Upgrading from V3
Version 4 of
yq is quite different from previous versions (and I apologise for that) - however it will be very familiar if you have used jq before as it now uses a similar syntax. Most commands that you could do in v3 are longer in v4 as a result of having a more expressive syntax language.Note that
v4 by default now:- prints all documents of a yaml file.
- prints in color (when outputting to a terminal).
- document separators are printed out by default
In
v3 yq had seperate commands for reading/writing/deleting and more. In v4 all these have been embedded into a single expression you specify to either the eval command (which runs the expression against each yaml document for each file given in sequence) or the eval-all command, which reads all documents of all files, and runs the given expression once.Many flags from
v3 have been put into the expression language, for instance stripComments allowing you to specify which nodes to strip comments from instead of only being able to apply the flag to the entire document.Lets have a look at the commands for the most common tasks:
v3:
yq r sample.yaml 'a.b.c'
v4:
yq '.a.b.c' sample.yaml
v3:
yq r sample.yaml --defaultValue frog path.not.there
yq '.path.not.there // "frog"' sample.yaml
v3:
yq r sample.yaml 'a.(b.d==cat).f'
v4:
yq '.a | select(.b.d == "cat") | .f' sample.yaml
v3:
yq r sample.yaml 'thing.**.name'
v4:
yq '.thing | .. | select(has("name"))' sample.yaml
v3:
yq r -d1 sample.yaml 'b.c'
v4 (via the document index operator):
yq 'select(documentIndex == 1) | .b.c' sample.yml
v3:
yq w sample.yaml 'a.b.c' fred
v4:
yq '.a.b.c = "fred"' sample.yaml
v3:
yq d sample.yaml 'a.b.c'
v4:
yq 'del(.a.b.c)' sample.yaml
Like
jq, merge is done via the multiply operator. In yq, the merge operator can take extra options to modify how it works.For
v3 compatability, use the n option to only merge in new fields.yq '. *n load("file2.yaml")' file1.yaml
v3:
yq p data1.yaml c.d
v4:
yq '{"c": {"d": . }}' data1.yml
Note that in v4 you can no longer run expressions against an empty file to populate it - because the file is empty, there are no matches for
yq to run through the expression pipeline - for what it's worth, this is what jq does as well. Instead use the --null-input/-n flag and pipe out the results to the file you want directly (see example below).v3:
yq n b.c cat
v4:
yq -n '.b.c = "cat"'
v3:
yq validate some.file
v4:
yq 'true' some.file > /dev/null
Note that passing 'true' as the expression saves having to reencode the yaml (only to pipe it to stdout). In v4 you can also do a slightly more sophisticated validation and assert the tag on the root level, so you can ensure the yaml file is a map or array at the top level:
yq --exit-status 'tag == "!!map" or tag== "!!seq"' some.file > /dev/null
v3:
yq compare --prettyPrint file1.yml file2.yml
v4:
In v4 there is no built in compare command, instead it relies on using diff. The downside is longer syntax, the upside is that you can use the full power of diff.
diff <(yq -P file1.yml) <(yq -P file2.yml)
v3 had a script feature that let you run an array of commands specified in a file in one go. The format for this looked like
- command: update
path: a.key1
value: things
- command: delete
path: a.ab.key2
V4 doesn't have a similar feature, however the fact that you can run multiple operations in a single expression makes it easier to come up with a shell script that does the same thing:
#!/bin/bash
yq '
.a.key1 = "things" |
del(.a.ab.key2)
' ./examples/data1.yaml
Read and and control yaml metadata better (e.g. tags, paths, document indexes, anchors and aliases, comments).
Work with multiple files (not just for merge)
The underlying expression language is much more powerful than
v3 so expect to see more features soon!Last modified 2mo ago