Module:PageSummary
武外梗百科 爱国好学自强图新的百科全书
更多操作
此模块的文档可以在Module:PageSummary/doc创建
local p = {}
-- 核心函数:通过计数平衡括号,彻底移除 [[File: ... ]] 及其嵌套内容
local function stripComplexTags(text)
if not text then return "" end
local result = text
local i = 1
while true do
-- 寻找可能的图片起始标签
local startPos = nil
local patterns = {"%[%[[Ff]ile:", "%[%[[Ii]mage:", "%[%[文件:", "%[%[图像:"}
local foundPat = nil
for _, pat in ipairs(patterns) do
local s = string.find(result, pat)
if s and (not startPos or s < startPos) then
startPos = s
foundPat = pat
end
end
if not startPos then break end -- 没有更多图片标签了
-- 从 startPos 开始寻找匹配的闭合括号
local count = 0
local endPos = nil
for j = startPos, mw.ustring.len(result) do
local char2 = mw.ustring.sub(result, j, j+1)
if char2 == "[[" then
count = count + 1
elseif char2 == "]]" then
count = count - 1
if count == 0 then
endPos = j + 1
break
end
end
end
if endPos then
-- 移除整段图片代码
result = mw.ustring.sub(result, 1, startPos - 1) .. mw.ustring.sub(result, endPos + 1)
else
-- 如果没找到闭合(语法错误),强制跳过这个起始点,防止死循环
result = mw.ustring.sub(result, 1, startPos - 1) .. mw.ustring.sub(result, startPos + 2)
end
end
return result
end
function p.getSummaryAndImage(frame)
local pageName = frame.args[1] or ""
local title = mw.title.new(pageName)
if not title or not title.exists then return "页面不存在" end
local content = title:getContent()
-- 1. 提取第一张图(在清理前提取)
local firstImage = string.match(content, "%[%[([Ff]ile:.-)[|%]]") or
string.match(content, "%[%[([Ii]mage:.-)[|%]]") or
string.match(content, "%[%[(文件:.-)[|%]]") or
string.match(content, "%[%[(图像:.-)[|%]]")
-- 2. 处理内容:预加载 1500 字符(处理长条目)
local rawText = mw.ustring.sub(content, 1, 1500)
-- A. 使用递归逻辑剥离图片
rawText = stripComplexTags(rawText)
-- B. 移除模板 {{...}} - 同样使用平衡计数逻辑处理嵌套
while true do
local s = string.find(rawText, "{{")
if not s then break end
local count = 0
local e = nil
for j = s, mw.ustring.len(rawText) do
local c2 = mw.ustring.sub(rawText, j, j+1)
if c2 == "{{" then count = count + 1
elseif c2 == "}}" then
count = count - 1
if count == 0 then e = j + 1 break end
end
end
if e then
rawText = mw.ustring.sub(rawText, 1, s - 1) .. mw.ustring.sub(rawText, e + 1)
else
rawText = mw.ustring.sub(rawText, 1, s - 1) .. mw.ustring.sub(rawText, s + 2)
end
end
-- C. 移除标题符号、列表符、引用
rawText = string.gsub(rawText, "\n==+.-==+", " ")
rawText = string.gsub(rawText, "\n%s*[*#:]+", " ")
rawText = string.gsub(rawText, "<ref.-</ref>", "")
rawText = string.gsub(rawText, "<ref.->", "")
rawText = string.gsub(rawText, "<!%-%-.-%-%->", "")
rawText = string.gsub(rawText, "&[Nn][Bb][Ss][Pp];", " ")
rawText = string.gsub(rawText, "\n", " ")
rawText = string.gsub(rawText, "%s+", " ")
local cleanText = mw.text.trim(rawText)
-- 3. 截断逻辑:150 字
local targetLen = 150
local summary = ""
if mw.ustring.len(cleanText) <= targetLen then
summary = cleanText
else
summary = mw.ustring.sub(cleanText, 1, targetLen)
local restText = mw.ustring.sub(cleanText, targetLen + 1)
-- 链接延展逻辑
local lastOpen = 0
local lastClose = 0
local tempPos = 0
while true do
local found = string.find(summary, "%[%[", tempPos + 1)
if not found then break end
lastOpen = found
tempPos = found
end
tempPos = 0
while true do
local found = string.find(summary, "%]%]", tempPos + 1)
if not found then break end
lastClose = found
tempPos = found
end
if lastOpen > lastClose then
local endOfLink = string.find(restText, "%]%]")
if endOfLink then
summary = summary .. string.sub(restText, 1, endOfLink + 1)
end
end
summary = summary .. "..."
end
-- 4. 渲染
local container = mw.html.create('div'):css({['margin']='15px 0', ['display']='flow-root'})
if firstImage then
container:wikitext('[[' .. firstImage .. '|150px|right|link=' .. pageName .. ']]')
end
container:tag('div')
:css({['font-size']='1.4em', ['font-weight']='bold', ['margin-bottom']='8px'})
:wikitext('[[' .. pageName .. ']]')
container:tag('div')
:css({['line-height']='1.6', ['color']='#202122'})
:wikitext(summary)
return tostring(container)
end
return p