Mercurial > farol > farolluz
diff farolluz/parsers/cpe.py @ 40:1d9b2b06067e
Add a CPE parser (and tests)
author | Benoît Allard <benoit.allard@greenbone.net> |
---|---|
date | Mon, 29 Dec 2014 14:30:39 +0100 |
parents | |
children | bb1dd2a55643 |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/farolluz/parsers/cpe.py Mon Dec 29 14:30:39 2014 +0100 @@ -0,0 +1,259 @@ +"""\ +a cpe class to ease the creation of a producttree based on cpe + +This is based on: + + NIST Interagency Report 7695 + Common Platform Enumeration: Naming Specification Version 2.3 + +CPE is a trademark of The MITRE Corporation. + +""" + +import re + +PCT_MAP ={'!': "%21", '"': "%22", '#': "%23", '$': "%24", '%': "%25", '&': "%26", + "'": "%27", '(': "%28", ')': "%29", '*': "%2a", '+': "%2b", ',': "%2c", + '/': "%2f", ':': "%3a", ';': "%3b", '<': "%3c", "=": "%3d", '>': "%3e", + '?': "%3f", '@': "%40", '[': "%5b", '\\': "%5c","]": "%5d", '^': "%5e", + '`': "%60", '{': "%7b", '|': "%7c", '}': "%7d", "~": "%7e"} + +PCT_MAP_i = dict((v, k) for k, v in PCT_MAP.iteritems()) + +def pct_encode(c): + """ Returns the right percent-encoding of c """ + if c in "-.": + return c + return PCT_MAP[c] + return {'!': "%21", '"': "%22", '#': "%23", '$': "%24", '%': "%25", '&': "%26", + "'": "%27", '(': "%28", ')': "%29", '*': "%2a", '+': "%2b", ',': "%2c", + "-": c, '.': c, '/': "%2f", ':': "%3a", ';': "%3b", '<': "%3c", + "=": "%3d", '>': "%3e", '?': "%3f", '@': "%40", '[': "%5b", '\\': "%5c", + "]": "%5d", '^': "%5e", '`': "%60", '{': "%7b", '|': "%7c", '}': "%7d", + "~": "%7e"}[c] + +def decode(s): + if s == '': + return ANY + if s == '-': + return NA + s = s.lower() + res = "" + idx = 0 + embedded = False + while idx < len(s): + c = s[idx] + if c in ".-~": + res += "\\" + c + embedded = True + elif c != '%': + res += c + embedded = True + else: + form = s[idx:idx+3] + if form == "%01": + if (((idx == 0) or (idx == (len(s) - 3))) or + ( not embedded and (s[idx-4:idx-1] == "%01")) or + (embedded and (len(s) > idx + 6) and (s[idx+3:idx+6] == "%01"))): + res += '?' + else: + raise ValueError + elif form == "%02": + if (idx == 0) or (idx == len(s) - 3): + res += '*' + else: + raise ValueError + else: + res += '\\' + PCT_MAP_i[form] + embedded = True + idx += 2 + idx += 1 + return CPEAttribute(res) + +def unbind_value_fs(s): + if s == '*': + return ANY + if s == '-': + return NA + res = "" + idx = 0 + embedded = False + while idx < len(s): + c = s[idx] + if re.match("[a-zA-Z0-9_]", c) is not None: + res += c + embedded = True + elif c == "\\": + res += s[idx:idx+2] + embedded = True + idx += 1 + elif c == "*": + if (idx == 0) or (idx == (len(s) - 1)): + res += c + embedded = True + else: + raise ValueError + elif c == "?": + if (((idx == 0) or (idx == (len(s) - 1))) or + (not embedded and (s[idx - 1] == "?")) or + (embedded and (s[idx + 1] == "?"))): + res += c + embedded = False + else: + raise ValueError + else: + res += "\\" + c + embedded = True + idx += 1 + return CPEAttribute(res) + +class CPEAttribute(object): + """ We need a special class to deal with ANY / NA / "string" """ + + def __init__(self, value=None, any=False, na=False): + self.any = any + self.na = na + self.value = value + + def bind_for_URI(self): +# print self.any, self.na, self.value + if self.any: + return "" + if self.na: + return '-' + return self.transform_for_uri() + + def transform_for_uri(self): + res = "" + idx = 0 + while idx < len(self.value): + c = self.value[idx] + if