I have been using puppeteer and always wondered to have something like this project in clj world, so thanks a lot!

When I try to port of my project from puppeteer to clj-chrome-devtools, there are essentially several missing pieces that are blocking me:

1. When connected to a browser, clj-chrome-devtools always use of the first page listed in the CDP websocket endpoint.
2. Some handy features are missing, e.g. page.title(), page.url() etc.
3. Querying by XPATH seems not supported, only css selectors are allowed.

Does it make sense to add these supports to this project? If so I may try to contribute to it.

Thanks for a great library:

I have one suggestion that might be useful for someone to get started with the project:

git clone https://github.com/tatut/clj-chrome-devtools.git
cd clj-chrome-devtools
git submodule init
git submodule update
# then follow the guide in the README to start the REPL
# e.g. 
lein repl 
# or 
lein test

Cloning: https://github.com/tatut/clj-chrome-devtools.git
Checking out: https://github.com/tatut/clj-chrome-devtools.git at a2c07289a55afef7c2ea81bf42cc76c70c9ebc63
Error building classpath. Checkout conflict with file: resources/devtools-protocol
org.eclipse.jgit.api.errors.JGitInternalException: Checkout conflict with file: resources/devtools-protocol

I have the following in my deps.edn:

   :aliases {:test {
                 :extra-deps {com.cognitect/test-runner {:git/url "https://github.com/cognitect-labs/test-runner.git"
                                                         :sha "209b64504cb3bd3b99ecfec7937b358a879f55c1"}
                              tatut/clj-chrome-devtools
                                                        {:git/url "https://github.com/tatut/clj-chrome-devtools.git"
                                                         :sha "a2c07289a55afef7c2ea81bf42cc76c70c9ebc63"}
                              }
                 :main-opts ["-m" "cognitect.test-runner"]}}

Hey all. Just started exploring this library, thanks for all the awesome work!

So this by alone does not work

(def conn (launcher/launch-automation {:headless? false}))

and results in the following exception after 30s

Execution error (ExceptionInfo) at clj-chrome-devtools.impl.connection/wait-for-remote-debugging-port (connection.clj:74).
Chrome remote debugging port not up

The only workaround I found for this was to first launch with headless? true and then launch another automation with headless? false

So this works properly

;; need this hack to run in GUI
(launcher/launch-automation {:headless? true})
(def conn (launcher/launch-automation {:headless? false}))

I've gotten clj-chrome-devtools.commands.page/print-to-pdf tantalizingly close to working, but clearly I'm doing something not quite right. Happy to help with steps-to-reproduce or a PR, but I'll need some help shaping these.

Steps to reproduce (maybe):

• install chrome on ubuntu 20.04
• run chrome in a shell with google-chrome --disable-gpu --headless --remote-debugging-port=9222
• connect to chrome in clojure and try to generate a PDF file per

(require
 '[clj-chrome-devtools.core          :as chrome]
 '[clj-chrome-devtools.commands.page :as page])

(defn b64-decode
  "https://stackoverflow.com/a/39188819/53790"
  [to-decode]
  (String. (.decode (Base64/getDecoder) to-decode)))

(def c (chrome/connect "localhost" 9222))
(page/navigate c {:url "https://github.com"})
(let [response (page/print-to-pdf c {:page-ranges "1-2"})
      pdf-data (b64-decode (:data response))]
  (spit "test.pdf" pdf-data))

Expected:

Should get a 2-page letter-size (ie. the default) PDF of the Github site as test.pdf.

Actual:

There is a 2-page letter-size PDF created at test.pdf, but its pages are blank.

Notes:

• I've tried different URLs and the file size bounces around, suggesting that there's different data in them.
• I can successfully generate PDFs using (clojure.java.shell/sh "google-chrome" "--headless" "--disable-gpu" "--print-to-pdf=test.pdf" "https://github.com") so I don't believe it's a problem with the Chrome installation.
• I can successfully use npm puppeteer-core with this Chrome installation and running this script as node test-script.js, so I'm suspecting the problem lies somewhere between Chrome and my code:

const puppeteer = require("puppeteer-core");

(async () => {
  const browser = await puppeteer.connect({
    // from "webSocketDebuggerUrl" at http://127.0.0.1:9222/json/version
    browserWSEndpoint:
      "ws://127.0.0.1:9222/devtools/browser/d6687d16-aabb-4eb5-9f39-38275307c218",
  });
  const page = await browser.newPage();
  await page.goto("https://github.com", {
    waitUntil: "networkidle2",
  });
  await page.pdf({ path: "test.pdf" });
})();

Hello guys,

I'm getting the following exception when trying to execute the clj-chrome-devtools.chrome-test/simple-page-load with lein eftest command (or just from IntelliJ Idea / Cursive plugin).

CompilerException java.lang.IllegalArgumentException: Cannot open <nil> as a Reader., compiling:(clj_chrome_devtools/commands/dom.clj:4:1)

Could you please help me to figure it out how can I execute the test example?
Thank you.

First big 👍 from me on this repo. I have been doing work with the devtools protocol in JS land recently and daydreaming about using it from Clojure.

I'm curious about your vision for the project. The beginnings of a rudimentary higher level API in clj-chrome-devtools.automation specifically. Is that intended to grow into a full high level API or to serve as quick start examples when building you own higher level interface lib?

If you spawn a headless chrome and connect right away it isn't ready and the quoted code above will run. This will throw an AssertionError where a normal RuntimeException or some subclass would be much more appropriate.

Clicks etc from clj-chrome-devtools.automation seem broken. For example:

(a/start!)
(a/click "div.button")

Throws:

1. Unhandled clojure.lang.ExceptionInfo
   Error in command DOM.querySelector: Invalid parameters
   {:request
    {:id 1,
     :method "DOM.querySelector",
     :params {"nodeId" nil, "selector" "div.button"}},
    :error
    {:code -32602,
     :message "Invalid parameters",
     :data "nodeId: integer value expected"}}
                   dom.clj:    4  clj-chrome-devtools.commands.dom/query-selector
                   dom.clj:    4  clj-chrome-devtools.commands.dom/query-selector
            automation.clj:  179  clj-chrome-devtools.automation/sel1/evaluate--auto--
            automation.clj:  178  clj-chrome-devtools.automation/sel1
            automation.clj:  174  clj-chrome-devtools.automation/sel1

I've traced this down to the fact that the library caches the root-node ID. If monkey-patch root-node-id so that resolves the ID on every call, the code above works!

(in-ns 'clj-chrome-devtools.automation)
(defn- root-node-id
  ([] (root-node-id @current-automation))
  ([ctx]
   (-> ctx :connection (clj-chrome-devtools.commands.dom/get-document {}) :root :node-id)))
(in-ns 'test.browser)

Current clojure.test integration has support for running cljs tests as part of a clojure test suite and supports leiningen projects.

Add support for building and running tests for deps.edn/figwheel-main based projects.

hi. with this version calling /json/list returns an empty array. previous versions returned an about:blank page that clj-chome-devtools found as the first-debuggable-page. without a found page, an exception is thrown.

the only reference I have found to others hitting this is quarto-dev/quarto-cli#4653