Split printer!

This reverts commit 4ce9433468.

.github/workflows/release.yml

@@ -11,11 +11,37 @@ jobs:
       - uses: actions/checkout@v2
       - uses: actions/setup-go@v2
         with:
-          go-version: '^1.15'
+          go-version: '^1.17.2'
+      - name: Compile man page markup
+        id: gen-man-page-md
+        run: |
+          ./scripts/generate-man-page-md.sh
+          echo "::set-output name=man-page-md::man.md"
+
+      - name: Get the version
+        id: get_version
+        run: echo ::set-output name=VERSION::${GITHUB_REF##*/}
+        
+      - name: Generate man page
+        uses: docker://pandoc/core:2.14.2
+        id: gen-man-page
+        with:
+          args: >-
+            --standalone
+            --to man
+            --variable=title:"YQ"
+            --variable=section:"1"
+            --variable=header:"yq (https://github.com/mikefarah/yq/) version ${{ steps.get_version.outputs.VERSION }}"
+            --variable=author:"Mike Farah"
+            --output=yq.1
+            ${{ steps.gen-man-page-md.outputs.man-page-md }}
+        
       - name: Cross compile
         run: |
           sudo apt-get install rhash -y
           go get github.com/mitchellh/gox
+          mkdir -p build
+          cp yq.1 build/yq.1
           ./scripts/xcompile.sh
 
       - name: Create Release

.github/workflows/test-yq.yml

@@ -0,0 +1,19 @@
+name: Test Yq Action
+on: [push]
+jobs:
+
+  build:
+    name: Build
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v2
+      - name: Get test
+        uses: mikefarah/yq@master
+        with:
+          cmd: yq eval '.a' examples/data1.yaml
+      - name: Write inplace test
+        id: lookupSdkVersion
+        uses: mikefarah/yq@master
+        with:
+          cmd: yq eval -i '.a.b = 5' examples/data1.yaml

.gitignore

@@ -38,3 +38,14 @@ parts/
 prime/
 .snapcraft/
 yq*.snap
+
+test.yml
+test*.yml
+0.yml
+1.yml
+2.yml
+
+
+# man page
+man.md
+yq.1

Dockerfile

@@ -1,18 +1,16 @@
-FROM golang:1.15 as builder
+FROM golang:1.17 as builder
 
 WORKDIR /go/src/mikefarah/yq
 
-# cache devtools
-COPY ./scripts/devtools.sh /go/src/mikefarah/yq/scripts/devtools.sh
-RUN ./scripts/devtools.sh
-
 COPY . /go/src/mikefarah/yq
 
-RUN CGO_ENABLED=0 make local build
+RUN CGO_ENABLED=0 go build .
+RUN ./scripts/test.sh
+RUN ./scripts/acceptance.sh
 
 # Choose alpine as a base image to make this useful for CI, as many
 # CI tools expect an interactive shell inside the container
-FROM alpine:3.13.5 as production
+FROM alpine:3 as production
 
 RUN mkdir /home/yq/
 RUN addgroup -g 1000 yq && \

Dockerfile.dev

@@ -1,4 +1,4 @@
-FROM golang:1.15
+FROM golang:1.17
 
 COPY scripts/devtools.sh /opt/devtools.sh
 

README.md

@@ -7,12 +7,43 @@ a lightweight and portable command-line YAML processor. `yq` uses [jq](https://g
 
 yq is written in go - so you can download a dependency free binary for your platform and you are good to go! If you prefer there are a variety of package managers that can be used as well as docker, all listed below.
 
-## V4 released!
-V4 is now officially released, it's quite different from V3 (sorry for the migration), however it is much more similar to ```jq```, using a similar expression syntax and therefore support much more complex functionality! 
+## Quick Usage Guide
 
-If you've been using v3 and want/need to upgrade, checkout the [upgrade guide](https://mikefarah.gitbook.io/yq/v/v4.x/upgrading-from-v3).
+Read a value:
+```bash
+yq e '.a.b[0].c' file.yaml
+```
 
-Support for v3 will cease August 2021, until then, critical bug and security fixes will still get applied if required.
+Pipe from STDIN:
+```bash
+cat file.yaml | yq e '.a.b[0].c' -
+```
+
+Update a yaml file, inplace
+```bash
+yq e -i '.a.b[0].c = "cool"' file.yaml
+```
+
+Update using environment variables
+```bash
+NAME=mike yq e -i '.a.b[0].c = strenv(NAME)' file.yaml
+```
+
+Merge multiple files
+```
+yq ea '. as $item ireduce ({}; . * $item )' path/to/*.yml
+```
+
+Multiple updates to a yaml file
+```bash
+yq e -i '
+  .a.b[0].c = "cool" |
+  .x.y.z = "foobar" |
+  .person.name = strenv(NAME)
+' file.yaml
+```
+
+See the [documentation](https://mikefarah.gitbook.io/yq/) for more.
 
 ## Install
 
@@ -87,6 +118,14 @@ rm /etc/myfile.tmp
 docker run --rm -v "${PWD}":/workdir mikefarah/yq <command> [flags] [expression ]FILE...
 ```
 
+#### Pipe in via STDIN:
+
+You'll need to pass the `-i\--interactive` flag to docker:
+
+```bash
+cat myfile.yml | docker run -i --rm mikefarah/yq e . -
+```
+
 #### Run commands interactively:
 
 ```bash
@@ -101,9 +140,10 @@ yq() {
 }
 ```
 
+
 #### Running as root:
 
-`yq`'s docker image no longer runs under root (https://github.com/mikefarah/yq/pull/860) - so if you'd like to install more things in the docker image you'll need to either:
+`yq`'s docker image no longer runs under root (https://github.com/mikefarah/yq/pull/860). If you'd like to install more things in the docker image, or you're having permissions issues when attempting to read/write files you'll need to either:
 
 
 ```
@@ -120,6 +160,16 @@ RUN apk add bash
 USER yq
 ```
 
+### GitHub Action
+```
+  - name: Set foobar to cool
+    uses: mikefarah/yq@master
+    with:
+      cmd: yq eval -i '.foo.bar = "cool"' 'config.yml'
+```
+
+See https://mikefarah.gitbook.io/yq/usage/github-action for more.
+
 ### Go Get:
 ```
 GO111MODULE=on go get github.com/mikefarah/yq/v4
@@ -176,6 +226,7 @@ Supported by @rmescandon (https://launchpad.net/~rmescandon/+archive/ubuntu/yq)
 - Written in portable go, so you can download a lovely dependency free binary
 - Uses similar syntax as `jq` but works with YAML and JSON files
 - Fully supports multi document yaml files
+- Supports yaml [front matter](https://mikefarah.gitbook.io/yq/usage/front-matter) blocks (e.g. jekyll/assemble)
 - Colorized yaml output
 - [Deeply traverse yaml](https://mikefarah.gitbook.io/yq/operators/traverse-read)
 - [Sort yaml by keys](https://mikefarah.gitbook.io/yq/operators/sort-keys)
@@ -183,7 +234,8 @@ Supported by @rmescandon (https://launchpad.net/~rmescandon/+archive/ubuntu/yq)
 - [Update yaml inplace](https://mikefarah.gitbook.io/yq/v/v4.x/commands/evaluate#flags)
 - [Complex expressions to select and update](https://mikefarah.gitbook.io/yq/operators/select#select-and-update-matching-values-in-map)
 - Keeps yaml formatting and comments when updating (though there are issues with whitespace)
-- [Convert to/from json to yaml](https://mikefarah.gitbook.io/yq/v/v4.x/usage/convert)
+- [Convert to/from json](https://mikefarah.gitbook.io/yq/v/v4.x/usage/convert)
+- [Convert to properties](https://mikefarah.gitbook.io/yq/v/v4.x/usage/properties)
 - [Pipe data in by using '-'](https://mikefarah.gitbook.io/yq/v/v4.x/commands/evaluate)
 - [General shell completion scripts (bash/zsh/fish/powershell)](https://mikefarah.gitbook.io/yq/v/v4.x/commands/shell-completion)
 - [Reduce](https://mikefarah.gitbook.io/yq/operators/reduce) to merge multiple files or sum an array or other fancy things.
@@ -199,34 +251,32 @@ Usage:
   yq [command]
 
 Available Commands:
-  eval             Apply expression to each document in each yaml file given in sequence
+  eval             Apply the expression to each document in each yaml file in sequence
   eval-all         Loads _all_ yaml documents of _all_ yaml files and runs expression once
   help             Help about any command
   shell-completion Generate completion script
 
 Flags:
-  -C, --colors        force print with colors
-  -e, --exit-status   set exit status if there are no matches or null or false is returned
-  -h, --help          help for yq
-  -I, --indent int    sets indent level for output (default 2)
-  -i, --inplace       update the yaml file inplace of first yaml file given.
-  -M, --no-colors     force print with no colors
-  -N, --no-doc        Don't print document separators (---)
-  -n, --null-input    Don't read input, simply evaluate the expression given. Useful for creating yaml docs from scratch.
-  -P, --prettyPrint    pretty print, shorthand for '... style = ""' 
-  -j, --tojson        output as json. Set indent to 0 to print json in one line.
-  -v, --verbose       verbose mode
-  -V, --version       Print version information and quit
+  -C, --colors                force print with colors
+  -e, --exit-status           set exit status if there are no matches or null or false is returned
+  -f, --front-matter string   (extract|process) first input as yaml front-matter. Extract will pull out the yaml content, process will run the expression against the yaml content, leaving the remaining data intact
+      --header-preprocess     Slurp any header comments and seperators before processing expression. This is a workaround for go-yaml to persist header content (default true)
+  -h, --help                  help for yq
+  -I, --indent int            sets indent level for output (default 2)
+  -i, --inplace               update the yaml file inplace of first yaml file given.
+  -M, --no-colors             force print with no colors
+  -N, --no-doc                Don't print document separators (---)
+  -n, --null-input            Don't read input, simply evaluate the expression given. Useful for creating yaml docs from scratch.
+  -o, --output-format string  [yaml|y|json|j|props|p] output format type. (default "yaml")
+  -P, --prettyPrint           pretty print, shorthand for '... style = ""'
+      --unwrapScalar          unwrap scalar, print the value with no quotes, colors or comments (default true)
+  -v, --verbose               verbose mode
+  -V, --version               Print version information and quit
 
