0 -a
   A a1
   A a2
   A b1
   A b2
   $ hg status -mru
   M b2
   R a2
   ? c3
   $ fileset -r0 'added() and revs("wdir()", modified() or removed() or unknown())'
   b2
   a2
   $ fileset -r0 'added() or revs("wdir()", added())'
   a1
   a2
   b1
   b2
   c1
 
 Test files properties
 
   >>> file('bin', 'wb').write('\0a')
   $ fileset 'binary()'
   $ fileset 'binary() and unknown()'
   bin
   $ echo '^bin$' >> .hgignore
   $ fileset 'binary() and ignored()'
   bin
   $ hg add bin
   $ fileset 'binary()'
   bin
 
   $ fileset 'grep("b{1}")'
   b2
   c1
   b1
   $ fileset 'grep("missingparens(")'
   hg: parse error: invalid match pattern: unbalanced parenthesis
   [255]
 
 #if execbit
   $ chmod +x b2
   $ fileset 'exec()'
   b2
 #endif
 
 #if symlink
   $ ln -s b2 b2link
   $ fileset 'symlink() and unknown()'
   b2link
   $ hg add b2link
 #endif
 
 #if no-windows
   $ echo foo > con.xml
   $ fileset 'not portable()'
   con.xml
   $ hg --config ui.portablefilenames=ignore add con.xml
 #endif
 
   >>> file('1k', 'wb').write(' '*1024)
   >>> file('2k', 'wb').write(' '*2048)
   $ hg add 1k 2k
   $ fileset 'size("bar")'
   hg: parse error: couldn't parse size: bar
   [255]
   $ fileset '(1k, 2k)'
   hg: parse error: can't use a list in this context
   (see hg help "filesets.x or y")
   [255]
   $ fileset 'size(1k)'
   1k
   $ fileset '(1k or 2k) and size("< 2k")'
   1k
   $ fileset '(1k or 2k) and size("<=2k")'
   1k
   2k
   $ fileset '(1k or 2k) and size("> 1k")'
   2k
   $ fileset '(1k or 2k) and size(">=1K")'
   1k
   2k
   $ fileset '(1k or 2k) and size(".5KB - 1.5kB")'
   1k
   $ fileset 'size("1M")'
   $ fileset 'size("1 GB")'
 
 Test merge states
 
   $ hg ci -m manychanges
   $ hg up -C 0
   * files updated, 0 files merged, * files removed, 0 files unresolved (glob)
   $ echo c >> b2
   $ hg ci -m diverging b2
   created new head
   $ fileset 'resolved()'
   $ fileset 'unresolved()'
   $ hg merge
   merging b2
   warning: conflicts while merging b2! (edit, then use 'hg resolve --mark')
   * files updated, 0 files merged, 1 files removed, 1 files unresolved (glob)
   use 'hg resolve' to retry unresolved file merges or 'hg update -C .' to abandon
   [1]
   $ fileset 'resolved()'
   $ fileset 'unresolved()'
   b2
   $ echo e > b2
   $ hg resolve -m b2
   (no more unresolved files)
   $ fileset 'resolved()'
   b2
   $ fileset 'unresolved()'
   $ hg ci -m merge
 
 Test subrepo predicate
 
   $ hg init sub
   $ echo a > sub/suba
   $ hg -R sub add sub/suba
   $ hg -R sub ci -m sub
   $ echo 'sub = sub' > .hgsub
   $ hg init sub2
   $ echo b > sub2/b
   $ hg -R sub2 ci -Am sub2
   adding b
   $ echo 'sub2 = sub2' >> .hgsub
   $ fileset 'subrepo()'
   $ hg add .hgsub
   $ fileset 'subrepo()'
   sub
   sub2
   $ fileset 'subrepo("sub")'
   sub
   $ fileset 'subrepo("glob:*")'
   sub
   sub2
   $ hg ci -m subrepo
 
 Test that .hgsubstate is updated as appropriate during a conversion.  The
 saverev property is enough to alter the hashes of the subrepo.
 
