evilgooinajar
/
htmlq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fix: proper printing of document

main
evilgooinajar 2025-01-18 18:52:52 +01:00
parent 82c0038580
commit 86f8f67562
1 changed files with 43 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
html.lua
View File

@ -523,23 +523,40 @@ function M.clean_text_nodes(node)
end end
function M.print_document(node, indent) function M.tostring(node, indent, include_internal_pseudoelements)
-- Default indentation is 0 (root level) -- Default indentation is 0 (root level)
indent = indent or 0 indent = indent or 0
include_internal_pseudoelements = include_internal_pseudoelements or false
local is_pseudo_element = (node.tag_name or ":root"):sub(1,1) == ":"
local indent_level_str = " " local indent_level_str = " "
-- Create the indentation string (e.g., " " for each level) -- Create the indentation string (e.g., " " for each level)
local indent_str = string.rep(indent_level_str, indent) local indent_str = string.rep(indent_level_str, indent)
if node.tag_name == ":text" then if node.tag_name == ":text" then
print(indent_str .. "<:text>\n" .. node.content .. "\n" .. indent_str .. "</:text>") local str = ""
return
if include_internal_pseudoelements then
str = str .. "<:text>"
end
str = str .. node.content
if include_internal_pseudoelements then
str = str .. "</:text>"
end
return str
end end
local node_name = "" local node_name = ""
if not is_pseudo_element or include_internal_pseudoelements then
-- Print the current node's tag name -- Print the current node's tag name
node_name = node_name .. indent_str .. "<" .. (node.tag_name or ":root") node_name = node_name .. "\n" .. indent_str .. "<" .. (node.tag_name or ":root")
end
-- Print attributes if any -- Print attributes if any
if next(node.attributes) ~= nil then if next(node.attributes) ~= nil then
@ -557,23 +574,40 @@ function M.print_document(node, indent)
end end
end end
if not is_pseudo_element or include_internal_pseudoelements then
node_name = node_name .. ">" node_name = node_name .. ">"
end
print( node_name ) --print( node_name )
local next_indent = indent + 1
if is_pseudo_element and not include_internal_pseudoelements then
next_indent = indent
end
-- Recursively print children -- Recursively print children
for _, child in ipairs(node.children) do for _, child in ipairs(node.children) do
M.print_document(child, indent + 1) node_name = node_name .. M.tostring(child, next_indent, include_internal_pseudoelements)
end end
if not VOID_TAGS[node.tag_name] and ( not is_pseudo_element or include_internal_pseudoelements ) then
-- Print the closing tag -- Print the closing tag
print(indent_str .. "</" .. (node.tag_name or ":root") .. ">") local end_indent = ""
local closing_text_tag = "</:text>"
if node_name:sub(#node_name, #node_name) == ">" and node_name:sub(#node_name - #closing_text_tag + 1, #node_name) ~= closing_text_tag then
end_indent = "\n" .. indent_str
end
node_name = node_name .. end_indent .. "</" .. (node.tag_name or ":root") .. ">"
end
return node_name
end end
function M.parse( html_string ) function M.parse( html_string )
local clean_html = M.preprocess( html_string ) local clean_html = M.preprocess( html_string )