mirrors/lua-htmlparser
1
0
Fork 0
mirror of https://github.com/TangentFoxy/lua-htmlparser.git synced 2025-07-28 19:12:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

parse valueless attributes (e.g. itemscope)

Browse Source 
And allow nested tags, which was a newly encountered bug
This commit is contained in:
Wouter Scherphof
2013-03-21 14:03:17 +01:00
parent 67be34285e
commit 82dbddfd19
2 changed files with 32 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
htmlparser.lua
View File

@@ -16,14 +16,18 @@ local function parse(text)
local tagst, apos = tag:gettext(), 1
while true do
local start, k, quote, v
start, apos, k, quote = string.find(tagst, "%s+([^%s=]+)=(['\"]?)", apos)
local start, k, eq, quote, v
start, apos, k, eq, quote = string.find(tagst, "%s+([^%s=]+)(=?)(['\"]?)", apos)
if not k then break end
local pattern = "=([^%s'\">]*)"
if quote ~= '' then
pattern = quote .. "([^" .. quote .. "]*)" .. quote
if eq == "" then
v = ""
else
local pattern = "=([^%s'\">]*)"
if quote ~= '' then
pattern = quote .. "([^" .. quote .. "]*)" .. quote
end
start, apos, v = string.find(tagst, pattern, apos)
end
start, apos, v = string.find(tagst, pattern, apos)
tag:addattribute(k, v)
end
@@ -31,17 +35,16 @@ local function parse(text)
descend = false
tag:close()
else
opentags[tag.name] = tag
opentags[tag.name] = opentags[tag.name] or {}
table.insert(opentags[tag.name], tag)
end
local closeend = tpos
while true do
local closestart, closing, closename
closestart, closeend, closing, closename = string.find(root._text, "[^<]*<(/?)(%w+)", closeend)
closing = closing and closing ~= ''
if not closing then break end
tag = opentags[closename]
opentags[closename] = nil
if not closing or closing == "" then break end
tag = table.remove(opentags[closename])
closestart = string.find(root._text, "<", closestart)
tag:close(closestart, closeend + 1)
node = tag.parent