fix: 收口批次D并修复本地垃圾箱清空

2026-05-21 15:45:11 +08:00
parent e709c0705c
commit dc18ec0a60
14 changed files with 525 additions and 72 deletions
@@ -78,18 +78,76 @@ struct MarkdownInlineStyles {
 }

 pub fn parse_markdown_page(markdown: &str, file_name: &str) -> ParsedLocalMarkdownPage {
-    let (_, body) = split_frontmatter(markdown);
-    let title = file_stem_title(file_name);
+    let (frontmatter, body) = split_frontmatter(markdown);
+    let body_owned = body.to_string();
+    // 标题优先级：frontmatter title > 正文第一条 H1 > 文件名。
+    let title = parse_frontmatter_title(frontmatter.as_deref())
+        .or_else(|| find_first_h1(&body_owned))
+        .unwrap_or_else(|| file_stem_title(file_name));
    ParsedLocalMarkdownPage {
        title,
-        body: body.to_string(),
+        body: body_owned,
    }
 }

+/// 从 frontmatter 中提取简单 `title:` 字段。
+fn parse_frontmatter_title(frontmatter: Option<&str>) -> Option<String> {
+    let text = frontmatter?;
+    for line in text.lines() {
+        let trimmed = line.trim();
+        if let Some(value) = trimmed.strip_prefix("title:") {
+            let raw = value.trim();
+            if raw.is_empty() {
+                continue;
+            }
+            let unquoted = if (raw.starts_with('"') && raw.ends_with('"'))
+                || (raw.starts_with('\'') && raw.ends_with('\''))
+            {
+                &raw[1..raw.len() - 1]
+            } else {
+                raw
+            };
+            let title = unquoted.trim().to_string();
+            if !title.is_empty() {
+                return Some(title);
+            }
+        }
+    }
+    None
+}
+
+/// 通过 AST 找正文第一条 H1，避免把代码块里的 `# ...` 误判为标题。
+fn find_first_h1(body: &str) -> Option<String> {
+    let arena = Arena::new();
+    let root = parse_document(&arena, body, &markdown_options());
+    for node in root.descendants() {
+        let value = node.data.borrow();
+        if let NodeValue::Heading(heading) = &value.value {
+            if heading.level != 1 {
+                continue;
+            }
+            let title = collect_plain_text(node).trim().to_string();
+            return (!title.is_empty()).then_some(title);
+        }
+    }
+    None
+}
+
 pub fn markdown_to_blocks(markdown: &str) -> Value {
    markdown_ast_document_to_blocks(&parse_markdown_ast_document(markdown))
 }

+fn markdown_options() -> Options<'static> {
+    let mut options = Options::default();
+    options.extension.table = true;
+    options.extension.tasklist = true;
+    options.extension.strikethrough = true;
+    options.extension.autolink = true;
+    options.extension.front_matter_delimiter = Some("---".to_string());
+    options.parse.tasklist_in_table = true;
+    options
+}
+
 pub fn parse_markdown_attachment_link(trimmed: &str) -> Option<(String, String)> {
    let value = trimmed
        .strip_prefix('!')
@@ -130,14 +188,7 @@ pub fn parse_markdown_attachment_link(trimmed: &str) -> Option<(String, String)>

 fn parse_markdown_ast_document(markdown: &str) -> MarkdownAstDocument {
    let arena = Arena::new();
-    let mut options = Options::default();
-    options.extension.table = true;
-    options.extension.tasklist = true;
-    options.extension.strikethrough = true;
-    options.extension.autolink = true;
-    options.extension.front_matter_delimiter = Some("---".to_string());
-    options.parse.tasklist_in_table = true;
-
+    let options = markdown_options();
    let root = parse_document(&arena, markdown, &options);
    let mut blocks = Vec::new();
    for node in root.children() {
@@ -730,7 +781,7 @@ pub(crate) fn file_stem_title(file_name: &str) -> String {

 #[cfg(test)]
 mod tests {
-    use super::markdown_to_blocks;
+    use super::{markdown_to_blocks, parse_markdown_page};

