io.pedestal.app.render.push.templates/update-template uses dom/set-html! to re-render its data. This is a problem if you have previously placed event listeners on the html element that was previously there, it gets wiped out.

[domina :as dom]
[domina.css :as dc]
[domina.events :as de]

(de/listen! (dc/sel "h1") :dblclick 
     (fn [e] (dom/set-html! (dc/sel "#stuff") "I am new html")))
(de/listen! (dom/by-id "stuff") :dblclick 
      (fn [e] (.log js/console "I was clicked on")))

<h1>Header</h1>
<div id="stuff">Click on me</div>

You can use the event handler for #stuff until you double click the header, then it will wipe out #stuff. It no longer responds to double click events.

Consider a the slightly modified dataflow from the skeleton app that contains a single transform:

(def example-app
  {;; There are currently 2 versions (formats) for dataflow
   ;; descritpion: the original version (version 1) and the current
   ;; version (version 2). If the version is not specified, the
   ;; descritpion will be assumed to be version 1 and an attempt
   ;; will be made to convert it to version 2.
   :version 2
   :transform [[:set-value [:greeting :hello] set-value-transform]]
   })

This results in a data-model of:

{:greeting {:hello "x"}}

However, io.pedestal.app/ensure-default-emitter generates:

[[#{[:*]} (default-emitter [])]]

The resulting app-model ends up being this:

[:node-create [] :map]
[:node-create [:greeting] :map]
[:value [:greeting] nil {:hello x}]

when one would expect:

[:node-create [] :map]
[:node-create [:greeting] :map]
[:node-create [:greeting :hello] :map]
[:value [:greeting :hello] nil x]

get-path only supports :* and not :** constructs, so it halts at the top level. Ideally it would support both and ensure-default-emitter would essentially mirror the data-model (which I think should be the default expectation).

As a developer working on a pedestal-app application
In order to more easily switch between test and implementation namespaces in my editor
I need test files generated by app-template to have the correct namespace format ({{namespace}}.behavior-test)

Acceptance Criteria:

• Test namespaces in app-template follow the {{namespace}}.behavior-test scheme and directory structure.

Generate an app: lein new pedestal-service heroku-pedestal

Modify project.lj to mention :min-lein-version "2.0.0"

Deploy and we see this:

$ git push heroku master
Counting objects: 12, done.
Delta compression using up to 8 threads.
Compressing objects: 100% (10/10), done.
Writing objects: 100% (10/10), 884 bytes, done.
Total 10 (delta 6), reused 0 (delta 0)

-----> Clojure (Leiningen 2) app detected
-----> Installing OpenJDK 1.6...done
-----> Installing Leiningen
       Downloading: leiningen-2.0.0-standalone.jar
       Writing: lein script
-----> Building with Leiningen
       Running: lein with-profile production compile :all
       Retrieving lein-standalone-repl/lein-standalone-repl/0.1.4/lein-standalone-repl-0.1.4.pom from clojars
       Retrieving lein-standalone-repl/lein-standalone-repl/0.1.4/lein-standalone-repl-0.1.4.jar from clojars
       Performing task 'compile' with profile(s): 'production'
       Compiling heroku-pedestal.server
       Exception in thread "main" java.lang.IllegalArgumentException: No implementation of method: :handler-interceptor of protocol: #'io.pedestal.service.http.route.definition.verbose/HandlerFn found for class: io.pedestal.service.impl.interceptor.Interceptor, compiling:
(server.clj:12:1)
        at clojure.lang.Compiler$InvokeExpr.eval(Compiler.java:3463)
        at clojure.lang.Compiler.compile1(Compiler.java:7153)
        at clojure.lang.Compiler.compile1(Compiler.java:7143)
        at clojure.lang.Compiler.compile(Compiler.java:7219)
        at clojure.lang.RT.compile(RT.java:398)
        at clojure.lang.RT.load(RT.java:438)

Running the command that failed above lein with-profile production compile :all locally doesn't cause any errors.

I suspected :main and removed it but that didn't change the outcome.

cheshire is significantly faster than clojure.data.json.

