是否有人用思源实现了生词本？

直接制卡看起来就是对一个句子挖空。

我希望能够记录单词，然后对释义制卡。

主要功能是通过 sql 查询挂件来实现的，相关的 sql 如下，因为我这个是韩语的生词本，所以针对韩语词尾变形作了优化，英语、其它语言的话，相应修改就行。

单词本

WITH RECURSIVE
  hanyi_blocks(id, parent_id, root_id, content, type, subtype, level) AS (
    SELECT id, parent_id, root_id, content, type, subtype, 0
    FROM blocks
    WHERE type = 'h' AND content = '汉义' AND subtype = 'h3'
  
    UNION ALL
  
    SELECT b.id, b.parent_id, b.root_id, b.content, b.type, b.subtype, hb.level + 1
    FROM blocks b
    JOIN hanyi_blocks hb ON b.parent_id = hb.id
    WHERE hb.level < 5  -- 限制递归深度，防止查询过大
  ),
  ciyuan_blocks(id, parent_id, root_id, content, type, subtype, level) AS (
    SELECT id, parent_id, root_id, content, type, subtype, 0
    FROM blocks
    WHERE type = 'h' AND content = '词源' AND subtype = 'h3'
  
    UNION ALL
  
    SELECT b.id, b.parent_id, b.root_id, b.content, b.type, b.subtype, hb.level + 1
    FROM blocks b
    JOIN ciyuan_blocks hb ON b.parent_id = hb.id
    WHERE hb.level < 5  -- 限制递归深度，防止查询过大
  )
SELECT
    '[' || b.content || '](siyuan://blocks/' || b.id || ')' AS __1____pre__单词,
    (
        SELECT GROUP_CONCAT(cb.content, '
')
        FROM hanyi_blocks cb
        WHERE cb.root_id = b.id AND cb.type = 'p'  -- 只选择段落类型的块
        ORDER BY cb.id
    ) AS __2____pre__汉义,
    (
        SELECT GROUP_CONCAT(cb.content, '
')
        FROM ciyuan_blocks cb
        WHERE cb.root_id = b.id AND cb.type = 'p'  -- 只选择段落类型的块
        ORDER BY cb.id
    ) AS __3____pre__词源,
    b.created AS __4____date__创建日期
FROM
    blocks AS b
WHERE
    b.type = 'd'
    AND b.content REGEXP '^[\x{AC00}-\x{D7AF}\s]+$'  -- 筛选文档标题为纯韩语字符（允许空格）
    AND EXISTS (
        SELECT 1
        FROM blocks hb
        WHERE hb.parent_id = b.id AND hb.type = 'h' AND hb.subtype = 'h3' AND hb.content = '汉义'
    )
ORDER BY
    b.created DESC
LIMIT 1000

例句

WITH RECURSIVE
current_document AS (
    SELECT 
        content AS title,
        CASE 
            WHEN content LIKE '%다' THEN substr(content, 1, length(content) - 1)
            ELSE content
        END AS root
    FROM blocks
    WHERE id = '.root{.id}'  -- 获取当前文档的ID
),
vowel_changes(from_vowel, to_vowel) AS (
    SELECT '이', '여' UNION ALL
    SELECT '으', '어' UNION ALL
    SELECT '아', '어' UNION ALL
    SELECT '오', '와' UNION ALL
    SELECT '우', '워' UNION ALL
    SELECT '이', '에' UNION ALL
    SELECT '지', '져' UNION ALL
    SELECT '치', '쳐' UNION ALL
    SELECT '비', '벼' UNION ALL
    SELECT '피', '펴' UNION ALL
    SELECT '기', '겨' UNION ALL
    SELECT '키', '켜' UNION ALL
    SELECT '시', '셔' UNION ALL
    SELECT '디', '뎌' UNION ALL
    SELECT '티', '텨' UNION ALL
    SELECT '히', '혀' UNION ALL
    SELECT '미', '며' UNION ALL
    SELECT '니', '녀' UNION ALL
    SELECT '리', '려'
),
possible_changes(original, changed, level) AS (
    SELECT root, root, 0
    FROM current_document
  
    UNION ALL
  
    SELECT 
        pc.original,
        substr(pc.changed, 1, instr(pc.changed, vc.from_vowel) - 1) || 
        vc.to_vowel || 
        substr(pc.changed, instr(pc.changed, vc.from_vowel) + length(vc.from_vowel)),
        pc.level + 1
    FROM possible_changes pc
    JOIN vowel_changes vc
    WHERE pc.changed LIKE '%' || vc.from_vowel || '%'
    AND pc.level < 2  -- 限制递归深度
)
SELECT DISTINCT
    '[' || replace(b.content, pc.changed, '==' || pc.changed || '==') || '](siyuan://blocks/' || b.id || ')' AS __1____pre__例句,
	b.hpath AS __2__文档路径,
    cd.title AS __3__搜索词,
    pc.changed AS __4__匹配形式
FROM
    blocks AS b,
    current_document cd,
    possible_changes pc
WHERE
    b.type = 'p'
    AND b.hpath LIKE '%韩语%'
    AND (
        b.content LIKE '%' || pc.changed || '%'
        OR b.content LIKE '%' || pc.changed || '다%'
        OR b.content LIKE '%' || pc.changed || '어%'
        OR b.content LIKE '%' || pc.changed || '아%'
        OR b.content LIKE '%' || pc.changed || '여%'
        OR b.content LIKE '%' || pc.changed || '야%'
    )
    AND b.root_id != '.root{.id}'  -- 排除当前文档
GROUP BY b.id
ORDER BY
    b.path
LIMIT 30

能够自己选文章，自己选单词然后记录

我印象中欧路词典以及一些背单词的 APP 就有这个功能，你可以上小红书搜搜看

我的灵感来自于集市里的单词本模板，在那个基础上进行了自己的修改，可以参考一下。
每篇文章中，自动汇总这篇文档的生词（原理是搜寻这篇文章的子文章）

每个生词就是一页，可以自动全局搜索文章中的例句：

有一个单词本的页面进行所有生词的汇总：