diff --git a/src/htmlparser.lua b/src/htmlparser.lua index 209332c..1f05ef9 100644 --- a/src/htmlparser.lua +++ b/src/htmlparser.lua @@ -23,21 +23,19 @@ local function parse(text) local start, k, eq, quote, v start, apos, k, eq, quote = string.find(tagst, "%s+" .. -- some uncaptured space - "([^%s=]+)" .. -- k = an unspaced string up to an optional "=" + "([^%s=/>]+)" .. -- k = an unspaced string up to an optional "=" or the "/" or ">" "(=?)" .. -- eq = the optional; "=", else "" "(['\"]?)", -- quote = an optional "'" or '"' following the "=", or "" apos) if not k or k == "/>" or k == ">" then break end - if eq == "" then - v = "" - else + if eq == "=" then local pattern = "=([^%s>]*)" - if quote ~= '' then + if quote ~= "" then pattern = quote .. "([^" .. quote .. "]*)" .. quote end start, apos, v = string.find(tagst, pattern, apos) end - tag:addattribute(k, v) + tag:addattribute(k, v or "") end if voidelements[string.lower(tag.name)] then diff --git a/src/htmlparser/ElementNode.lua b/src/htmlparser/ElementNode.lua index d6c99e8..e58454c 100644 --- a/src/htmlparser/ElementNode.lua +++ b/src/htmlparser/ElementNode.lua @@ -51,9 +51,8 @@ function ElementNode:addattribute(k, v) self.attributes[k] = v if string.lower(k) == "id" then self.id = v - end -- class attribute contains "space-separated tokens", each of which we'd like quick access to - if string.lower(k) == "class" then + elseif string.lower(k) == "class" then for class in string.gmatch(v, "%S+") do table.insert(self.classes, class) end @@ -98,18 +97,20 @@ local function select(self, s) local sets = {[""] = self.deeperelements, ["["] = self.deeperattributes, ["#"] = self.deeperids, ["."] = self.deeperclasses} local function match(t, w) - local m, v - if t == "[" then w, m, v = string.match(w, + local m, e, v + if t == "[" then w, m, e, v = string.match(w, "([^=|%*~%$!%^]+)" .. -- w = 1 or more characters up to a possible "=", "|", "*", "~", "$", "!", or "^" "([|%*~%$!%^]?)" .. -- m = an optional "|", "*", "~", "$", "!", or "^", preceding the optional "=" - "=?" .. -- an optional uncaptured "=" + "(=?)" .. -- e = the optional "=" "(.*)" -- v = anything following the "=", or else "" ) end local matched = Set:new(sets[t][w]) -- attribute value selectors - if v and v ~= "" then + if e == "=" then + if #v < 2 then v = "'" .. v .. "'" end -- values should be quoted v = string.sub(v, 2, #v - 1) -- strip quotes + if m == "!" then matched = Set:new(self.deepernodes) end -- include those without that attribute for node in pairs(matched) do local a = node.attributes[w] -- equals diff --git a/tst/init.lua b/tst/init.lua index 8f1a53f..1349809 100644 --- a/tst/init.lua +++ b/tst/init.lua @@ -6,10 +6,9 @@ local lunitx = require("lunitx") module("html", lunitx.testcase, package.seeall) local htmlparser = require("htmlparser") -local tree, sel function test_void() - tree = htmlparser.parse([[ + local tree = htmlparser.parse([[



@@ -27,14 +26,62 @@ function test_void() assert_equal(4, #n.nodes, "deeper level") else assert_equal("br", n.name, "name") - assert_equal(0, #n.attributes, "attributes") assert_equal("", n:getcontent(), "content") + for _ in pairs(n.attributes) do + fail("should not have attributes") + end end end end +function test_attr() + local tree = htmlparser.parse([[ + + + + ]]) + assert_equal(3, #tree.nodes, "top level") + local n + for _,v in ipairs(tree.nodes) do + if v.name == "n" then n = v break end + end + assert(tree("[a1]")[n], "a1") + assert(tree("[a2]")[n], "a2") + assert(tree("[a3]")[n], "a3") + assert(tree("[a4]")[n], "a4") + assert(tree("[a5]")[n], "a5") + assert(tree("[a6]")[n], "a6") + assert(tree("[a7]")[n], "a7") + assert(tree("[a8]")[n], "a8") + assert(tree("[a1='']")[n], "a1=''") + assert(tree("[a2='']")[n], "a2=''") + assert(tree("[a3='']")[n], "a3=''") + assert(tree("[a4='']")[n], "a4=''") + assert(tree("[a5='a\"5\"']")[n], "a5='a\"5\"'") + assert(tree("[a6=\"a'6'\"]")[n], "a6=\"a'6'\"") + assert(tree("[a8='a=8']")[n], "a8='a=8'") + assert_equal(1, tree("[a10=]"):len(), "a10=") + assert_equal(1, tree("[a10='']"):len(), "a10=''") + assert_equal(2, tree("[a10!='enen']"):len(), "a10!='enen'") + assert_equal(2, tree("[a10!='']"):len(), "a10!=''") + assert_equal(3, tree("[a0!='']"):len(), "a0!=''") + assert_equal(0, tree("[a0='']"):len(), "a0=''") + assert_equal(2, tree("[a9|='en']"):len(), "a9|='en'") + assert_equal(3, tree("[a9^='en']"):len(), "a9^='en'") + assert_equal(1, tree("[a9$='en']"):len(), "a9$='en'") + assert_equal(1, tree("[a11~='two']"):len(), "a1~='two'") + assert_equal(2, tree("[a11~='three']"):len(), "a1~='three'") + assert_equal(1, tree("[a11~='four']"):len(), "a1~='four'") + assert_equal(1, tree("[a7*='7']"):len(), "a7*='7'") + assert_equal(1, tree("[a11*='f']"):len(), "a11*='f'") +end + function test_descendants() - tree = htmlparser.parse([[ + local tree = htmlparser.parse([[ 1 1 2 @@ -57,12 +104,12 @@ function test_descendants() not ]]) - sel = tree("parent child") + local sel = tree("parent child") assert_equal(8, sel:len(), 'parent child') end function test_children() - tree = htmlparser.parse([[ + local tree = htmlparser.parse([[ 1 1 2 @@ -85,6 +132,6 @@ function test_children() not ]]) - sel = tree("parent > child") + local sel = tree("parent > child") assert_equal(4, sel:len(), 'parent > child') end \ No newline at end of file