Upgrade body parsing and JSON response generation to use cheshire. Make sure the API is consuming input data directly from a request's input stream and producing output data directly to a response's output stream without making a copy in a string in either direction.

• Update body-params json parser
• Update json-response emitter
• Keywordize JSON by default.
• Keywordize JSON can be turned off trivially.
• Add any new tests that may be necessary.
• Bump minor version (this may slightly break some existing apps we know of that rely on existing behavior)

Obviously, all tests should remain passing.

Now that we have updated ClojureScript support in pedestal-app, we should switch to using namespaced keyword shorthands (::msg/topic instead of msg/topic)

Switch from msg/topic and msg/type symbols to ::msg/topic and ::msg/type keywords.

• Make the change in app
• Make the change in app-template
• Make the change in samples (create card there as needed.)
• Make the change in app-tutorial (create card there as needed.)

Hi! I am new to clojure and just ran across this project and am hoping to use it. One question immediately would be how can I, when I make changes, have those changes reflected immediately on a refresh instead of having to restart the server? It seems by default that is how it is. Thanks!

Gathering by the current state of route.clj, it should be possible to eliminate the local definitions of url-for from service-template (and any samples, see pedestal/samples#33)

Resources found in the filesystem under a pedestal-service project are served as a mime-type based on their file extension.

If I move the resources into the classpath of a jar file that is included in my project, then the bytes of the resource are served, but the Content-Type header is set to application/octet-stream regardless of the file extension of the resource in the jar.

I think http-kit is cool. I also think pedestal is cool and so I want to use them together. After a brief read through server.clj it seems like a server constructed from a ring handler (not a servlet) should be possible.

Is it? Are there other components that rely on the Servlet API which cannot be avoided? Any help or direction would be appreciated.

Here is my attempt at an HTTP adapter I modeled after io.pedestal.service.http.jetty.

io.pedestal.service.http.http-kit.

I think it would be helpful to change the log methods for the application to be a macro instead of function calls as they are right now. This would allow for the line number output in Chrome console, etc to map to the correct file and location of where the logging output was generated.

Right now all of the logging output references console_log.js

Previously, it was the case that running lein new pedestal-service helloworld and making src/helloworld/service.clj match the following would result in the / route serving "Hello world" as normal browser-friendly response.

(ns helloworld.service
    (:require [io.pedestal.service.http :as bootstrap]
              [io.pedestal.service.http.route.definition :refer [defroutes]]
              [ring.util.response :refer [response]]))

(defn home-page
  [request]
  (response "Hello World!"))

(defroutes routes
  [[["/" {:get home-page}]]])

;; Consumed by helloworld.server/create-server
(def service {:env :prod
              ::bootstrap/routes routes
              ::bootstrap/type :jetty
              ::bootstrap/port 8080})

Now, without the addition of an html-body interceptor as below the / route is served as application/octet-stream.

(defroutes routes
 [[["/" {:get home-page} ^:interceptors [(body-params/body-params) bootstrap/html-body]]]])

It should probably be the case that by-default content is not served as an octet-stream.

This is a lot less clear than it should be.

As brought up in #105, sometimes we write code that can't actually be represented in ClojureScript (character manipulation, for example.) In these cases, app-tools generally serves a blank page instead of ones applications, and it is quite difficult to track down the source of these errors without scanning through log output.

We should degrade more gracefully in these cases. Ideally we might display an error page a la the Rails stack trace page:

EDIT(@rkneufeld): I've co-opted this issue to serve as the issue for adding an option to disable directory flattening.

Original Issue Report:

The decision of how to name compiled/copied files should be left to the :application configuration map, which already handles where to generate them.

The current method is automatic directory flattening. It might make more sense for the user to have to opt-in to flattening, and default to a less opinionated method.

See https://groups.google.com/forum/#!topic/pedestal-users/0bat90hlsCM

There is a potential problem with how the default emitter handles updates to models when the value changes from a true value to a false value. For example, let us say you have a model {:tasks {task1 {task-id: task1}}

If you update the tasks model so that it's now:

{:tasks {task1 {task-id: task1 :completed true}}

It will output the new value, [:value [:tasks task1 :completed] true]

However, if you change it from a true value to a false one, such as nil or false, then it won't output the new value. For example, I would expect the following to be output:

[:value [:tasks task1 :completed] false]

That's not what happens though. Nothing gets output.

