markdown的图片语法为

1
{% image src title %}

pandoc 3.x，将markdown导出为docx时，会把图片的caption默认设置为alt文本

见issue

但是目前的笔记软件，比如思源笔记、语雀是用markdown图片语法中的title作为图片的标题的

所以我需要尝试修改pandoc的默认行为，使得docx图片标题与笔记软件里的预览效果一样

参考这个issue提供的代码进行改进，Lua filter on image captions does not work. · Issue #8974 · jgm/pandoc

改进的代码

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
-- image-title-to-caption.lua
-- A Pandoc Lua filter to set image title as caption
-- Format: ![alt](src "title")

-- Function to create a new caption from text
function create_caption(text)
    -- For simple captions, we can just use a Str element
    if text:find("^%s*$") then
        -- Empty caption
        return {}
    else
        -- Non-empty caption
        return {pandoc.Str(text)}
    end
end

-- Function to process the image and set title as caption
function process_image(img)

    -- Check if title exists and is not empty
    if img.title and img.title ~= "" then
        
        -- Set the title as the image caption
        img.caption = create_caption(img.title)
        
        -- Clear the title to avoid duplication
        img.title = ""
        
        -- Return the modified image
        return img
    end
    
    -- If no title, return the image unchanged
    return img
end

-- Handler for standalone images
function Image(img)
    return process_image(img)
end

-- Handler for Figure blocks (which may contain images)
function Figure(fig)
    -- Check if the figure has an image
    for i, block in ipairs(fig.content) do
        if block.t == "Plain" then
            for j, inline in ipairs(block.content) do
                if inline.t == "Image" then
                    -- Process the image
                    local processed_image = process_image(inline)
                    -- Update the image in the figure
                    block.content[j] = processed_image
                    
                    -- Also update the figure caption
                    if processed_image.caption and #processed_image.caption > 0 then
                        -- Extract the caption text
                        local caption_text = pandoc.utils.stringify(processed_image.caption)
                        -- Update the figure caption
                        fig.caption = create_caption(caption_text)
                    end
                end
            end
        end
    end
    
    return fig
end

-- Return the filter
return {{Image = Image}, {Figure = Figure}}

给思源笔记的文献引用插件也贡献了这个代码：导出图片标题使用title而不是alt文本 by Achuan-2 · Pull Request #104 · WingDr/siyuan-plugin-citation

笔记

  • Lua filter on image captions does not work. · Issue #8974 · jgm/pandoc提供的根据![Caption text|width](image.png)中的alt文本修改图片大小修改caption的代码

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    42
    43
    44
    45
    46
    47
    48
    49
    50
    51
    52
    53
    54
    55
    56
    57
    58
    59
    60
    61
    62
    63
    64
    65
    66
    67
    68
    69
    70
    71
    72
    73
    74
    75
    76
    77
    78
    79
    80
    81
    82
    83
    84
    85
    86
    87
    88
    89
    90
    91
    92
    -- image-width.lua
    -- A Pandoc Lua filter to adjust image widths based on caption format
    -- Format: ![Caption text|width](image.png)
    -- Function to create a new caption from text
    function create_caption(text)
        -- For simple captions, we can just use a Str element
        if text:find("^%s*$") then
            -- Empty caption
            return {}
        else
            -- Non-empty caption
            return pandoc.Inlines(pandoc.Str(text))
        end
    end

    -- Function to process the image and extract width from caption
    function process_image(img)

        -- Convert the caption to a single string
        local caption_text = pandoc.utils.stringify(img.caption)

        -- Check if the caption contains a pipe followed by a number
        -- Using a more flexible pattern that looks for a pipe character followed by digits
        local pipe_pos = caption_text:find("|")

        if pipe_pos then

            -- Extract the parts before and after the pipe
            local new_caption = caption_text:sub(1, pipe_pos - 1)
            local width_part = caption_text:sub(pipe_pos + 1)

            -- Extract the width number from the width part
            local width = width_part:match("(%d+)")

            if width then

                -- Update the image caption without the width part
                img.caption = create_caption(new_caption)

                -- Set the width attribute for the image (in pixels)
                img.attributes.width = width .. "px"

                -- Log the attributes
                for k, v in pairs(img.attributes) do end

                -- Return the modified image
                return img
            else
            end
        else
        end

        -- If no width specification was found, return the image unchanged
        return img
    end

    -- Handler for standalone images
    function Image(img) return process_image(img) end

    -- Handler for Figure blocks (which may contain images)
    function Figure(fig)

        -- Check if the figure has an image
        for i, block in ipairs(fig.content) do
            if block.t == "Plain" then
                for j, inline in ipairs(block.content) do
                    if inline.t == "Image" then
                        -- Process the image
                        local processed_image = process_image(inline)
                        -- Update the image in the figure
                        block.content[j] = processed_image

                        -- Also update the figure caption if needed
                        if processed_image.caption then
                            -- Extract the caption text
                            local caption_text =
                                pandoc.utils.stringify(processed_image.caption)
                            -- Update the figure caption
                            fig.caption = create_caption(caption_text)
                        end
                    end
                end
            end
        end

        return fig
    end

    -- Log when the filter is loaded

    -- Return the filter
    return {{Image = Image}, {Figure = Figure}}

  • pandoc 3新增了Figure对象，所以不能仅仅用img.caption来修改图片标题

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    function Image(img)
      -- 打印调试信息（可选，用于调试）
      -- print("Image found:", img.src, "Title:", img.title, "Caption:", pandoc.utils.stringify(img.caption))
      
      -- 检查是否存在 title 属性且不为空
      if img.title and img.title ~= "" then
        -- 创建新的 caption，使用 pandoc.Inlines 包装
        local new_caption = pandoc.Inlines({pandoc.Str(img.title)})
        
        -- 替换原有的 caption
        img.caption = new_caption
        
        -- 清空 title 以避免重复显示
        img.title = ""
      end
      
      return img
    end

  • 发现pandoc有时候没有生成图片标题，是因为图片前后没有空一行，像img1就没有caption

    1
    2
    3
    4
    5
      测试
    {% image preview.png title1 %}

      测试
    {% image preview.png title2 %}
许可协议

本文采用 署名-非商业性使用-相同方式共享 4.0 国际 许可协议，转载请注明出处。

快来参与讨论吧~