From bf0f068938dd7100d739adb8ef461719fba73935 Mon Sep 17 00:00:00 2001
From: Rafael Silva <xfalcox@gmail.com>
Date: Fri, 13 Dec 2024 14:14:13 -0300
Subject: [PATCH] Word Cloud report

---
 .../rewind/action/word_cloud.rb               | 73 +++++++++++++++++++
 1 file changed, 73 insertions(+)
 create mode 100644 app/services/discourse_rewind/rewind/action/word_cloud.rb

diff --git a/app/services/discourse_rewind/rewind/action/word_cloud.rb b/app/services/discourse_rewind/rewind/action/word_cloud.rb
new file mode 100644
index 0000000..6959575
--- /dev/null
+++ b/app/services/discourse_rewind/rewind/action/word_cloud.rb
@@ -0,0 +1,73 @@
+# frozen_string_literal: true
+
+# User Word Cloud
+module DiscourseRewind
+  class Rewind::Action::WordCloud < Action::Base
+    def call
+      words = DB.query(<<~SQL, user_id: user.id, date_start: date.first, date_end: date.last)
+        WITH popular_words AS (
+          SELECT
+            *
+          FROM
+            ts_stat(
+              $INNERSQL$
+                SELECT
+                  search_data
+                FROM
+                  post_search_data
+                INNER JOIN
+                  posts ON posts.id = post_search_data.post_id
+                WHERE
+                  posts.user_id = :user_id
+                  AND posts.created_at BETWEEN :date_start AND :date_end
+              $INNERSQL$
+            ) AS search_data
+          ORDER BY
+            nentry DESC,
+            ndoc DESC,
+            word
+          LIMIT
+            100
+        ), lex AS (
+          SELECT
+            DISTINCT ON (lexeme) to_tsvector('english', word) as lexeme,
+            word as original_word
+          FROM
+            ts_stat ($INNERSQL$
+              SELECT
+                to_tsvector('simple', raw)
+              FROM
+                posts AS p
+              WHERE
+                p.created_at BETWEEN :date_start AND :date_end
+                AND p.user_id = :user_id
+            $INNERSQL$)
+        ), ranked_words AS (
+          SELECT
+            popular_words.*, lex.original_word,
+            ROW_NUMBER() OVER (PARTITION BY word ORDER BY LENGTH(original_word)) AS rn
+          FROM
+            popular_words
+          INNER JOIN
+            lex ON lex.lexeme @@ to_tsquery('english', popular_words.word)
+        )
+        SELECT
+          word,
+          ndoc,
+          nentry,
+          original_word
+        FROM
+          ranked_words
+        WHERE
+          rn = 1
+        ORDER BY
+          ndoc + nentry DESC
+        LIMIT 100
+      SQL
+
+      word_score = words.map { [_1.original_word, _1.ndoc + _1.nentry] }.to_h
+
+      { data: word_score, identifier: "word-cloud" }
+    end
+  end
+end