 Use "yq [command] --help" for more information about a command.
 ```
-
-Simple Example:
-
-```bash
-yq e '.a.b | length' f1.yml f2.yml 
-```
-
 ## Known Issues / Missing Features
 - `yq` attempts to preserve comment positions and whitespace as much as possible, but it does not handle all scenarios (see https://github.com/go-yaml/yaml/tree/v3 for details)
-- Powershell has its own...opinions: https://mikefarah.gitbook.io/yq/usage/tips-and-tricks#quotes-in-windows-powershell
+- Powershell has its own...opinions: https://mikefarah.gitbook.io/yq/usage/tips-and-tricks#quotes-in-windows-powershell
+
+See [tips and tricks](https://mikefarah.gitbook.io/yq/usage/tips-and-tricks) for more common problems and solutions.

acceptance_tests/basic.sh

@@ -0,0 +1,72 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+}
+
+testBasicEvalRoundTrip() {
+  ./yq e -n ".a = 123" > test.yml
+  X=$(./yq e '.a' test.yml)
+  assertEquals 123 "$X"
+}
+
+testBasicUpdateInPlaceSequence() {
+  cat >test.yml <<EOL
+a: 0
+EOL
+  ./yq e -i ".a = 10" test.yml
+  X=$(./yq e '.a' test.yml)
+  assertEquals "10" "$X"
+}
+
+testBasicUpdateInPlaceSequenceEvalAll() {
+  cat >test.yml <<EOL
+a: 0
+EOL
+  ./yq ea -i ".a = 10" test.yml
+  X=$(./yq e '.a' test.yml)
+  assertEquals "10" "$X"
+}
+
+testBasicNoExitStatus() {
+  echo "a: cat" > test.yml
+  X=$(./yq e '.z' test.yml)
+  assertEquals "null" "$X"
+}
+
+testBasicExitStatus() {
+  echo "a: cat" > test.yml
+  X=$(./yq e -e '.z' test.yml 2&>/dev/null)
+  assertEquals 1 "$?"
+}
+
+testBasicExtractFieldWithSeperator() {
+    cat >test.yml <<EOL
+---
+name: chart-name
+version: 1.2.3
+EOL
+  X=$(./yq e '.name' test.yml)
+  assertEquals "chart-name" "$X"
+}
+
+testBasicExtractMultipleFieldWithSeperator() {
+    cat >test.yml <<EOL
+---
+name: chart-name
+version: 1.2.3
+---
+name: thing
+version: 1.2.3
+EOL
+
+read -r -d '' expected << EOM
+chart-name
+---
+thing
+EOM
+  X=$(./yq e '.name' test.yml)
+  assertEquals "$expected" "$X"
+}
+
+source ./scripts/shunit2

acceptance_tests/empty.sh

@@ -0,0 +1,83 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+  cat >test.yml <<EOL
+# comment
+EOL
+}
+
+testEmptyEval() {
+  X=$(./yq e test.yml)
+  expected=$(cat test.yml)
+  assertEquals 0 $?
+  assertEquals "$expected" "$X"
+}
+
+testEmptyEvalNoNewLine() {
+  echo -n "#comment" >test.yml
+  X=$(./yq e test.yml)
+  expected=$(cat test.yml)
+  assertEquals 0 $?
+  assertEquals "$expected" "$X"
+}
+
+testEmptyEvalNoNewLineWithExpression() {
+  echo -n "# comment" >test.yml
+  X=$(./yq e '.apple = "tree"' test.yml)
+  read -r -d '' expected << EOM
+# comment
+apple: tree
+EOM
+  assertEquals "$expected" "$X"
+}
+
+testEmptyEvalPipe() {
+  X=$(./yq e - < test.yml)
+  assertEquals 0 $?
+}
+
+testEmptyCommentsWithExpressionEval() {
+  read -r -d '' expected << EOM
+# comment
+apple: tree
+EOM
+
+  X=$(./yq e '.apple="tree"' test.yml)
+
+  assertEquals "$expected" "$X"
+}
+
+testEmptyCommentsWithExpressionEvalAll() {
+  read -r -d '' expected << EOM
+# comment
+apple: tree
+EOM
+
+  X=$(./yq ea '.apple="tree"' test.yml)
+
+  assertEquals "$expected" "$X"
+}
+
+testEmptyWithExpressionEval() {
+  rm test.yml
+  touch test.yml
+  expected="apple: tree"
+
+  X=$(./yq e '.apple="tree"' test.yml)
+
+  assertEquals "$expected" "$X"
+}
+
+testEmptyWithExpressionEvalAll() {
+  rm test.yml
+  touch test.yml
+  expected="apple: tree"
+
+  X=$(./yq ea '.apple="tree"' test.yml)
+
+  assertEquals "$expected" "$X"
+}
+
+
+source ./scripts/shunit2

acceptance_tests/front-matter.sh

@@ -0,0 +1,76 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+  cat >test.yml <<EOL
+---
+a: apple
+b: cat
+---
+not yaml
+c: at
+EOL
+}
+
+testFrontMatterProcessEval() {
+  read -r -d '' expected << EOM
+---
+a: apple
+b: dog
+---
+not yaml
+c: at
+EOM
+  ./yq e --front-matter="process" '.b = "dog"' test.yml -i
+  assertEquals "$expected" "$(cat test.yml)"
+}
+
+testFrontMatterProcessEvalAll() {
+  read -r -d '' expected << EOM
+---
+a: apple
+b: dog
+---
+not yaml
+c: at
+EOM
+  ./yq ea --front-matter="process" '.b = "dog"' test.yml -i
+  assertEquals "$expected" "$(cat test.yml)"
+}
+
+testFrontMatterExtractEval() {
+    cat >test.yml <<EOL
+a: apple
+b: cat
+---
+not yaml
+c: at
+EOL
+
+  read -r -d '' expected << EOM
+a: apple
+b: dog
+EOM
+  ./yq e --front-matter="extract" '.b = "dog"' test.yml -i
+  assertEquals "$expected" "$(cat test.yml)"
+}
+
+testFrontMatterExtractEvalAll() {
+  cat >test.yml <<EOL
+a: apple
+b: cat
+---
+not yaml
+c: at
+EOL
+
+  read -r -d '' expected << EOM
+a: apple
+b: dog
+EOM
+  ./yq ea --front-matter="extract" '.b = "dog"' test.yml -i
+  assertEquals "$expected" "$(cat test.yml)"
+}
+
+
+source ./scripts/shunit2

acceptance_tests/leading-seperator.sh

@@ -0,0 +1,450 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+  cat >test.yml <<EOL
+---
+a: test
+EOL
+}
+
+testLeadingSeperatorWithDoc() {
+  cat >test.yml <<EOL
+# hi peeps
+# cool
+---
+a: test
+---
+b: cool
+EOL
+
+  read -r -d '' expected << EOM
+# hi peeps
+# cool
+---
+a: thing
+---
+b: cool
+EOM
+
+  X=$(./yq e '(select(di == 0) | .a) = "thing"' - < test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorPipeIntoEvalSeq() {
+  X=$(./yq e - < test.yml)
+  expected=$(cat test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorExtractField() {
+  X=$(./yq e '.a' - < test.yml)
+  assertEquals "test" "$X"
+}
+
+testLeadingSeperatorExtractFieldWithCommentsAfterSep() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+EOL
+  X=$(./yq e '.a' test.yml)
+  assertEquals "test" "$X"
+}
+
+testLeadingSeperatorExtractFieldWithCommentsBeforeSep() {
+  cat >test.yml <<EOL
+# hi peeps
+# cool
+---
+a: test
+EOL
+  X=$(./yq e '.a' test.yml)
+  assertEquals "test" "$X"
+}
+
+
+testLeadingSeperatorExtractFieldMultiDoc() {
+  cat >test.yml <<EOL
+---
+a: test
+---
+a: test2
+EOL
+
+  read -r -d '' expected << EOM
+test
+---
+test2
+EOM
+  X=$(./yq e '.a' test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorExtractFieldMultiDocWithComments() {
+  cat >test.yml <<EOL
+# here
+---
+# there
+a: test
+# whereever
+---
+# you are
+a: test2
+# woop
+EOL
+
+  read -r -d '' expected << EOM
+test
+---
+test2
+EOM
+  X=$(./yq e '.a' test.yml)
+  assertEquals "$expected" "$X"
+}
+
+
+testLeadingSeperatorEvalSeq() {
+  X=$(./yq e test.yml)
+  expected=$(cat test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorPipeIntoEvalAll() {
+  X=$(./yq ea - < test.yml)
+  expected=$(cat test.yml)
+  assertEquals "$expected" "$X"
+}
+
+
+testLeadingSeperatorEvalAll() {
+  X=$(./yq ea test.yml)
+  expected=$(cat test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalSimple() {
+  read -r -d '' expected << EOM
+---
+a: test
+---
+version: 3
+application: MyApp
+EOM
+
+
+  X=$(./yq e '.' test.yml examples/order.yaml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocInOneFile() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+---
+b: things
+EOL
+  expected=$(cat test.yml)
+  X=$(./yq e '.' test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocInOneFileEvalAll() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+---
+b: things
+EOL
+  expected=$(cat test.yml)
+  X=$(./yq ea '.' test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalComments() {
+  cat >test.yml <<EOL
+# hi peeps
+# cool
+a: test
+EOL
+
+cat >test2.yml <<EOL
+# this is another doc
+# great
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+# hi peeps
+# cool
+a: test
+---
+# this is another doc
+# great
+b: sane
+EOM
+
+
+  X=$(./yq e '.' test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalCommentsTrailingSep() {
+  cat >test.yml <<EOL
+# hi peeps
+# cool
+---
+a: test
+EOL
+
+cat >test2.yml <<EOL
+# this is another doc
+# great
+---
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+# hi peeps
+# cool
+---
+a: test
+---
+# this is another doc
+# great
+---
+b: sane
+EOM
+
+
+  X=$(./yq e '.' test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiMultiDocEvalCommentsTrailingSep() {
+  cat >test.yml <<EOL
+# hi peeps
+# cool
+---
+a: test
+---
+a1: test2
+EOL
+
+cat >test2.yml <<EOL
+# this is another doc
+# great
+---
+b: sane
+---
+b2: cool
+EOL
+
+  read -r -d '' expected << EOM
+# hi peeps
+# cool
+---
+a: test
+---
+a1: test2
+---
+# this is another doc
+# great
+---
+b: sane
+---
+b2: cool
+EOM
+
+
+  X=$(./yq e '.' test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalCommentsLeadingSep() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+EOL
+
+cat >test2.yml <<EOL
+---
+# this is another doc
+# great
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+---
+# hi peeps
+# cool
+a: test
+---
+# this is another doc
+# great
+b: sane
+EOM
+
+
+  X=$(./yq e '.' test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+# https://github.com/mikefarah/yq/issues/919
+testLeadingSeparatorDoesNotBreakCommentsOnOtherFiles() {
+  cat >test.yml <<EOL
+# a1
+a: 1
+# a2
+EOL
+
+cat >test2.yml <<EOL
+# b1
+b: 2
+# b2
+EOL
+
+  read -r -d '' expected << EOM
+# a1
+a: 1
+# a2
+
+# b1
+b: 2
+# b2
+EOM
+
+
+  X=$(./yq ea 'select(fi == 0) * select(fi == 1)' test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalCommentsStripComments() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+---
+# this is another doc
+# great
+b: sane
+EOL
+
+  # it will be hard to remove that top level separator
+  read -r -d '' expected << EOM
+a: test
+---
+b: sane
+EOM
+
+  X=$(./yq e '... comments=""'  test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalCommentsLeadingSepNoDocFlag() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+---
+# this is another doc
+# great
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+# hi peeps
+# cool
+a: test
+# this is another doc
+# great
+b: sane
+EOM
+
+
+  X=$(./yq e '.' --no-doc test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalJsonFlag() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+EOL
+
+cat >test2.yml <<EOL
+---
+# this is another doc
+# great
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+{
+  "a": "test"
+}
+{
+  "b": "sane"
+}
+EOM
+
+
+  X=$(./yq e '.' -j test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalAllJsonFlag() {
