User:Havard

2008-09-15T20:16:58Z

Havard:

John Havard. I wrote the crappy blog software known as [[Havard's Crappy Blog]].

Havard's Crappy Blog

2008-09-15T20:13:53Z

Havard:

Havard's Crappy Blog is an aptly-named piece of software written by [[User:Havard|John Havard]] to power one of his web sites. It's a simple blog-style web app that runs under our favorite web application server. It was initially a couple of <tt>.adp</tt> scripts, but with a little prodding from [[User:Dossy|Dossy]], it was whipped into shape to be insanely fast. Since most blogs are relatively small, and RAM is cheap, everything is cached.

All the code is kept in the [http://aolserver.cvs.sourceforge.net/aolserver/ AOLserver CVS repository] as the <tt>havardblog</tt> package. Be warned that the code is likely full of WTF-isms. It's bad enough that I considered contributing it anonymously to avoid tarnishing my family's good name.

Havard's Crappy Blog

2008-09-15T20:12:51Z

Havard:

User:Havard

2008-09-15T20:07:38Z

Havard: New page: I wrote the crappy blog software known as Havard's Crappy Blog.

I wrote the crappy blog software known as [[Havard's Crappy Blog]].

Havard's Crappy Blog

2008-09-09T17:38:59Z

Havard: New page: I whipped up a quick bloggish app in php, but decided to port it over to AOLserver for good measure. It's not very complete, has little to no sanity checking, was written in a little unde...

I whipped up a quick bloggish app in php, but decided to port it over to AOLserver for good measure. It's not very complete, has little to no sanity checking, was written in a little under an hour without any thought or planning, and is barely useful, but it does work.

First thing's first, you'll need to create a pool for your PostgreSQL database. Let's assume for this project the name of the pool is <tt>pool1</tt>.

== schema.pgsql ==

<code>
CREATE DOMAIN AUTHOR_ID AS VARCHAR(16) NOT NULL;
CREATE DOMAIN EMAIL AS VARCHAR(128) NOT NULL;
CREATE DOMAIN SLUG AS VARCHAR(32) NOT NULL;

CREATE TABLE AUTHOR (
ID AUTHOR_ID PRIMARY KEY,
PERSONAL_NAME VARCHAR(64) NOT NULL,
DISPLAY_EMAIL EMAIL,
REAL_EMAIL EMAIL,
PASSWORD VARCHAR(64) NOT NULL
);

CREATE TABLE ARTICLE (
SLUG SLUG NOT NULL,
PUBDATE DATE NOT NULL,
UPDATED TIMESTAMP NOT NULL,
AUTHOR AUTHOR_ID REFERENCES AUTHOR(ID) NOT NULL,
LEAD TEXT NOT NULL,
BODY TEXT NOT NULL,
TITLE VARCHAR(128) NOT NULL,
PRIMARY KEY (SLUG, PUBDATE)
);
</code>

== init.tcl ==

<code>
proc do_article {} {
set mainStuff ""
# /article/YYYY/MM/DD/slug/junk
set urlv [ns_conn urlv]
set qslug [lindex $urlv 4]
set qpubdate [lindex $urlv 1]-[lindex $urlv 2]-[lindex $urlv 3]
set qry "SELECT ARTICLE.*,AUTHOR.PERSONAL_NAME FROM ARTICLE,AUTHOR WHERE AUTHOR.ID=ARTICLE.AUTHOR AND PUBDATE = '${qpubdate}' AND SLUG = '${qslug}'"
catch {
set hdl [ns_db gethandle pool1]
set res [ns_db 1row $hdl $qry]

#[ns_db getrow $hdl $res]
set slug [subst [ns_set get $res slug ]]
set pubdate [subst [ns_set get $res pubdate]]
set updated [subst [ns_set get $res updated]]
set author [subst [ns_set get $res author ]]
set title [subst [ns_set get $res title ]]
set lead [subst [ns_set get $res lead ]]
set body [subst [ns_set get $res body ]]
ns_db releasehandle $hdl
}

if { [string compare $slug $qslug] == 0 } {
set mainStuff [ns_adp_parse -file template/article.adp]
set rPole [ns_adp_parse -file template/rpole.adp]
ns_return 200 "text/html" [ns_adp_parse -file template/template.adp]
} else {
ns_returnfile 404 "text/html" "global/file-not-found.html"
}
}

