Module:MarkdownToWikitext

Documentation for this module may be created at Module:MarkdownToWikitext/doc

-- =================================================================
-- Module:Markdown
-- Description: Converts a string of Markdown text into wikitext.
-- Author: Gemini
-- License: CC BY-SA 4.0
-- =================================================================

local p = {}

-- Helper function to trim whitespace from a string
local function trim(s)
    return s:match('^%s*(.-)%s*$')
end

-- Helper function to escape characters that have special meaning in wikitext patterns.
-- This is crucial for safely inserting user-provided text (like URLs or link text)
-- into the replacement part of a gsub.
local function escapePattern(s)
    return s:gsub("([%(%)%.%%%+%-%*%?%[%^%$%]])", "%%%1")
end

-- =================================================================
-- Core Conversion Functions
-- =================================================================

-- Function to strip any existing HTML tags from the input to prevent
-- conflicts and potential security issues.
local function stripHtml(text)
    -- This removes anything that looks like an HTML tag.
    return text:gsub('<[^>]->', '')
end

-- Converts Markdown-style headers to wikitext headers.
-- e.g., ## Header 2 -> == Header 2 ==
local function convertHeaders(text)
    for i = 6, 1, -1 do
        local h_md = string.rep('#', i)
        local h_wiki = string.rep('=', i)
        -- Pattern: ^(#{i})%s+(.-)%s*$
        -- Captures the start of a line, the hash marks, and the header text.
        text = text:gsub('^' .. h_md .. '%s+(.-)%s*$', h_wiki .. ' %1 ' .. h_wiki)
    end
    return text
end

-- Converts bold and italic syntax.
-- Handles nested cases by doing bold first.
local function convertEmphasis(text)
    -- Bold: **text** or __text__ -> '''text'''
    text = text:gsub('%*%*([^\n]+)%*%*', "'''%1'''")
    text = text:gsub('__([^\n]+)__', "'''%1'''")
    -- Italic: *text* or _text_ -> ''text''
    text = text:gsub('%*([^\n]+)%*', "''%1''")
    text = text:gsub('_([^\n]+)_', "''%1''")
    return text
end

-- Converts Markdown links and images.
-- e.g., [text](url) -> [url text]
-- e.g., ![alt](url) -> [[File:url|alt]]
local function convertLinksAndImages(text)
    -- Images must be processed first, as they are a superset of the link syntax.
    -- Pattern: !%[(.-)%]%((.-)%)
    -- Captures alt text and URL for images.
    text = text:gsub('!%[([^\n%]]*)%]%((.-)%)', '[[File:%2|%1]]')

    -- Links
    -- Pattern: %[(.-)%]%((.-)%)
    -- Captures link text and URL.
    text = text:gsub('%[([^\n%]]+)%]%((.-)%)', '[%2 %1]')
    return text
end

-- Converts inline code blocks.
-- e.g., `code` -> <code>code</code>
local function convertInlineCode(text)
    -- Pattern: `(.-)`
    -- Lazily captures text between backticks.
    return text:gsub('`([^`\n]+)`', '<code>%1</code>')
end

-- Converts Markdown lists to wikitext lists.
-- Handles both ordered (1.) and unordered (*, -) lists.
local function convertLists(text)
    local lines = {}
    for line in text:gmatch("([^\n]*)\n?") do
        -- Unordered lists
        line = line:gsub('^%s*[%*%-]%s+', '* ')
        -- Ordered lists
        line = line:gsub('^%s*%d+%.%s+', '# ')
        table.insert(lines, line)
    end
    return table.concat(lines, '\n')
end

-- =================================================================
-- Main Public Function
-- =================================================================

---
-- The main function to be called from wikitext.
-- It takes a string of Markdown text and converts it to wikitext.
--
-- Usage in wikitext:
-- {{#invoke:Markdown|markdown|source=...}}
-- or when used with ExternalData:
-- {{#invoke:Markdown|markdown|1={{{1}}}}}
---
function p.markdown(frame)
    -- Get the input text. It can be passed as the first argument (e.g., from ExternalData)
    -- or from a named 'source' argument.
    local inputText = frame.args[1] or frame.args.source or ''

    -- 1. Initial sanitization: Strip any pre-existing HTML.
    local wikitext = stripHtml(inputText)

    -- 2. Block-level conversions (Headers, Lists)
    wikitext = convertHeaders(wikitext)
    wikitext = convertLists(wikitext)

    -- 3. Inline conversions (Emphasis, Links, Images, Code)
    -- These are processed line by line to avoid issues with multi-line matching.
    local lines = {}
    for line in wikitext:gmatch("([^\n]*)\n?") do
        line = convertEmphasis(line)
        line = convertLinksAndImages(line)
        line = convertInlineCode(line)
        table.insert(lines, line)
    end
    wikitext = table.concat(lines, '\n')

    -- 4. Handle paragraph breaks. In Markdown, two newlines create a paragraph.
    -- In wikitext, a blank line does the same. Here, we'll replace double
    -- newlines with a single one to preserve paragraph structure, and then
    -- we can use <br> for single newlines if needed, but standard wikitext
    -- paragraph handling should suffice. Let's ensure paragraphs are separated.
    wikitext = wikitext:gsub('\n\n+', '\n\n') -- Normalize multiple blank lines to one
    
    -- 5. Handle single line breaks within a paragraph. Markdown requires two spaces
    -- at the end of a line for a <br>. We will convert single newlines that are not
    -- part of a list or header into <br /> tags.
    local finalLines = {}
    local inList = false
    for line in wikitext:gmatch("([^\n]*)\n?") do
        local isListOrHeader = line:match('^[%*#=]')
        if #line > 0 and not isListOrHeader and not finalLines[#finalLines]:match('^$') and #finalLines > 0 then
             -- This is a line break within a paragraph, not a new paragraph.
             -- However, standard wikitext handles paragraphs separated by blank lines well.
             -- Let's stick to that and not add <br> tags unless necessary, as it can
             -- lead to messy formatting. The double newline replacement above should be sufficient.
        end
        table.insert(finalLines, line)
    end
    wikitext = table.concat(finalLines, '\n')


    return wikitext
end

return p