+  cat >test.yml <<EOL
+---
+# hi peeps
+# cool
+a: test
+EOL
+
+cat >test2.yml <<EOL
+---
+# this is another doc
+# great
+b: sane
+EOL
+
+  read -r -d '' expected << EOM
+{
+  "a": "test"
+}
+{
+  "b": "sane"
+}
+EOM
+
+
+  X=$(./yq ea '.' -j test.yml test2.yml)
+  assertEquals "$expected" "$X"
+}
+
+testLeadingSeperatorMultiDocEvalAll() {
+  read -r -d '' expected << EOM
+---
+a: test
+---
+version: 3
+application: MyApp
+EOM
+
+
+  X=$(./yq ea '.' test.yml examples/order.yaml)
+  assertEquals "$expected" "$X"
+}
+
+source ./scripts/shunit2

acceptance_tests/output-format.sh

@@ -0,0 +1,105 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+}
+
+testOutputJsonDeprecated() {
+  cat >test.yml <<EOL
+a: {b: ["cat"]}
+EOL
+
+  read -r -d '' expected << EOM
+{
+  "a": {
+    "b": [
+      "cat"
+    ]
+  }
+}
+EOM
+
+  X=$(./yq e -j test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea -j test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testOutputJson() {
+  cat >test.yml <<EOL
+a: {b: ["cat"]}
+EOL
+
+  read -r -d '' expected << EOM
+{
+  "a": {
+    "b": [
+      "cat"
+    ]
+  }
+}
+EOM
+
+  X=$(./yq e --output-format=json test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --output-format=json test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testOutputJsonShort() {
+  cat >test.yml <<EOL
+a: {b: ["cat"]}
+EOL
+
+  read -r -d '' expected << EOM
+{
+  "a": {
+    "b": [
+      "cat"
+    ]
+  }
+}
+EOM
+
+  X=$(./yq e -o=j test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea -o=j test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testOutputProperties() {
+  cat >test.yml <<EOL
+a: {b: {c: ["cat"]}}
+EOL
+
+  read -r -d '' expected << EOM
+a.b.c.0 = cat
+EOM
+
+  X=$(./yq e --output-format=props test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --output-format=props test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testOutputPropertiesShort() {
+  cat >test.yml <<EOL
+a: {b: {c: ["cat"]}}
+EOL
+
+  read -r -d '' expected << EOM
+a.b.c.0 = cat
+EOM
+
+  X=$(./yq e -o=p test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea -o=p test.yml)
+  assertEquals "$expected" "$X"
+}
+
+source ./scripts/shunit2

acceptance_tests/pretty-print.sh

@@ -0,0 +1,176 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+}
+
+testPrettyPrintWithBooleans() {
+  cat >test.yml <<EOL
+leaveUnquoted: [yes, no, on, off, y, n, true, false]
+leaveQuoted: ["yes", "no", "on", "off", "y", "n", "true", "false"]
+
+EOL
+
+  read -r -d '' expected << EOM
+leaveUnquoted:
+  - yes
+  - no
+  - on
+  - off
+  - y
+  - n
+  - true
+  - false
+leaveQuoted:
+  - "yes"
+  - "no"
+  - "on"
+  - "off"
+  - "y"
+  - "n"
+  - "true"
+  - "false"
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintWithBooleansCapitals() {
+  cat >test.yml <<EOL
+leaveUnquoted: [YES, NO, ON, OFF, Y, N, TRUE, FALSE]
+leaveQuoted: ["YES", "NO", "ON", "OFF", "Y", "N", "TRUE", "FALSE"]
+EOL
+
+  read -r -d '' expected << EOM
+leaveUnquoted:
+  - YES
+  - NO
+  - ON
+  - OFF
+  - Y
+  - N
+  - TRUE
+  - FALSE
+leaveQuoted:
+  - "YES"
+  - "NO"
+  - "ON"
+  - "OFF"
+  - "Y"
+  - "N"
+  - "TRUE"
+  - "FALSE"
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintOtherStringValues() {
+  cat >test.yml <<EOL
+leaveUnquoted: [yesSir, hellno, bonapite]
+makeUnquoted: ["yesSir", "hellno", "bonapite"]
+EOL
+
+  read -r -d '' expected << EOM
+leaveUnquoted:
+  - yesSir
+  - hellno
+  - bonapite
+makeUnquoted:
+  - yesSir
+  - hellno
+  - bonapite
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintKeys() {
+  cat >test.yml <<EOL
+"removeQuotes": "please"
+EOL
+
+  read -r -d '' expected << EOM
+removeQuotes: please
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintOtherStringValues() {
+  cat >test.yml <<EOL
+leaveUnquoted: [yesSir, hellno, bonapite]
+makeUnquoted: ["yesSir", "hellno", "bonapite"]
+EOL
+
+  read -r -d '' expected << EOM
+leaveUnquoted:
+  - yesSir
+  - hellno
+  - bonapite
+makeUnquoted:
+  - yesSir
+  - hellno
+  - bonapite
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintStringBlocks() {
+  cat >test.yml <<EOL
+"removeQuotes": | 
+  "please"
+EOL
+
+  read -r -d '' expected << EOM
+removeQuotes: |
+  "please"
+EOM
+
+  X=$(./yq e --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+testPrettyPrintWithExpression() {
+  cat >test.yml <<EOL
+a: {b: {c: ["cat"]}}
+EOL
+
+  read -r -d '' expected << EOM
+b:
+  c:
+    - cat
+EOM
+
+  X=$(./yq e '.a' --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+
+  X=$(./yq ea '.a' --prettyPrint test.yml)
+  assertEquals "$expected" "$X"
+}
+
+source ./scripts/shunit2

acceptance_tests/split-printer.sh

@@ -0,0 +1,157 @@
+#!/bin/bash
+
+setUp() {
+  rm test*.yml || true
+}
+
+testBasicSplitWithName() {
+  cat >test.yml <<EOL
+a: test_doc1
+--- 
+a: test_doc2
+EOL
+
+  ./yq e test.yml -s ".a"
+
+  doc1=$(cat test_doc1.yml)
+  
+  assertEquals "a: test_doc1" "$doc1"
+
+  doc2=$(cat test_doc2.yml)
+  read -r -d '' expectedDoc2 << EOM
+---
+a: test_doc2
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+}
+
+testBasicSplitWithNameEvalAll() {
+  cat >test.yml <<EOL
+a: test_doc1
+--- 
+a: test_doc2
+EOL
+
+  ./yq ea test.yml -s ".a"
+
+  doc1=$(cat test_doc1.yml)
+  
+  assertEquals "a: test_doc1" "$doc1"
+
+  doc2=$(cat test_doc2.yml)
+  read -r -d '' expectedDoc2 << EOM
+---
+a: test_doc2
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+}
+
+testBasicSplitWithIndex() {
+  cat >test.yml <<EOL
+a: test_doc1
+--- 
+a: test_doc2
+EOL
+
+  ./yq e test.yml -s '"test_" + $index'
+
+  doc1=$(cat test_0.yml)
+  
+  assertEquals "a: test_doc1" "$doc1"
+
+  doc2=$(cat test_1.yml)
+  read -r -d '' expectedDoc2 << EOM
+---
+a: test_doc2
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+}
+
+testBasicSplitWithIndexEvalAll() {
+  cat >test.yml <<EOL
+a: test_doc1
+--- 
+a: test_doc2
+EOL
+
+  ./yq ea test.yml -s '"test_" + $index'
+
+  doc1=$(cat test_0.yml)
+  
+  assertEquals "a: test_doc1" "$doc1"
+
+  doc2=$(cat test_1.yml)
+  read -r -d '' expectedDoc2 << EOM
+---
+a: test_doc2
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+}
+
+
+testArraySplitWithNameNoSeparators() {
+  cat >test.yml <<EOL
+- name: test_fred
+  age: 35
+- name: test_catherine
+  age: 37
+EOL
+
+  ./yq e --no-doc -s ".name"  ".[]" test.yml 
+
+  doc1=$(cat test_fred.yml)
+  read -r -d '' expectedDoc1 << EOM
+name: test_fred
+age: 35
+EOM
+
+  assertEquals "$expectedDoc1" "$doc1"
+
+  doc2=$(cat test_catherine.yml)
+  read -r -d '' expectedDoc2 << EOM
+name: test_catherine
+age: 37
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+}
+
+testArraySplitWithNameNoSeparatorsEvalAll() {
+  cat >test.yml <<EOL
+- name: test_fred
+  age: 35
+- name: test_catherine
+  age: 37
+EOL
+
+cat >test2.yml <<EOL
+- name: test_mike
+  age: 564
+EOL
+
+  ./yq ea --no-doc -s ".name"  ".[]" test.yml test2.yml
+
+  doc1=$(cat test_fred.yml)
+  read -r -d '' expectedDoc1 << EOM
+name: test_fred
+age: 35
+EOM
+
+  assertEquals "$expectedDoc1" "$doc1"
+
+  doc2=$(cat test_catherine.yml)
+  read -r -d '' expectedDoc2 << EOM
+name: test_catherine
+age: 37
+EOM
+  assertEquals "$expectedDoc2" "$doc2"
+
+
+  doc3=$(cat test_mike.yml)
+  read -r -d '' expectedDoc3 << EOM
+name: test_mike
+age: 564
+EOM
+  assertEquals "$expectedDoc3" "$doc3"
+}
+
+source ./scripts/shunit2

cmd/constant.go

@@ -1,9 +1,11 @@
 package cmd
 
+var leadingContentPreProcessing = true
 var unwrapScalar = true
 
 var writeInplace = false
 var outputToJSON = false
+var outputFormat = "yaml"
 
 var exitStatus = false
 var forceColor = false
@@ -16,4 +18,9 @@ var verbose = false
 var version = false
 var prettyPrint = false
 
+// can be either "" (off), "extract" or "process"
+var frontMatter = ""
+
+var splitFileExp = ""
+
 var completedSuccessfully = false

cmd/evaluate_all_command.go

@@ -15,13 +15,27 @@ func createEvaluateAllCommand() *cobra.Command {
 		Aliases: []string{"ea"},
 		Short:   "Loads _all_ yaml documents of _all_ yaml files and runs expression once",
 		Example: `
-# merges f2.yml into f1.yml (inplace)
-yq eval-all --inplace 'select(fileIndex == 0) * select(fileIndex == 1)' f1.yml f2.yml		
+# Merge f2.yml into f1.yml (inplace)
+yq eval-all --inplace 'select(fileIndex == 0) * select(fileIndex == 1)' f1.yml f2.yml
+## the same command and expression using shortened names:
+yq ea -i 'select(fi == 0) * select(fi == 1)' f1.yml f2.yml
 
-# use '-' as a filename to read from STDIN
+
+# Merge all given files
+yq ea '. as $item ireduce ({}; . * $item )' file1.yml file2.yml ...
+
+# Pipe from STDIN
+## use '-' as a filename to pipe from STDIN
 cat file2.yml | yq ea '.a.b' file1.yml - file3.yml
 `,
-		Long: "Evaluate All:\nUseful when you need to run an expression across several yaml documents or files. Consumes more memory than eval",
+		Long: `yq is a portable command-line YAML processor (https://github.com/mikefarah/yq/) 
+See https://mikefarah.gitbook.io/yq/ for detailed documentation and examples.
+
+## Evaluate All ##
+This command loads _all_ yaml documents of _all_ yaml files and runs expression once
+Useful when you need to run an expression across several yaml documents or files (like merge).
+Note that it consumes more memory than eval.
+`,
 		RunE: evaluateAll,
 	}
 	return cmdEvalAll
@@ -56,6 +70,10 @@ func evaluateAll(cmd *cobra.Command, args []string) error {
 		return fmt.Errorf("Write inplace flag only applicable when giving an expression and at least one file")
 	}
 
+	if writeInplace && splitFileExp != "" {
+		return fmt.Errorf("Write inplace cannot be used with split file")
+	}
+
 	if writeInplace {
 		// only use colors if its forced
 		colorsEnabled = forceColor
@@ -72,26 +90,53 @@ func evaluateAll(cmd *cobra.Command, args []string) error {
 	if nullInput && len(args) > 1 {
 		return errors.New("Cannot pass files in when using null-input flag")
 	}
+	// backwards compatibilty
+	if outputToJSON {
+		outputFormat = "json"
+	}
 
-	printer := yqlib.NewPrinter(out, outputToJSON, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
+	format, err := yqlib.OutputFormatFromString(outputFormat)
+	if err != nil {
+		return err
+	}
+
