Parse EAPI with pattern from PMS section 7.3.1.

This implements the specification that was approved in Gentoo's council
meeting on May 8, 2012 (see bug #402167). The parse-eapi-ebuild-head
FEATURES setting is now enabled by default, and causes non-conformant
ebuilds to be treated as invalid. This behavior will soon become
enabled unconditionally.

12 files changed, 132 insertions, 98 deletions

diff --git a/bin/ebuild.sh b/bin/ebuild.sh
index 32dc64f0c..d161bc65d 100755
--- a/bin/ebuild.sh
+++ b/bin/ebuild.sh
@@ -534,7 +534,7 @@ if ! has "$EBUILD_PHASE" clean cleanrm ; then
 		# In order to ensure correct interaction between ebuilds and
 		# eclasses, they need to be unset before this process of
 		# interaction begins.
-		unset DEPEND RDEPEND PDEPEND INHERITED IUSE REQUIRED_USE \
+		unset EAPI DEPEND RDEPEND PDEPEND INHERITED IUSE REQUIRED_USE \
 			ECLASS E_IUSE E_REQUIRED_USE E_DEPEND E_RDEPEND E_PDEPEND
 
 		if [[ $PORTAGE_DEBUG != 1 || ${-/x/} != $- ]] ; then
@@ -551,7 +551,7 @@ if ! has "$EBUILD_PHASE" clean cleanrm ; then
 			rm "$PORTAGE_BUILDDIR/.ebuild_changed"
 		fi
 
-		[[ -n $EAPI ]] || EAPI=0
+		[ "${EAPI+set}" = set ] || EAPI=0
 
 		if has "$EAPI" 0 1 2 3 3_pre2 ; then
 			export RDEPEND=${RDEPEND-${DEPEND}}
@@ -670,7 +670,7 @@ if [[ $EBUILD_PHASE = depend ]] ; then
 		PROPERTIES DEFINED_PHASES UNUSED_05 UNUSED_04
 		UNUSED_03 UNUSED_02 UNUSED_01"
 
-	[ -n "${EAPI}" ] || EAPI=0
+	[ "${EAPI+set}" = set ] || EAPI=0
 
 	# The extra $(echo) commands remove newlines.
 	if [ -n "${dbkey}" ] ; then
diff --git a/bin/repoman b/bin/repoman
index cfe4b8e01..3d1a4b0c1 100755
--- a/bin/repoman
+++ b/bin/repoman
@@ -322,7 +322,7 @@ qahelp={
 	"LICENSE.virtual":"Virtuals that have a non-empty LICENSE variable",
 	"DESCRIPTION.missing":"Ebuilds that have a missing or empty DESCRIPTION variable",
 	"DESCRIPTION.toolong":"DESCRIPTION is over %d characters" % max_desc_len,
-	"EAPI.definition":"EAPI definition does not conform to PMS section 8.3.1 (first non-comment, non-blank line)",
+	"EAPI.definition":"EAPI definition does not conform to PMS section 7.3.1 (first non-comment, non-blank line)",
 	"EAPI.deprecated":"Ebuilds that use features that are deprecated in the current EAPI",
 	"EAPI.incompatible":"Ebuilds that use features that are only available with a different EAPI",
 	"EAPI.unsupported":"Ebuilds that have an unsupported EAPI version (you must upgrade portage)",
diff --git a/cnf/make.globals b/cnf/make.globals
index 975726ab6..1717baf68 100644
--- a/cnf/make.globals
+++ b/cnf/make.globals
@@ -52,7 +52,7 @@ FETCHCOMMAND_SFTP="bash -c \"x=\\\${2#sftp://} ; host=\\\${x%%/*} ; port=\\\${ho
 