   $ hg init ../converted
   $ hg --config extensions.convert= convert --config convert.hg.saverev=True  \
   >      sub ../converted/sub
   initializing destination ../converted/sub repository
   scanning source...
   sorting...
   converting...
   0 sub
   $ hg clone -U sub2 ../converted/sub2
   $ hg --config extensions.convert= convert --config convert.hg.saverev=True  \
   >      . ../converted
   scanning source...
   sorting...
   converting...
   4 addfiles
   3 manychanges
   2 diverging
   1 merge
   0 subrepo
   no ".hgsubstate" updates will be made for "sub2"
   $ hg up -q -R ../converted -r tip
   $ hg --cwd ../converted cat sub/suba sub2/b -r tip
   a
   b
   $ oldnode=`hg log -r tip -T "{node}\n"`
   $ newnode=`hg log -R ../converted -r tip -T "{node}\n"`
   $ [ "$oldnode" != "$newnode" ] || echo "nothing changed"
 
 Test with a revision
 
   $ hg log -G --template '{rev} {desc}\n'
   @  4 subrepo
   |
   o    3 merge
   |\
   | o  2 diverging
   | |
   o |  1 manychanges
   |/
   o  0 addfiles
   
   $ echo unknown > unknown
   $ fileset -r1 'modified()'
   b2
   $ fileset -r1 'added() and c1'
   c1
   $ fileset -r1 'removed()'
   a2
   $ fileset -r1 'deleted()'
   $ fileset -r1 'unknown()'
   $ fileset -r1 'ignored()'
   $ fileset -r1 'hgignore()'
   b2
   bin
   $ fileset -r1 'binary()'
   bin
   $ fileset -r1 'size(1k)'
   1k
   $ fileset -r3 'resolved()'
   $ fileset -r3 'unresolved()'
 
 #if execbit
   $ fileset -r1 'exec()'
   b2
 #endif
 
 #if symlink
   $ fileset -r1 'symlink()'
   b2link
 #endif
 
 #if no-windows
   $ fileset -r1 'not portable()'
   con.xml
   $ hg forget 'con.xml'
 #endif
 
   $ fileset -r4 'subrepo("re:su.*")'
   sub
   sub2
   $ fileset -r4 'subrepo("sub")'
   sub
   $ fileset -r4 'b2 or c1'
   b2
   c1
 
   >>> open('dos', 'wb').write("dos\r\n")
   >>> open('mixed', 'wb').write("dos\r\nunix\n")
   >>> open('mac', 'wb').write("mac\r")
   $ hg add dos mixed mac
 
 (remove a1, to examine safety of 'eol' on removed files)
   $ rm a1
 
   $ fileset 'eol(dos)'
   dos
   mixed
   $ fileset 'eol(unix)'
   mixed
   .hgsub
   .hgsubstate
   b1
   b2
   c1
   $ fileset 'eol(mac)'
   mac
 
 Test safety of 'encoding' on removed files
 
   $ fileset 'encoding("ascii")'
   dos
   mac
   mixed
   .hgsub
   .hgsubstate
   1k
   2k
   b1
   b2
   b2link (symlink !)
   bin
   c1
 
 Test detection of unintentional 'matchctx.existing()' invocation
 
   $ cat > $TESTTMP/existingcaller.py <<EOF
   > from mercurial import registrar
   > 
   > filesetpredicate = registrar.filesetpredicate()
   > @filesetpredicate('existingcaller()', callexisting=False)
   > def existingcaller(mctx, x):
   >     # this 'mctx.existing()' invocation is unintentional
   >     return [f for f in mctx.existing()]
   > EOF
 
   $ cat >> .hg/hgrc <<EOF
   > [extensions]
   > existingcaller = $TESTTMP/existingcaller.py
   > EOF
 
   $ fileset 'existingcaller()' 2>&1 | tail -1
   AssertionError: unexpected existing() invocation
 
 Test 'revs(...)'
 ================
 
 small reminder of the repository state
 
   $ hg log -G
   @  changeset:   4:* (glob)
   |  tag:         tip
   |  user:        test
   |  date:        Thu Jan 01 00:00:00 1970 +0000
   |  summary:     subrepo
   |
   o    changeset:   3:* (glob)
   |\   parent:      2:55b05bdebf36
   | |  parent:      1:* (glob)
   | |  user:        test
   | |  date:        Thu Jan 01 00:00:00 1970 +0000
   | |  summary:     merge
   | |
   | o  changeset:   2:55b05bdebf36
   | |  parent:      0:8a9576c51c1f
   | |  user:        test
   | |  date:        Thu Jan 01 00:00:00 1970 +0000
   | |  summary:     diverging
   | |
   o |  changeset:   1:* (glob)
   |/   user:        test
   |    date:        Thu Jan 01 00:00:00 1970 +0000
   |    summary:     manychanges
   |
   o  changeset:   0:8a9576c51c1f
      user:        test
      date:        Thu Jan 01 00:00:00 1970 +0000
      summary:     addfiles
   
