Build Logseq Datalog Query

Generate an optimized Datalog query for Logseq DB graphs based on the user's description.

Input: $ARGUMENTS

Query Building Process

1. Parse the request - Understand what data the user wants to find
2. Identify entities - Pages, blocks, tasks, or custom classes
3. Determine filters - Properties, tags, dates, relationships
4. Choose output format - Pull syntax, aggregations, or simple values
5. Optimize - Put selective clauses first, minimize wildcards

Query Components

Find Clause

• (pull ?e [*]) - All attributes
• (pull ?e [:block/title :user.property/rating]) - Specific attributes
• ?title ?author - Multiple values
• (count ?e) - Aggregations

Where Clause Patterns

• [?b :block/tags ?t] - Tag/class membership
• [?t :block/title "Book"] - String match
• [?b :user.property/rating ?r] - Property access
• [(>= ?r 4)] - Comparisons
• (not [...]) - Negation
• (or [...] [...]) - Disjunction

Output Format

Provide:

1. The Datalog query - Ready to use
2. Explanation - What each clause does
3. Usage - How to run it (query block or API)
4. Variations - Alternative versions if applicable

Example Transformations

Natural Language	Query Pattern
"Find all books"	[?b :block/tags ?t] [?t :block/title "Book"]
"Books rated 5 stars"	Add [?b :user.property/rating 5]
"Unread books"	Add [?b :user.property/status "To Read"]
"Books by Stephen King"	Add [?b :user.property/author "Stephen King"]
"Count books per author"	Use (count ?b) with grouping

Response Template

## Query: [Description]

### Datalog Query
\`\`\`clojure
[:find (pull ?b [:block/title ...])
 :where
 ...]
\`\`\`

### Explanation
- Line 1: ...
- Line 2: ...

### Usage

**In Query Block:**
\`\`\`
#+BEGIN_QUERY
{:title "..."
 :query [:find ...]}
#+END_QUERY
\`\`\`

**In Plugin (JavaScript):**
\`\`\`javascript
const results = await logseq.DB.datascriptQuery(\`...\`)
\`\`\`

### Variations
- To also include X: add `[...]`
- To exclude Y: add `(not [...])`

Generate the query now based on: $ARGUMENTS