+	printerWriter := configurePrinterWriter(format, out)
+
+	printer := yqlib.NewPrinter(printerWriter, format, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
+
+	if frontMatter != "" {
+		frontMatterHandler := yqlib.NewFrontMatterHandler(args[firstFileIndex])
+		err = frontMatterHandler.Split()
+		if err != nil {
+			return err
+		}
+		args[firstFileIndex] = frontMatterHandler.GetYamlFrontMatterFilename()
+
+		if frontMatter == "process" {
+			reader := frontMatterHandler.GetContentReader()
+			printer.SetAppendix(reader)
+			defer yqlib.SafelyCloseReader(reader)
+		}
+		defer frontMatterHandler.CleanUp()
+	}
 
 	allAtOnceEvaluator := yqlib.NewAllAtOnceEvaluator()
 	switch len(args) {
 	case 0:
 		if pipingStdIn {
-			err = allAtOnceEvaluator.EvaluateFiles(processExpression(""), []string{"-"}, printer)
+			err = allAtOnceEvaluator.EvaluateFiles(processExpression(""), []string{"-"}, printer, leadingContentPreProcessing)
 		} else {
 			cmd.Println(cmd.UsageString())
 			return nil
 		}
 	case 1:
 		if nullInput {
-			err = yqlib.NewStreamEvaluator().EvaluateNew(processExpression(args[0]), printer)
+			err = yqlib.NewStreamEvaluator().EvaluateNew(processExpression(args[0]), printer, "")
 		} else {
-			err = allAtOnceEvaluator.EvaluateFiles(processExpression(""), []string{args[0]}, printer)
+			err = allAtOnceEvaluator.EvaluateFiles(processExpression(""), []string{args[0]}, printer, leadingContentPreProcessing)
 		}
 	default:
-		err = allAtOnceEvaluator.EvaluateFiles(processExpression(args[0]), args[1:], printer)
+		err = allAtOnceEvaluator.EvaluateFiles(processExpression(args[0]), args[1:], printer, leadingContentPreProcessing)
 	}
 
 	completedSuccessfully = err == nil

cmd/evalute_sequence_command.go

@@ -13,35 +13,42 @@ func createEvaluateSequenceCommand() *cobra.Command {
 	var cmdEvalSequence = &cobra.Command{
 		Use:     "eval [expression] [yaml_file1]...",
 		Aliases: []string{"e"},
-		Short:   "Apply expression to each document in each yaml file given in sequence",
+		Short:   "Apply the expression to each document in each yaml file in sequence",
 		Example: `
-# runs the expression against each file, in series
-yq e '.a.b | length' f1.yml f2.yml 
+# Reads field under the given path for each file
+yq e '.a.b' f1.yml f2.yml 
 
-# prints out the file
+# Prints out the file
 yq e sample.yaml 
 
-# use '-' as a filename to read from STDIN
+# Pipe from STDIN
+## use '-' as a filename to pipe from STDIN
 cat file2.yml | yq e '.a.b' file1.yml - file3.yml
 
-# prints a new yaml document
+# Creates a new yaml document
+## Note that editing an empty file does not work.
 yq e -n '.a.b.c = "cat"' 
 
-
-# updates file.yaml directly
+# Update a file inplace
 yq e '.a.b = "cool"' -i file.yaml 
 `,
-		Long: "Evaluate Sequence:\nIterate over each yaml document, apply the expression and print the results, in sequence.",
+		Long: `yq is a portable command-line YAML processor (https://github.com/mikefarah/yq/) 
+See https://mikefarah.gitbook.io/yq/ for detailed documentation and examples.
+
+## Evaluate Sequence ##
+This command iterates over each yaml document from each given file, applies the 
+expression and prints the result in sequence.`,
 		RunE: evaluateSequence,
 	}
 	return cmdEvalSequence
 }
 
 func processExpression(expression string) string {
+	var prettyPrintExp = `(... | (select(tag != "!!str"), select(tag == "!!str") | select(test("(?i)^(y|yes|n|no|on|off)$") | not))  ) style=""`
 	if prettyPrint && expression == "" {
-		return `... style=""`
+		return prettyPrintExp
 	} else if prettyPrint {
-		return fmt.Sprintf("%v | ... style= \"\"", expression)
+		return fmt.Sprintf("%v | %v", expression, prettyPrintExp)
 	}
 	return expression
 }
@@ -76,6 +83,10 @@ func evaluateSequence(cmd *cobra.Command, args []string) error {
 		return fmt.Errorf("Write inplace flag only applicable when giving an expression and at least one file")
 	}
 
+	if writeInplace && splitFileExp != "" {
+		return fmt.Errorf("Write inplace cannot be used with split file")
+	}
+
 	if writeInplace {
 		// only use colors if its forced
 		colorsEnabled = forceColor
@@ -89,7 +100,19 @@ func evaluateSequence(cmd *cobra.Command, args []string) error {
 		defer func() { writeInPlaceHandler.FinishWriteInPlace(completedSuccessfully) }()
 	}
 
-	printer := yqlib.NewPrinter(out, outputToJSON, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
+	// backwards compatibilty
+	if outputToJSON {
+		outputFormat = "json"
+	}
+
+	format, err := yqlib.OutputFormatFromString(outputFormat)
+	if err != nil {
+		return err
+	}
+
+	printerWriter := configurePrinterWriter(format, out)
+
+	printer := yqlib.NewPrinter(printerWriter, format, unwrapScalar, colorsEnabled, indent, !noDocSeparators)
 
 	streamEvaluator := yqlib.NewStreamEvaluator()
 
@@ -97,22 +120,38 @@ func evaluateSequence(cmd *cobra.Command, args []string) error {
 		return errors.New("Cannot pass files in when using null-input flag")
 	}
 
+	if frontMatter != "" {
+		frontMatterHandler := yqlib.NewFrontMatterHandler(args[firstFileIndex])
+		err = frontMatterHandler.Split()
+		if err != nil {
+			return err
+		}
+		args[firstFileIndex] = frontMatterHandler.GetYamlFrontMatterFilename()
+
+		if frontMatter == "process" {
+			reader := frontMatterHandler.GetContentReader()
+			printer.SetAppendix(reader)
+			defer yqlib.SafelyCloseReader(reader)
+		}
+		defer frontMatterHandler.CleanUp()
+	}
+
 	switch len(args) {
 	case 0:
 		if pipingStdIn {
-			err = streamEvaluator.EvaluateFiles(processExpression(""), []string{"-"}, printer)
+			err = streamEvaluator.EvaluateFiles(processExpression(""), []string{"-"}, printer, leadingContentPreProcessing)
 		} else {
 			cmd.Println(cmd.UsageString())
 			return nil
 		}
 	case 1:
 		if nullInput {
-			err = streamEvaluator.EvaluateNew(processExpression(args[0]), printer)
+			err = streamEvaluator.EvaluateNew(processExpression(args[0]), printer, "")
 		} else {
-			err = streamEvaluator.EvaluateFiles(processExpression(""), []string{args[0]}, printer)
+			err = streamEvaluator.EvaluateFiles(processExpression(""), []string{args[0]}, printer, leadingContentPreProcessing)
 		}
 	default:
-		err = streamEvaluator.EvaluateFiles(processExpression(args[0]), args[1:], printer)
+		err = streamEvaluator.EvaluateFiles(processExpression(args[0]), args[1:], printer, leadingContentPreProcessing)
 	}
 	completedSuccessfully = err == nil
 

cmd/printer_utils.go

@@ -0,0 +1,28 @@
+package cmd
+
+import (
+	"io"
+
+	"github.com/mikefarah/yq/v4/pkg/yqlib"
+)
+
+func configurePrinterWriter(format yqlib.PrinterOutputFormat, out io.Writer) yqlib.PrinterWriter {
+
+	var printerWriter yqlib.PrinterWriter
+
+	if splitFileExp == "i" {
+		colorsEnabled = forceColor
+		printerWriter = yqlib.NewMultiPrinterWriter(nil, format)
+
+	} else if splitFileExp != "" {
+		colorsEnabled = forceColor
+		splitExp, err := yqlib.NewExpressionParser().ParseExpression(splitFileExp)
+		if err != nil {
+			return nil
+		}
+		printerWriter = yqlib.NewMultiPrinterWriter(splitExp, format)
+	} else {
+		printerWriter = yqlib.NewSinglePrinterWriter(out)
+	}
+	return printerWriter
+}

cmd/root.go

@@ -11,7 +11,8 @@ func New() *cobra.Command {
 	var rootCmd = &cobra.Command{
 		Use:   "yq",
 		Short: "yq is a lightweight and portable command-line YAML processor.",
-		Long:  `yq is a lightweight and portable command-line YAML processor. It aims to be the jq or sed of yaml files.`,
+		Long: `yq is a portable command-line YAML processor (https://github.com/mikefarah/yq/) 
+See https://mikefarah.gitbook.io/yq/ for detailed documentation and examples.`,
 		RunE: func(cmd *cobra.Command, args []string) error {
 			if version {
 				cmd.Print(GetVersionDisplay())
@@ -40,7 +41,14 @@ func New() *cobra.Command {
 	}
 
 	rootCmd.PersistentFlags().BoolVarP(&verbose, "verbose", "v", false, "verbose mode")
-	rootCmd.PersistentFlags().BoolVarP(&outputToJSON, "tojson", "j", false, "output as json. Set indent to 0 to print json in one line.")
+
+	rootCmd.PersistentFlags().BoolVarP(&outputToJSON, "tojson", "j", false, "(deprecated) output as json. Set indent to 0 to print json in one line.")
+	err := rootCmd.PersistentFlags().MarkDeprecated("tojson", "please use -o=json instead")
+	if err != nil {
+		panic(err)
+	}
+
+	rootCmd.PersistentFlags().StringVarP(&outputFormat, "output-format", "o", "yaml", "[yaml|y|json|j|props|p] output format type.")
 	rootCmd.PersistentFlags().BoolVarP(&nullInput, "null-input", "n", false, "Don't read input, simply evaluate the expression given. Useful for creating yaml docs from scratch.")
 	rootCmd.PersistentFlags().BoolVarP(&noDocSeparators, "no-doc", "N", false, "Don't print document separators (---)")
 
@@ -53,6 +61,11 @@ func New() *cobra.Command {
 
 	rootCmd.PersistentFlags().BoolVarP(&forceColor, "colors", "C", false, "force print with colors")
 	rootCmd.PersistentFlags().BoolVarP(&forceNoColor, "no-colors", "M", false, "force print with no colors")
+	rootCmd.PersistentFlags().StringVarP(&frontMatter, "front-matter", "f", "", "(extract|process) first input as yaml front-matter. Extract will pull out the yaml content, process will run the expression against the yaml content, leaving the remaining data intact")
+	rootCmd.PersistentFlags().BoolVarP(&leadingContentPreProcessing, "header-preprocess", "", true, "Slurp any header comments and separators before processing expression. This is a workaround for go-yaml to persist header content properly.")
+
+	rootCmd.PersistentFlags().StringVarP(&splitFileExp, "split-exp", "s", "", "print each result (or doc) into a file named (exp). [exp] argument must return a string. You can use $index in the expression as the result counter.")
+
 	rootCmd.AddCommand(
 		createEvaluateSequenceCommand(),
 		createEvaluateAllCommand(),

cmd/version.go

@@ -11,7 +11,7 @@ var (
 	GitDescribe string
 
 	// Version is main version number that is being run at the moment.
-	Version = "4.9.8"
+	Version = "4.13.5"
 
 	// VersionPrerelease is a pre-release marker for the version. If this is "" (empty string)
 	// then it means that it is a final release. Otherwise, this is a pre-release
@@ -25,7 +25,7 @@ const ProductName = "yq"