 # Default user options
 FEATURES="assume-digests binpkg-logs distlocks ebuild-locks
-          fixlafiles news parallel-fetch protect-owned
+          fixlafiles news parallel-fetch parse-eapi-ebuild-head protect-owned
           sandbox sfperms strict unknown-features-warn unmerge-logs
           unmerge-orphans userfetch"
 
diff --git a/man/make.conf.5 b/man/make.conf.5
index b7108945c..9897073c6 100644
--- a/man/make.conf.5
+++ b/man/make.conf.5
@@ -407,9 +407,9 @@ parallelization. For additional parallelization, disable
 \fIebuild\-locks\fR.
 .TP
 .B parse\-eapi\-ebuild\-head
-Parse \fBEAPI\fR from the head of the ebuild (first 30 lines). This feature
-is only intended for experimental purposes and should not be enabled under
-normal circumstances.
+Parse \fBEAPI\fR from the head of the ebuild as specified in PMS section
+7.3.1, and treat non\-conformant ebuilds as invalid. This feature is
+enabled by default, and will soon become enabled unconditionally.
 .TP
 .B prelink\-checksums
 If \fBprelink\fR(8) is installed then use it to undo any prelinks on files
diff --git a/man/repoman.1 b/man/repoman.1
index 0e0715c9d..4305ce7d7 100644
--- a/man/repoman.1
+++ b/man/repoman.1
@@ -146,7 +146,7 @@ Syntax error in DEPEND (usually an extra/missing space/parenthesis)
 Ebuilds that have a missing or empty DESCRIPTION variable
 .TP
 .B EAPI.definition
-EAPI definition does not conform to PMS section 8.3.1 (first
+EAPI definition does not conform to PMS section 7.3.1 (first
 non\-comment, non\-blank line)
 .TP
 .B EAPI.deprecated
diff --git a/pym/_emerge/EbuildMetadataPhase.py b/pym/_emerge/EbuildMetadataPhase.py
index a34542d43..f89077ed1 100644
--- a/pym/_emerge/EbuildMetadataPhase.py
+++ b/pym/_emerge/EbuildMetadataPhase.py
@@ -9,9 +9,13 @@ from portage import os
 from portage import _encodings
 from portage import _unicode_decode
 from portage import _unicode_encode
+from portage.dep import _repo_separator
+from portage.elog import elog_process
+from portage.elog.messages import eerror
 import errno
 import fcntl
 import io
+import textwrap
 
 class EbuildMetadataPhase(SubProcess):
 
@@ -20,9 +24,9 @@ class EbuildMetadataPhase(SubProcess):
 	used to extract metadata from the ebuild.
 	"""
 
-	__slots__ = ("cpv", "eapi", "ebuild_hash", "fd_pipes", "metadata_callback",
-		"metadata", "portdb", "repo_path", "settings") + \
-		("_raw_metadata",)
+	__slots__ = ("cpv", "ebuild_hash", "fd_pipes",
+		"metadata_callback", "metadata", "portdb", "repo_path", "settings") + \
+		("_eapi", "_eapi_lineno", "_raw_metadata",)
 
 	_file_names = ("ebuild",)
 	_files_dict = slot_dict_class(_file_names, prefix="")
@@ -33,26 +37,31 @@ class EbuildMetadataPhase(SubProcess):
 		settings.setcpv(self.cpv)
 		ebuild_path = self.ebuild_hash.location
 
-		# the caller can pass in eapi in order to avoid
-		# redundant _parse_eapi_ebuild_head calls
-		eapi = self.eapi
-		if eapi is None and \
-			'parse-eapi-ebuild-head' in settings.features:
-			with io.open(_unicode_encode(ebuild_path,
-				encoding=_encodings['fs'], errors='strict'),
-				mode='r', encoding=_encodings['repo.content'],
-				errors='replace') as f:
-				eapi = portage._parse_eapi_ebuild_head(f)
-
-		if eapi is not None:
-			if not portage.eapi_is_supported(eapi):
-				self.metadata = self.metadata_callback(self.cpv,
-					self.repo_path, {'EAPI' : eapi}, self.ebuild_hash)
-				self._set_returncode((self.pid, os.EX_OK << 8))
-				self.wait()
-				return
-
-			settings.configdict['pkg']['EAPI'] = eapi
+		with io.open(_unicode_encode(ebuild_path,
+			encoding=_encodings['fs'], errors='strict'),
+			mode='r', encoding=_encodings['repo.content'],
+			errors='replace') as f:
+			self._eapi, self._eapi_lineno = portage._parse_eapi_ebuild_head(f)
+
+		parsed_eapi = self._eapi
+		if parsed_eapi is None:
+			parsed_eapi = "0"
+
+		if not parsed_eapi:
+			# An empty EAPI setting is invalid.
+			self._eapi_invalid(None)
+			self._set_returncode((self.pid, 1 << 8))
+			self.wait()
+			return
+
+		if not portage.eapi_is_supported(parsed_eapi):
+			self.metadata = self.metadata_callback(self.cpv,
+				self.repo_path, {'EAPI' : parsed_eapi}, self.ebuild_hash)
+			self._set_returncode((self.pid, os.EX_OK << 8))
+			self.wait()
+			return
+
+		settings.configdict['pkg']['EAPI'] = parsed_eapi
 
