Add Donnie's quote check, refactor other ebuild-content checks into classes and get them out of repoman in an attempt to start moving other checks out as well. Again wonder at the price of StringIO and N passes, is the readability worth it?

svn path=/main/trunk/; revision=7895

1 files changed, 34 insertions, 65 deletions

diff --git a/bin/repoman b/bin/repoman
index 9e78a964f..28b68dab8 100755
--- a/bin/repoman
+++ b/bin/repoman
@@ -20,6 +20,11 @@ try:
 except ImportError:
 	import pickle
 
+try:
+	import cStringIO as StringIO
+except ImportError:
+	import StringIO
+
 if not hasattr(__builtins__, "set"):
 	from sets import Set as set
 
@@ -38,10 +43,19 @@ try:
 	import portage
 except ImportError:
 	from os import path as osp
-	sys.path.insert(0, osp.join(osp.dirname(osp.dirname(osp.realpath(__file__))), "pym"))
+	sys.path.insert(0, osp.join(osp.dirname(osp.dirname(osp.realpath(__file__))), 'pym'))
 	import portage
 del os.environ["PORTAGE_LEGACY_GLOBALS"]
 
+try:
+	from repoman.checks import EbuildWhitespaceCheck, EbuildHeaderCheck, EbuildQuoteCheck, \
+		EbuildAssignmentCheck
+except ImportError:
+	from os import path as osp
+	sys.path.insert(0, osp.join(osp.dirname(osp.dirname(osp.realpath(__file__))), 'pym'))
+	from repoman.checks import EbuildWhitespaceCheck, EbuildHeaderCheck, EbuildQuoteCheck, \
+		EbuildAssignmentCheck
+
 import portage.checksum
 import portage.const
 import portage.dep
@@ -1322,71 +1336,26 @@ for x in scanlist:
 				stats["RESTRICT.invalid"] += len(mybadrestrict)
 				for mybad in mybadrestrict:
 					fails["RESTRICT.invalid"].append(x+"/"+y+".ebuild: %s" % mybad)
+		# Syntax Checks
+		path = checkdir + '/' + y + '.ebuild'
+		myear = time.gmtime(os.stat(path)[ST_MTIME])[0]
+		file_contents = open(path, 'rb').readlines()
+		contents = StringIO.StringIO(''.join(file_contents))
+		for check in (EbuildWhitespaceCheck, EbuildQuoteCheck, EbuildAssignmentCheck):
+			c = check(contents)
+			errors = c.Run()
+			for e in errors:
+				stats[c.repoman_check_name] += 1
+				fails[c.repoman_check_name].append(x + '/' + y + '.ebuild: %s' % e[1] % e[0])
+			contents.seek(0) # move fp to the beginning of the StringIO Object
+		del check
+		check = EbuildHeaderCheck(contents, str(myear))
+		errors = check.Run()
+		for e in errors:
+			stats[check.repoman_check_name] += 1
+			fails[check.repoman_check_name].append(x + '/' + y + '.ebuild: %s' % e[1] % e[0])
+		del check, errors
 
-		#syntax checks
-		myear = time.gmtime(os.stat(checkdir+"/"+y+".ebuild")[ST_MTIME])[0]
-		gentoo_copyright = re.compile(r'^# Copyright ((1999|200\d)-)?' + str(myear) + r' Gentoo Foundation')
-		gentoo_license = re.compile(r'^# Distributed under the terms of the GNU General Public License v2$')
-		cvs_header = re.compile(r'^#\s*\$Header.*\$$')
-		ignore_line = re.compile(r'(^$)|(^(\t)*#)')
-		leading_spaces = re.compile(r'^[\S\t]')
-		trailing_whitespace = re.compile(r'.*([\S]$)')
-		readonly_assignment = re.compile(r'^\s*(export\s+)?(A|CATEGORY|P|PV|PN|PR|PVR|PF|D|WORKDIR|FILESDIR|FEATURES|USE)=')
-		line_continuation = re.compile(r'([^#]*\S)(\s+|\t)\\$')
-		linenum=0
-		previous_line = None
-		for line in input(checkdir+"/"+y+".ebuild"):
-			linenum += 1
-			# Gentoo copyright check
-			if linenum == 1:
-				match = gentoo_copyright.match(line)
-				if not match:
-					myerrormsg = "Copyright header Error. Possibly date related."
-					stats["ebuild.badheader"] +=1
-					fails["ebuild.badheader"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-			# Gentoo license check
-			elif linenum == 2:
-				match = gentoo_license.match(line)
-				if not match:
-					myerrormsg = "Gentoo License Error."
-					stats["ebuild.badheader"] +=1
-					fails["ebuild.badheader"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-			# CVS Header check
-			elif linenum == 3:
-				match = cvs_header.match(line)
-				if not match:
-					myerrormsg = "CVS Header Error."
-					stats["ebuild.badheader"] +=1
-					fails["ebuild.badheader"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-			else:
-				match = ignore_line.match(line)
-				if not match:
-					# Excluded Blank lines and full line comments. Good!
-					# Leading Spaces Check
-					match = leading_spaces.match(line)
-					if not match:
-						#Line has got leading spaces. Bad!
-						myerrormsg = "Leading Space Syntax Error. Line %d" % linenum
-						stats["ebuild.minorsyn"] +=1
-						fails["ebuild.minorsyn"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-					# Trailing whitespace check
-					match = trailing_whitespace.match(line)
-					if not match:
-						#Line has got trailing whitespace. Bad!
-						myerrormsg = "Trailing whitespace Syntax Error. Line %d" % linenum
-						stats["ebuild.minorsyn"] +=1
-						fails["ebuild.minorsyn"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-					# Readonly variable assignment check
-					match = readonly_assignment.match(line)
-					# The regex can give a false positive for continued lines,
-					# so we check the previous line to see if it was continued.
-					if match and (not previous_line or not line_continuation.match(previous_line)):
-						# invalid assignment, very bad!
-						myerrormsg = "Readonly variable assignment to %s on line %d" % (match.group(2), linenum)
-						stats["variable.readonly"] += 1
-						fails["variable.readonly"].append(x+"/"+y+".ebuild: %s" % myerrormsg)
-			previous_line = line
-		del previous_line
 
 		if "--force" in myoptions:
 			# The dep_check() calls are the most expensive QA test. If --force