For a integration test, I need to download a CSV file using poltergeist driver with Capybara. In selenium(for example firefox/chrom webdriver), I can specify download directory and it works fine. But in poltergeist, is there a way to specify the download directory or any special configuration?. Basically I need to know how download stuff works using poltergeist,Capybara, Phantomjs.
I can read server response header as Hash using ruby but can not read the server response to get the file content.Any clue? or help please.
Finally I solved the download part by simply using CURL inside Ruby code without using any webdriver. The idea is simple, first of all, I submitted the login form via CURL and saved the cookie into my server and then submitted(via CURL) the CVS Export form using the saved cookie like this
post_data = "p1=d1&p2=d2&p3=d3"
`curl -c cookie.txt -d "userName=USERNAME&password=PASSWORD" LOGIN SUBMIT_URL`
csv_data = `curl -X POST -b cookie.txt -d '#{post_data}' SUBMIT_URL_FOR_DOWNLOAD_CSV`
Related
I've managed to locate the correct element on the web page and I can click it. I see the download prompt come up, asking me whether I should open the pdf, open it in Firefox, or save it.
My code looks like this:
my $profile = Selenium::Firefox::Profile->new;
$profile->set_preference(
"browser.download.dir" => "/Users/john/Downloads",
"browser.download.folderList" => "1",
"browser.helperapps.neverAsk.SaveToDisk" => "application/pdf"
);
my $driver = Selenium::Firefox->new(
'binary' => "/usr/local/bin/geckodriver",
'firefox_profile' => $profile
);
[...]
$driver->find_child_element($driver->find_element_by_id('secHead_1'), "./a[\#class='phoDirFont']")->click();
My understanding is that if I've set up the correct preferences, then this file would save without the prompt. That's not happening though. I've dug down into it with dev tools, and it does seem to be serving up the pdf file with "application/pdf" as the mime-type. Firefox certainly recognizes it as one (offering to open it in Firefox, not just with the registered app).
If there is another method (perhaps by sending keystrokes to the prompts), that would be acceptable too. Though I've been using Firefox (trying to get away from Google products in my personal life), if Chrome would make a difference I could switch to that instead.
Looking at my about:config (with the Selenium-opened window), the preferences settings seem to have taken. However, it still prompts for the file.
According to https://www.selenium.dev/documentation/en/worst_practices/file_downloads/ file downloads are not supported in selenium.
You could try using curl instead eg:
system("curl -s -o output.pdf -L $URL");
where -s stands for silent, -o stands for where to save the file and -L tells curl to follow redirect
If you need cookies you can obtain them with:
my #cookies = $driver->get_all_cookies();
extract whichever cookies you need and then pass them to curl with a --cookie parameter, like so:
system("curl -s -o output.pdf -L --cookie $cookie_one --cookie $another_cookie $URL");
I'm trying to kick-off an import of RDF files into a GraphDB repository via the workbench REST API.
It works fine when the file is in the {graphdb.workbench.importDirectory} directory and the request specifies "filenames": [ "file1.owl" ].
However, if the file is in a subdirectory (eg. {graphdb.workbench.importDirectory}/top/) and the request uses "filenames": [ "top/file1.owl" ], no such luck - nor does "/top/file1.owl" work. The Workbench Import UI shows the entire collection of eligible files under the {graphdb.workbench.importDirectory} directory. The file in question imports when the Workbench UI is used to initiate the import.
My question is: does the REST API support importing server files that are located is such child directories? And if so, what simple syntax am I missing out? any chance I have to specify any other property (eg. "baseURI":"file:/home/steve/graphdb-import/top/file1.owl")
Many thanks for any feedback.
If you have started GDB with -Dgraphdb.workbench.importDirectory=<path_to_the_import_directory> in "Server files" tab you should be able to see listed all files in this directory and in the child directories, which are located in the <path_to_the_import_directory> in following manner:
I've started GDB with
-Dgraphdb.workbench.importDirectory=/home/sava/Videos/data_for_import and in this directory I have subDirectory "movieDB" with two files "movieDB.brf" and "movieDB.brf.gz" and both are shown in the tab like "movieDB/movieDB.brf" and "movieDB/movieDB.brf.gz".
