Documentation Source Text

the page type.  For the following computations, let U be the usable size
of a database page, the total page size less the reserved space at the
end of each page.  And let P be the payload size.</p>

<blockquote><dl>
<dt>Table B-Tree Leaf Cell:</dt>
<dd><p>
^If the payload size P is less than or equal to U-35 then
the entire payload is stored on the b-tree leaf page.  
^(Let M be ((U-12)*32/255)-23.  If P is greater than U-35
then the number of byte stored on the b-tree leaf page is the smaller of
M+((P-M)%(U-4)) and U-35.)^
^(Note that number of bytes stored on the leaf page is never less than M.)^
</p></dd>

<dt>Table B-Tree Interior Cell:</dt>
<dd><p>
Interior pages of table b-trees have no payload and so there is never
any payload to spill.
</p></dd>

<dt>Index B-Tree Leaf Or Interior Cell:</dt>
<dd><p>
^(Let X be ((U-12)*64/255)-23).  If the payload size P is less than
or equal to X then the entire payload is stored on the b-tree page.)^
^(Let M be ((U-12)*32/255)-23.  If P is greater than X then the number
of byte stored on the b-tree page is the smaller of
M+((P-M)%(U-4)) and X.)^
^(Note that number of bytes stored on the index page is never less than M.)^
</p></dd>
</dl></blockquote>

<p>The overflow thresholds are designed to give a minimum fanout of
4 for index b-trees and to make sure enough of the payload
is on the b-tree page that the record header can usually be accessed
without consulting an overflow page.  In hindsight, the designers of
the SQLite b-tree logic realize that these thresholds could have been
made much simpler.  However, the computations cannot be changed
without resulting in an incompatible file format.  And the current computations
work well, even if they are a little complex.</p>

<h3>1.6 Cell Payload Overflow Pages</h3>

<p>^When the payload of a b-tree cell is too large for the b-tree page,
the surplus is spilled onto overflow pages.  ^Overflow pages form a linked