I have a curl request in below format
curl -v -H "Content-Type:application/json" -H "x-user-id:xxx" -H "x-api-key:yyy" --data '{"logs":"'"${TEST_OUTPUT}"'","pass":"true | false"}' https://razeedash.one.qqq.cloud.com/api/v1/clusters/zzz/api/test_results
This works fine while I do from my MAC terminal. But the same command throws
13:49:26 {
13:49:26 "status": "error",
13:49:26 "message": "Invalid credentials"
13:49:26 }
I saw this post but not sure how else would I send a json body without curly braces. I know that we can save it as a file.json and use the file as body.But for some reasons that cannot be implemented in my scenario
In general, you should avoid trying to build JSON using string interpolation. Use a tool like jq to handle any necessary quoting.
jq -n --argson o "$TEST_OUTPUT" '{logs: $o, pass: "true | false"}' |
curl -v -H "Content-Type:application/json" \
-H "x-user-id:xxx" \
-H "x-api-key:yyy" \
--data #- \
https://razeedash.one.qqq.cloud.com/api/v1/clusters/zzz/api/test_results
However, if you can manage to correctly generate your JSON as you are now, you can just replace the jq command with echo:
echo '{"logs": ...' | curl ...
The #- argument to --data says to read from standard input.
Related
I need to make API call with payload (everything needs to be exactly like that)
{"file": "//'HLQ.DATASET(MEMBER)'"}
in Jenkins pipeline. I can't figure correct escaping of the payload. Problem is with the round brackets, single escape - Groovy complains, double escape - one of the slashes bubbles all the way into curl call.
def String job = """{\\"file\\": \\"\\'HLQ.DATASET\\(MEMBER\\)\\'\\"}"""
...
script {
def String response = sh(script: " curl -X PUT -w %{http_code} -v --header 'Content-Type: application/json' --cookie cookies.txt --header 'X-CSRF-ZOSMF-HEADER: dummy' --header 'X-IBM-Notification-URL: ${hook.getURL()}' https://.../zosmf/restjobs/jobs --data '$job'", returnStdout: true).trim()
}
If you are checking the Jenkins console output to determine whether the message is correctly sent it would mislead you. What you see in the console output is not always the interpreted string.
Can you try something like the below? Also inorder to check what Curl is sending out you can use a flag like --trace
def job = "{\"file\": \"//'HLQ.DATASET(MEMBER)'\"}"
writeFile(file: 'payload.txt', text: job)
sh 'cat payload.txt'
def String response = sh(script: "curl -X PUT -w %{http_code} -v --header 'Content-Type: application/json' --cookie cookies.txt --header 'X-CSRF-ZOSMF-HEADER: dummy' --header 'X-IBM-Notification-URL: ${hook.getURL()}' https://.../zosmf/restjobs/jobs --data #payload.txt", returnStdout: true).trim()
I am hitting a wall trying to build a script to save myself quite a good bit of time. I am working in a system in which I need to run a curl POST against a list of values. The list is about 400 lines long, so I am hoping to find a way of scripting this in Bash instead of running that call manually for each entry. Below are some details to help understand what I'm trying to accomplish:
If I were to be doing this task manually, each call would be formatted like the below:
curl -X POST --header "Content-Type: application/json" -v 'http://www.website.com:8081/cc/membership' -d #json_payload.json
This points to my JSON in the listed file which shows as the below:
{
"groupId": "12345678987654321",
"type": "serial",
"memberInfo": "apple"
}
If I run the above, the call works, and the expected operation occurs. The issue is that I need to run this against roughly 400 values for that "memberInfo" field in the JSON payload. I'm trying to identify a way to run a single bash script, which will run this curl command over and over, and update the JSON payload to use each row in a file as the below:
memberList.txt
apple
banana
peach
pear
orange
.
.
And then maybe insert a pointer in my JSON for the "memberInfo" field over to this file.
Any and all help/suggestions are greatly appreciated!
.
This will do as you intend. Its a little convoluted but you might polish it a bit.