 // GetVersionDisplay composes the parts of the version in a way that's suitable
 // for displaying to humans.
 func GetVersionDisplay() string {
-	return fmt.Sprintf("%s version %s\n", ProductName, getHumanVersion())
+	return fmt.Sprintf("yq (https://github.com/mikefarah/yq/) version %s\n", getHumanVersion())
 }
 
 func getHumanVersion() string {

cmd/version_test.go

@@ -3,7 +3,7 @@ package cmd
 import "testing"
 
 func TestGetVersionDisplay(t *testing.T) {
-	var expectedVersion = ProductName + " version " + Version
+	var expectedVersion = ProductName + " (https://github.com/mikefarah/yq/) version " + Version
 	if VersionPrerelease != "" {
 		expectedVersion = expectedVersion + "-" + VersionPrerelease
 	}

debian/changelog

@@ -1,3 +1,19 @@
+yq (4.13.0) focal; urgency=medium
+
+  * New `with` operator for making multiple changes to a given path
+  * New `contains` operator, works like the `jq` equivalent
+  * Subtract operator now supports subtracting elements from arrays!
+  * Fixed Swapping values using variables #934
+  * Github Action now properly supports multiline output #936, thanks @pjxiao
+  * Fixed missing closing bracket validation #932
+  * Fix processing of hex numbers #929
+  * Fixed alternative and union operator issues #930
+  * Can now convert yaml to properties properties format (`-o=props`), See [docs](https://mikefarah.gitbook.io/yq/v/v4.x/usage/properties) for more info.
+  * Fixed document header/footer comment handling when merging (https://github.com/mikefarah/yq/issues/919)
+  * pretty print yaml 1.1 compatibility  (https://github.com/mikefarah/yq/issues/914)
+
+ -- Roberto Mier Escandon <rmescandon@gmail.com>  Thu, 16 Sep 2021 20:58:30 +0200
+
 yq (4.9.6) focal; urgency=medium
 
   * Added darwin/arm64 build, thanks @alecthomas

debian/rules

@@ -45,7 +45,9 @@ override_dh_auto_build:
 	(cd ${SRCDIR} && go install -buildmode=pie ./...)
 	
 override_dh_auto_test:
+ifeq (,$(filter nocheck,$(DEB_BUILD_OPTIONS)))
 	(cd ${SRCDIR} && go test -v ./...)
+endif
 
 override_dh_auto_install:
 	cp ${GOBIN}/yq ${DESTDIR}/${BINDIR}

examples/data1.yaml

@@ -1 +1,5 @@
-{"a":{"b":1}}
+Resources:
+  S3Bucket:
+    Type: AWS::CloudFormation::Stack
+    Properties:
+      BucketName: !Ref MyBucketNameA

examples/data2.yaml

@@ -1,7 +1,3 @@
-a: other # better than the original
-b: [3, 4]
-c:
-  toast: leave
-  test: 1
-  tell: 1
-  tasty.taco: cool
+#b1 
+b: 2
+#b2

examples/empty.yaml

@@ -1,2 +1 @@
-# a: apple
-# b: cat
+# comment

examples/front-matter.yaml

@@ -0,0 +1,6 @@
+---
+a: apple
+b: bannana
+---
+hello there
+apples: great

examples/leading-seperator.yaml

@@ -0,0 +1,2 @@
+---
+a: test

github-action/Dockerfile

@@ -1,5 +1,8 @@
-FROM mikefarah/yq:4.9.8
+FROM mikefarah/yq:4.13.5
 
 COPY entrypoint.sh /entrypoint.sh
 
+# this seems to be the default user in github actions
+USER 1001
+
 ENTRYPOINT ["/entrypoint.sh"]

github-action/entrypoint.sh

@@ -1,6 +1,9 @@
 #!/bin/sh -l
-
+set -e
 echo "::debug::\$cmd: $1"
 RESULT=$(eval "$1")
+RESULT="${RESULT//'%'/'%25'}"
+RESULT="${RESULT//$'\n'/'%0A'}"
+RESULT="${RESULT//$'\r'/'%0D'}"
 echo "::debug::\$RESULT: $RESULT"
 echo ::set-output name=result::"$RESULT"

go.mod

@@ -5,12 +5,21 @@ require (
 	github.com/fatih/color v1.10.0
 	github.com/goccy/go-yaml v1.8.9
 	github.com/jinzhu/copier v0.2.8
+	github.com/magiconair/properties v1.8.5
 	github.com/spf13/cobra v1.1.3
-	github.com/timtadh/data-structures v0.5.3 // indirect
 	github.com/timtadh/lexmachine v0.2.2
-	golang.org/x/sys v0.0.0-20210510120138-977fb7262007 // indirect
 	gopkg.in/op/go-logging.v1 v1.0.0-20160211212156-b2cb9fa56473
 	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b
 )
 
-go 1.15
+require (
+	github.com/inconshreveable/mousetrap v1.0.0 // indirect
+	github.com/mattn/go-colorable v0.1.8 // indirect
+	github.com/mattn/go-isatty v0.0.12 // indirect
+	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/timtadh/data-structures v0.5.3 // indirect
+	golang.org/x/sys v0.0.0-20210809222454-d867a43fc93e // indirect
+	golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 // indirect
+)
+
+go 1.17

go.sum

@@ -119,6 +119,8 @@ github.com/kr/text v0.1.0 h1:45sCR5RtlFHMR4UwH9sdQ5TC8v0qDQCHnXt+kaKSTVE=
 github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI=
 github.com/leodido/go-urn v1.2.0/go.mod h1:+8+nEpDfqqsY+g338gtMEUOtuK+4dEMhiQEgxpxOKII=
 github.com/magiconair/properties v1.8.1/go.mod h1:PppfXfuXeibc/6YijjN8zIbojt8czPbwD3XqdrwzmxQ=
+github.com/magiconair/properties v1.8.5 h1:b6kJs+EmPFMYGkow9GiUyCyOvIwYetYJ3fSaWak/Gls=
+github.com/magiconair/properties v1.8.5/go.mod h1:y3VJvCyxH9uVvJTWEGAELF3aiYNyPKd5NZ3oSwXrF60=
 github.com/mattn/go-colorable v0.0.9/go.mod h1:9vuHe8Xs5qXnSaW/c/ABM9alt+Vo+STaOChaDxuIBZU=
 github.com/mattn/go-colorable v0.1.8 h1:c1ghPdyEDarC70ftn0y+A/Ee++9zz8ljHG1b13eJ0s8=
 github.com/mattn/go-colorable v0.1.8/go.mod h1:u6P/XSegPjTcexA+o6vUJrdnUu04hMope9wVRipJSqc=
@@ -256,8 +258,8 @@ golang.org/x/sys v0.0.0-20190606165138-5da285871e9c/go.mod h1:h1NjWce9XRLGQEsW7w
 golang.org/x/sys v0.0.0-20190624142023-c5567b49c5d0/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200116001909-b77594299b42/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200223170610-d5e6a3e2c0ae/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/sys v0.0.0-20210510120138-977fb7262007 h1:gG67DSER+11cZvqIMb8S8bt0vZtiN6xWYARwirrOSfE=
-golang.org/x/sys v0.0.0-20210510120138-977fb7262007/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
+golang.org/x/sys v0.0.0-20210809222454-d867a43fc93e h1:WUoyKPm6nCo1BnNUvPGnFG3T5DUVem42yDJZZ4CNxMA=
+golang.org/x/sys v0.0.0-20210809222454-d867a43fc93e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.1-0.20180807135948-17ff2d5776d2/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk=

pkg/yqlib/all_at_once_evaluator.go

@@ -8,7 +8,7 @@ import (
 
 // A yaml expression evaluator that runs the expression once against all files/nodes in memory.
 type Evaluator interface {
-	EvaluateFiles(expression string, filenames []string, printer Printer) error
+	EvaluateFiles(expression string, filenames []string, printer Printer, leadingContentPreProcessing bool) error
 
 	// EvaluateNodes takes an expression and one or more yaml nodes, returning a list of matching candidate nodes
 	EvaluateNodes(expression string, nodes ...*yaml.Node) (*list.List, error)
@@ -46,15 +46,21 @@ func (e *allAtOnceEvaluator) EvaluateCandidateNodes(expression string, inputCand
 	return context.MatchingNodes, nil
 }
 
-func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string, printer Printer) error {
+func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string, printer Printer, leadingContentPreProcessing bool) error {
 	fileIndex := 0
+	firstFileLeadingContent := ""
 
 	var allDocuments *list.List = list.New()
 	for _, filename := range filenames {
-		reader, err := readStream(filename)
+		reader, leadingContent, err := readStream(filename, fileIndex == 0 && leadingContentPreProcessing)
 		if err != nil {
 			return err
 		}
+
+		if fileIndex == 0 {
+			firstFileLeadingContent = leadingContent
+		}
+
 		fileDocuments, err := readDocuments(reader, filename, fileIndex)
 		if err != nil {
 			return err
@@ -62,6 +68,19 @@ func (e *allAtOnceEvaluator) EvaluateFiles(expression string, filenames []string
 		allDocuments.PushBackList(fileDocuments)
 		fileIndex = fileIndex + 1
 	}
+
+	if allDocuments.Len() == 0 {
+		candidateNode := &CandidateNode{
+			Document:  0,
+			Filename:  "",
+			Node:      &yaml.Node{Kind: yaml.DocumentNode, HeadComment: firstFileLeadingContent, Content: []*yaml.Node{{Tag: "!!null", Kind: yaml.ScalarNode}}},
+			FileIndex: 0,
+		}
+		allDocuments.PushBack(candidateNode)
+	} else {
+		allDocuments.Front().Value.(*CandidateNode).Node.HeadComment = firstFileLeadingContent
+	}
+
 	matches, err := e.EvaluateCandidateNodes(expression, allDocuments)
 	if err != nil {
 		return err

pkg/yqlib/candidate_node.go

@@ -1,6 +1,7 @@
 package yqlib
 
 import (
+	"container/list"
 	"fmt"
 
 	"github.com/jinzhu/copier"
@@ -28,6 +29,12 @@ func (n *CandidateNode) GetKey() string {
 	return fmt.Sprintf("%v%v - %v", keyPrefix, n.Document, n.Path)
 }
 
+func (n *CandidateNode) AsList() *list.List {
+	elMap := list.New()
+	elMap.PushBack(n)
+	return elMap
+}
+
 func (n *CandidateNode) CreateChild(path interface{}, node *yaml.Node) *CandidateNode {
 	return &CandidateNode{
 		Node:      node,

pkg/yqlib/context.go

@@ -2,8 +2,10 @@ package yqlib
 
 import (
 	"container/list"
+	"fmt"
 
 	"github.com/jinzhu/copier"
+	logging "gopkg.in/op/go-logging.v1"
 )
 
 type Context struct {
@@ -41,19 +43,52 @@ func (n *Context) SetVariable(name string, value *list.List) {
 }
 
 func (n *Context) ChildContext(results *list.List) Context {
+	clone := Context{DontAutoCreate: n.DontAutoCreate}
+	clone.Variables = make(map[string]*list.List)
+	if len(n.Variables) > 0 {
+		err := copier.Copy(&clone.Variables, n.Variables)
+		if err != nil {
+			log.Error("Error cloning context :(")
+			panic(err)
+		}
+	}
+	clone.MatchingNodes = results
+	return clone
+}
+
+func (n *Context) ToString() string {
+	if !log.IsEnabledFor(logging.DEBUG) {
+		return ""
+	}
+	result := fmt.Sprintf("Context\nDontAutoCreate: %v\n", n.DontAutoCreate)
+	return result + NodesToString(n.MatchingNodes)
+}
+
+func (n *Context) DeepClone() Context {
 	clone := Context{}
 	err := copier.Copy(&clone, n)
+	// copier doesn't do lists properly for some reason
+	clone.MatchingNodes = list.New()
+	for el := n.MatchingNodes.Front(); el != nil; el = el.Next() {
+		clonedNode, err := el.Value.(*CandidateNode).Copy()
+		if err != nil {
+			log.Error("Error cloning context :(")
+			panic(err)
+		}
+		clone.MatchingNodes.PushBack(clonedNode)
+	}
+
 	if err != nil {
 		log.Error("Error cloning context :(")
 		panic(err)
 	}
-	clone.MatchingNodes = results
 	return clone
 }
 
 func (n *Context) Clone() Context {
 	clone := Context{}
 	err := copier.Copy(&clone, n)
+
 	if err != nil {
 		log.Error("Error cloning context :(")
 		panic(err)

pkg/yqlib/doc/Anchor and Alias Operators.md

@@ -299,3 +299,33 @@ foobar:
   thing: foobar_thing
 ```
 