 		debug = settings.get("PORTAGE_DEBUG") == "1"
 		master_fd = None
@@ -144,7 +153,58 @@ class EbuildMetadataPhase(SubProcess):
 				# number of lines is incorrect.
 				self.returncode = 1
 			else:
-				metadata = zip(portage.auxdbkeys, metadata_lines)
-				self.metadata = self.metadata_callback(self.cpv,
-					self.repo_path, metadata, self.ebuild_hash)
+				metadata_valid = True
+				metadata = dict(zip(portage.auxdbkeys, metadata_lines))
+				parsed_eapi = self._eapi
+				if parsed_eapi is None:
+					parsed_eapi = "0"
+				if portage.eapi_is_supported(metadata["EAPI"]) and \
+					metadata["EAPI"] != parsed_eapi:
+					self._eapi_invalid(metadata)
+					if 'parse-eapi-ebuild-head' in self.settings.features:
+						metadata_valid = False
+
+				if metadata_valid:
+					self.metadata = self.metadata_callback(self.cpv,
+						self.repo_path, metadata, self.ebuild_hash)
+				else:
+					self.returncode = 1
+
+	def _eapi_invalid(self, metadata):
+
+		repo_name = self.portdb.getRepositoryName(self.repo_path)
 
+		msg = []
+		msg.extend(textwrap.wrap(("EAPI assignment in ebuild '%s%s%s' does not"
+			" conform with PMS section 7.3.1:") %
+			(self.cpv, _repo_separator, repo_name), 70))
+
+		if not self._eapi:
+			# None means the assignment was not found, while an
+			# empty string indicates an (invalid) empty assingment.
+			msg.append(
+				"\tvalid EAPI assignment must"
+				" occur on or before line: %s" %
+				self._eapi_lineno)
+		else:
+			msg.append(("\tbash returned EAPI '%s' which does not match "
+				"assignment on line: %s") %
+				(metadata["EAPI"], self._eapi_lineno))
+
+		if 'parse-eapi-ebuild-head' in self.settings.features:
+			msg.extend(textwrap.wrap(("NOTE: This error will soon"
+				" become unconditionally fatal in a future version of Portage,"
+				" but at this time, it can by made non-fatal by setting"
+				" FEATURES=-parse-eapi-ebuild-head in"
+				" make.conf."), 70))
+		else:
+			msg.extend(textwrap.wrap(("NOTE: This error will soon"
+				" become unconditionally fatal in a future version of Portage."
+				" At the earliest opportunity, please enable"
+				" FEATURES=parse-eapi-ebuild-head in make.conf in order to"
+				" make this error fatal."), 70))
+
+		for line in msg:
+			eerror(line, phase="other", key=self.cpv)
+		elog_process(self.cpv, self.settings,
+			phasefilter=("other",))
diff --git a/pym/_emerge/actions.py b/pym/_emerge/actions.py
index 9c8712050..62f3ff79d 100644
--- a/pym/_emerge/actions.py
+++ b/pym/_emerge/actions.py
@@ -28,7 +28,7 @@ portage.proxy.lazyimport.lazyimport(globals(),
 from portage.localization import _
 from portage import os
 from portage import shutil
-from portage import _unicode_decode
+from portage import eapi_is_supported, _unicode_decode
 from portage.cache.cache_errors import CacheError
 from portage.const import GLOBAL_CONFIG_PATH
 from portage.const import _ENABLE_DYN_LINK_MAP, _ENABLE_SET_CONFIG
@@ -1716,9 +1716,6 @@ def action_metadata(settings, portdb, myopts, porttrees=None):
 	if onProgress is not None:
 		onProgress(maxval, curval)
 
