luasocket/etc/check-links.lua

88 lines
2.6 KiB
Lua
Raw Normal View History

2003-03-28 22:08:50 +01:00
-----------------------------------------------------------------------------
-- Little program that checks links in HTML files
-- LuaSocket sample files
-- Author: Diego Nehab
-- RCS ID: $Id$
2003-03-28 22:08:50 +01:00
-----------------------------------------------------------------------------
local http = require("socket.http")
local url = require("socket.url")
2004-06-04 17:15:45 +02:00
http.TIMEOUT = 10
2001-09-27 22:02:58 +02:00
function readfile(path)
2004-06-04 17:15:45 +02:00
path = url.unescape(path)
2004-03-26 01:18:41 +01:00
local file, error = io.open(path, "r")
2001-09-27 22:02:58 +02:00
if file then
2004-03-26 01:18:41 +01:00
local body = file:read("*a")
file:close()
2001-09-27 22:02:58 +02:00
return body
else return nil, error end
end
2004-06-04 17:15:45 +02:00
function getstatus(u)
local parsed = url.parse(u, {scheme = "file"})
2001-09-27 22:02:58 +02:00
if parsed.scheme == "http" then
2004-06-17 00:51:04 +02:00
local r, c, h, s = http.request{url = u, method = "HEAD"}
if c ~= 200 then return s or c end
2001-09-27 22:02:58 +02:00
elseif parsed.scheme == "file" then
2004-06-04 17:15:45 +02:00
local file, error = io.open(url.unescape(parsed.path), "r")
2004-06-17 00:51:04 +02:00
if file then file:close()
else return error end
else return string.format("unhandled scheme '%s'", parsed.scheme) end
2001-09-27 22:02:58 +02:00
end
2004-06-04 17:15:45 +02:00
function retrieve(u)
local parsed = url.parse(u, { scheme = "file" })
2004-03-26 01:18:41 +01:00
local body, headers, code, error
2004-06-04 17:15:45 +02:00
local base = u
2001-09-27 22:02:58 +02:00
if parsed.scheme == "http" then
2004-06-17 00:51:04 +02:00
body, code, headers = http.request(u)
2004-03-26 01:18:41 +01:00
if code == 200 then
base = base or headers.location
2001-09-27 22:02:58 +02:00
end
2004-06-17 00:51:04 +02:00
if not body then
error = code
end
2001-09-27 22:02:58 +02:00
elseif parsed.scheme == "file" then
body, error = readfile(parsed.path)
else error = string.format("unhandled scheme '%s'", parsed.scheme) end
2001-09-27 22:02:58 +02:00
return base, body, error
end
function getlinks(body, base)
-- get rid of comments
body = string.gsub(body, "%<%!%-%-.-%-%-%>", "")
2001-09-27 22:02:58 +02:00
local links = {}
-- extract links
2004-03-26 01:18:41 +01:00
body = string.gsub(body, '[Hh][Rr][Ee][Ff]%s*=%s*"([^"]*)"', function(href)
2004-06-04 17:15:45 +02:00
table.insert(links, url.absolute(base, href))
2001-09-27 22:02:58 +02:00
end)
2004-03-26 01:18:41 +01:00
body = string.gsub(body, "[Hh][Rr][Ee][Ff]%s*=%s*'([^']*)'", function(href)
2004-06-04 17:15:45 +02:00
table.insert(links, url.absolute(base, href))
2001-09-27 22:02:58 +02:00
end)
2004-03-26 01:18:41 +01:00
string.gsub(body, "[Hh][Rr][Ee][Ff]%s*=%s*(.-)>", function(href)
2004-06-04 17:15:45 +02:00
table.insert(links, url.absolute(base, href))
2001-09-27 22:02:58 +02:00
end)
return links
end
2004-06-04 17:15:45 +02:00
function checklinks(u)
local base, body, error = retrieve(u)
2001-09-27 22:02:58 +02:00
if not body then print(error) return end
local links = getlinks(body, base)
for _, l in ipairs(links) do
io.stderr:write("\t", l, "\n")
local err = getstatus(l)
if err then io.stderr:write('\t', l, ": ", err, "\n") end
2001-09-27 22:02:58 +02:00
end
end
arg = arg or {}
if table.getn(arg) < 1 then
print("Usage:\n luasocket check-links.lua {<url>}")
2001-09-27 22:02:58 +02:00
exit()
end
for _, a in ipairs(arg) do
print("Checking ", a)
2004-06-04 17:15:45 +02:00
checklinks(url.absolute("file:", a))
2001-09-27 22:02:58 +02:00
end