Here is how everything gets done:

(let [updates (->> inputs
                   dataflow/updated-inputs
                   (filter-changes inputs))]
  (mapv (fn [[k v]] [:value (prefixed k prefix) v]) updates))

It then goes from updated-inputs into changed-inputs, which calls (input-map inputs)

(defn input-map [{:keys [new-model input-paths]}]
  (into {} (for [path input-paths
                 [k v] (get-path new-model path)
                 :when v]
             [k v])))

The problem comes from the fact that if v is false or nil, it's not getting included because of the when clause. I could see excluding nil, but I think false is a valid value that a map can hold.

Perhaps a :when (nil? v) clause instead?

Problem

If you run the chat sample in the "production" mode, the client app fails as the window.EventSource gets obfuscated by the closure compiler with advanced compilation mode. This isn't limited to the chat sample, but that is just an easy way to replicate it.

I haven't been able to properly pass in an extern file in the config to resolve this. Is there some way to pass in externs in the config and have them passed through to the Closure compiler?

Long term, I think that it might be useful to include the extern file from Google https://code.google.com/p/closure-compiler/source/browse/contrib/externs/w3c_eventsource.js in Pedestal in order to make the production mode work by default.

Update template so lein run does not use tiered compilation. Ping @timewald for more information on this if this isn't entirely clear.

There's a function called 'update-constraints' in the file called 'io/pedestal/service/http/route/definition/verbose.clj' and it has a bug.

Line 93 looks like that:

(update-in dna [:path-constraints] merge (map capture-constraint constraints))

But it should be probably something like that:

(update-in dna [:path-constraints] merge (into {} (map capture-constraint constraints)))

google appengine's servlet implementation doesn't have a .isAsyncSupported method for HttpServletRequest

this method was added in the 3.0 version of the servlet spec, GAE is still on 2.5

In app-tools.build/split-path (copied below) we're creating a regular expression out of the raw java.io.File/separator.

(defn- split-path [s]
  (string/split s (re-pattern File/separator)))

On *nix systems, this is "/" and creates a re-pattern that works correctly.

On Windows, this is "\\" and creates a re-pattern that will probably raise a java.util.regex.Pattern.error. You would instead have to escape this separator as "\\\\" for it to split a string "C:\some\dir" correctly.

There are a couple of really nasty ways to fix this ((clojure.string/escape "\\" {\\ "\\\\"}) for one).

Can we come up with a better solution and some tests to stress this?

With the following definition in behavior.clj

(def letter (char 75)); \K

the data-ui in pedestal-app fails to render. The number supplied to char doesn't matter. I do not know if the app continues to work elsewhere.

I created a gist here: https://gist.github.com/JouisWalker/5991163

behavior.clj can be copy-pasted into a project created by

lein new pedestal-app broken-character

Arch Linux; 
Linux 3.9.9-1-ARCH x86_64
Leiningen 2.2.0 on Java 1.7.0_40 OpenJDK 64-Bit Server VM
[io.pedestal/pedestal.app "0.1.10"]
[io.pedestal/pedestal.app-tools "0.1.10"]

This occurs in firefox 22.0-1 and chromium 28.0.1500.71-1, so I assume it to be browser independent.

In io.pedestal.service.http.route/url-for line 396:

(throw (ex-info "*url-for* not bound"))

must include a map in the call, like:

(throw (ex-info "*url-for* not bound" {}))

How Can I load css files so that my html templates files can use them. My css files are under /resources/public/css

$ lein deps
Retrieving org/clojure/clojure/1.3.0-alpha6/clojure-1.3.0-alpha6.jar from central
Failing dependency resolution because:

[ch.qos.logback/logback-classic "1.0.7"] -> [org.slf4j/slf4j-api "1.6.6"]
is overrulling
[org.slf4j/jcl-over-slf4j "1.7.2"] -> [org.slf4j/slf4j-api "1.7.2"]

Please use [ch.qos.logback/logback-classic "1.0.7" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.7.2"] or use [org.slf4j/jcl-over-slf4j "1.7.2" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.6.6"].