   $ hg status --change 0
   A a1
   A a2
   A b1
   A b2
   $ hg status --change 1
   M b2
   A 1k
   A 2k
   A b2link (no-windows !)
   A bin
   A c1
   A con.xml (no-windows !)
   R a2
   $ hg status --change 2
   M b2
   $ hg status --change 3
   M b2
   A 1k
   A 2k
   A b2link (no-windows !)
   A bin
   A c1
   A con.xml (no-windows !)
   R a2
   $ hg status --change 4
   A .hgsub
   A .hgsubstate
   $ hg status
   A dos
   A mac
   A mixed
   R con.xml (no-windows !)
   ! a1
   ? b2.orig
   ? c3
   ? unknown
 
 Test files at -r0 should be filtered by files at wdir
 -----------------------------------------------------
 
   $ fileset -r0 '* and revs("wdir()", *)'
   a1
   b1
   b2
 
 Test that "revs()" work at all
 ------------------------------
 
   $ fileset "revs('2', modified())"
   b2
 
 Test that "revs()" work for file missing in the working copy/current context
 ----------------------------------------------------------------------------
 
 (a2 not in working copy)
 
   $ fileset "revs('0', added())"
   a1
   a2
   b1
   b2
 
 (none of the file exist in "0")
 
   $ fileset -r 0 "revs('4', added())"
   .hgsub
   .hgsubstate
 
 Call with empty revset
 --------------------------
 
   $ fileset "revs('2-2', modified())"
 
 Call with revset matching multiple revs
 ---------------------------------------
 
   $ fileset "revs('0+4', added())"
   a1
   a2
   b1
   b2
   .hgsub
   .hgsubstate
 
 overlapping set
 
   $ fileset "revs('1+2', modified())"
   b2
 
 test 'status(...)'
 =================
 
 Simple case
 -----------
 
   $ fileset "status(3, 4, added())"
   .hgsub
   .hgsubstate
 
 use rev to restrict matched file
 -----------------------------------------
 
   $ hg status --removed --rev 0 --rev 1
   R a2
   $ fileset "status(0, 1, removed())"
   a2
   $ fileset "* and status(0, 1, removed())"
   $ fileset -r 4 "status(0, 1, removed())"
   a2
   $ fileset -r 4 "* and status(0, 1, removed())"
   $ fileset "revs('4', * and status(0, 1, removed()))"
   $ fileset "revs('0', * and status(0, 1, removed()))"
   a2
 
 check wdir()
 ------------
 
   $ hg status --removed  --rev 4
   R con.xml (no-windows !)
   $ fileset "status(4, 'wdir()', removed())"
   con.xml (no-windows !)
 
   $ hg status --removed --rev 2
   R a2
   $ fileset "status('2', 'wdir()', removed())"
   a2
 
 test backward status
 --------------------
 
   $ hg status --removed --rev 0 --rev 4
   R a2
   $ hg status --added --rev 4 --rev 0
   A a2
   $ fileset "status(4, 0, added())"
   a2
 
 test cross branch status
 ------------------------
 
   $ hg status --added --rev 1 --rev 2
   A a2
   $ fileset "status(1, 2, added())"
   a2
 
 test with multi revs revset
 ---------------------------
   $ hg status --added --rev 0:1 --rev 3:4
   A .hgsub
   A .hgsubstate
   A 1k
   A 2k
   A b2link (no-windows !)
   A bin
   A c1
   A con.xml (no-windows !)
   $ fileset "status('0:1', '3:4', added())"
   .hgsub
   .hgsubstate
   1k
   2k
   b2link (no-windows !)
   bin
   c1
   con.xml (no-windows !)
 
 tests with empty value
 ----------------------
 
 Fully empty revset
 
   $ fileset "status('', '4', added())"
   hg: parse error: first argument to status must be a revision
   [255]
   $ fileset "status('2', '', added())"
   hg: parse error: second argument to status must be a revision
   [255]
 
 Empty revset will error at the revset layer
 
   $ fileset "status(' ', '4', added())"
   hg: parse error at 1: not a prefix: end
   [255]
   $ fileset "status('2', ' ', added())"
   hg: parse error at 1: not a prefix: end
   [255]