wesnoth/data/tools/wmlvalidator
Severin Glöckner 06d9909c86 Changed Shebang line of python scripts.
It's mostly about making the scripts run if python defaults to python3.
Has been tested for each script.
2016-11-25 09:57:25 +01:00

282 lines
13 KiB
Python
Executable file

#!/usr/bin/env python2
"""
wmltest -- tool to validate the syntax and semantics of WML.
Use --help to see usage.
"""
#TODO:
#-define verbosity levels better
from __future__ import print_function
import wesnoth.wmldata as wmldata
import wesnoth.wmlparser as wmlparser
import wesnoth.wmlgrammar as wmlgrammar
import re
def print_indent(string, depth, char=' '):
print("%s%s" % (depth * char, string))
class Validator:
"""
The class that takes a wmlgrammar object to validate wml trees with
"""
def __init__(self, schema, verbosity=0):
self.schema = wmlgrammar.Grammar(schema)
self.verbosity = verbosity
self.validate_result = {}
def validate_result_add(self, from_file, line, origin, message):
if not from_file in self.validate_result:
self.validate_result[from_file] = []
self.validate_result[from_file].append({'line': line, 'origin': origin, 'message': message})
def validate_result_print(self):
normal = '\033[0m'
bold = '\033[1m'
underline = '\033[4m'
for k, v in self.validate_result.iteritems():
print("%s%s%s" % (bold, k, normal))
for i in v:
print("%s#%d: %s%s %s" % (underline, i['line'], i['origin'], normal, i['message']))
def validate(self, node, depth=0, name=None):
"""
Validate the given DataSub node.
depth indicates how deep we've recursed into the tree,
name is a mechanism for overwriting the node name to look up in the schema, used for overloaded names.
"""
if not name or name == node.name:
name = node.name
verbosename = name
else:
verbosename = "%s (%s)" % (node.name, name)
if self.verbosity > 1:
print_indent(node.name, depth)
try:
schema = self.schema.get_element(name)
except KeyError:
print("No valid schema found for %s" % verbosename)
return
# Validate the attributes
for attribute in schema.get_attributes():
matches = node.get_texts(attribute.name)
# Check frequency
nummatches = len(matches)
if attribute.freq == wmlgrammar.REQUIRED and nummatches != 1:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Should appear exactly once, not %d times" % nummatches)
elif attribute.freq == wmlgrammar.OPTIONAL and nummatches > 1:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Should appear at most once, not %d times" % nummatches)
elif attribute.freq == wmlgrammar.FORBIDDEN and nummatches > 0:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Should not appear. It appears %d times" % nummatches)
# Check use
for match in matches:
if 'translatable' in attribute.optionals and match.is_translatable() == False:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Value is translatable, but haven't _ at the beginning")
elif 'translatable' not in attribute.optionals and 'optional-translatable' not in attribute.optionals and match.is_translatable() == True:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Value isn't translatable, but have a _ at the beginning")
def check_attribute_value(value, pos=None):
def gerate_message_with_pos():
if pos is None:
return ""
else:
return " (At position %d)" % pos
if not attribute.validate(value):
self.validate_result_add(node.file, node.line, "Attribute [%s] %s%s" % (verbosename, attribute.name, gerate_message_with_pos()), "Value should be %s, found: %s" % (attribute.type, value))
regex_limit = re.compile(ur'^limit\((\d+.\d+|\d+),(\d+.\d+|\d+)\)$')
check_limit = [i for i in attribute.optionals if regex_limit.search(i)]
if len(check_limit):
check_limit = check_limit[0]
number_min, number_max = regex_limit.search(check_limit).groups()
if float(value) > float(number_max) or float(value) < float(number_min):
self.validate_result_add(node.file, node.line, "Attribute [%s] %s%s" % (verbosename, attribute.name, gerate_message_with_pos()), "Value must be between %s and %s, found : %s" % (number_min, number_max, value))
regex_limit_lower = re.compile(ur'^limit-lower\((\d+.\d+|\d+)\)$')
check_limit_lower = [i for i in attribute.optionals if regex_limit_lower.search(i)]
if len(check_limit_lower):
check_limit_lower = check_limit_lower[0]
number = regex_limit_lower.search(check_limit_lower).group(1)
if float(value) < float(number):
self.validate_result_add(node.file, node.line, "Attribute [%s] %s%s" % (verbosename, attribute.name, gerate_message_with_pos()), "Value needs to be at least %s, found : %s" % (number, value))
regex_limit_max = re.compile(ur'^limit-max\((\d+.\d+|\d+)\)$')
check_limit_max = [i for i in attribute.optionals if regex_limit_max.search(i)]
if len(check_limit_max):
check_limit_max = check_limit_max[0]
number = regex_limit_max.search(check_limit_max).group(1)
if float(value) > float(number):
self.validate_result_add(node.file, node.line, "Attribute [%s] %s%s" % (verbosename, attribute.name, gerate_message_with_pos()), "Value needs to be at max %s, found : %s" % (number, value))
regex_file_exist = re.compile(ur'^need-file-in\(([\w.\-\/]+)\)$')
check_file_exist = [i for i in attribute.optionals if regex_file_exist.search(i)]
if len(check_file_exist):
check_file_exist = check_file_exist[0]
directory = regex_file_exist.search(check_file_exist).group(1)
value_directory, value_file = re.search(re.compile(ur'(?:(.*)?\/)?(.+)'), value).groups()
import glob
if directory == '.':
sub_directory = os.path.dirname(node.file) + '/'
else:
sub_directory = os.path.dirname(node.file) + '/' + directory + '/'
if not value_directory is None:
sub_directory += value_directory + '/'
files_from_sub_directory = glob.glob(sub_directory + '*')