ns_register_proc GET /article do_article

ns_log warning "We processed init.tcl!"
</code>

== index.adp ==
<%
set mainStuff ""
set qry "SELECT ARTICLE.*,AUTHOR.PERSONAL_NAME FROM ARTICLE,AUTHOR WHERE AUTHOR.ID=ARTICLE.AUTHOR ORDER BY PUBDATE DESC, UPDATED DESC LIMIT 10"
catch {
set hdl [ns_db gethandle pool1]
set res [ns_db select $hdl $qry]

while {[ns_db getrow $hdl $res]} {
set slug [subst [ns_set get $res slug ]]
set pubdate [subst [ns_set get $res pubdate]]
set updated [subst [ns_set get $res updated]]
set author [subst [ns_set get $res author ]]
set title [subst [ns_set get $res title ]]
set lead [subst [ns_set get $res lead ]]
set body [subst [ns_set get $res body ]]
set pubdate [string map {- /} $pubdate]
set mainStuff "$mainStuff [ns_adp_parse -file template/entry.adp]"
}
}
ns_adp_puts [ns_adp_parse -file template/template.adp]
%>

== /admin/post.adp ==
<code>
<html>
<head>
<title>Post an article</title>
</head>
<body>
<table>
<form method="POST" action="do_post.adp">
<tr>
<td>Title</td>
<td><input name="title" size="80" value=""/></td>
</tr>
<tr>
<td>username</td>
<td><input name="username" value=""/></td>
</tr>
<tr>
<td>password</td>
<td><input type="password" name="password" value=""/></td>
</tr>
<tr>
<td>date (YYYY/MM/DD)</td>
<td><input name="date" value=""/></td>
</tr>
<tr>
<td>slug</td>
<td><input name="slug" value=""/></td>
</tr>
<tr>
<td>lead</td>
<td><textarea name="lead" cols="80" rows="5"></textarea>
</tr>
<tr>
<td>body</td>
<td><textarea name="body" cols="80" rows="12"></textarea>
</tr>
<tr>
<td><input type="submit"></td>
<td><input type="reset"></td>
</tr>
</form>
</table>
</body>
</html>
</code>

== /admin/do_post.adp ==
<code>
<%
set title [ns_dbquotevalue [ns_queryget title]]
set user [ns_dbquotevalue [ns_queryget username]]
set pass [ns_dbquotevalue [ns_queryget password]]
set date [ns_dbquotevalue [ns_queryget date]]
set slug [ns_dbquotevalue [ns_queryget slug]]
set lead [ns_dbquotevalue [ns_queryget lead]]
set body [ns_dbquotevalue [ns_queryget body]]

set qry "INSERT INTO ARTICLE VALUES ($slug, $date, now(), (SELECT ID FROM AUTHOR WHERE ID = $user and PASSWORD = $pass), $lead, $body, $title)"

set hdl [ns_db gethandle pool1]

if {[catch {ns_db dml $hdl $qry} res]} {
ns_adp_puts "post probably failed, hit back and try again"
ns_adp_puts "$res"
} else {
ns_adp_puts "post was a success!"
ns_adp_puts "$res"
}
%>
</code>
== notes ==

You'll need a <tt>template</tt> directory off your pageroot. The files <tt>entry.adp</tt> and <tt>article.adp</tt> are virtually the same except that <tt>entry.adp</tt> is used for the home page listing, while <tt>article.adp</tt> is used for the full article page. You'll generally want to use the lead in <tt>entry.adp</tt> and body in <tt>article.adp</tt>. The main template is <tt>template.adp</tt>. The main content section is filled with the <tt>mainStuff</tt> variable.

Your links to articles will be in the form of <tt>/article/YYYY/MM/DD/slug/whatever.html</tt>. If you'll notice, pubdate has dashes substituted for slashes. This makes forming the URL nice and easy. What I use to form the <em>Read More</em> link is:

<code>
<a href="/article/<%= $pubdate %>/<%= $slug %>/article.html">Read More</a>
</code>

AOLserver Wiki - User contributions [en]

User:Havard

Havard's Crappy Blog

Havard's Crappy Blog

User:Havard

Havard's Crappy Blog