[ch.qos.logback/logback-classic "1.0.7"] -> [org.slf4j/slf4j-api "1.6.6"]
is overrulling
[io.pedestal/pedestal.service "0.1.6-SNAPSHOT"] -> [org.slf4j/slf4j-api "1.7.2"]

Please use [ch.qos.logback/logback-classic "1.0.7" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.7.2"] or use [io.pedestal/pedestal.service "0.1.6-SNAPSHOT" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.6.6"].

[ch.qos.logback/logback-classic "1.0.7"] -> [org.slf4j/slf4j-api "1.6.6"]
is overrulling
[org.slf4j/log4j-over-slf4j "1.7.2"] -> [org.slf4j/slf4j-api "1.7.2"]

Please use [ch.qos.logback/logback-classic "1.0.7" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.7.2"] or use [org.slf4j/log4j-over-slf4j "1.7.2" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.6.6"].

[ch.qos.logback/logback-classic "1.0.7"] -> [org.slf4j/slf4j-api "1.6.6"]
is overrulling
[org.slf4j/jul-to-slf4j "1.7.2"] -> [org.slf4j/slf4j-api "1.7.2"]

Please use [ch.qos.logback/logback-classic "1.0.7" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.7.2"] or use [org.slf4j/jul-to-slf4j "1.7.2" :exclusions [org.slf4j/slf4j-api]] to get [org.slf4j/slf4j-api "1.6.6"].

Hi Relevance: Would it be possible to update the PDF for the Contributor Agreement to use actual PDF form fields please? That would enable a faster and easier submission of the form for those of us who don't have a printer and scanner: we could type the required data in, and either type-to-sign or use one of the various finger-signing apps for iOS/Android. Thanks!

Consider for example the following template:

<li template="todo" field="id:id" field="class:status">
    <div class="view">
        <input class="toggle" type="checkbox" checked>
        <label field="content:text">Create a TodoMVC template</label>
        <button class="destroy"></button>
    </div>
    <input class="edit" value="Create a TodoMVC template" field="value:text">
</li>

In current app version, the last occurence of the "something:text" overwrites all previous occurences, so for example in template above, the content of label tag is not bound to the text data field.

I already solved the issue in my fork of pedestal (janherich@18c40d5), i will create a pull request as soon as i sign CA.

Tried creating a pedestal project, using the steps mentioned https://github.com/pedestal/pedestal/tree/master/app
But getting the below error, looks like the jars are missing in repo,

Please suggest.
Retrieving org/clojure/clojure/1.5.1/clojure-1.5.1.pom (6k)
from http://repo1.maven.org/maven2/
Could not find artifact io.pedestal:pedestal.app-tools:pom:0.1.1-SNAPSHOT in central (http://repo1.maven.org/maven2)
Could not find artifact io.pedestal:pedestal.app-tools:pom:0.1.1-SNAPSHOT in clojars (https://clojars.org/repo/)
Could not find artifact io.pedestal:pedestal.app-tools:jar:0.1.1-SNAPSHOT in central (http://repo1.maven.org/maven2)
Retrieving org/clojure/clojure/1.5.1/clojure-1.5.1.jar (3501k)
from http://repo1.maven.org/maven2/
Could not find artifact io.pedestal:pedestal.app-tools:jar:0.1.1-SNAPSHOT in clojars (https://clojars.org/repo/)
Could not find artifact io.pedestal:pedestal.app-tools:jar:0.1.1-SNAPSHOT in central (http://repo1.maven.org/maven2)
Check :dependencies and :repositories for typos.
It's possible the specified jar is not in any repository.
If so, see "Free-floating Jars" under http://j.mp/repeatability

Take for example the following html_templates.clj content:

(defmacro todo-mvc-templates
  []
  {:todo-mvc-page (dtfn (tnodes "todo-mvc.html" "todoapp" [[:#todo-list] [:#footer]]) #{})
   :todo-item (dtfn (tnodes "todo-mvc.html" "todo-item") #{:id})
   :count (dtfn (tnodes "todo-mvc.html" "count") #{})
   :filters (dtfn (tnodes "todo-mvc.html" "filters") #{})
   :clear-completed (dtfn (tnodes "todo-mvc.html" "clear-completed") #{})})

and this template html:

<li class="completed" template="todo-item" field="class:status" >
  <div class="view" field="id:id">
    <input class="toggle" type="checkbox">
    <label field="content:content">Create a TodoMVC template</label>
    <button class="destroy"></button>
  </div>
  <input class="edit" value="Create a TodoMVC template" field="value:content">
</li>

I'd like to hook up "status" in some way to <input class="toggle" type="checkbox">, but setting the checked field to anything can cause issues. In this case I'd like checked="checked" to appear for true, and no attribute at all to appear for false.

Current behavior is to spam the console with a flood of stacktraces, requiring a kill/restart.

Additionally, it appears to corrupt the files under ./out, requiring a manual remove of that directory.

(daemian@headcrab:~/relevance/samples/todo-mvc)       (todo-mvc)       (08:46:25)
> lein repl
nREPL server started on port 64904
REPL-y 0.1.0-beta10
Clojure 1.5.0
    Exit: Control+D or (exit) or (quit)
Commands: (user/help)
    Docs: (doc function-name-here)
          (find-doc "part-of-name-here")
  Source: (source function-name-here)
          (user/sourcery function-name-here)
 Javadoc: (javadoc java-object-or-class-here)
Examples from clojuredocs.org: [clojuredocs or cdoc]
          (user/clojuredocs name-here)
          (user/clojuredocs "ns-here" "name-here")
user=> (dev)
#<Namespace dev>
dev=> (watch :developmentt)
watching [:todo-mvc] / :developmentt
:ok
dev=> ERROR io.pedestal.app-tools.build - {:line 203}
java.io.FileNotFoundException: out/public/generated-js (Is a directory)
        at java.io.FileOutputStream.open(Native Method) ~[na:1.7.0_15]
        at java.io.FileOutputStream.<init>(FileOutputStream.java:212) ~[na:1.7.0_15]
        at clojure.java.io$fn__8664.invoke(io.clj:234) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$fn__8588$G__8564__8595.invoke(io.clj:73) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$fn__8676.invoke(io.clj:267) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$fn__8588$G__8564__8595.invoke(io.clj:73) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$fn__8638.invoke(io.clj:170) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$fn__8614$G__8568__8621.invoke(io.clj:73) ~[clojure-1.5.0.jar:na]
        at clojure.java.io$writer.doInvoke(io.clj:123) ~[clojure-1.5.0.jar:na]
        at clojure.lang.RestFn.invoke(RestFn.java:410) ~[clojure-1.5.0.jar:na]
        at clojure.lang.AFn.applyToHelper(AFn.java:161) ~[clojure-1.5.0.jar:na]
        at clojure.lang.RestFn.applyTo(RestFn.java:132) ~[clojure-1.5.0.jar:na]
        at clojure.core$apply.invoke(core.clj:619) ~[clojure-1.5.0.jar:na]
        at clojure.core$spit.doInvoke(core.clj:6291) ~[clojure-1.5.0.jar:na]
        at clojure.lang.RestFn.invoke(RestFn.java:425) ~[clojure-1.5.0.jar:na]
        at cljs.closure$output_one_file.invoke(closure.clj:781) ~[na:na]
        at cljs.closure$output_deps_file.invoke(closure.clj:785) ~[na:na]
        at cljs.closure$output_unoptimized.doInvoke(closure.clj:841) ~[na:na]
        at clojure.lang.RestFn.applyTo(RestFn.java:139) ~[clojure-1.5.0.jar:na]
        at clojure.core$apply.invoke(core.clj:619) ~[clojure-1.5.0.jar:na]
        at cljs.closure$build.invoke(closure.clj:915) ~[na:na]
        at io.pedestal.app_tools.compile$build_sources_BANG_.invoke(compile.clj:124) ~[na:na]
        at io.pedestal.app_tools.compile$compile_BANG_.invoke(compile.clj:199) ~[na:na]
        at io.pedestal.app_tools.build$build_BANG_.invoke(build.clj:193) ~[na:na]
        at io.pedestal.app_tools.build$thread_safe_build_BANG_$fn__8069$fn__8070.invoke(build.clj:201) ~[na:na]
        at io.pedestal.app_tools.build$thread_safe_build_BANG_$fn__8069.invoke(build.clj:201) ~[na:na]
        at clojure.core$binding_conveyor_fn$fn__4115.invoke(core.clj:1839) ~[clojure-1.5.0.jar:na]
        at clojure.lang.AFn.applyToHelper(AFn.java:161) ~[clojure-1.5.0.jar:na]
        at clojure.lang.RestFn.applyTo(RestFn.java:132) ~[clojure-1.5.0.jar:na]
        at clojure.lang.Agent$Action.doRun(Agent.java:114) [clojure-1.5.0.jar:na]
        at clojure.lang.Agent$Action.run(Agent.java:163) [clojure-1.5.0.jar:na]
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) [na:1.7.0_15]
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) [na:1.7.0_15]
        at java.lang.Thread.run(Thread.java:722) [na:1.7.0_15]
^CERROR

