Commit 29e0551b authored by Wandenberg's avatar Wandenberg

docs update

parent efc4bc36
......@@ -99,6 +99,7 @@ one terminal and start playing http pubsub:
h1(#examples). Some Examples <a name="examples" href="#">&nbsp;</a>
* "Curl examples":curl
* "Forever (hidden) iFrame":forever_iframe
* "Event Source":event_source
* "WebSocket":websocket
......@@ -111,6 +112,16 @@ h1(#FAQ). FAQ <a names="faq" href="#">&nbsp;</a>
Doubts?! Check the "FAQ":wiki.
h1(#bug_report). Bug report <a name="bug_report" href="#">&nbsp;</a>
To report a bug, please provide the following information when applicable
# Which push stream module version is been used (commit sha1)?
# Which nginx version is been used?
# Nginx configuration in use
# "nginx -V" command outuput
# Core dump indicating a failure on the module code. Check "here":nginx_debugging how to produce one.
# Step by step description to reproduce the error.
h1(#who). Who is using the module? <a names="faq" href="#">&nbsp;</a>
......@@ -226,6 +237,7 @@ h1(#contributors). Contributors
[repository]https://github.com/wandenberg/nginx-push-stream-module
[contributors]https://github.com/wandenberg/nginx-push-stream-module/contributors
[changelog]CHANGELOG.textile
[curl]docs/examples/curl.textile#curl
[forever_iframe]docs/examples/forever_iframe.textile#forever_iframe
[event_source]docs/examples/event_source.textile#event_source
[websocket]docs/examples/websocket.textile#websocket
......@@ -267,3 +279,4 @@ h1(#contributors). Contributors
[push_stream_allowed_origins]docs/directives/subscribers.textile#push_stream_allowed_origins
[push_stream_websocket_allow_publish]docs/directives/subscribers.textile#push_stream_websocket_allow_publish
[wiki]https://github.com/wandenberg/nginx-push-stream-module/wiki/_pages
[nginx_debugging]http://wiki.nginx.org/Debugging
h1(#curl). Curl examples <a name="curl" href="#">&nbsp;</a>
Some commands to explain how the module protocol is implemented, and help to do your own client ;)
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
*Server:*
<pre>
location /channels-stats {
# activate channels statistics mode for this location
push_stream_channels_statistics;
# query string based channel id
push_stream_channels_path $arg_id;
}
location /pub {
# activate publisher (admin) mode for this location
push_stream_publisher admin;
# query string based channel id
push_stream_channels_path $arg_id;
}
location ~ /sub/(.*) {
# activate long-polling mode for this location
push_stream_subscriber long-polling;
# positional channel path
push_stream_channels_path $1;
# message template
push_stream_message_template "{\"id\":~id~,\"channel\":\"~channel~\",\"text\":\"~text~\"}";
# connection timeout
push_stream_longpolling_connection_ttl 30s;
}
</pre>
h2(#common_operations). Common operations
<pre>
# Subscribe to a channel
curl -s -v --no-buffer 'http://localhost/sub/my_channel_1'
curl -s -v --no-buffer 'http://localhost/sub/your_channel_1'
curl -s -v --no-buffer 'http://localhost/sub/your_channel_2'
# Publish messages
curl -s -v -X POST 'http://localhost/pub?id=my_channel_1' -d 'Hello World!'
curl -s -v -X POST 'http://localhost/pub?id=your_channel_1' -d 'Hi everybody!'
curl -s -v -X POST 'http://localhost/pub?id=your_channel_2' -d 'Goodbye!'
# Channels Stats for publisher (json format)
curl -s -v 'http://localhost/pub?id=my_channel_1'
# All Channels Stats summarized (json format)
curl -s -v 'http://localhost/channels-stats'
# All Channels Stats detailed (json format)
curl -s -v 'http://localhost/channels-stats?id=ALL'
# Prefixed Channels Stats detailed (json format)
curl -s -v 'http://localhost/channels-stats?id=your_channel_*'
# Channels Stats (json format)
curl -s -v 'http://localhost/channels-stats?id=my_channel_1'
# Delete Channels
curl -s -v -X DELETE 'http://localhost/pub?id=my_channel_1'
</pre>
h2(#getting_old_messages). Getting old messages
To get old messages you can use a backtrack, an event id or a time in the past to specify a start point.
All control methods use some HTTP headers by default, except for backtrack.
But they can use other methods also, like URL parameters.
To deliver old messages it's necessary to properly configure the directives "push_stream_store_messages", "push_stream_max_messages_stored_per_channel" and "push_stream_message_ttl".
*Using backtrack:*
<pre>
# To get the last 4 messages from channelA, the last 2 messages from channelC and no old messages from channelB
curl -s -v --no-buffer 'http://localhost/sub/channelA.b4/channelB/channelC.b2'
</pre>
*Using time in the past:*
When a message is published it receives a time and a tag value.
The tag is used to untie messages published on the same second.
These values are available to the message template using the ==~time~== and ==~tag~== patterns, and also on headers "Last-Modified" and "Etag" when on long-polling mode.
