Dataview

Dataview is not required. Hypomnema does not try to be a dashboard: it writes plain text, properties and tasks that Dataview already knows how to read. The recipes below assume you have extracted comments at least once (Extract comments from this note, or right-click a folder → Extract comments from this folder).

What an extraction note contains

---
hypomnema-extract: true
hypomnema-source:
  - "[[Why we forget what we read]]"
hypomnema-date: 2026-10-31
tags:
  - idea
  - to-verify
kanban-plugin: board        # Kanban format only
---

## Idea

- [ ] Comment text — “quoted passage” [↩](obsidian://hypomnema?vault=…&file=…&id=a1)

The properties make the note easy to find with WHERE hypomnema-extract. The ↩ link reopens the source note and selects the commented passage. It also identifies the comment: when you extract again with Add only new comments, a comment whose link is already in the note is skipped, wherever you moved it.

Minimal recipe: list your extraction notes

```dataview
TABLE dateformat(hypomnema-date, "yyyy-MM-dd") AS "Extracted", length(hypomnema-source) AS "Notes", tags AS "Categories"
WHERE hypomnema-extract
SORT hypomnema-date DESC
```

Open items across every board

Kanban cards and list items are ordinary tasks, so Dataview can query them:

```dataview
TASK
WHERE hypomnema-extract AND !completed
GROUP BY file.link
```

Variant: every comment in the vault, without extracting

This DataviewJS query reads the %% annotations block of each note directly, using the storage format. It requires Enable JavaScript queries in Dataview’s settings.

```dataviewjs
const ENTRY = /^\[(a\d+)\]\s+(\d{4}-\d{2}-\d{2})(?: #([^\s#]+))?(?: (.*))?$/;
const rows = [];
for (const page of dv.pages()) {
  const file = app.vault.getAbstractFileByPath(page.file.path);
  const text = await app.vault.cachedRead(file);
  const block = text.match(/^%% annotations[ \t]*\n([\s\S]*?)^%%[ \t]*$/m);
  if (!block) continue;
  for (const line of block[1].split("\n")) {
    const m = ENTRY.exec(line);
    if (m && m[4]) rows.push([page.file.link, m[3] ? "#" + m[3] : "", m[2], m[4]]);
  }
}
rows.sort((a, b) => b[2].localeCompare(a[2]));
dv.table(["Note", "Category", "Date", "Comment"], rows);
```

Only the first line of each comment is shown, and replies are left out. A filter in the loop (for instance if (m[3] !== "to-verify") continue;) gives a “to verify” list.

Variant: through the API

Note

Available from version 0.5.0.

The public API reads threads for you, replies included, and filters by folder, category and date:

```dataviewjs
const api = app.plugins.plugins.hypomnema?.api;
if (!api) { dv.paragraph("Hypomnema is not enabled."); }
else {
  const since = moment().subtract(30, "days").format("YYYY-MM-DD");
  const items = await api.allThreads({ categories: ["to-verify"], since });
  dv.table(["Note", "Date", "Comment", "Replies"],
    items.map(({ file, thread }) =>
      [dv.fileLink(file.path), thread.root.date, thread.root.text, thread.replies.length]));
}
```

A thread holds the root comment (thread.root, with date, tag and text), its replies and the commented passage (thread.quote). See the API page.