Documentation Source Text

Check-in [ce71f82b22]
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Fix minor typo. Forum post 1af006ebaf.
Downloads: Tarball | ZIP archive
Timelines: family | ancestors | descendants | both | branch-3.39
Files: files | file ages | folders
SHA3-256: ce71f82b22326a8c11a0460d08899cc33e43e6c0d44d1141436d0de355ae923c
User & Date: drh 2022-10-26 20:02:18.017
Context
2022-11-01
22:57
Fix a typo in the vtab documentation. (check-in: a7a20e71da user: drh tags: branch-3.39)
2022-10-26
20:02
Fix minor typo. Forum post 1af006ebaf. (check-in: ce71f82b22 user: drh tags: branch-3.39)
06:01
anti-robot JS code: handle case that HTML element #mtimelink is missing (triggered on download.html). (check-in: 5f822ec052 user: stephan tags: trunk)
Changes
Unified Diff Ignore Whitespace Patch
Changes to pages/lang_select.in.
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
... FROM t1, t2 NATURAL FULL JOIN t3 ...
</pre></blockquote>

<p>In standard SQL, the FULL JOIN between t2 and t3 would occur first, and
then the result of the left join would be cross-joined against t1.  But SQLite
always handles all joins from left to right.  Thus, SQLite will do a cross
join on t1 and t2 first, then the result of that cross join will feed into the FULL JOIN
with t3.  Inner joins are inheriently associative, so the difference is only
evident if your FROM clause contains one or more outer joins.

<p>You can work around this, and make your SQL statements portable across all
systems, by observing the following stylistic rules:

<ul>
<li><p> Do not mix comma-joins with the JOIN keyword.  It is fine to use comma-joins,







|







674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
... FROM t1, t2 NATURAL FULL JOIN t3 ...
</pre></blockquote>

<p>In standard SQL, the FULL JOIN between t2 and t3 would occur first, and
then the result of the left join would be cross-joined against t1.  But SQLite
always handles all joins from left to right.  Thus, SQLite will do a cross
join on t1 and t2 first, then the result of that cross join will feed into the FULL JOIN
with t3.  Inner joins are inherently associative, so the difference is only
evident if your FROM clause contains one or more outer joins.

<p>You can work around this, and make your SQL statements portable across all
systems, by observing the following stylistic rules:

<ul>
<li><p> Do not mix comma-joins with the JOIN keyword.  It is fine to use comma-joins,