-	from portage import eapi_is_supported, \
-		_validate_cache_for_unsupported_eapis
-
 	# TODO: Display error messages, but do not interfere with the progress bar.
 	# Here's how:
 	#  1) erase the progress bar
@@ -1758,8 +1755,7 @@ def action_metadata(settings, portdb, myopts, porttrees=None):
 				eapi = eapi.lstrip('-')
 				eapi_supported = eapi_is_supported(eapi)
 				if not eapi_supported:
-					if not _validate_cache_for_unsupported_eapis:
-						continue
+					continue
 
 				dest = None
 				try:
diff --git a/pym/portage/__init__.py b/pym/portage/__init__.py
index 3495b96ec..31d580742 100644
--- a/pym/portage/__init__.py
+++ b/pym/portage/__init__.py
@@ -435,29 +435,25 @@ def eapi_is_supported(eapi):
 		return False
 	return eapi <= portage.const.EAPI
 
-# Generally, it's best not to assume that cache entries for unsupported EAPIs
-# can be validated. However, the current package manager specification does not
-# guarantee that the EAPI can be parsed without sourcing the ebuild, so
-# it's too costly to discard existing cache entries for unsupported EAPIs.
-# Therefore, by default, assume that cache entries for unsupported EAPIs can be
-# validated. If FEATURES=parse-eapi-* is enabled, this assumption is discarded
-# since the EAPI can be determined without the incurring the cost of sourcing
-# the ebuild.
-_validate_cache_for_unsupported_eapis = True
-
-_parse_eapi_ebuild_head_re = re.compile(r'^EAPI=[\'"]?([^\'"#]*)')
-_parse_eapi_ebuild_head_max_lines = 30
+# This pattern is specified by PMS section 7.3.1.
+_pms_eapi_re = re.compile(r"^[ \t]*EAPI=(['\"]?)([A-Za-z0-9+_.-]*)\1[ \t]*(#.*)?$")
+_comment_or_blank_line = re.compile(r"^\s*(#.*)?$")
 
 def _parse_eapi_ebuild_head(f):
-	count = 0
+	eapi = None
+	eapi_lineno = None
+	lineno = 0
 	for line in f:
-		m = _parse_eapi_ebuild_head_re.match(line)
-		if m is not None:
-			return m.group(1).strip()
-		count += 1
-		if count >= _parse_eapi_ebuild_head_max_lines:
+		lineno += 1
+		m = _comment_or_blank_line.match(line)
+		if m is None:
+			eapi_lineno = lineno
+			m = _pms_eapi_re.match(line)
+			if m is not None:
+				eapi = m.group(2)
 			break
-	return '0'
+
+	return (eapi, eapi_lineno)
 
 def _movefile(src, dest, **kwargs):
 	"""Calls movefile and raises a PortageException if an error occurs."""
diff --git a/pym/portage/dbapi/porttree.py b/pym/portage/dbapi/porttree.py
index bdb87f13e..16962b566 100644
--- a/pym/portage/dbapi/porttree.py
+++ b/pym/portage/dbapi/porttree.py
@@ -485,38 +485,22 @@ class portdbapi(dbapi):
 				raise KeyError(mycpv)
 
 			self.doebuild_settings.setcpv(mycpv)
-			eapi = None
-
-			if eapi is None and \
-				'parse-eapi-ebuild-head' in self.doebuild_settings.features:
-				with io.open(_unicode_encode(myebuild,
-					encoding=_encodings['fs'], errors='strict'),
-					mode='r', encoding=_encodings['repo.content'],
-					errors='replace') as f:
-					eapi = portage._parse_eapi_ebuild_head(f)
-
-			if eapi is not None:
-				self.doebuild_settings.configdict['pkg']['EAPI'] = eapi
-
-			if eapi is not None and not portage.eapi_is_supported(eapi):
-				mydata = self._metadata_callback(
-					mycpv, mylocation, {'EAPI':eapi}, ebuild_hash)
-			else:
-				proc = EbuildMetadataPhase(cpv=mycpv, eapi=eapi,
-					ebuild_hash=ebuild_hash,
-					metadata_callback=self._metadata_callback, portdb=self,
-					repo_path=mylocation,
-					scheduler=PollScheduler().sched_iface,
-					settings=self.doebuild_settings)
 
