Patterns

Author: pchemguy

SM ASCII IDs.md renamed to Patterns ASCII IDs.md

@@ -1,9 +1,9 @@
 ---
 layout: default
 title: ASCII IDs
-nav_order: 4
-parent: String Manipulation
-permalink: /strings/ascii-id
+nav_order: 7
+parent: Design Patterns
+permalink: /patterns/ascii-id
 ---
 
 The following query generates random eight-character-long ASCII IDs (*id_counts*.*id_counter* pieces) using the specified alphabet (*char_templates*). The dash and underscore characters extend the alphanumeric alphabet to 64 characters allowing the selection of the six least significant bits from a random 64-bit number via bitwise AND (instead of *mod*). The *ids* block generates a 64-bit integer representation of the same ID. Saving character code before grouping in *ascii_ids* would not work because the query optimizer tends to rerun the query, executing the *random* function and generating the character and character code independently. The intermediate table aggregation operation prevents this optimization.
@@ -16,20 +16,20 @@ WITH
     ascii_ids AS (
         SELECT group_concat(substr(char_template, (random() & 63) + 1, 1), '') AS ascii_id, "key"/8 AS counter
         FROM char_templates, json_templates, json_each(json_templates.json_template) AS terms
-		GROUP BY counter
-	),
-	ids AS (
-		SELECT counter, ascii_id,
-			   (unicode(substr(ascii_id, 1, 1)) << 8*7) +
-			   (unicode(substr(ascii_id, 2, 1)) << 8*6) +
-			   (unicode(substr(ascii_id, 3, 1)) << 8*5) +
-			   (unicode(substr(ascii_id, 4, 1)) << 8*4) +
-			   (unicode(substr(ascii_id, 5, 1)) << 8*3) +
-			   (unicode(substr(ascii_id, 6, 1)) << 8*2) +
-			   (unicode(substr(ascii_id, 7, 1)) << 8*1) +
-			   (unicode(substr(ascii_id, 8, 1)) << 8*0) AS bin_id
-		FROM ascii_ids
-	)
+        GROUP BY counter
+    ),
+    ids AS (
+        SELECT counter, ascii_id,
+               (unicode(substr(ascii_id, 1, 1)) << 8*7) +
+               (unicode(substr(ascii_id, 2, 1)) << 8*6) +
+               (unicode(substr(ascii_id, 3, 1)) << 8*5) +
+               (unicode(substr(ascii_id, 4, 1)) << 8*4) +
+               (unicode(substr(ascii_id, 5, 1)) << 8*3) +
+               (unicode(substr(ascii_id, 6, 1)) << 8*2) +
+               (unicode(substr(ascii_id, 7, 1)) << 8*1) +
+               (unicode(substr(ascii_id, 8, 1)) << 8*0) AS bin_id
+        FROM ascii_ids
+    )
 SELECT * FROM ids;
 ~~~
 

SM Fixed-Width Values.md renamed to Patterns Fixed-Width Values.md

@@ -1,9 +1,9 @@
 ---
 layout: default
 title: Fixed-Width Values
-nav_order: 3
-parent: String Manipulation
-permalink: /strings/split-fwv
+nav_order: 6
+parent: Design Patterns
+permalink: /patterns/split-fwv
 ---
 
 ### Context
@@ -18,7 +18,7 @@ where the key term is node ASCII ID, and the value term is node path/prefix, not
 
 ### Splitting fixed-width value strings
 
-Like in the case of [DSV][] strings, the following query iscontext-specific, but the notes below should help adapt it to other needs.
+The following query is context-specific, but the notes below should help adapt it to other needs.
 
 <a name="FWV-Query"></a>
 ~~~sql