#!/bin/bash
function getString(){
echo $1 | python3 -c '
import json
import sys
payload="""
{
"groupId": "12345678987654321",
"type": "serial",
"memberInfo": ""
}
"""
obj = json.loads(payload)
obj["memberInfo"] = sys.stdin.read().strip()
print(json.dumps(obj, indent = " "))
'
}
while read member
do
getString "$member" > json_payload.json
curl -X POST --header "Content-Type: application/json" -v 'http://www.website.com:8081/cc/membership' -d #json_payload.json
done <<< "$( cat fruits.txt )"
Hope it helps!
while read member; do
curl -X POST --header "Content-Type: application/json" -v 'http://www.website.com:8081/cc/membership' -d '{"groupId": "12345678987654321","type": "serial","memberInfo": "$member"}'
done <members.txt
This will work if you only care about the memberInfo field, another method could be writing your json line by line to payloads.txt file.
payloads.txt
{"groupId": "12345678987455432","type": "stereo","memberInfo": "apple"}
{"groupId": "34532453453453465","type": "serial","memberInfo": "banana"}
...
then use this as the script
while read payload; do
curl -X POST --header "Content-Type: application/json" -v 'http://www.website.com:8081/cc/membership' -d '$payload'
done <payloads.txt
here is a collection of bash scripting common uses I've had to use
https://github.com/felts94/advanced-bash/blob/master/bash_learn.sh
I have a bash script that sends a curl request and displays the response.
#!/bin/bash
token=$(curl -k -X GET \
'https://v.mytesting.io/oauth/token?grant_type=password&username=user1&password=123' \
-H 'Authorization: Basic 12345678' \
-H 'Host: v.mytesting.io.io')
v=$( jq -r ".access_token" <<<"$token" )
ts=$(curl -k -X POST \
https://timeseries.mytimeseries.io/v5/time_series/query \
-H 'Authorization: Bearer '"$v" \
-H 'Content-Type: application/json' \
-H 'Host: timeseries.mytimeseries.io' \
-H 'tenant: 123-123-123' \
-d '{"operation" : "raw","responseFormat" : "kairosDB","startTime": "1d-ago","stopTime": "now","tagList" : [ {"tagId" : "V.S.23164117.AVG.10M"}]}')
p=$(jq '.queries[].sample_size, .queries[].results[].name' <<<"$ts")
echo "$p"
My current output is just a value and the name of the tagId.
My query only allows for 1 tagId ( you can see above )
I want to be able to set a list of tagId's.
Then when I run this script it should loop through the list of tagId's and execute the curl request replacing the V.S.23164117.AVG.10M with each value
in the list.
Then output the entire list of results into a file.
list would be like so - (I would love to be able to enter this list into a seperate file and the bash script calls that file. Sometimes this list can be a few hundred lines.
V.S.23164117.AVG.10M
V.S.23164118.AVG.10M
V.S.23164119.AVG.10M
V.S.23164115.AVG.10M
V.S.23164114.AVG.10M
output would like look so.
value tagId
value tagId
value tagId
100 V.S.23164117.AVG.10M
etc..
thank you for any help
You can loop over list of tags using a small script. I'm not 100% clean of the output format. You can change the 'echo' to match the required format.
Note minor change to quotes to allow variable expansion in the body.
The tags will be stored in a file, for examples, tags.txt
V.S.23164117.AVG.10M
V.S.23164118.AVG.10M
V.S.23164119.AVG.10M
And the script will be use the file
#! /bin/bash
# Use user defined list of tags
tags=tags.txt
token=$(curl -k -X GET \
'https://v.mytesting.io/oauth/token?grant_type=password&username=user1&password=123' \
-H 'Authorization: Basic 12345678' \
-H 'Host: v.mytesting.io.io')
v=$( jq -r ".access_token" <<<"$token" )
for tag in $(<$tags) ; do
ts=$(curl -k -X POST \
https://timeseries.mytimeseries.io/v5/time_series/query \
-H 'Authorization: Bearer '"$v" \
-H 'Content-Type: application/json' \
-H 'Host: timeseries.mytimeseries.io' \
-H 'tenant: 123-123-123' \
-d '{"operation" : "raw","responseFormat" : "kairosDB","startTime": "1d-ago","stopTime": "now","tagList" : [ {"tagId" : "'"$tag"'"}]}')
p=$(jq '.queries[].sample_size, .queries[].results[].name' <<<"$ts")
echo "$tag $p"
done
#!/bin/bash
curl -v \
--request PATCH \
--data "$(
printf '{"files": {"somefile.json": {"content": " {"field": "value"} "}}}' \
)" \
--user x:x \
https://api.github.com/gists/x
Tried adding --header "Content-Type: application/json", no luck.
