In Gatling, I'd like to perform a check on some JSON included in an HTML response like below:
<!doctype html>
<html lang="fr">
<head>
<script>
var documentLoaded = performance.now();
</script>
<link rel="stylesheet" href="/styles/main.f14d8fab5a7e.css">
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<link rel="icon" type="image/png" sizes="16x16" href="/favicon-16x16.png">
<link rel="manifest" href="/manifest.json">
<link rel="preconnect" href="https://www.gstatic.com">
<title data-react-helmet="true">Asus Discount</title>
<meta data-react-helmet="true" name="description" content="Asus discount”/><meta data-react-helmet="true" name="keywords" content="Asus"/>
</head>
<body>
<div>Some content</div>
<script>
var parseStart = performance.now();
</script>
<script>
window.__INITIAL_STATE__ = {some JSON}; <!-- This is what I need -->
window.__ENV_VARIABLES__ = {some other JSON};
window.renderTime = '76';
window.fetchTime = '349';
</script>
<script type="text/javascript" charset="utf-8" src="/vendor.e33d9940372.js"></script>
<script type="application/ld+json" src="/schema.fr.json"></script>
</body>
</html>
My actual solution (which is working) looks as follow:
def loadPageJsonInHTML(requestName: String, link: String): ChainBuilder ={
exec(
http(requestName)
.get(link)
.check(regex("""window[.]__INITIAL_STATE__ = ([^;]+)""").find.transform(s => parseSToProdList(s)).saveAs("prod_list")
)
)
doIf("${prod_list.size()}" == 0){
exec{session => session.markAsFailed}
}
}
def parseSToProdList(jsonString: String): Seq[String] ={
val jsonMap = jsonStrToMap(jsonString)
val buffer = mutable.Buffer.empty[String]
jsonMap("products").asInstanceOf[Map[String, Any]].foreach{f =>
if(f._2.asInstanceOf[Map[String, Any]].keySet.exists(_ == "code"))
buffer.append(f._2.asInstanceOf[Map[String, Any]]("code").asInstanceOf[String])
}
buffer.toSeq
}
def jsonStrToMap(jsonStr: String): Map[String, Any] = {
implicit val formats = org.json4s.DefaultFormats
parse(jsonStr).extract[Map[String, Any]]
}
However, this solution has several drawbacks:
What I'd like to achieve is something that would look a bit like this:
def loadPageJsonInHTML(requestName: String, link: String): ChainBuilder ={
exec(
http(requestName)
.get(link)
.check(jsonPath("""$.products.*.code""").findAll.saveAs("prod_list")
)
Or
def loadPageJsonInHTML(requestName: String, link: String): ChainBuilder ={
exec(
http(requestName)
.get(link)
.check(jsonpJsonPath("""$.products.*.code""").findAll.saveAs("prod_list")
)
Of course, jsonPath doesn't work since most of the answer is HTML. jsonpJsonPath doesn't work either as there's several Json strings in the response.
Any good input as to how could I do this more effectively (and nicely) while avoiding a regex on some HTML? Thanks in advance
So, after some digging I found a workaround using ".transformResponse", in order to extract the string before it's actually checked, giving it a default value that is parsable in Json. Then, to make sure we actually did find the regex, we make sure that it's not our default value :
def loadPageJsonInHTML(requestName: String, link: String): ChainBuilder = {
exec(
http(requestName)
.get(link)
.transformResponse{(session, response) =>
response.copy(body = new StringResponseBody(
(for(m <- """window[.]__INITIAL_STATE__ = ([^;]+)""".r
.findFirstMatchIn(response.body.string)
) yield m.group(1)
).getOrElse("""{"error":"chain not found"}"""),
UTF_8
)
)
}
.check(bodyString.not("""{"error":"chain not found"}"""))
.check(jsonPath("""$.products.*.code""").findAll.saveAs("prod_list")
)
)
}