+## Dereference and update a field
+`Use explode with multiply to dereference an object
+
+Given a sample.yml file of:
+```yaml
+item_value: &item_value
+  value: true
+thingOne:
+  name: item_1
+  !!merge <<: *item_value
+thingTwo:
+  name: item_2
+  !!merge <<: *item_value
+```
+then
+```bash
+yq eval '.thingOne |= explode(.) * {"value": false}' sample.yml
+```
+will output
+```yaml
+item_value: &item_value
+  value: true
+thingOne:
+  name: item_1
+  value: false
+thingTwo:
+  name: item_2
+  !!merge <<: *item_value
+```
+

pkg/yqlib/doc/Assign (Update).md

@@ -136,7 +136,9 @@ a:
   b: frog
 ```
 
-## Update selected results
+## Update deeply selected results
+Note that the LHS is wrapped in brackets! This is to ensure we dont first filter out the yaml and then update the snippet.
+
 Given a sample.yml file of:
 ```yaml
 a:

pkg/yqlib/doc/Contains.md

@@ -0,0 +1,85 @@
+
+## Array contains array
+Array is equal or subset of
+
+Given a sample.yml file of:
+```yaml
+- foobar
+- foobaz
+- blarp
+```
+then
+```bash
+yq eval 'contains(["baz", "bar"])' sample.yml
+```
+will output
+```yaml
+true
+```
+
+## Object included in array
+Given a sample.yml file of:
+```yaml
+"foo": 12
+"bar":
+  - 1
+  - 2
+  - "barp": 12
+    "blip": 13
+```
+then
+```bash
+yq eval 'contains({"bar": [{"barp": 12}]})' sample.yml
+```
+will output
+```yaml
+true
+```
+
+## Object not included in array
+Given a sample.yml file of:
+```yaml
+"foo": 12
+"bar":
+  - 1
+  - 2
+  - "barp": 12
+    "blip": 13
+```
+then
+```bash
+yq eval 'contains({"foo": 12, "bar": [{"barp": 15}]})' sample.yml
+```
+will output
+```yaml
+false
+```
+
+## String contains substring
+Given a sample.yml file of:
+```yaml
+foobar
+```
+then
+```bash
+yq eval 'contains("bar")' sample.yml
+```
+will output
+```yaml
+true
+```
+
+## String equals string
+Given a sample.yml file of:
+```yaml
+meow
+```
+then
+```bash
+yq eval 'contains("meow")' sample.yml
+```
+will output
+```yaml
+true
+```
+

pkg/yqlib/doc/Encoder and Decoder.md

@@ -0,0 +1,177 @@
+Encode operators will take the piped in object structure and encode it as a string in the desired format. The decode operators do the opposite, they take a formatted string and decode it into the relevant object structure.
+
+Note that you can optionally pass an indent value to the encode functions (see below).
+
+These operators are useful to process yaml documents that have stringified embeded yaml/json/props in them.
+## Encode value as yaml string
+Indent defaults to 2
+
+Given a sample.yml file of:
+```yaml
+a:
+  cool:
+    bob: dylan
+```
+then
+```bash
+yq eval '.b = (.a | to_yaml)' sample.yml
+```
+will output
+```yaml
+a:
+  cool:
+    bob: dylan
+b: |
+  cool:
+    bob: dylan
+```
+
+## Encode value as yaml string, with custom indentation
+You can specify the indentation level as the first parameter.
+
+Given a sample.yml file of:
+```yaml
+a:
+  cool:
+    bob: dylan
+```
+then
+```bash
+yq eval '.b = (.a | to_yaml(8))' sample.yml
+```
+will output
+```yaml
+a:
+  cool:
+    bob: dylan
+b: |
+  cool:
+          bob: dylan
+```
+
+## Encode value as yaml string, using toyaml
+Does the same thing as to_yaml, matching jq naming convention.
+
+Given a sample.yml file of:
+```yaml
+a:
+  cool: thing
+```
+then
+```bash
+yq eval '.b = (.a | to_yaml)' sample.yml
+```
+will output
+```yaml
+a:
+  cool: thing
+b: |
+  cool: thing
+```
+
+## Encode value as json string
+Given a sample.yml file of:
+```yaml
+a:
+  cool: thing
+```
+then
+```bash
+yq eval '.b = (.a | to_json)' sample.yml
+```
+will output
+```yaml
+a:
+  cool: thing
+b: |
+  {
+    "cool": "thing"
+  }
+```
+
+## Encode value as json string, on one line
+Pass in a 0 indent to print json on a single line.
+
+Given a sample.yml file of:
+```yaml
+a:
+  cool: thing
+```
+then
+```bash
+yq eval '.b = (.a | to_json(0))' sample.yml
+```
+will output
+```yaml
+a:
+  cool: thing
+b: '{"cool":"thing"}'
+```
+
+## Encode value as props string
+Given a sample.yml file of:
+```yaml
+a:
+  cool: thing
+```
+then
+```bash
+yq eval '.b = (.a | to_props)' sample.yml
+```
+will output
+```yaml
+a:
+  cool: thing
+b: |
+  cool = thing
+```
+
+## Decode a yaml encoded string
+Given a sample.yml file of:
+```yaml
+a: 'foo: bar'
+```
+then
+```bash
+yq eval '.b = (.a | from_yaml)' sample.yml
+```
+will output
+```yaml
+a: 'foo: bar'
+b:
+  foo: bar
+```
+
+## Update a multiline encoded yaml string
+Given a sample.yml file of:
+```yaml
+a: |
+  foo: bar
+  baz: dog
+
+```
+then
+```bash
+yq eval '.a |= (from_yaml | .foo = "cat" | to_yaml)' sample.yml
+```
+will output
+```yaml
+a: |
+  foo: cat
+  baz: dog
+```
+
+## Update a single line encoded yaml string
+Given a sample.yml file of:
+```yaml
+a: 'foo: bar'
+```
+then
+```bash
+yq eval '.a |= (from_yaml | .foo = "cat" | to_yaml)' sample.yml
+```
+will output
+```yaml
+a: 'foo: cat'
+```
+