# We just want the names of the files from directory, but...
if os.path.splitext(value_file)[1] == '':
# ... if without extension in the value_file, then it is implied. In this case, we do want extensions
files_from_sub_directory = [re.sub(re.compile(r'^.*\/(.*)\..*'), r'\1', i) for i in files_from_sub_directory]
else:
files_from_sub_directory = [re.sub(re.compile(r'^.*\/(.*)'), r'\1', i) for i in files_from_sub_directory]
if not value_file in files_from_sub_directory:
self.validate_result_add(node.file, node.line, "Attribute [%s] %s%s" % (verbosename, attribute.name, gerate_message_with_pos()), "The file %s not exist in directory %s" % (value_file, sub_directory))
if 'list' in attribute.optionals:
pos = 1
for i in match.data.split(","):
if i[0] == ' ': i = i[1:]
check_attribute_value(i, pos=pos)
pos += 1
else:
check_attribute_value(match.data)
node.remove(match) # Get rid of these so we can see what's left
for attribute in node.get_all_text():
self.validate_result_add(node.file, node.line, "Attribute [%s] %s" % (verbosename, attribute.name), "Found, which has no meaning there")
# Validate the elements
for element in schema.get_elements():
matches = node.get_subs(element.name)
# Check frequency
nummatches = len(matches)
if element.freq == wmlgrammar.REQUIRED and nummatches != 1:
self.validate_result_add(node.file, node.line, "Element [%s] [%s]" % (verbosename, element.name), "Should appear exactly once, not %d times" % nummatches)
elif element.freq == wmlgrammar.OPTIONAL and nummatches > 1:
self.validate_result_add(node.file, node.line, "Element [%s] [%s]" % (verbosename, element.name), "Should appear at most once, not %d times" % nummatches)
elif element.freq == wmlgrammar.FORBIDDEN and nummatches > 0:
self.validate_result_add(node.file, node.line, "Element [%s] [%s]" % (verbosename, element.name), "Should not appear. It appears %d times" % nummatches)
# Check sub
for match in matches:
self.validate(match, depth+1, element.subname)
node.remove(match)
for element in node.get_all_subs():
self.validate_result_add(node.file, node.line, "Element [%s] [%s]" % (verbosename, element.name), "Found, which has no meaning there")
if __name__ == '__main__':
import argparse, subprocess, os, codecs, sys
# Ugly hack to force the output of UTF-8.
# This prevents us from crashing when we're being verbose
# and encounter a non-ascii character.
sys.stdout = codecs.getwriter('utf-8')(sys.stdout)
ap = argparse.ArgumentParser("Usage: %(prog)s [options]")
ap.add_argument("-p", "--path",
help = "Specify Wesnoth's data dir",
dest = "path")
ap.add_argument("-u", "--userpath",
help = "Specify user data dir",
dest = "userpath")
ap.add_argument("-s", "--schema",
help = "Specify WML schema",
dest = "schema")
ap.add_argument("-v", "--verbose",
action = "count",
dest = "verbose",
help = "Increase verbosity, 4 is the maximum.")
ap.add_argument("-D", "--define",
action = "append",
dest = "defines",
default = [],
help = "Define (empty) preprocessor macros, for campaign/multiplayer inclusion.")
ap.add_argument("filename",
nargs = "*",
help = "Files to validate or directory. If it is a directory, get all the cfg files from directory")
args = ap.parse_args()
if args.path:
args.path = os.path.expanduser(args.path)
if args.userpath:
args.userpath = os.path.expanduser(args.userpath)
if args.filename:
args.filename = [os.path.expanduser(i) for i in args.filename]
if not args.path:
try:
p = subprocess.Popen(["wesnoth", "--path"], stdout = subprocess.PIPE)
path = p.stdout.read().strip()
args.path = os.path.join(path, "data")
sys.stderr.write("No Wesnoth path given.\nAutomatically found '%s'\n" % (args.path, ) )
except OSError:
args.path = '.'
sys.stderr.write("Could not determine Wesnoth path.\nAssuming '%s'\n" % (args.path, ) )
list_files_analyze = []
if len(args.filename) < 1:
list_files_analyze.append(os.path.join(args.path, '_main.cfg'))
for i in args.filename:
if os.path.isdir(i):
if i[-1] != '/':
i += '/'
cfg_from_dir = [i + cfg for cfg in os.listdir(i) if cfg[-3:] == 'cfg']
list_files_analyze += cfg_from_dir
else:
list_files_analyze.append(i)
if args.verbose > 1:
print("Args: %s\n"% (args, ))
if not args.schema:
args.schema = os.path.join(args.path, 'schema.cfg')
# Parse the schema
parser = wmlparser.Parser(args.path)
if args.verbose > 3:
parser.verbose = True
parser.parse_file(args.schema)
schema = wmldata.DataSub("schema")
parser.parse_top(schema)
# Construct the validator
validator = Validator(schema, args.verbose)
# Parse the WML
parser = wmlparser.Parser(args.path, args.userpath)
if args.verbose > 3:
parser.verbose = True
if args.userpath:
parser.parse_text("{~add-ons}")
for file in list_files_analyze:
parser.parse_file(file)
for macro in args.defines:
parser.parse_text("#define %s \n#enddef" % (macro, ) )
data = wmldata.DataSub("root")
parser.parse_top(data)
# Validate
validator.validate(data)
validator.validate_result_print()
# vim: tabstop=4: shiftwidth=4: expandtab: softtabstop=4: autoindent: