7.1 KiB
Multiply (Merge)
Like the multiple operator in jq, depending on the operands, this multiply operator will do different things. Currently numbers, arrays and objects are supported.
Objects and arrays - merging
Objects are merged deeply matching on matching keys. By default, array values override and are not deeply merged.
Note that when merging objects, this operator returns the merged object (not the parent). This will be clearer in the examples below.
Merge Flags
You can control how objects are merged by using one or more of the following flags. Multiple flags can be used together, e.g. .a *+? .b
. See examples below
+
append arraysd
deeply merge arrays?
only merge existing fieldsn
only merge new fields
Merge two files together
This uses the load operator to merge file2 into file1.
yq '. *= load("file2.yml")' file1.yml
Merging all files
Note the use of eval-all
to ensure all documents are loaded into memory.
yq eval-all '. as $item ireduce ({}; . * $item )' *.yml
{% hint style="warning" %} Note that versions prior to 4.18 require the 'eval/e' command to be specified.
yq e <exp> <file>
{% endhint %}
Multiply integers
Given a sample.yml file of:
a: 3
b: 4
then
yq '.a *= .b' sample.yml
will output
a: 12
b: 4
Merge objects together, returning merged result only
Given a sample.yml file of:
a:
field: me
fieldA: cat
b:
field:
g: wizz
fieldB: dog
then
yq '.a * .b' sample.yml
will output
field:
g: wizz
fieldA: cat
fieldB: dog
Merge objects together, returning parent object
Given a sample.yml file of:
a:
field: me
fieldA: cat
b:
field:
g: wizz
fieldB: dog
then
yq '. * {"a":.b}' sample.yml
will output
a:
field:
g: wizz
fieldA: cat
fieldB: dog
b:
field:
g: wizz
fieldB: dog
Merge keeps style of LHS
Given a sample.yml file of:
a: {things: great}
b:
also: "me"
then
yq '. * {"a":.b}' sample.yml
will output
a: {things: great, also: "me"}
b:
also: "me"
Merge arrays
Given a sample.yml file of:
a:
- 1
- 2
- 3
b:
- 3
- 4
- 5
then
yq '. * {"a":.b}' sample.yml
will output
a:
- 3
- 4
- 5
b:
- 3
- 4
- 5
Merge, only existing fields
Given a sample.yml file of:
a:
thing: one
cat: frog
b:
missing: two
thing: two
then
yq '.a *? .b' sample.yml
will output
thing: two
cat: frog
Merge, only new fields
Given a sample.yml file of:
a:
thing: one
cat: frog
b:
missing: two
thing: two
then
yq '.a *n .b' sample.yml
will output
thing: one
cat: frog
missing: two
Merge, appending arrays
Given a sample.yml file of:
a:
array:
- 1
- 2
- animal: dog
value: coconut
b:
array:
- 3
- 4
- animal: cat
value: banana
then
yq '.a *+ .b' sample.yml
will output
array:
- 1
- 2
- animal: dog
- 3
- 4
- animal: cat
value: banana
Merge, only existing fields, appending arrays
Given a sample.yml file of:
a:
thing:
- 1
- 2
b:
thing:
- 3
- 4
another:
- 1
then
yq '.a *?+ .b' sample.yml
will output
thing:
- 1
- 2
- 3
- 4
Merge, deeply merging arrays
Merging arrays deeply means arrays are merge like objects, with indexes as their key. In this case, we merge the first item in the array, and do nothing with the second.
Given a sample.yml file of:
a:
- name: fred
age: 12
- name: bob
age: 32
b:
- name: fred
age: 34
then
yq '.a *d .b' sample.yml
will output
- name: fred
age: 34
- name: bob
age: 32
Merge arrays of objects together, matching on a key
This is a fairly complex expression - you can use it as is by providing the environment variables as seen in the example below.
It merges in the array provided in the second file into the first - matching on equal keys.
Explanation:
The approach, at a high level, is to reduce into a merged map (keyed by the unique key) and then convert that back into an array.
First the expression will create a map from the arrays keyed by the idPath, the unique field we want to merge by. The reduce operator is merging '({}; . * $item )', so array elements with the matching key will be merged together.
Next, we convert the map back to an array, using reduce again, concatenating all the map values together.
Finally, we set the result of the merged array back into the first doc.
Thanks Kev from stackoverflow
Given a sample.yml file of:
myArray:
- a: apple
b: appleB
- a: kiwi
b: kiwiB
- a: banana
b: bananaB
something: else
And another sample another.yml file of:
newArray:
- a: banana
c: bananaC
- a: apple
b: appleB2
- a: dingo
c: dingoC
then
idPath=".a" originalPath=".myArray" otherPath=".newArray" yq eval-all '
(
(( (eval(strenv(originalPath)) + eval(strenv(otherPath))) | .[] | {(eval(strenv(idPath))): .}) as $item ireduce ({}; . * $item )) as $uniqueMap
| ( $uniqueMap | to_entries | .[]) as $item ireduce([]; . + $item.value)
) as $mergedArray
| select(fi == 0) | (eval(strenv(originalPath))) = $mergedArray
' sample.yml another.yml
will output
myArray:
- a: apple
b: appleB2
- a: kiwi
b: kiwiB
- a: banana
b: bananaB
c: bananaC
- a: dingo
c: dingoC
something: else
Merge to prefix an element
Given a sample.yml file of:
a: cat
b: dog
then
yq '. * {"a": {"c": .a}}' sample.yml
will output
a:
c: cat
b: dog
Merge with simple aliases
Given a sample.yml file of:
a: &cat
c: frog
b:
f: *cat
c:
g: thongs
then
yq '.c * .b' sample.yml
will output
g: thongs
f: *cat
Merge copies anchor names
Given a sample.yml file of:
a:
c: &cat frog
b:
f: *cat
c:
g: thongs
then
yq '.c * .a' sample.yml
will output
g: thongs
c: &cat frog
Merge with merge anchors
Given a sample.yml file of:
foo: &foo
a: foo_a
thing: foo_thing
c: foo_c
bar: &bar
b: bar_b
thing: bar_thing
c: bar_c
foobarList:
b: foobarList_b
!!merge <<:
- *foo
- *bar
c: foobarList_c
foobar:
c: foobar_c
!!merge <<: *foo
thing: foobar_thing
then
yq '.foobar * .foobarList' sample.yml
will output
c: foobarList_c
!!merge <<:
- *foo
- *bar
thing: foobar_thing
b: foobarList_b
Custom types: that are really numbers
When custom tags are encountered, yq will try to decode the underlying type.
Given a sample.yml file of:
a: !horse 2
b: !goat 3
then
yq '.a = .a * .b' sample.yml
will output
a: !horse 6
b: !goat 3
Custom types: that are really maps
Custom tags will be maintained.
Given a sample.yml file of:
a: !horse
cat: meow
b: !goat
dog: woof
then
yq '.a = .a * .b' sample.yml
will output
a: !horse
cat: meow
dog: woof
b: !goat
dog: woof