In a pedestal-app, if you have a page named child.html that rests _within some parent.html, changing parent.html and then re-visiting child.html will not trigger recompilation of that page. These changes will only be seen when child.html is modified next.

• Find out if it is possible to trigger re-compile in this scenario
• Trigger re-compile in this scenario

Hi Pedestal Team,

I get the following exception when reading a custom EDN tag with
the body-params interceptor:

java.lang.RuntimeException: No reader function for tag geo/point
    at clojure.lang.EdnReader$TaggedReader.readTagged(EdnReader.java:739) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader$TaggedReader.invoke(EdnReader.java:721) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader$DispatchReader.invoke(EdnReader.java:507) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader.readDelimitedList(EdnReader.java:695) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader$MapReader.invoke(EdnReader.java:638) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader.read(EdnReader.java:148) ~[clojure-1.5.1.jar:na]
    at clojure.lang.EdnReader.read(EdnReader.java:114) ~[clojure-1.5.1.jar:na]
    at clojure.edn$read.invoke(edn.clj:35) ~[clojure-1.5.1.jar:na]
    at io.pedestal.service.http.body_params$edn_parser.invoke(body_params.clj:62) ~[na:na]
    at io.pedestal.service.http.body_params$parse_content_type.invoke(body_params.clj:30) ~[na:na]

In my project I defined a data_readers.clj file sitting in my
resources directory, so I expect pedestal to honor my custom tags
as well. My data_readers.clj looks like this.

{geo/line-string geo.postgis/read-line-string
 geo/multi-point geo.postgis/read-multi-point
 geo/multi-polygon geo.postgis/read-multi-polygon
 geo/point geo.postgis/read-point
 inst datumbazo.io/read-instant-date-time}

I thing the fix is to call edn/read in the edn-parser like this:

(edn/read {:eof nil :readers *data-readers*})

I can sign the CA tomorrow and send a patch tomorrow if you are
fine with that solution.

I had this problem with:

Linux bombaclaat 3.9.8-1-ARCH #1 SMP PREEMPT Thu Jun 27 21:37:31 CEST 2013 x86_64 GNU/Linux
Leiningen 2.1.3 on Java 1.7.0_40 OpenJDK 64-Bit Server VM
[io.pedestal/pedestal.service "0.1.9"]
[io.pedestal/pedestal.jetty "0.1.9"]

Thanks, Roman.

In the app_tools application-host generation, relative paths are used for the generated-javascript files:

https://github.com/pedestal/pedestal/blob/master/app-tools/src/io/pedestal/app_tools/host_page.clj#L127

Could these be made absolute, like the other links of the host page? I have a dynamic landing page for my app, so that when I include the content via ring/file-response, the resources won't be located.

Cheers,
Si

I can't find the above mentioned namespace in the distribution shipped for 0.1.2. But I can see the namespace from the source. Am I missing proper usage?

