[gentoo-portage-dev] [RFC/PATCH v2] New objects cpv, pv and version to be used instead of raw strings.

[Ali Polatel]

Thanks for the comments.

Version 2, updates:
* Use "import sys" instead of "from sys import hexversion"
* Use super() to call functions from parent class.
* Fix initialization, used to pass m.groups() to parent object which is a
leftover from older versions where I used to subclass list instead of str.
Removed __str__ functions and __str which are useless as well.
* Added readonly attributes cvs, main_version and revision to version, pv and
cpv, package to pv, cpv and category to cpv. I think this is a better way than
making __parts available so it's guaranteed that there'll be no unwanted change
in internal state.
* version uses __init__ now. pv and cpv still use __new__ because to fit code
easily they return None when the argument isn't a valid PV or CPV instead of
raising TypeError.

Please comment.

---
 pym/portage/versions.py |  332 +++++++++++++++++++++++++++++++++++++++++-----
 1 files changed, 296 insertions(+), 36 deletions(-)

diff --git a/pym/portage/versions.py b/pym/portage/versions.py
index 261fa9d..507d7a1 100644
--- a/pym/portage/versions.py
+++ b/pym/portage/versions.py
@@ -4,6 +4,7 @@
 # $Id$
 
 import re
+import warnings
 
 ver_regexp = re.compile("^(cvs\\.)?(\\d+)((\\.\\d+)*)([a-z]?)((_(pre|p|beta|alpha|rc)\\d*)*)(-r(\\d+))?$")
 suffix_regexp = re.compile("^(alpha|beta|rc|pre|p)(\\d*)$")
@@ -12,6 +13,245 @@ endversion_keys = ["pre", "p", "alpha", "beta", "rc"]
 
