andy-gg's picture
Upload 2477 files
7e9dc27 verified
--[[------------------------------------------------------
# Very fast xml parser for Lua <a href="https://travis-ci.org/lubyk/xml"><img src="https://travis-ci.org/lubyk/xml.png" alt="Build Status"></a>
This parser uses [RapidXML](http://rapidxml.sourceforge.net/) to parse XML
content.
<html><a href="https://github.com/lubyk/xml"><img style="position: absolute; top: 0; right: 0; border: 0;" src="https://s3.amazonaws.com/github/ribbons/forkme_right_green_007200.png" alt="Fork me on GitHub"></a></html>
*MIT license* &copy Marcin Kalicinski 2006, 2009, Gaspard Bucher 2014.
## Installation
With [luarocks](http://luarocks.org):
$ luarocks install xml
## Usage example
local data = xml.load(some_xml)
local xml_string = xml.dump(some_table)
--]]-----------------------------------------------------
local lub = require 'lub'
local lib = lub.Autoload 'xml'
local ipairs, pairs, insert, type,
match, tostring =
ipairs, pairs, table.insert, type,
string.match, tostring
local parser = lib.Parser()
-- Current version respecting [semantic versioning](http://semver.org).
lib.VERSION = '1.1.2'
lib.DEPENDS = { -- doc
-- Compatible with Lua 5.1 to 5.3 and LuaJIT
'lua >= 5.1, < 5.4',
-- Uses [Lubyk base library](http://doc.lubyk.org/lub.html)
'lub >= 1.0.3, < 2',
}
-- nodoc
lib.DESCRIPTION = {
summary = "Very fast xml parser based on RapidXML",
detailed = [[
This module is part of the Lubyk project.
Main features are:
- Fast and easy to use
- Complete documentation
- Based on proven code (RapidXML)
- Full test coverage
Read the documentation at http://doc.lubyk.org/xml.html.
]],
homepage = "http://doc.lubyk.org/"..lib.type..".html",
author = "Gaspard Bucher",
license = "MIT",
}
-- nodoc
lib.BUILD = {
github = 'lubyk',
includes = {'include', 'src/bind', 'src/vendor'},
platlibs = {
linux = {'stdc++'},
macosx = {'stdc++'},
},
-- FIXME: Implement platform flags for lut.Builder and see how it works with
-- luarocks /EHsc is needed for exception handling.
platflags = {
win32 = {'EHsc'},
},
}
--[[
# Lua table format
This xml library uses string keys in Lua tables to store
attributes and numerical keys for sub-nodes. Since the 'xml'
attribute is not allowed in XML, we use this key to store the
tag. Here is an example of Lua content:
## Lua
{xml='document',
{xml = 'article',
{xml = 'p', 'This is the first paragraph.'},
{xml = 'h2', class = 'opt', 'Title with opt style'},
},
{xml = 'article',
{xml = 'p', 'Some ', {xml = 'b', 'important'}, ' text.'},
},
}
## XML
And the equivalent xml:
#txt
<document>
<article>
<p>This is the first paragraph.</p>
<h2 class='opt'>Title with opt style</h2>
</article>
<article>
<p>Some <b>important</b> text.</p>
</article>
</document>
# Notes on speed
RapidXML is a very fast parser that uses in-place modification of the input
text. Since Lua strings are immutable, we have to make a copy except for the
xml.Parser.NonDestructive and xml.Parser.Fastest settings. With these types
some xml entities such as `&amp;lt;` are not translated.
See [RapidXML](http://rapidxml.sourceforge.net/) for details.
--]]------------------------------------------------------
-- # Class methods
-- Parse a `string` containing xml content and return a table. Uses
-- xml.Parser with the xml.Parser.Default type.
function lib.load(string)
return parser:load(string)
end
-- Parse the XML content of the file at `path` and return a lua table. Uses
-- xml.Parser with the xml.Parser.Default type.
function lib.loadpath(path)
return parser:load(lub.content(path))
end
local function escape(v)
if type(v) == 'boolean' then
return v and 'true' or 'false'
else
return v:gsub('&','&amp;'):gsub('>','&gt;'):gsub('<','&lt;'):gsub("'",'&apos;')
end
end
local function tagWithAttributes(data)
local res = data.xml or 'table'
for k,v in pairs(data) do
if k ~= 'xml' and type(k) == 'string' then
res = res .. ' ' .. k .. "='" .. escape(v) .. "'"
end
end
return res
end
local function doDump(data, indent, output, last, depth, max_depth)
if depth > max_depth then
error(string.format("Could not dump table to XML. Maximal depth of %i reached.", max_depth))
end
if data[1] then
insert(output, (last == 'n' and indent or '')..'<'..tagWithAttributes(data)..'>')
last = 'n'
local ind = indent..' '
for _, child in ipairs(data) do
local typ = type(child)
if typ == 'table' then
doDump(child, ind, output, last, depth + 1, max_depth)
last = 'n'
elseif typ == 'number' then
insert(output, tostring(child))
else
local s = escape(child)
insert(output, s)
last = 's'
end
end
insert(output, (last == 'n' and indent or '')..'</'..(data.xml or 'table')..'>')
last = 'n'
else
-- no children
insert(output, (last == 'n' and indent or '')..'<'..tagWithAttributes(data)..'/>')
last = 'n'
end
end
-- Dump a lua table in the format described above and return an XML string. The
-- `max_depth` parameter is used to avoid infinite recursion in case a table
-- references one of its ancestors.
--
-- Default maximal depth is 3000.
function lib.dump(table, max_depth)
local max_depth = max_depth or 3000
local res = {}
doDump(table, '\n', res, 's', 1, max_depth)
return lub.join(res, '')
end
local function doRemoveNamespace(data, prefix)
data.xml = match(data.xml, prefix .. ':(.*)') or data.xml
for _, sub in ipairs(data) do
if type(sub) == 'table' then
doRemoveNamespace(sub, prefix)
end
end
end
-- This function finds the `xmlns:NAME='KEY'` declaration and removes `NAME:` from
-- the tag names.
--
-- Example:
--
-- local data = xml.load [[
-- <foo:document xmlns:foo='bar'>
-- <foo:name>Blah</foo:name>
-- </foo:document>
-- ]]
--
-- xml.removeNamespace(data, 'bar')
--
-- -- Result
-- {xml = 'document', ['xmlns:foo'] = 'bar',
-- {xml = 'name', 'Blah'},
-- }
function lib.removeNamespace(data, key)
local nm
for k, v in pairs(data) do
if v == key then
nm = match(k, 'xmlns:(.*)')
if nm == '' then
-- error
return
else
doRemoveNamespace(data, nm)
end
end
end
end
-- Recursively find the first table with a tag equal to `tag`. This
-- search uses [lub.search](lub.html#search) to do an [Iterative deepening depth-first search](https://en.wikipedia.org/wiki/Iterative_deepening_depth-first_search)
-- because we usually search for elements close to the surface.
--
-- For more options, use [lub.search](lub.html#search) directly with a custom
-- function.
--
-- You can also pass an attribute key and attribute value to further filter the
-- searched node. This gives this function the same arguments as LuaXML's find
-- function.
--
-- Usage examples:
--
-- local sect = xml.find(elem, 'simplesect', 'kind', 'section')
--
-- print(xml.find(sect, 'title'))
function lib.find(data, tag, attr_key, attr_value)
if attr_key then
return lub.search(data, function(node)
if node.xml == tag and node[attr_key] == attr_value then
return node
end
end)
else
return lub.search(data, function(node)
if node.xml == tag then
return node
end
end)
end
end
-- # Classes
return lib