I'm using this because the content is actually a command output but right now I'm testing the basics because this is not workig.
I believe is somethig related to double quote escaping in bash, tried for a couple of hours with no luck. This is a nightmare.
Any tip is welcomed. Thanks.
It looks like you have too many quotation marks. If you want the value of the "content" element to be an object, then instead of this:
"content": " {"field": "value"} "
try this:
"content": {"field": "value"}
On the off chance that you want it to be a string, then try this:
"content": " {\"field\": \"value\"} "
Instead of fighting against quote escaping, you could write your payload to a file and tell curl to use that file as data like so:
curl -v \
--request PATCH \
--data #/tmp/some/file \
--user x:x \
https://api.github.com/gists/x
Note the # sign in the --data argument, which tells curl that the rest of the argument is a file name to read data from.
Depending on how you create your payload, you could also pipe it to curl using - as filename (echo payload | curl --data #- ...)
I have a curl command:
curl -u ${USER_ID}:${PASSWORD} -X GET 'http://blah.gso.woo.com:8080/rest/job-execution/job-details/${job_id}'
The variable job_id has a value in it, say, 1160. When I execute the curl command in shell it gives me the following error:
{"message":"Sorry. An unexpected error occured.", "stacktrace":"Bad Request. The request could not be understood by the server due to malformed syntax."}
If I pass the number '1160' directly in the command, as shown below, the curl command works.
curl -u ${USER_ID}:${PASSWORD} -X GET 'http://blah.gso.woo.com:8080/rest/job-execution/job-details/1160'
I want to be able to pass the value of the variable in the curl command.
When using variables in shell, you can only use doubles quotes, not single quotes : the variables inside single quotes are not expanded.
Learn the difference between ' and " and `. See http://mywiki.wooledge.org/Quotes and http://wiki.bash-hackers.org/syntax/words
I ran into this problem with passing as well, it was solved by using ' " $1 " '
See connection.uri below
curl -X POST -H "Content-Type: application/json" --data '
{"name": "mysql-atlas-sink",
"config": {
"connector.class":"com.mongodb.kafka.connect.MongoSinkConnector",
"tasks.max":"1",
"topics":"mysqlstock.Stocks.StockData",
"connection.uri":"'"$1"'",
"database":"Stocks",
"collection":"StockData",
"key.converter":"io.confluent.connect.avro.AvroConverter",
"key.converter.schema.registry.url":"http://schema-registry:8081",
"value.converter":"io.confluent.connect.avro.AvroConverter",
"value.converter.schema.registry.url":"http://schema-registry:8081",
"transforms": "ExtractField",
"transforms.ExtractField.type":"org.apache.kafka.connect.transforms.ExtractField$Value",
"transforms.ExtractField.field":"after"
}}' http://localhost:8083/connectors -w "\n"
How to pass json to curl with shell variable(s):
myvar=foobar
curl -H "Content-Type: application/json" --data #/dev/stdin<<EOF
{ "xkey": "$myvar" }
EOF
With the switch -d or --data, the POST request is implicit
use variable in a double-quote single-quote "' $variable '"
#!/usr/bin/bash
token=xxxxxx
curl --location --request POST 'http://127.0.0.1:8009/submit/expense/' \
--form 'token="'$token'"' \
--form 'text="'$1'"' \
--form 'amount="'$2'"'
userdetails="$username:$apppassword"
base_url_part='https://api.XXX.org/2.0/repositories'
path="/$teamName/$repoName/downloads/$filename"
base_url="$base_url_part$path"**strong text**
curl -L -u "$userdetails" "$base_url" -o "$downloadfilename"