Module:PageSummary

此模块的文档可以在Module:PageSummary/doc创建

local p = {}

-- 核心递归清理函数：通过平衡括号计数，彻底移除 [[File: ... ]] 或 {{模板}} 及其嵌套内容
local function stripBalancedTags(text, openTag, closeTag)
    if not text then return "" end
    local result = text
    
    while true do
        local startPos = string.find(result, openTag, 1, true)
        if not startPos then break end
        
        local count = 0
        local endPos = nil
        local len = mw.ustring.len(result)
        
        for j = startPos, len do
            local char2 = mw.ustring.sub(result, j, j + 1)
            if char2 == openTag then
                count = count + 1
            elseif char2 == closeTag then
                count = count - 1
                if count == 0 then
                    endPos = j + 1
                    break
                end
            end
        end
        
        if endPos then
            result = mw.ustring.sub(result, 1, startPos - 1) .. mw.ustring.sub(result, endPos + 1)
        else
            -- 容错：若未闭合则移除起始符防止死循环
            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（用于单独显示）
    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 500 字符
    local rawText = mw.ustring.sub(content, 1, 500)
    
    -- A. 剥离图片和模板（递归处理嵌套链接和长描述）
    rawText = stripBalancedTags(rawText, "[[", "]]") -- 移除所有链接（含图片），后文再截断
    -- 注意：因为上面已经移除了所有 [[ ]]，摘要里的链接会暂时丢失。
    -- 如果要保留链接，我们需要更复杂的正则。但为了“删干净”图片描述，
    -- 我们重新从 content 获取一个不含图片但保留链接的版本。
    
    -- 重新逻辑：仅针对图片前缀进行剥离，保留普通链接
    local function stripOnlyImages(text)
        local prefixes = {"[[File:", "[[Image:", "[[文件:", "[[图像:", "[[file:", "[[image:"}
        for _, pre in ipairs(prefixes) do
            while string.find(text, pre, 1, true) do
                local s = string.find(text, pre, 1, true)
                local count = 0
                local e = nil
                for j = s, mw.ustring.len(text) do
                    local c2 = mw.ustring.sub(text, 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 text = mw.ustring.sub(text, 1, s - 1) .. mw.ustring.sub(text, e + 1)
                else text = mw.ustring.sub(text, 1, s - 1) .. mw.ustring.sub(text, s + 2) end
            end
        end
        return text
    end
    
    local processedText = mw.ustring.sub(content, 1, 500)
    processedText = stripOnlyImages(processedText) -- 仅删图片，保留链接
    processedText = stripBalancedTags(processedText, "{{", "}}") -- 删模板

    -- B. 移除格式符号
    processedText = string.gsub(processedText, "\n==+.-==+", " ") -- 删标题
    processedText = string.gsub(processedText, "\n%s*[*#:]+", " ") -- 删列表符
    processedText = string.gsub(processedText, "<ref.-</ref>", "")
    processedText = string.gsub(processedText, "<ref.->", "")
    processedText = string.gsub(processedText, "<!%-%-.-%-%->", "")
    processedText = string.gsub(processedText, "&[Nn][Bb][Ss][Pp];", " ")
    processedText = string.gsub(processedText, "\n", " ")
    processedText = string.gsub(processedText, "%s+", " ")
    
    local cleanText = mw.text.trim(processedText)

    -- 3. 截断逻辑：150 字并处理链接补全
    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
        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