John

This one's a little light on description. If you can manage w/o, go for it, otherwise ping @brentonashworth.

I couldn't find any instruction about lein new pedestal-app <name> on http://pedestal.io/ .
Probably, we should add client side "Getting Started" like http://pedestal.io/documentation/hello-world-service/ .

Running:

lein marg

in /service, as per documentation on http://pedestal.io/documentation/ fails fails with:

Could not transfer artifact io.pedestal:pedestal.jetty:pom:0.1.1-SNAPSHOT from/to relevance-private (http://nexus.thinkrelevance.com/nexus/content/groups/private/): Not authorized, ReasonPhrase:Unauthorized.

We had this working before with INFO but it seems we switched logback.xml to log at WARN or above. If we don't want to change this, we should update the template to not mention "By default, app logs to stdout" and the comment in logback.xml. @aredington: Thoughts? I think you made the change.

It's not clear to me how to get rid of control panel div with all aspect links from config.clj in case i don't want it in some aspect (for example :production). I can see that this div with links is appended to host-page by default dev-server (and in turn service) interceptor add-control-panel which is defined in the namespace io.pedestal.app-tools.host-page. But i think it should be possible to somehow override this behavior in a config.clj file for each aspect, for example with the key :control-panel set to false... Or am i missing some point and it's already possible ?

As a project
In order to be responsible
We need to properly attribute third-party code Pedestal utilizes in a NOTICES file.

@sumbach and @brentonashworth have been noted as having info on this.

When we moved from v1 to v2 dataflows we renamed a number of dataflow function types.

• Review names in app and dataflow namespace to ensure that all the old names have been removed
  • change msg/output to msg/effect

@brentonashworth are there any other likely suspects contributors should look out for?

Using a app tree like this:

[{:app
{:todos #{}}}]

Causes the following exception in the browser:

Uncaught Error: No matching clause: :unknown tree.js:146
(anonymous function) tree.js:146
(anonymous function) tree.js:149
cljs.core.apply.apply__2 core.js:7659
cljs.core.apply.apply core.js:7744
do_dispatch core.js:19568
cljs.core.MultiFn.cljs$core$IMultiFn$dispatch$arity$2 core.js:19663
dispatch core.js:19544
cljs.core.MultiFn.call.G__18053__delegate core.js:19669
cljs.core.MultiFn.call.G__18053 core.js:19676
cljs.core.seq_reduce.seq_reduce__3 core.js:4569
cljs.core.seq_reduce.seq_reduce core.js:4592
cljs.core.reduce.reduce__3 core.js:4673
cljs.core.reduce.reduce core.js:4681
update_tree tree.js:431
apply_deltas tree.js:666
cljs.core.swap_BANG.swap_BANG___3 core.js:18508
cljs.core.swap_BANG.swap_BANG_ core.js:18547
io.pedestal.app.render.consume_app_model_queue render.js:9
process_next_item queue.js:20
io.pedestal.app.queue.process_next_item

https://github.com/ptaoussanis/timbre

It would be nice to have a concise way of plugging in more than one renderer for a path.

Sometimes you have two views over the same data and you don't want to do them all in one function. Presently it seems the only way to go about it is to make it one function that calls the dedicated renderer for each of the views:

(defn render-simulation [renderer delta transmitter]
  (render-simulation-gfx renderer delta transmitter)
  (render-simulation-labels renderer delta transmitter))

It would be nice if I could do it right in the renderer config, like so:

[:value [:simulation-state] [render-simulation-gfx render-simulation-labels]]

Or:

[:value [:simulation-state] render-simulation-gfx]
[:value [:simulation-state] render-simulation-labels]