<pre>
# publish a message on t1
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '1'
# publish another message on t2
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '2'
# publish another message on t2
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '3'
# publish another message on t3
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '4'
# Get the messages published after the t2 time, tag 1
curl -s -v --no-buffer 'http://localhost/sub/channelA' -H 'If-Modified-Since: t2' -H 'If-None-Match: 1'
# t2, must be on the format "%a, %d %b %Y %T %Z", for instance "Thu, 1 Jan 1970 00:00:00 GMT"
# this subscriber will receive the messages "3" and "4"
</pre>
*Using time in the past (not using headers):*
Adding the directives "push_stream_last_received_message_time", "push_stream_last_received_message_tag" to subscriber location,
is possible to set the values for getting old messages using other methods, like URL parameters or a piece of the path.
modified server:
<pre>
location ~ /sub/(.*) {
push_stream_subscriber long-polling;
push_stream_channels_path $1;
push_stream_last_received_message_time $arg_time;
push_stream_last_received_message_tag $arg_tag;
push_stream_message_template "{\"id\":~id~,\"channel\":\"~channel~\",\"text\":\"~text~\", \"time\":\"~time~\", \"tag\":\"~tag~\"}";
push_stream_longpolling_connection_ttl 30s;
}
</pre>
using the same published messages on the above example:
<pre>
# Get the messages published after the t2 time, tag 2
curl -s -v --no-buffer 'http://localhost/sub/channelA?tag=2&time=t2'
# t2, must be on the format "%a, %d %b %Y %T %Z", for instance "Thu, 1 Jan 1970 00:00:00 GMT"
# this subscriber will receive only the message "4"
</pre>
*Using EventId:*
When a message is published with an Event-Id header this value can be used as a start point to get old messages.
It's available to the message template using the ==~event-id~== pattern.
<pre>
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '1'
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '2' -H 'Event-Id: some_special_event'
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '3'
curl -s -v -X POST 'http://localhost/pub?id=channelA' -d '4'
# Get the messages published after that event, in this example messages '3' and '4'
curl -s -v --no-buffer 'http://localhost/sub/channelA' -H 'Last-Event-Id: some_special_event'
</pre>
*Using EventId (not using headers):*
Adding the directive "push_stream_last_event_id" to subscriber location,
is possible to set the value for getting old messages using other methods, like URL parameters or a piece of the path.
modified server:
<pre>
location ~ /sub/(.*) {
push_stream_subscriber long-polling;
push_stream_channels_path $1;
push_stream_last_event_id $arg_last_id;
push_stream_message_template "{\"id\":~id~,\"channel\":\"~channel~\",\"text\":\"~text~\", \"event\":\"~event-id~\"}";
push_stream_longpolling_connection_ttl 30s;
}
</pre>
using the same published messages on the above example:
<pre>
# Get the messages published after the event 'some_special_event'
curl -s -v --no-buffer 'http://localhost/sub/channelA?last_id=some_special_event'
# this subscriber will receive the messages '3' and '4'
</pre>
......@@ -3,7 +3,7 @@ h1(#event_source). Event Source <a name="event_source" href="#">&nbsp;</a>
Using EventSource to receive the messages.
*This example uses the PushStream class present in _misc/js/pushstream.js_ file, copy it to your server htdocs.*
Configure your server like suggested bellow. You should complete this configuration with other directives according with target application.
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
Create a html page with the content on **Client** part, access it from browser and try with the command *curl http://localhost/pub?id=ch1 -d =="Some Text"==* .
*Server:*
......
......@@ -3,7 +3,7 @@ h1(#forever_iframe). Forever Iframe <a name="forever_iframe" href="#">&nbsp;</a>
Using an invisible iFrame on the page to receive the messages and pass them to main page.
*This example uses the PushStream class present in _misc/js/pushstream.js_ file, copy it to your server htdocs.*
Configure your server like suggested bellow. You should complete this configuration with other directives according with target application.
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
Create a html page with the content on **Client** part, access it from browser and try with the command *curl http://localhost/pub?id=ch1 -d =="Some Text"==* .
*Server:*
......
......@@ -3,7 +3,7 @@ h1(#long_polling). Long Polling <a name="long_polling" href="#">&nbsp;</a>
Using PushStream to receive the messages through long polling technique.
*This example uses the PushStream class present in _misc/js/pushstream.js_ file, copy it to your server htdocs.*
Configure your server like suggested bellow. You should complete this configuration with other directives according with target application.
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
Create a html page with the content on **Client** part, access it from browser and try with the command *curl http://localhost/pub?id=ch1 -d =="Some Text"==* .
*Server:*
......@@ -214,7 +214,7 @@ h1(#jsonp). JSONP <a name="jsonp" href="#">&nbsp;</a>
Using PushStream to receive the messages through JSONP technique.
*This example uses the PushStream class present in _misc/js/pushstream.js_ file, copy it to your server htdocs.*
Configure your server like suggested bellow. You should complete this configuration with other directives according with target application.
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
Create a html page with the content on **Client** part, access it from browser and try with the command *curl http://localhost/pub?id=ch1 -d "Some Text"*.
_The configuration in the example is the same used on long polling, just forcing the use of JSONP, this is automatic when the domains are different_
......
......@@ -3,7 +3,7 @@ h1(#websocket). WebSocket <a name="websocket" href="#">&nbsp;</a>
Using WebSocket to receive the messages.
*This example uses the PushStream class present in _misc/js/pushstream.js_ file, copy it to your server htdocs.*
Configure your server like suggested bellow. You should complete this configuration with other directives according with target application.
Configure your server like suggested bellow. You should complete this configuration with other directives according to the target application.
Create a html page with the content on **Client** part, access it from browser and try with the command *curl http://localhost/pub?id=ch1 -d =="Some Text"==* .
*Server:*
......
......@@ -120,7 +120,7 @@ begin
task :watch do
puts "watching for changes on textile files"
Dir.chdir(project_dir) do
FileWatcher.new(["*.textile"]).watch do |file, event|
FileWatcher.new(["**/*.textile"]).watch do |file, event|
if(event != :delete)
generate_preview_for(file, project_dir, base_path)
end
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment