[checkmk-commits] Simplified htmllib.Encoder()

Module: check_mk Branch: master Commit: 49f1cbeb134a80d888354a1619a0cd505303c77b URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=49f1cbeb134a80… Author: Lars Michelsen &lt;lm(a)mathias-kettner.de&gt; Date: Wed May 23 08:59:23 2018 +0200 Simplified htmllib.Encoder() * Replaced Check_MK specific encoding logic with standard urllib calls * html.urlencode() now uses html.urlencode_plus() logic * Cleaned up html.urlencode_plus() Change-Id: I4838ca293d508ba4ebf50b238293bb2409b42cb1 --- tests/unit/web/test_htmllib_Encoder.py | 28 +++--------------- web/htdocs/htmllib.py | 54 +++++++++------------------------- web/plugins/views/painters.py | 2 +- 3 files changed, 19 insertions(+), 65 deletions(-) diff --git a/tests/unit/web/test_htmllib_Encoder.py b/tests/unit/web/test_htmllib_Encoder.py index b0bd054..74dfa83 100644 --- a/tests/unit/web/test_htmllib_Encoder.py +++ b/tests/unit/web/test_htmllib_Encoder.py @@ -10,7 +10,6 @@ def test_htmllib_integration(register_builtin_html): assert html.urlencode_vars([]) == "" assert html.urlencode("") == "" - assert html.urlencode_plus("") == "" @pytest.mark.parametrize("inp,out", [ @@ -21,8 +20,8 @@ def test_htmllib_integration(register_builtin_html): ([("a", "_-.")], "a=_-."), ([("a", "#")], "a=%23"), ([("a", "+")], "a=%2B"), - ([("a", " ")], "a=%20"), - ([("a", "/")], "a=/"), + ([("a", " ")], "a=+"), + ([("a", "/")], "a=%2F"), ]) def test_urlencode_vars(inp, out): result = htmllib.Encoder().urlencode_vars(inp) @@ -31,24 +30,6 @@ def test_urlencode_vars(inp, out): @pytest.mark.parametrize("inp,out", [ - (u"välue", "v%c3%a4lue"), - # TODO: None / int handling inconsistent with urlencode_vars() - (None, ""), - ("ä", "%c3%a4"), - ("_-.", "_-."), - ("#", "%23"), - ("+", "%2b"), - # TODO: Why + instead of %20? - (" ", "+"), - ("/", "/"), -]) -def test_urlencode(inp, out): - result = htmllib.Encoder().urlencode(inp) - assert type(result) == str - assert result == out - - -(a)pytest.mark.parametrize(&quot;inp,out&quot;,out", [ (u"välue", "v%C3%A4lue"), # TODO: None / int handling inconsistent with urlencode_vars() (None, ""), @@ -56,11 +37,10 @@ def test_urlencode(inp, out): ("_-.", "_-."), ("#", "%23"), ("+", "%2B"), - # TODO: Why + instead of %20? (" ", "+"), ("/", "%2F"), ]) -def test_urlencode_plus(inp, out): - result = htmllib.Encoder().urlencode_plus(inp) +def test_urlencode(inp, out): + result = htmllib.Encoder().urlencode(inp) assert type(result) == str assert result == out diff --git a/web/htdocs/htmllib.py b/web/htdocs/htmllib.py index 75e3cc7..b97610f 100644 --- a/web/htdocs/htmllib.py +++ b/web/htdocs/htmllib.py @@ -165,20 +165,15 @@ class Escaper(object): # '----------------------------------------------------------------------' class Encoder(object): - # This function returns a str object, never unicode! - # Beware: this code is crucial for the performance of Multisite! - # Changing from the self coded urlencode to urllib.quote - # is saving more then 90% of the total HTML generating time - # on more complex pages! - # - # TODO: Cleanup self.urlencode_vars, self.urlencode and self.urlencode_plus. - # urlencode_vars() should directly use urlencode or urlencode_vars and - # not fallback to self.urlencode on it's own. self.urlencode() should - # work for a single value exacly as urlencode_vars() does for multiple - # TODO: Refactor to use urllib.urlencode() def urlencode_vars(self, vars): + """Convert a mapping object or a sequence of two-element tuples to a “percent-encoded” string + + This function returns a str object, never unicode! + Note: This should be changed once we change everything to + unicode internally. + """ assert type(vars) == list - output = [] + pairs = [] for varname, value in sorted(vars): assert type(varname) == str @@ -189,35 +184,18 @@ class Encoder(object): #assert type(value) == str, "%s: %s" % (varname, value) - try: - # urllib is not able to encode non-Ascii characters. Yurks - output.append(varname + '=' + urllib.quote(value)) - except: - output.append(varname + '=' + self.urlencode(value)) # slow but working + pairs.append((varname, value)) - return '&'.join(output) + return urllib.urlencode(pairs) def urlencode(self, value): - if type(value) == unicode: - value = value.encode("utf-8") - elif value == None: - return "" + """Replace special characters in string using the %xx escape. - assert type(value) == str - - ret = "" - for c in value: - if c == " ": - c = "+" - elif ord(c) <= 32 or ord(c) > 127 or c in [ '#', '+', '"', "'", "=", "&", ":", "%" ]: - c = "%%%02x" % ord(c) - ret += c - return ret - - - # Like urllib.quote() but also replaces spaces and / - def urlencode_plus(self, value): + This function returns a str object, never unicode! + Note: This should be changed once we change everything to + unicode internally. + """ if type(value) == unicode: value = value.encode("utf-8") elif value == None: @@ -1387,10 +1365,6 @@ class html(HTMLGenerator): def urlencode(self, value): return self.encoder.urlencode(value) - # TODO: Cleanup all call sites to self.encoder.* - def urlencode_plus(self, value): - return self.encoder.urlencode_plus(value) - # # escaping - deprecated functions # diff --git a/web/plugins/views/painters.py b/web/plugins/views/painters.py index 2c12297..07ffb00 100644 --- a/web/plugins/views/painters.py +++ b/web/plugins/views/painters.py @@ -402,7 +402,7 @@ def replace_action_url_macros(url, what, row): for key, val in macros.items(): url = url.replace("$%s$" % key, val) - url = url.replace("$%s_URL_ENCODED$" % key, html.urlencode_plus(val)) + url = url.replace("$%s_URL_ENCODED$" % key, html.urlencode(val)) return url