[gentoo-portage-dev] [PATCH 3/4] repoman: Use XML Schema for metadata.xml validation

[Michał Górny]

---
 pym/repoman/_xml.py                              | 16 +++++-----
 pym/repoman/metadata.py                          | 39 ++++++++++++------------
 pym/repoman/modules/scan/metadata/__init__.py    |  2 +-
 pym/repoman/modules/scan/metadata/pkgmetadata.py |  6 ++--
 pym/repoman/scanner.py                           |  8 ++---
 5 files changed, 36 insertions(+), 35 deletions(-)

diff --git a/pym/repoman/_xml.py b/pym/repoman/_xml.py
index d55dda5..33a536a 100644
--- a/pym/repoman/_xml.py
+++ b/pym/repoman/_xml.py
@@ -12,7 +12,7 @@ from portage import os
 from portage.output import red
 from portage.process import find_binary
 
-from repoman.metadata import fetch_metadata_dtd
+from repoman.metadata import fetch_metadata_xsd
 from repoman._subprocess import repoman_getstatusoutput
 
 
@@ -53,12 +53,12 @@ class _MetadataTreeBuilder(xml.etree.ElementTree.TreeBuilder):
 
 class XmlLint(object):
 
-	def __init__(self, options, repoman_settings, metadata_dtd=None):
-		self.metadata_dtd = (metadata_dtd or
-			os.path.join(repoman_settings["DISTDIR"], 'metadata.dtd'))
+	def __init__(self, options, repoman_settings, metadata_xsd=None):
+		self.metadata_xsd = (metadata_xsd or
+			os.path.join(repoman_settings["DISTDIR"], 'metadata.xsd'))
 		self.options = options
 		self.repoman_settings = repoman_settings
-		self._is_capable = metadata_dtd is not None
+		self._is_capable = metadata_xsd is not None
 		self.binary = None
 		self._check_capable()
 
@@ -69,7 +69,7 @@ class XmlLint(object):
 		if not self.binary:
 			print(red("!!! xmllint not found. Can't check metadata.xml.\n"))
 		elif not self._is_capable:
-			if not fetch_metadata_dtd(self.metadata_dtd, self.repoman_settings):
+			if not fetch_metadata_xsd(self.metadata_xsd, self.repoman_settings):
 				sys.exit(1)
 			# this can be problematic if xmllint changes their output
 			self._is_capable = True
@@ -93,8 +93,8 @@ class XmlLint(object):
 		# xmlint can produce garbage output even on success, so only dump
 		# the ouput when it fails.
 		st, out = repoman_getstatusoutput(
-			self.binary + " --nonet --noout --dtdvalid %s %s" % (
-				portage._shell_quote(self.metadata_dtd),
+			self.binary + " --nonet --noout --schema %s %s" % (
+				portage._shell_quote(self.metadata_xsd),
 				portage._shell_quote(
 					os.path.join(checkdir, "metadata.xml"))))
 		if st != os.EX_OK:
diff --git a/pym/repoman/metadata.py b/pym/repoman/metadata.py
index e95ad41..7a514dc 100644
--- a/pym/repoman/metadata.py
+++ b/pym/repoman/metadata.py
@@ -33,8 +33,9 @@ metadata_xml_declaration = '<?xml version="1.0" encoding="%s"?>' \
 	% (metadata_xml_encoding,)
 metadata_doctype_name = 'pkgmetadata'
 metadata_dtd_uri = 'http://www.gentoo.org/dtd/metadata.dtd'
+metadata_xsd_uri = 'http://www.gentoo.org/xml-schema/metadata.xsd'
 # force refetch if the local copy creation time is older than this
-metadata_dtd_ctime_interval = 60 * 60 * 24 * 7  # 7 days
+metadata_xsd_ctime_interval = 60 * 60 * 24 * 7  # 7 days
 
 
 def parse_metadata_use(xml_tree):
@@ -86,36 +87,36 @@ def parse_metadata_use(xml_tree):
 	return uselist
 
 
-def fetch_metadata_dtd(metadata_dtd, repoman_settings):
+def fetch_metadata_xsd(metadata_xsd, repoman_settings):
 	"""
-	Fetch metadata.dtd if it doesn't exist or the ctime is older than
-	metadata_dtd_ctime_interval.
+	Fetch metadata.xsd if it doesn't exist or the ctime is older than
+	metadata_xsd_ctime_interval.
 	@rtype: bool
 	@return: True if successful, otherwise False
 	"""
 
 	must_fetch = True
-	metadata_dtd_st = None
+	metadata_xsd_st = None
 	current_time = int(time.time())
 	try:
-		metadata_dtd_st = os.stat(metadata_dtd)
+		metadata_xsd_st = os.stat(metadata_xsd)
 	except EnvironmentError as e:
 		if e.errno not in (errno.ENOENT, errno.ESTALE):
 			raise
 		del e
 	else:
-		# Trigger fetch if metadata.dtd mtime is old or clock is wrong.
-		if abs(current_time - metadata_dtd_st.st_ctime) \
-			< metadata_dtd_ctime_interval:
+		# Trigger fetch if metadata.xsd mtime is old or clock is wrong.
+		if abs(current_time - metadata_xsd_st.st_ctime) \
+			< metadata_xsd_ctime_interval:
 			must_fetch = False
 