If you want to import these files using cURL use server import URL with method POST or:
curl -H POST 'http://localhost:7200/rest/data/import/server/w1' -H 'Accept: application/json, text/plain, /' -H 'Content-Type: application/json;charset=UTF-8' --data-binary '{"importSettings":{"name":"movieDB/movieDB.brf","status":"NONE","message":"","context":"","replaceGraphs":[],"baseURI":null,"forceSerial":false,"type":"file","format":null,"data":null,"timestamp":1608016179633,"parserSettings":{"preserveBNodeIds":false,"failOnUnknownDataTypes":false,"verifyDataTypeValues":false,"normalizeDataTypeValues":false,"failOnUnknownLanguageTags":false,"verifyLanguageTags":true,"normalizeLanguageTags":false,"stopOnError":true},"requestIdHeadersToForward":null},"fileNames":["movieDB/movieDB.brf"]}'
I'm using Jmeter for testing APIs and I want to parametrize the project's path from the terminal and then I want to use this parameter in JMeter. I set testurl = test.com in basic terminal and i want to get this url by using testurl. The parameter that I've sent via Command Line : ./jmeter -n -t your_script.jmx -l -Jurl=$testurl in homebrew terminal. The parameter that I've used in httpsRequest --> Server name or IP; ${__P(url)}. But when I run my automation in the homebrew terminal, my test scripts are not going to URL that's been defined. Please help me!! Thanks.
From first impression it seems what you are trying should work, but devil is in the detail. I would suggest you try:
Verify if the environment variable is set correctly use export testurl=test.com (removed spaces). Try verify using echo $testurl
Try debug sampler which should help you verify if JMeter is picking up the var correctly: https://www.blazemeter.com/blog/how-debug-your-apache-jmeter-script/
Hope this helps.
I'm writing a script to automatically setup a repository starting from a clean GraphDB running in a Docker container.
I have a config.ttl file containing repository configuration, the namespace and a dump in a file init.nq
I have successfully created the repository using the config.ttf and updated namespace but I cannot understand how to load the init.nq file.
This operation is extremely simple from web interface: Import -> RFD -> Upload, but I'm not able to understand how to perform it using Curl. I suppose that the correct API should be
post /repositories/{repositoryID}/statements
but the dump is to huge to pass it as simple text (~44MB).
This should work:
curl -X POST -H "Content-Type:application/n-quads" -T init.nq 'http://localhost:7200/repositories/test/statements'
I am trying to download all PDFs from http://www.fayette-pva.com/.
I believe the problem is that when hovering over the link to download the PDF chrome shows the URL in the bottom left hand corner without a .pdf file extension. I saw and used another forum answer similar to this but the .pdf extension was used for the URL when hovering over the PDF link with my cursor. I have tried the same code that is in the link below but it doesn't pick up the PDF files.
Here is the code I have been testing with:
wget --no-directories -e robots=off -A.pdf -r -l1 \
http://www.fayette-pva.com/sales-reports/salesreport03-feb-09feb2015/
I am using this on a single page of which I know that it has a PDF on it.
The complete code should be something like
wget --no-directories -e robots=off -A.pdf -r http://www.fayette-pva.com/
Related answer: WGET problem downloading pdfs from website
I am not sure if downloading the entire website would work and if it wouldn't take forever. How do I get around this and download only the PDFs?
Yes, the problem is precisely what you stated: The URLs do not contain regular or absolute filenames, but are calls to a script/servlet/... which hands out the actual files.
The solution is to use the --content-disposition option, which tells wget to honor the Content-Disposition field in the HTTP response, which carries the actual filename:
HTTP/1.1 200 OK
(...)
Content-Disposition: attachment; filename="SalesIndexThru09Feb2015.pdf"
(...)
Connection: close
This option is supported in wget at least since version 1.11.4, which is already 7 years old.
So you would do the following:
wget --no-directories --content-disposition -e robots=off -A.pdf -r \
http://www.fayette-pva.com/