 from portage.exception import InvalidData
 
+# builtin all() is new in Python-2.5
+# TODO Move compatibility stuff to a new module portage.compat
+# and import from it like from portage.compat import all
+import sys
+if sys.hexversion < 0x02050000:
+	def all(iterable):
+		for i in iterable:
+			if not bool(i):
+				return False
+		return True
+
+def needs_version(func):
+	"""Decorator for functions that require non-keyword arguments of type version."""
+	def func_proxy(*args, **kwargs):
+		if not all([isinstance(arg, version) for arg in args]):
+			raise TypeError("Not all non-keyword arguments are of type version")
+		return func(*args, **kwargs)
+	func_proxy.__doc__ = func.__doc__
+	return func_proxy
+
+def needs_pv(func):
+	"""Decorator for functions that require non-keyword arguments of type pv."""
+	def func_proxy(*args, **kwargs):
+		if not all([isinstance(arg, pv) for arg in args]):
+			raise TypeError("Not all non-keyword arguments are of type pv")
+		return func(*args, **kwargs)
+	func_proxy.__doc__ = func.__doc__
+	return func_proxy
+
+def needs_cpv(func):
+	"""Decorator for functions that require non-keyword arguments of type cpv."""
+	def func_proxy(*args, **kwargs):
+		if not all([isinstance(arg, cpv) for arg in args]):
+			raise TypeError("Not all non-keyword arguments are of type cpv")
+		return func(*args, **kwargs)
+	func_proxy.__doc__ = func.__doc__
+	return func_proxy
+
+class version(str):
+	"""Represents a package version"""
+
+	__hash = None
+	__parts = ()
+
+	def __init__(self, value):
+		m = ver_regexp.match(value)
+		if m is None:
+			raise TypeError("Syntax error in version: %s" % value)
+		else:
+			super(version, self).__init__(value)
+			self.__hash = hash(m.groups()) + hash(value)
+			self.__parts = m.groups()
+
+	def __repr__(self):
+		return "<%s object at 0x%x: %s>" % (self.__class__.__name__,
+						id(self), self)
+
+	def __hash__(self):
+		return self.__hash
+
+	def __getitem__(self, i):
+		return self.__parts[i]
+
+	def __getslice__(self, i, j):
+		return self.__parts[i:j]
+
+	def __len__(self):
+		return len(self.__parts)
+
+	@needs_version
+	def __cmp__(self, y):
+		return vercmp(self, y)
+
+	@needs_version
+	def __eq__(self, y):
+		return vercmp(self, y) == 0
+
+	@needs_version
+	def __ne__(self, y):
+		return vercmp(self, y) != 0
+
+	@needs_version
+	def __lt__(self, y):
+		return vercmp(self, y) < 0
+
+	@needs_version
+	def __le__(self, y):
+		return vercmp(self, y) <= 0
+
+	@needs_version
+	def __gt__(self, y):
+		return vercmp(self, y) > 0
+
+	@needs_version
+	def __ge__(self, y):
+		return vercmp(self, y) >= 0
+
+	@property
+	def cvs(self):
+		return self.__parts[0]
+
+	@property
+	def main_version(self):
+		mv = self.__parts[1:3]
+		mv += self.__parts[4:6]
+		return "".join(mv)
+
+	@property
+	def revision(self):
+		return self.__parts[8]
+
+class pv(str):
+	"""Represents a pv"""
+
+	__hash = None
+	__parts = ()
+
+	def __new__(cls, value):
+		parts = pkgsplit(value)
+		if parts is None:
+			# Ideally a TypeError should be raised here.
+			# But to fit code using this easily, fail silently.
+			return None
+		else:
+			new_pv = super(pv, cls).__new__(cls, value)
+			new_pv.__hash = hash(parts) + hash(value)
+			new_pv.__parts = (parts[0], version("-".join(parts[1:])))
+
+			return new_pv
+
+	def __repr__(self):
+		return "<%s object at 0x%x: %s>" % (self.__class__.__name__,
+						id(self), self)
+
+	def __hash__(self):
+		return self.__hash
+
+	def __getitem__(self, i):
+		return self.__parts[i]
+
+	def __getslice__(self, i, j):
+		return self.__parts[i:j]
+
+	def __len__(self):
+		return len(self.__parts)
+
+	@needs_pv
+	def __cmp__(self, y):
+		if self.__parts[0] != y.__parts[0]:
+			return None
+		else:
+			return cmp(self[1], y[1])
+
+	@property
+	def package(self):
+		return self.__parts[0]
+
+	@property
+	def version(self):
+		return self.__parts[1]
+
+	@property
+	def cvs(self):
+		return self.__parts[1].cvs
+
+	@property
+	def main_version(self):
+		return self.__parts[1].main_version
+
+	@property
+	def revision(self):
+		return self.__parts[1].revision
+
+class cpv(str):
+	"""Represents a cpv"""
+
+	__hash = None
+	__parts = ()
+
+	def __new__(cls, value):
+		parts = catpkgsplit(value)
+		if parts is None:
+			# Ideally a TypeError should be raised here.
+			# But to fit code using this easily, fail silently.
+				return None
+		else:
+			new_cpv = super(cpv, cls).__new__(cls, value)
+			new_cpv.__hash = hash(parts) + hash(value)
+			new_cpv.__parts = (parts[0], pv("-".join(parts[1:])))
+
+			return new_cpv
+
+	def __repr__(self):
+		return "<%s object at 0x%x: %s>" % (self.__class__.__name__,
+						id(self), self)
+
+	def __hash__(self):
+		return self.__hash
+
+	def __getitem__(self, i):
+		return self.__parts[i]
+
+	def __getslice__(self, i, j):
+		return self.__parts[i:j]
+
+	def __len__(self):
+		return len(self.__parts)
+
+	@needs_cpv
+	def __cmp__(self, y):
+		if self[0] != y[0]:
+			return None
+		else:
+			return cmp(self[1], y[1])
+
+	@property
+	def category(self):
+		return self.__parts[0]
+
+	@property
+	def package(self):
+		return self.__parts[1]
+
+	@property
+	def version(self):
+		return self.__parts[1].version
+
+	@property
+	def cvs(self):
+		return self.__parts[1].cvs
+
+	@property
+	def main_version(self):
+		return self.__parts[1].main_version
+
+	@property
+	def revision(self):
+		return self.__parts[1].revision
+
 def ververify(myver, silent=1):
 	if ver_regexp.match(myver):
 		return 1
@@ -45,43 +285,63 @@ def vercmp(ver1, ver2, silent=1):
 	4. None if ver1 or ver2 are invalid (see ver_regexp in portage.versions.py)
 	"""
 