 	if must_fetch:
 		print()
 		print(
-			"%s the local copy of metadata.dtd "
+			"%s the local copy of metadata.xsd "
 			"needs to be refetched, doing that now" % green("***"))
 		print()
-		parsed_url = urlparse(metadata_dtd_uri)
+		parsed_url = urlparse(metadata_xsd_uri)
 		setting = 'FETCHCOMMAND_' + parsed_url.scheme.upper()
 		fcmd = repoman_settings.get(setting)
 		if not fcmd:
@@ -125,29 +126,29 @@ def fetch_metadata_dtd(metadata_dtd, repoman_settings):
 				return False
 
 		destdir = repoman_settings["DISTDIR"]
-		fd, metadata_dtd_tmp = tempfile.mkstemp(
-			prefix='metadata.dtd.', dir=destdir)
+		fd, metadata_xsd_tmp = tempfile.mkstemp(
+			prefix='metadata.xsd.', dir=destdir)
 		os.close(fd)
 
 		try:
 			if not portage.getbinpkg.file_get(
-				metadata_dtd_uri, destdir, fcmd=fcmd,
-				filename=os.path.basename(metadata_dtd_tmp)):
+				metadata_xsd_uri, destdir, fcmd=fcmd,
+				filename=os.path.basename(metadata_xsd_tmp)):
 				logging.error(
-					"failed to fetch metadata.dtd from '%s'" % metadata_dtd_uri)
+					"failed to fetch metadata.xsd from '%s'" % metadata_xsd_uri)
 				return False
 
 			try:
 				portage.util.apply_secpass_permissions(
-					metadata_dtd_tmp,
+					metadata_xsd_tmp,
 					gid=portage.data.portage_gid, mode=0o664, mask=0o2)
 			except portage.exception.PortageException:
 				pass
 
-			shutil.move(metadata_dtd_tmp, metadata_dtd)
+			shutil.move(metadata_xsd_tmp, metadata_xsd)
 		finally:
 			try:
-				os.unlink(metadata_dtd_tmp)
+				os.unlink(metadata_xsd_tmp)
 			except OSError:
 				pass
 
diff --git a/pym/repoman/modules/scan/metadata/__init__.py b/pym/repoman/modules/scan/metadata/__init__.py
index 90981d8..b2463fc 100644
--- a/pym/repoman/modules/scan/metadata/__init__.py
+++ b/pym/repoman/modules/scan/metadata/__init__.py
@@ -18,7 +18,7 @@ module_spec = {
 			'functions': ['check'],
 			'func_desc': {
 			},
-			'mod_kwargs': ['repo_settings', 'qatracker', 'options', 'metadata_dtd',
+			'mod_kwargs': ['repo_settings', 'qatracker', 'options', 'metadata_xsd',
 			],
 			'func_kwargs': {
 			},
diff --git a/pym/repoman/modules/scan/metadata/pkgmetadata.py b/pym/repoman/modules/scan/metadata/pkgmetadata.py
index 030cbca..fa73bb5 100644
--- a/pym/repoman/modules/scan/metadata/pkgmetadata.py
+++ b/pym/repoman/modules/scan/metadata/pkgmetadata.py
@@ -46,17 +46,17 @@ class PkgMetadata(ScanBase):
 		@param repo_settings: settings instance
 		@param qatracker: QATracker instance
 		@param options: argparse options instance
-		@param metadata_dtd: path of metadata.dtd
+		@param metadata_xsd: path of metadata.xsd
 		'''
 		super(PkgMetadata, self).__init__(**kwargs)
 		repo_settings = kwargs.get('repo_settings')
 		self.qatracker = kwargs.get('qatracker')
 		self.options = kwargs.get('options')
-		metadata_dtd = kwargs.get('metadata_dtd')
+		metadata_xsd = kwargs.get('metadata_xsd')
 		self.repoman_settings = repo_settings.repoman_settings
 		self.musedict = {}
 		self.xmllint = XmlLint(self.options, self.repoman_settings,
-			metadata_dtd=metadata_dtd)
+			metadata_xsd=metadata_xsd)
 
 	def check(self, **kwargs):
 		'''Performs the checks on the metadata.xml for the package
diff --git a/pym/repoman/scanner.py b/pym/repoman/scanner.py
index e9a8e20..0d17f74 100644
--- a/pym/repoman/scanner.py
+++ b/pym/repoman/scanner.py
@@ -54,11 +54,11 @@ class Scanner(object):
 			portage.util.stack_lists([self.categories], incremental=1))
 		self.categories = self.repo_settings.repoman_settings.categories
 
-		metadata_dtd = None
+		metadata_xsd = None
 		for path in reversed(self.repo_settings.repo_config.eclass_db.porttrees):
-			path = os.path.join(path, 'metadata/dtd/metadata.dtd')
+			path = os.path.join(path, 'metadata/xml-schema/metadata.xsd')
 			if os.path.exists(path):
-				metadata_dtd = path
+				metadata_xsd = path
 				break
 
 		self.portdb = repo_settings.portdb
@@ -184,7 +184,7 @@ class Scanner(object):
 			"qatracker": self.qatracker,
 			"vcs_settings": self.vcs_settings,
 			"options": self.options,
-			"metadata_dtd": metadata_dtd,
+			"metadata_xsd": metadata_xsd,
 			"uselist": uselist,
 			"checks": self.checks,
 			"repo_metadata": self.repo_metadata,
-- 
2.8.1