re.match("[a-zA-Z0-9_]", c) is not None: + res += c + elif c == '\\': + idx += 1 + c = self.value[idx] + res += pct_encode(c) + elif c == '?': + res += "%01" + elif c == '*': + res += "%02" + idx += 1 + return res + + def bind_for_fs(self): + if self.any: + return "*" + if self.na: + return "-" + return self.process_quoted_chars() + + def process_quoted_chars(self): + res = "" + idx = 0 + while idx < len(self.value): + c = self.value[idx] + if c != '\\': + res += c + else: + idx += 1 + c = self.value[idx] + if c in ".-_": + res += c + else: + res += '\\' + c + idx += 1 + return res + +ANY = CPEAttribute(any=True) +NA = CPEAttribute(na=True) + +class CPE(object): + + def __init__(self, part=None, vendor=None, product=None, version=None, update=None, edition=None, language=None, sw_edition=None, target_sw=None, target_hw=None, other=None): + self.part = part or CPEAttribute(any=True) + self.vendor = vendor or CPEAttribute(any=True) + self.product = product or CPEAttribute(any=True) + self.version = version or CPEAttribute(any=True) + self.update = update or CPEAttribute(any=True) + self.edition = edition or CPEAttribute(any=True) + self.language = language or CPEAttribute(any=True) + # Extended attributes: + self.sw_edition = sw_edition or CPEAttribute(any=True) + self.target_sw = target_sw or CPEAttribute(any=True) + self.target_hw = target_hw or CPEAttribute(any=True) + self.other = other or CPEAttribute(any=True) + + def bind_to_URI(self): + uri = 'cpe:/' + uri += ':'.join(a.bind_for_URI() for a in (self.part, self.vendor, self.product, self.version, self.update)) + # Special handling for edition + ed = self.edition.bind_for_URI() + sw_ed = self.sw_edition.bind_for_URI() + t_sw = self.target_sw.bind_for_URI() + t_hw = self.target_hw.bind_for_URI() + oth = self.other.bind_for_URI() + if sw_ed == "" and t_sw == "" and t_hw == "" and oth == "": + uri += ":" + ed + else: + uri += ":~" + '~'.join([ed, sw_ed, t_sw, t_hw, oth]) + uri += ':' + self.language.bind_for_URI() + return uri.rstrip(':') + + def unbind_URI(self, uri): + for idx, comp in enumerate(uri.split(':')): + if idx == 0: + continue + elif idx == 1: + self.part = decode(comp[1:]) + elif idx == 2: + self.vendor = decode(comp) + elif idx == 3: + self.product = decode(comp) + elif idx == 4: + self.version = decode(comp) + elif idx == 5: + self.update = decode(comp) + elif idx == 6: + if comp == "" or comp[0] != '~': + self.edition = decode(comp) + else: + ed, sw_ed, t_sw, t_hw, oth = comp[1:].split('~') + self.edition = decode(ed) + self.sw_edition = decode(sw_ed) + self.target_sw = decode(t_sw) + self.target_hw = decode(t_hw) + self.other = decode(oth) + elif idx == 7: + self.language = decode(comp) + + def bind_to_fs(self): + fs = 'cpe:2.3:' + fs += ':'.join(a.bind_for_fs() for a in (self.part, self.vendor, self.product, self.version, self.update, self.edition, self.language, self.sw_edition, self.target_sw, self.target_hw, self.other)) + return fs + + def unbind_fs(self, fs): + for idx, v in enumerate(fs.split(':')): + v = unbind_value_fs(v) + if idx == 2: + self.part = v + elif idx == 3: + self.vendor = v + elif idx == 4: + self.product = v + elif idx == 5: + self.version = v + elif idx == 6: + self.update = v + elif idx == 7: + self.edition = v + elif idx == 8: + self.language = v + elif idx == 9: + self.sw_edition = v + elif idx == 10: + self.target_sw = v + elif idx == 11: + self.target_hw = v + elif idx == 12: + self.other = v