When I create a new pedestal service using the following command:

lein new pedestal-service pedestal-test

And then run the tests:

cd pedestal-test && lein test

I get 2 failures:

lein test pedestal-test.service-test
INFO  i.p.s.http.impl.servlet-interceptor - {:line 362, :in :interceptor-service-fn, :context {:servlet #<FnServlet io.pedestal.service.http.servlet.FnServlet@180c20e9>, :servlet-config nil, :servlet-response #<test$test_servlet_response$reify__278 io.pedestal.service.test$test_servlet_response$reify__278@78cf88cf>, :servlet-request #<test$test_servlet_request$reify__268 io.pedestal.service.test$test_servlet_request$reify__268@1d4cd493>}}
INFO  io.pedestal.service.http - {:line 58, :msg "GET /about"}
INFO  i.p.s.http.impl.servlet-interceptor - {:line 362, :in :interceptor-service-fn, :context {:servlet #<FnServlet io.pedestal.service.http.servlet.FnServlet@7447d144>, :servlet-config nil, :servlet-response #<test$test_servlet_response$reify__278 io.pedestal.service.test$test_servlet_response$reify__278@16fc0300>, :servlet-request #<test$test_servlet_request$reify__268 io.pedestal.service.test$test_servlet_request$reify__268@7b2900c9>}}
INFO  io.pedestal.service.http - {:line 58, :msg "GET /about"}

lein test :only pedestal-test.service-test/about-page-test

FAIL in (about-page-test) (service_test.clj:22)
expected: (= (:headers (response-for service :get "/about")) {"Content-Type" "text/html"})
  actual: (not (= {"Content-Type" "text/html;charset=UTF-8"} {"Content-Type" "text/html"}))
INFO  i.p.s.http.impl.servlet-interceptor - {:line 362, :in :interceptor-service-fn, :context {:servlet #<FnServlet io.pedestal.service.http.servlet.FnServlet@2678b3c1>, :servlet-config nil, :servlet-response #<test$test_servlet_response$reify__278 io.pedestal.service.test$test_servlet_response$reify__278@3e6157a2>, :servlet-request #<test$test_servlet_request$reify__268 io.pedestal.service.test$test_servlet_request$reify__268@68c06cac>}}
INFO  io.pedestal.service.http - {:line 58, :msg "GET /"}
INFO  i.p.s.http.impl.servlet-interceptor - {:line 362, :in :interceptor-service-fn, :context {:servlet #<FnServlet io.pedestal.service.http.servlet.FnServlet@311ec5ee>, :servlet-config nil, :servlet-response #<test$test_servlet_response$reify__278 io.pedestal.service.test$test_servlet_response$reify__278@5b54a573>, :servlet-request #<test$test_servlet_request$reify__268 io.pedestal.service.test$test_servlet_request$reify__268@123a4fdd>}}
INFO  io.pedestal.service.http - {:line 58, :msg "GET /"}

lein test :only pedestal-test.service-test/home-page-test

FAIL in (home-page-test) (service_test.clj:14)
expected: (= (:headers (response-for service :get "/")) {"Content-Type" "text/html"})
  actual: (not (= {"Content-Type" "text/html;charset=UTF-8"} {"Content-Type" "text/html"}))

Ran 2 tests containing 4 assertions.
2 failures, 0 errors.
Tests failed.

OS: Ubuntu 13.04
Leiningen version: Leiningen 2.1.3 on Java 1.7.0_21 OpenJDK 64-Bit Server VM
Pedestal version: io.pedestal/pedestal.service "0.1.7"

Re: https://github.com/pedestal/pedestal/blob/master/service/src/io/pedestal/service/test.clj#L96-97

I spent a bit of time trying to pass a "content-type" header to io.pedestal.service.test/response-for before realizing I needed to pass a "Content-Type" header instead.

Do we suppose this is a sufficiently-common error that it's worth searching for arbitrarily-cased headers in the test-servlet-request fn?

If so, I'll submit a PR.

if you have a ring handler that responds with a map

{:status 200 :body "foo"}

pedestal.service gives you a 404 with a body of "Not Found"

adding a empty headers like below fixes it:

{:status 200 :body "foo" :headers {}}

When I created client side app by lein new pedestal-app <name>, Pedestal created a project with README.md. This README.md has old links don't exist any more:

## Links

* [Getting Started](https://github.com/relevance/platform/tree/master/app#usage)
* [Overview of how pedestal-client applications work](https://github.com/relevance/platform/wiki/Client-Application-Model)