-				proc.start()
-				proc.wait()
+			proc = EbuildMetadataPhase(cpv=mycpv,
+				ebuild_hash=ebuild_hash,
+				metadata_callback=self._metadata_callback, portdb=self,
+				repo_path=mylocation,
+				scheduler=PollScheduler().sched_iface,
+				settings=self.doebuild_settings)
+
+			proc.start()
+			proc.wait()
 
-				if proc.returncode != os.EX_OK:
-					self._broken_ebuilds.add(myebuild)
-					raise KeyError(mycpv)
+			if proc.returncode != os.EX_OK:
+				self._broken_ebuilds.add(myebuild)
+				raise KeyError(mycpv)
 
-				mydata = proc.metadata
+			mydata = proc.metadata
 
 		mydata["repository"] = self.repositories.get_name_for_location(mylocation)
 		mydata["_mtime_"] = ebuild_hash.mtime
diff --git a/pym/portage/package/ebuild/config.py b/pym/portage/package/ebuild/config.py
index 38f15c081..32d65367e 100644
--- a/pym/portage/package/ebuild/config.py
+++ b/pym/portage/package/ebuild/config.py
@@ -776,9 +776,6 @@ class config(object):
 			if bsd_chflags:
 				self.features.add('chflags')
 
-			if 'parse-eapi-ebuild-head' in self.features:
-				portage._validate_cache_for_unsupported_eapis = False
-
 			self._iuse_implicit_match = _iuse_implicit_match_cache(self)
 
 			self._validate_commands()
diff --git a/pym/portage/package/ebuild/doebuild.py b/pym/portage/package/ebuild/doebuild.py
index da30bda38..4f7d4a8a1 100644
--- a/pym/portage/package/ebuild/doebuild.py
+++ b/pym/portage/package/ebuild/doebuild.py
@@ -337,12 +337,14 @@ def doebuild_environment(myebuild, mydo, myroot=None, settings=None,
 	# when uninstalling a package that has corrupt EAPI metadata.
 	eapi = None
 	if mydo == 'depend' and 'EAPI' not in mysettings.configdict['pkg']:
-		if eapi is None and 'parse-eapi-ebuild-head' in mysettings.features:
+		if eapi is None:
 			with io.open(_unicode_encode(ebuild_path,
 				encoding=_encodings['fs'], errors='strict'),
 				mode='r', encoding=_encodings['content'],
 				errors='replace') as f:
-				eapi = _parse_eapi_ebuild_head(f)
+				eapi, eapi_lineno = _parse_eapi_ebuild_head(f)
+				if eapi is None:
+					eapi = "0"
 
 		if eapi is not None:
 			if not eapi_is_supported(eapi):
diff --git a/pym/repoman/checks.py b/pym/repoman/checks.py
index 733bbc3c1..77df603a2 100644
--- a/pym/repoman/checks.py
+++ b/pym/repoman/checks.py
@@ -8,6 +8,7 @@ and correctness of an ebuild."""
 import re
 import time
 import repoman.errors as errors
+import portage
 from portage.eapi import eapi_supports_prefix, eapi_has_implicit_rdepend, \
 	eapi_has_src_prepare_and_src_configure, eapi_has_dosed_dohard, \
 	eapi_exports_AA, eapi_exports_KV
@@ -284,14 +285,12 @@ class EbuildUselessCdS(LineCheck):
 
 class EapiDefinition(LineCheck):
 	"""
-	Check that EAPI assignment conforms to PMS section 8.3.1
+	Check that EAPI assignment conforms to PMS section 7.3.1
 	(first non-comment, non-blank line).
 	"""
 	repoman_check_name = 'EAPI.definition'
 	ignore_comment = True
-
-	# This pattern is specified by PMS section 8.3.1.
-	_eapi_re = re.compile(r"^[ \t]*EAPI=(['\"]?)([A-Za-z0-9+_.-]*)\1[ \t]*(#.*)?$")
+	_eapi_re = portage._pms_eapi_re
 
 	def new(self, pkg):
 		self._cached_eapi = pkg.metadata['EAPI']