pkg/yqlib/doc/Flatten.md

@@ -0,0 +1,70 @@
+This recursively flattens arrays.
+
+## Flatten
+Recursively flattens all arrays
+
+Given a sample.yml file of:
+```yaml
+- 1
+- - 2
+- - - 3
+```
+then
+```bash
+yq eval 'flatten' sample.yml
+```
+will output
+```yaml
+- 1
+- 2
+- 3
+```
+
+## Flatten with depth of one
+Given a sample.yml file of:
+```yaml
+- 1
+- - 2
+- - - 3
+```
+then
+```bash
+yq eval 'flatten(1)' sample.yml
+```
+will output
+```yaml
+- 1
+- 2
+- - 3
+```
+
+## Flatten empty array
+Given a sample.yml file of:
+```yaml
+- []
+```
+then
+```bash
+yq eval 'flatten' sample.yml
+```
+will output
+```yaml
+[]
+```
+
+## Flatten array of objects
+Given a sample.yml file of:
+```yaml
+- foo: bar
+- - foo: baz
+```
+then
+```bash
+yq eval 'flatten' sample.yml
+```
+will output
+```yaml
+- foo: bar
+- foo: baz
+```
+

pkg/yqlib/doc/Group By.md

@@ -0,0 +1,54 @@
+This is used to group items in an array by an expression.
+
+## Group by field
+Given a sample.yml file of:
+```yaml
+- foo: 1
+  bar: 10
+- foo: 3
+  bar: 100
+- foo: 1
+  bar: 1
+```
+then
+```bash
+yq eval 'group_by(.foo)' sample.yml
+```
+will output
+```yaml
+- - foo: 1
+    bar: 10
+  - foo: 1
+    bar: 1
+- - foo: 3
+    bar: 100
+```
+
+## Group by field, with nuls
+Given a sample.yml file of:
+```yaml
+- cat: dog
+- foo: 1
+  bar: 10
+- foo: 3
+  bar: 100
+- no: foo for you
+- foo: 1
+  bar: 1
+```
+then
+```bash
+yq eval 'group_by(.foo)' sample.yml
+```
+will output
+```yaml
+- - cat: dog
+  - no: foo for you
+- - foo: 1
+    bar: 10
+  - foo: 1
+    bar: 1
+- - foo: 3
+    bar: 100
+```
+

pkg/yqlib/doc/Multiply (Merge).md

@@ -357,7 +357,7 @@ yq eval '.foobar * .foobarList' sample.yml
 will output
 ```yaml
 c: foobarList_c
-<<:
+!!merge <<:
   - *foo
   - *bar
 thing: foobar_thing

pkg/yqlib/doc/String Operators.md

@@ -1,5 +1,47 @@
 # String Operators
 
+## RegEx
+This uses golangs native regex functions under the hood - See https://github.com/google/re2/wiki/Syntax for the supported syntax.
+
+
+# String blocks, bash and newlines
+Bash is notorious for chomping on precious trailing newline characters, making it tricky to set strings with newlines properly. In particular, the `$( exp )` _will trim trailing newlines_.
+
+For instance to get this yaml:
+
+```
+a: |
+  cat
+```
+
+Using `$( exp )` wont work, as it will trim the trailing new line.
+
+```
+m=$(echo "cat\n") yq e -n '.a = strenv(m)'
+a: cat
+```
+
+However, using printf works:
+```
+printf -v m "cat\n" ; m="$m" yq e -n '.a = strenv(m)'
+a: |
+  cat
+```
+
+As well as having multiline expressions:
+```
+m="cat
+"  yq e -n '.a = strenv(m)'
+a: |
+  cat
+```
+
+Similarly, if you're trying to set the content from a file, and want a trailing new line:
+
+```
+IFS= read -rd '' output < <(cat my_file)
+output=$output ./yq e '.data.values = strenv(output)' first.yml
+```
 ## Join strings
 Given a sample.yml file of:
 ```yaml
@@ -18,6 +60,171 @@ will output
 cat; meow; 1; ; true
 ```
 
+## Match string
+Given a sample.yml file of:
+```yaml
+foo bar foo
+```
+then
+```bash
+yq eval 'match("foo")' sample.yml
+```
+will output
+```yaml
+string: foo
+offset: 0
+length: 3
+captures: []
+```
+
+## Match string, case insensitive
+Given a sample.yml file of:
+```yaml
+foo bar FOO
+```
+then
+```bash
+yq eval '[match("(?i)foo"; "g")]' sample.yml
+```
+will output
+```yaml
+- string: foo
+  offset: 0
+  length: 3
+  captures: []
+- string: FOO
+  offset: 8
+  length: 3
+  captures: []
+```
+
+## Match with capture groups
+Given a sample.yml file of:
+```yaml
+abc abc
+```
+then
+```bash
+yq eval '[match("(abc)+"; "g")]' sample.yml
+```
+will output
+```yaml
+- string: abc
+  offset: 0
+  length: 3
+  captures:
+    - string: abc
+      offset: 0
+      length: 3
+- string: abc
+  offset: 4
+  length: 3
+  captures:
+    - string: abc
+      offset: 4
+      length: 3
+```
+
+## Match with named capture groups
+Given a sample.yml file of:
+```yaml
+foo bar foo foo  foo
+```
+then
+```bash
+yq eval '[match("foo (?P<bar123>bar)? foo"; "g")]' sample.yml
+```
+will output
+```yaml
+- string: foo bar foo
+  offset: 0
+  length: 11
+  captures:
+    - string: bar
+      offset: 4
+      length: 3
+      name: bar123
+- string: foo  foo
+  offset: 12
+  length: 8
+  captures:
+    - string: null
+      offset: -1
+      length: 0
+      name: bar123
+```
+
+## Capture named groups into a map
+Given a sample.yml file of:
+```yaml
+xyzzy-14
+```
+then
+```bash
+yq eval 'capture("(?P<a>[a-z]+)-(?P<n>[0-9]+)")' sample.yml
+```
+will output
+```yaml
+a: xyzzy
+n: "14"
+```
+
+## Match without global flag
+Given a sample.yml file of:
+```yaml
+cat cat
+```
+then
+```bash
+yq eval 'match("cat")' sample.yml
+```
+will output
+```yaml
+string: cat
+offset: 0
+length: 3
+captures: []
+```
+
+## Match with global flag
+Given a sample.yml file of:
+```yaml
+cat cat
+```
+then
+```bash
+yq eval '[match("cat"; "g")]' sample.yml
+```
+will output
+```yaml
+- string: cat
+  offset: 0
+  length: 3
+  captures: []
+- string: cat
+  offset: 4
+  length: 3
+  captures: []
+```
+
+## Test using regex
+Like jq'q equivalant, this works like match but only returns true/false instead of full match details
+
+Given a sample.yml file of:
+```yaml
+- cat
+- dog
+```
+then
+```bash
+yq eval '.[] | test("at")' sample.yml
+```
+will output
+```yaml
+true
+false
+```
+
 ## Substitute / Replace string
 This uses golang regex, described [here](https://github.com/google/re2/wiki/Syntax)
 Note the use of `|=` to run in context of the current string value.

pkg/yqlib/doc/Style.md

@@ -18,8 +18,6 @@ a:
 ```
 
 ## Update and set style of a particular node using path variables
-You can use a variable to re-use a path
-
 Given a sample.yml file of:
 ```yaml
 a:
@@ -28,7 +26,7 @@ a:
 ```
 then
 ```bash
-yq eval '.a.b as $x | $x = "new" | $x style="double"' sample.yml
+yq eval 'with(.a.b ; . = "new" | . style="double")' sample.yml
 ```
 will output
 ```yaml

pkg/yqlib/doc/Subtract.md

@@ -1,4 +1,43 @@
 
+## Array subtraction
+Running
+```bash
+yq eval --null-input '[1,2] - [2,3]'
+```
+will output
+```yaml
+- 1
+```
+
+## Array subtraction with nested array
+Running
+```bash
+yq eval --null-input '[[1], 1, 2] - [[1], 3]'
+```
+will output
+```yaml
+- 1
+- 2
+```
+
+## Array subtraction with nested object
+Note that order of the keys does not matter
+
+Given a sample.yml file of:
+```yaml
+- a: b
+  c: d
+- a: b
+```
+then
+```bash
+yq eval '. - [{"c": "d", "a": "b"}]' sample.yml
+```
+will output
+```yaml
+- a: b
+```
+
 ## Number subtraction - float
 If the lhs or rhs are floats then the expression will be calculated with floats.
 

pkg/yqlib/doc/Variable Operators.md

@@ -1,4 +1,6 @@
-For more complex scenarios, variables can be used to hold values of expression to be used in other expressions.
+Like the `jq` equivalents, variables are sometimes required for the more complex expressions (or swapping values between fields).
+
+Note that there is also an additional `ref` operator that holds a reference (instead of a copy) of the path, allowing you to make multiple changes to the same path.
 
 ## Single value variable
 Given a sample.yml file of:
@@ -56,3 +58,39 @@ title: A well-written article
 author: Person McPherson
 ```
 
+## Using variables to swap values
+Given a sample.yml file of:
+```yaml
+a: a_value
+b: b_value
+```
+then
+```bash
+yq eval '.a as $x  | .b as $y | .b = $x | .a = $y' sample.yml
+```
+will output
+```yaml
+a: b_value
+b: a_value
+```
+
+## Use ref to reference a path repeatedly
+Note: You may find the `with` operator more useful.
+
+Given a sample.yml file of:
+```yaml
+a:
+  b: thing
+  c: something
+```
+then
+```bash
+yq eval '.a.b ref $x | $x = "new" | $x style="double"' sample.yml
+```
+will output
+```yaml
+a:
+  b: "new"
+  c: something
+```
+

pkg/yqlib/doc/With.md

@@ -0,0 +1,62 @@
+Use the `with` operator to conveniently make multiple updates to a deeply nested path, or to update array elements relatively to each other. The first argument expression sets the root context, and the second expression runs against that root context.
+
+## Update and style
+Given a sample.yml file of:
+```yaml
+a:
+  deeply:
+    nested: value
+```
+then
+```bash
+yq eval 'with(.a.deeply.nested; . = "newValue" | . style="single")' sample.yml
+```
+will output
+```yaml
+a:
+  deeply:
+    nested: 'newValue'
+```
+
+## Update multiple deeply nested properties
+Given a sample.yml file of:
+```yaml
+a:
+  deeply:
+    nested: value
+    other: thing
+```
+then
+```bash
+yq eval 'with(.a.deeply; .nested = "newValue" | .other= "newThing")' sample.yml
+```
+will output
+```yaml
+a:
+  deeply:
+    nested: newValue
+    other: newThing
+```
+
+## Update array elements relatively
+The second expression runs with each element of the array as it's contextual root. This allows you to make updates relative to the element.
+
+Given a sample.yml file of:
+```yaml
+myArray:
+  - a: apple
+  - a: banana
+```
+then
+```bash
+yq eval 'with(.myArray[]; .b = .a + " yum")' sample.yml
+```
+will output
+```yaml
+myArray:
+  - a: apple
+    b: apple yum
+  - a: banana
+    b: banana yum
+```
+

pkg/yqlib/doc/aa.md

@@ -1,5 +1,3 @@
-# Operators
-
 In `yq` expressions are made up of operators and pipes. A context of nodes is passed through the expression and each operation takes the context as input and returns a new context as output. That output is piped in as input for the next operation in the expression. To begin with, the context is set to the first yaml document of the first yaml file (if processing in sequence using eval).
 