    #[test]
    fn markdown_image_parses_as_image_block() {
@@ -781,10 +832,14 @@ mod tests {
            .as_array()
            .expect("table content rows");
        // 表头行：两个单元格
-        let header_row = &content_rows[0]["content"].as_array().expect("header row cells");
+        let header_row = &content_rows[0]["content"]
+            .as_array()
+            .expect("header row cells");
        assert_eq!(header_row.len(), 2);
        // 表头第一个单元格（非空）
-        let h1_cell = &header_row[0]["content"].as_array().expect("header cell paragraphs");
+        let h1_cell = &header_row[0]["content"]
+            .as_array()
+            .expect("header cell paragraphs");
        assert!(!h1_cell.is_empty(), "表头第一个单元格不应为空");
        // 表头第二个单元格（空）：单元格的 content 为 [{type:"paragraph", content:[]}]
        // 段落层的 content 应为空数组
@@ -797,7 +852,9 @@ mod tests {
        );

        // 数据行
-        let data_row = &content_rows[1]["content"].as_array().expect("data row cells");
+        let data_row = &content_rows[1]["content"]
+            .as_array()
+            .expect("data row cells");
        assert_eq!(data_row.len(), 2);
        // 数据行第一个单元格（空）：段落层的 content 应为空数组
        let d1_par_content = &data_row[0]["content"][0]["content"];
@@ -812,4 +869,69 @@ mod tests {
        let d2_inline = d2_par_content.as_array().expect("data cell para content");
        assert!(!d2_inline.is_empty(), "非空数据单元格不应为空");
    }
+
+    // 标题优先级：frontmatter title > H1 > 文件名。
+
+    #[test]
+    fn parse_markdown_title_uses_frontmatter_title() {
+        let parsed = parse_markdown_page(
+            "---\ntitle: Frontmatter Title\n---\n# H1 Heading\nbody\n",
+            "file.md",
+        );
+        assert_eq!(parsed.title, "Frontmatter Title");
+    }
+
+    #[test]
+    fn parse_markdown_title_uses_first_h1_when_no_frontmatter_title() {
+        let parsed = parse_markdown_page("# H1 Title\nbody\n## Not H1\n", "file.md");
+        assert_eq!(parsed.title, "H1 Title");
+    }
+
+    #[test]
+    fn parse_markdown_title_falls_back_to_filename() {
+        let parsed = parse_markdown_page("plain text\nmore text\n", "My File.md");
+        assert_eq!(parsed.title, "My File");
+    }
+
+    #[test]
+    fn parse_markdown_title_prefers_frontmatter_over_h1() {
+        let parsed = parse_markdown_page("---\ntitle: FM Title\n---\n# H1 here\n", "file.md");
+        assert_eq!(parsed.title, "FM Title");
+    }
+
+    #[test]
+    fn parse_markdown_title_handles_quoted_double() {
+        let parsed = parse_markdown_page("---\ntitle: \"Quoted Double\"\n---\nbody\n", "file.md");
+        assert_eq!(parsed.title, "Quoted Double");
+    }
+
+    #[test]
+    fn parse_markdown_title_handles_quoted_single() {
+        let parsed = parse_markdown_page("---\ntitle: 'Single Quoted'\n---\nbody\n", "file.md");
+        assert_eq!(parsed.title, "Single Quoted");
+    }
+
+    #[test]
+    fn parse_markdown_title_uses_h1_when_frontmatter_has_no_title() {
+        let parsed = parse_markdown_page("---\nother: value\n---\n# From H1\n", "file.md");
+        assert_eq!(parsed.title, "From H1");
+    }
+
+    #[test]
+    fn parse_markdown_title_uses_filename_when_body_has_no_h1() {
+        let parsed = parse_markdown_page("---\ntags: foo\n---\nplain text\n", "NoH1.md");
+        assert_eq!(parsed.title, "NoH1");
+    }
+
+    #[test]
+    fn parse_markdown_title_handles_atx_closing_markers() {
+        let parsed = parse_markdown_page("# Heading With Closing #\n", "file.md");
+        assert_eq!(parsed.title, "Heading With Closing");
+    }
+
+    #[test]
+    fn parse_markdown_title_ignores_hash_inside_code_block() {
+        let parsed = parse_markdown_page("```md\n# Not A Title\n```\nplain text\n", "Fallback.md");
+        assert_eq!(parsed.title, "Fallback");
+    }
 }