This question already has answers here:
jq not working on tag name with dashes and numbers
(2 answers)
Closed 4 years ago.
I have a (not so complicated) json file and I need to extract its contents using bash. I want to use jq for the processing, it should be straightforward. The problem is that I'm getting a weird error in the processing that I don't know how to solve (because I don't know what is causing it).
A minimal sample causing me problems:
{
"E23763": {
"data": "information"
}
}
If I just run jq to pretty-print it, it works:
$ cat test.json | jq .
{
"E23763": {
"data": "information"
}
}
But if I try to extract the first field, it fails criptically:
$ cat test.json | jq .E23763
jq: error: Invalid numeric literal at EOF at line 1, column 7 (while parsing '.E23763') at <top-level>, line 1:
.E23763
jq: 1 compile error
The expected result would had been:
{
"data": "information"
}
Anyone found a similar issue? Why it is complaining about a numeric literal when he is really looking into a string?
Quotation didn't seem to matter here, same error.
Please refer to this issue on GitHub there are many responses posted here which might help you with your problem: https://github.com/stedolan/jq/issues/1526
I'll post one of the solutions here however:
jq '.["E23763"]' test.json
Another Solution as said by #Inian is:
jq '."E23763"' json
Without the [], in this case it was the correct solution but try both nonetheless
Basically the parser is buggy and treats .E as the beginning of a number.
Related
Consider I have the below JSON:
{
"id": "Ab12",
"details": "{\"timeValue\":null,\"lastModifiedIn\":\"PHX\"}",
"version": 3
}
I want to extract the value of 'lastModifiedIn' key without using the jq command.
Basically, the result I seek for is 'PHX'.
Is there a way to extract this using basic shell scripting?
It's sloppy but given the input and your requirement to not use jq this might be good enough for your purposes:
$ sed -n 's/.*"lastModifiedIn\\":\\"\([^\\]*\).*/\1/p' file
PHX
I tried accessing the object using both the generic object index and the shorthand version of it but it appears that the generic object index won't work. Can someone explain to me why?
I have the following json file jsonsample.txt
{
"id": "v4cw72hf3",
"output": {
"url": "//srv01.cloudconvert.com/download/~ugl5vnrpfO",
"downloads": 0
},
}
The Jq manual explained the generic syntax
Generic Object Index: .[<string>]
You can also look up fields of an object using syntax like .["foo"] (.foo above is a shorthand version of this, but only for identifier-like strings).
and I tried two ways to access the url field
jq '["output"].["url"]' jsonsample.txt
jq .output.url jsonsample.txt
But the first one doesn't give me the desired results
#Result for the first line
jq: error: syntax error, unexpected '[', expecting FORMAT or QQSTRING_START (Unix shell quoting issues?) at <top-level>, line 1:
.["output"].["url"]
jq: 1 compile error
shell returned 3
#Results for the second line
"//srv01.cloudconvert.com/download/~ugl5vnrpfO"
The input is not quite valid JSON, so the following assumes it has been fixed.
The basic form for a pipeline of array and/or object accessors is
.[<string-or-integer>] | .[<string-or-integer>] | ...
So you'd be safe with .["output"]|.["url"]
Certain abbreviations are allowed, but different versions of jq differ in the details. However, it's generally safe to remove an interior |., i.e. one would expect
.["output"]["url"]
to work, as indeed it does going back at least to jq 1.3.
The restricted .foo.bar notation is also supported going back at least to jq 1.3.
jq 1.4 added support for unrestricted dot-string notation wherein the key name can be any valid JSON string (i.e. with quotation marks), e.g.
."foo with space"."bar with space"
I have exported a lot of JSON files, almost 100.
They are basically huge one-liners.
Each file has "uid":"random".
How to cut only "uid":"random" part for all files from the master directory?
I've tried with grep {} but can't workaround the quotes on the uid and the value.
Also, I tried to cut it but still the quotes are the problem.
Line structure:
..."title":"Random title","uid":"r4nd0muid","version":X},"overwrite": true}
Don't use anything but JSON aware tools for processing JSON files and post a proper sample for testing but to workaround the quotes on the uid and the value on the posted string:
$ grep -o \"uid\":\"[^\"]*\" foodata
"uid":"r4nd0muid"
Basically "uid":"[^"]*", ie. after "uid":" all non-" and a ".
It's probably a lot easier to use jq which is a search and transformation tool for JSON files.
Given the file test.json:
{
"uuid": "whatever"
}
You can extract just the uuid field with:
jq '.uuid' test.json
# output: "whatever"
So, I am writing a shell script and I am running a command that gives me an output like:
{"a":"some_text","b":some_other_text","c":"even_more_text"}
Now, I am not sure how to parse it, I basically need the value of "c", i.e. "ever_more_text" in a variable, but finding out results on internet have not worked yet! TIA.
the output which you paste here is not valid json. Check with https://jsonformatter.curiousconcept.com/ There is missing first double quote in "some_other_text". If you add it, you can then easily parse with jq:
./your_script.sh | jq -r ".c"
This question already has answers here:
jq not working on tag name with dashes and numbers
(2 answers)
Closed 5 years ago.
I have a REST API, which returns something like this:
{
"foo": 1,
"bar": 2,
"foo-bar": 3
}
when I do `http /endpoint/url | jq '.foo-bar', it gave the following error:
jq: error (at <stdin>:1): null (null) and boolean (true) cannot be subtracted
it looks like jq thinks I'm trying to do arithmetic operation with foo-bar.
How do I correctly form this kind of path? Or this is a bug of jq?
In JSON text, JSON keys are always double-quoted. Perhaps your REST API was formatting it properly in double-quotes and your example in your last edit was incorrect. Because without the same jq cannot parse the syntax as a valid JSON.
As far the issue you are seeing, you need to put the field within quotes to let jq know that it is a single field foo-bar you are accessing and not as separate fields
jq '."foo-bar"'
Or more specifically use the array access operator as jq '.["foo-bar"]'