-	if ver1 == ver2:
-		return 0
-	mykey=ver1+":"+ver2
-	try:
-		return vercmp_cache[mykey]
-	except KeyError:
-		pass
-	match1 = ver_regexp.match(ver1)
-	match2 = ver_regexp.match(ver2)
-	
-	# checking that the versions are valid
-	if not match1 or not match1.groups():
-		if not silent:
-			print "!!! syntax error in version: %s" % ver1
-		return None
-	if not match2 or not match2.groups():
-		if not silent:
-			print "!!! syntax error in version: %s" % ver2
-		return None
+	if isinstance(ver1, version) and isinstance(ver2, version):
+		if ver1._str == ver2._str:
+			return 0
+		mykey = ver1._str+":"+ver2._str
+		if mykey in vercmp_cache:
+			return vercmp_cache[mykey]
+
+		group1 = ver1[:]
+		group2 = ver2[:]
+	elif isinstance(ver1, str) and isinstance(ver2, str):
+		## Backwards compatibility
+		msg = "vercmp(str,str) is deprecated use portage.version object instead"
+		warnings.warn(msg, DeprecationWarning)
+
+		if ver1 == ver2:
+			return 0
+		mykey=ver1+":"+ver2
+		try:
+			return vercmp_cache[mykey]
+		except KeyError:
+			pass
+		match1 = ver_regexp.match(ver1)
+		match2 = ver_regexp.match(ver2)
+
+		# checking that the versions are valid
+		if not match1 or not match1.groups():
+			if not silent:
+				print "!!! syntax error in version: %s" % ver1
+			return None
+		if not match2 or not match2.groups():
+			if not silent:
+				print "!!! syntax error in version: %s" % ver2
+			return None
+
+		group1 = match1.groups()
+		group2 = match2.groups()
+	else:
+		raise TypeError(
+			"Arguments aren't of type str,str or version,version")
 
 	# shortcut for cvs ebuilds (new style)
-	if match1.group(1) and not match2.group(1):
+	if group1[0] and not group2[0]:
 		vercmp_cache[mykey] = 1
 		return 1
-	elif match2.group(1) and not match1.group(1):
+	elif group2[0] and not group1[0]:
 		vercmp_cache[mykey] = -1
 		return -1
 	
 	# building lists of the version parts before the suffix
 	# first part is simple
-	list1 = [int(match1.group(2))]
-	list2 = [int(match2.group(2))]
+	list1 = [int(group1[1])]
+	list2 = [int(group2[1])]
 
 	# this part would greatly benefit from a fixed-length version pattern
-	if len(match1.group(3)) or len(match2.group(3)):
-		vlist1 = match1.group(3)[1:].split(".")
-		vlist2 = match2.group(3)[1:].split(".")
+	if len(group1[2]) or len(group2[2]):
+		vlist1 = group1[2][1:].split(".")
+		vlist2 = group2[2][1:].split(".")
 		for i in range(0, max(len(vlist1), len(vlist2))):
 			# Implcit .0 is given a value of -1, so that 1.0.0 > 1.0, since it
 			# would be ambiguous if two versions that aren't literally equal
@@ -111,10 +371,10 @@ def vercmp(ver1, ver2, silent=1):
 				list2.append(int(vlist2[i].ljust(max_len, "0")))
 
 	# and now the final letter
-	if len(match1.group(5)):
-		list1.append(ord(match1.group(5)))
-	if len(match2.group(5)):
-		list2.append(ord(match2.group(5)))
+	if len(group1[4]):
+		list1.append(ord(group1[4]))
+	if len(group2[4]):
+		list2.append(ord(group2[4]))
 
 	for i in range(0, max(len(list1), len(list2))):
 		if len(list1) <= i:
@@ -128,8 +388,8 @@ def vercmp(ver1, ver2, silent=1):
 			return list1[i] - list2[i]
 	
 	# main version is equal, so now compare the _suffix part
-	list1 = match1.group(6).split("_")[1:]
-	list2 = match2.group(6).split("_")[1:]
+	list1 = group1[5].split("_")[1:]
+	list2 = group2[5].split("_")[1:]
 	
 	for i in range(0, max(len(list1), len(list2))):
 		# Implicit _p0 is given a value of -1, so that 1 < 1_p0
@@ -154,12 +414,12 @@ def vercmp(ver1, ver2, silent=1):
 				return r1 - r2
 	
 	# the suffix part is equal to, so finally check the revision
-	if match1.group(10):
-		r1 = int(match1.group(10))
+	if group1[9]:
+		r1 = int(group1[9])
 	else:
 		r1 = 0
-	if match2.group(10):
-		r2 = int(match2.group(10))
+	if group2[9]:
+		r2 = int(group2[9])
 	else:
 		r2 = 0
 	vercmp_cache[mykey] = r1 - r2
-- 
Regards,
Ali Polatel