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

Moved some details to a function

Browse Source 
... regarding [attribute=value] and other matching.
Also explained (in comments) the different matching patterns.
And fixed a bug where /> would be listed as an attribute.
And added a few more tests.
This commit is contained in:
Wouter Scherphof
2013-03-25 13:55:35 +01:00
parent de746865be
commit 64f3eb4df3
4 changed files with 52 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
htmlparser.lua
View File

@@ -9,7 +9,11 @@ local function parse(text)
local node, descend, tpos, opentags = root, true, 1, {}
while true do
local openstart, name
openstart, tpos, name = string.find(root._text, "<(%w+)[^>]*>", tpos)
openstart, tpos, name = string.find(root._text,
"<" .. -- an uncaptured starting "<"
"(%w+)" .. -- name = the first word, directly following the "<"
"[^>]*>", -- include, but not capture everything up to the next ">"
tpos)
if not name then break end
local tag = ElementNode:new(name, node, descend, openstart, tpos)
node = tag
@@ -17,12 +21,17 @@ local function parse(text)
local tagst, apos = tag:gettext(), 1
while true do
local start, k, eq, quote, v
start, apos, k, eq, quote = string.find(tagst, "%s+([^%s=]+)(=?)(['\"]?)", apos)
if not k then break end
start, apos, k, eq, quote = string.find(tagst,
"%s+" .. -- some uncaptured space
"([^%s=]+)" .. -- k = an unspaced string up to an optional "="
"(=?)" .. -- eq = the optiona; "=", else ""
"(['\"]?)", -- quote = an optional "'" or '"' following the "=", or ""
apos)
if not k or k == "/>" then break end
if eq == "" then
v = ""
else
local pattern = "=([^%s'\">]*)"
local pattern = "=([^%s>]*)"
if quote ~= '' then
pattern = quote .. "([^" .. quote .. "]*)" .. quote
end