 Lets look at a couple of examples.

pkg/yqlib/doc/headers/Encoder and Decoder.md

@@ -0,0 +1,5 @@
+Encode operators will take the piped in object structure and encode it as a string in the desired format. The decode operators do the opposite, they take a formatted string and decode it into the relevant object structure.
+
+Note that you can optionally pass an indent value to the encode functions (see below).
+
+These operators are useful to process yaml documents that have stringified embeded yaml/json/props in them.

pkg/yqlib/doc/headers/Flatten.md

@@ -0,0 +1 @@
+This recursively flattens arrays.

pkg/yqlib/doc/headers/Group By.md

@@ -0,0 +1 @@
+This is used to group items in an array by an expression.

pkg/yqlib/doc/headers/Main.md

@@ -0,0 +1,61 @@
+# NAME
+  *yq* is a portable command-line YAML processor
+
+# SYNOPSIS 
+
+yq [eval/eval-all] [expression] files..
+
+eval/e  - Apply the expression to each document in each yaml file in sequence
+
+eval-all/ea - Loads all yaml documents of all yaml files and runs expression once
+
+# DESCRIPTION
+
+a lightweight and portable command-line YAML processor. `yq` uses [jq](https://github.com/stedolan/jq) like syntax but works with yaml files as well as json. It doesn't yet support everything `jq` does - but it does support the most common operations and functions, and more is being added continuously.
+
+This documentation is also available at https://mikefarah.gitbook.io/yq/
+# QUICK GUIDE 
+
+## Read a value:
+```bash
+yq e '.a.b[0].c' file.yaml
+```
+
+## Pipe from STDIN:
+```bash
+cat file.yaml | yq e '.a.b[0].c' -
+```
+
+## Update a yaml file, inplace
+```bash
+yq e -i '.a.b[0].c = "cool"' file.yaml
+```
+
+## Update using environment variables
+```bash
+NAME=mike yq e -i '.a.b[0].c = strenv(NAME)' file.yaml
+```
+
+## Merge multiple files
+```
+yq ea '. as $item ireduce ({}; . * $item )' path/to/*.yml
+```
+
+## Multiple updates to a yaml file
+```bash
+yq e -i '
+  .a.b[0].c = "cool" |
+  .x.y.z = "foobar" |
+  .person.name = strenv(NAME)
+' file.yaml
+```
+
+See the [documentation](https://mikefarah.gitbook.io/yq/) for more.
+
+# KNOWN ISSUES / MISSING FEATURES
+- `yq` attempts to preserve comment positions and whitespace as much as possible, but it does not handle all scenarios (see https://github.com/go-yaml/yaml/tree/v3 for details)
+- Powershell has its own...opinions: https://mikefarah.gitbook.io/yq/usage/tips-and-tricks#quotes-in-windows-powershell
+
+# BUGS / ISSUES / FEATURE REQUESTS
+
+Please visit the GitHub page https://github.com/mikefarah/yq/.

pkg/yqlib/doc/headers/String Operators.md

@@ -1 +1,44 @@
 # String Operators
+
+## RegEx
+This uses golangs native regex functions under the hood - See https://github.com/google/re2/wiki/Syntax for the supported syntax.
+
+
+# String blocks, bash and newlines
+Bash is notorious for chomping on precious trailing newline characters, making it tricky to set strings with newlines properly. In particular, the `$( exp )` _will trim trailing newlines_.
+
+For instance to get this yaml:
+
+```
+a: |
+  cat
+```
+
+Using `$( exp )` wont work, as it will trim the trailing new line.
+
+```
+m=$(echo "cat\n") yq e -n '.a = strenv(m)'
+a: cat
+```
+
+However, using printf works:
+```
+printf -v m "cat\n" ; m="$m" yq e -n '.a = strenv(m)'
+a: |
+  cat
+```
+
+As well as having multiline expressions:
+```
+m="cat
+"  yq e -n '.a = strenv(m)'
+a: |
+  cat
+```
+
+Similarly, if you're trying to set the content from a file, and want a trailing new line:
+
+```
+IFS= read -rd '' output < <(cat my_file)
+output=$output ./yq e '.data.values = strenv(output)' first.yml
+```

pkg/yqlib/doc/headers/Variable Operators.md

@@ -1 +1,3 @@
-For more complex scenarios, variables can be used to hold values of expression to be used in other expressions.
+Like the `jq` equivalents, variables are sometimes required for the more complex expressions (or swapping values between fields).
+
+Note that there is also an additional `ref` operator that holds a reference (instead of a copy) of the path, allowing you to make multiple changes to the same path.

pkg/yqlib/doc/headers/With.md

@@ -0,0 +1 @@
+Use the `with` operator to conveniently make multiple updates to a deeply nested path, or to update array elements relatively to each other. The first argument expression sets the root context, and the second expression runs against that root context.

pkg/yqlib/encoder_properties.go

@@ -0,0 +1,80 @@
+package yqlib
+
+import (
+	"fmt"
+	"io"
+	"strings"
+
+	"github.com/magiconair/properties"
+	yaml "gopkg.in/yaml.v3"
+)
+
+type propertiesEncoder struct {
+	destination io.Writer
+}
+
+func NewPropertiesEncoder(destination io.Writer) Encoder {
+	return &propertiesEncoder{destination}
+}
+
+func (pe *propertiesEncoder) Encode(node *yaml.Node) error {
+	mapKeysToStrings(node)
+	p := properties.NewProperties()
+	err := pe.doEncode(p, node, "")
+	if err != nil {
+		return err
+	}
+
+	_, err = p.WriteComment(pe.destination, "#", properties.UTF8)
+	return err
+}
+
+func (pe *propertiesEncoder) doEncode(p *properties.Properties, node *yaml.Node, path string) error {
+	p.SetComment(path,
+		strings.Replace(node.HeadComment, "#", "", 1)+
+			strings.Replace(node.LineComment, "#", "", 1))
+	switch node.Kind {
+	case yaml.ScalarNode:
+		_, _, err := p.Set(path, node.Value)
+		return err
+	case yaml.DocumentNode:
+		return pe.doEncode(p, node.Content[0], path)
+	case yaml.SequenceNode:
+		return pe.encodeArray(p, node.Content, path)
+	case yaml.MappingNode:
+		return pe.encodeMap(p, node.Content, path)
+	case yaml.AliasNode:
+		return pe.doEncode(p, node.Alias, path)
+	default:
+		return fmt.Errorf("Unsupported node %v", node.Tag)
+	}
+}
+
+func (pe *propertiesEncoder) appendPath(path string, key interface{}) string {
+	if path == "" {
+		return fmt.Sprintf("%v", key)
+	}
+	return fmt.Sprintf("%v.%v", path, key)
+}
+
+func (pe *propertiesEncoder) encodeArray(p *properties.Properties, kids []*yaml.Node, path string) error {
+	for index, child := range kids {
+		err := pe.doEncode(p, child, pe.appendPath(path, index))
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func (pe *propertiesEncoder) encodeMap(p *properties.Properties, kids []*yaml.Node, path string) error {
+	for index := 0; index < len(kids); index = index + 2 {
+		key := kids[index]
+		value := kids[index+1]
+		err := pe.doEncode(p, value, pe.appendPath(path, key.Value))
+		if err != nil {
+			return err
+		}
+	}
+	return nil
+}

pkg/yqlib/encoder_properties_test.go

@@ -0,0 +1,78 @@
+package yqlib
+
+import (
+	"bufio"
+	"bytes"
+	"strings"
+	"testing"
+
+	"github.com/mikefarah/yq/v4/test"
+)
+
+func yamlToProps(sampleYaml string) string {
+	var output bytes.Buffer
+	writer := bufio.NewWriter(&output)
+
+	var propsEncoder = NewPropertiesEncoder(writer)
+	inputs, err := readDocuments(strings.NewReader(sampleYaml), "sample.yml", 0)
+	if err != nil {
+		panic(err)
+	}
+	node := inputs.Front().Value.(*CandidateNode).Node
+	err = propsEncoder.Encode(node)
+	if err != nil {
+		panic(err)
+	}
+	writer.Flush()
+
+	return strings.TrimSuffix(output.String(), "\n")
+}
+
+func TestPropertiesEncoderSimple(t *testing.T) {
+	var sampleYaml = `a: 'bob cool'`
+
+	var expectedJson = `a = bob cool`
+	var actualProps = yamlToProps(sampleYaml)
+	test.AssertResult(t, expectedJson, actualProps)
+}
+
+func TestPropertiesEncoderSimpleWithComments(t *testing.T) {
+	var sampleYaml = `a: 'bob cool' # line`
+
+	var expectedJson = `# line
+a = bob cool`
+	var actualProps = yamlToProps(sampleYaml)
+	test.AssertResult(t, expectedJson, actualProps)
+}
+
+func TestPropertiesEncoderDeep(t *testing.T) {
+	var sampleYaml = `a: 
+  b: "bob cool"
+`
+
+	var expectedJson = `a.b = bob cool`
+	var actualProps = yamlToProps(sampleYaml)
+	test.AssertResult(t, expectedJson, actualProps)
+}
+
+func TestPropertiesEncoderDeepWithComments(t *testing.T) {
+	var sampleYaml = `a:  # a thing
+  b: "bob cool" # b thing
+`
+
+	var expectedJson = `# b thing
+a.b = bob cool`
+	var actualProps = yamlToProps(sampleYaml)
+	test.AssertResult(t, expectedJson, actualProps)
+}
+
+func TestPropertiesEncoderArray(t *testing.T) {
+	var sampleYaml = `a: 
+  b: [{c: dog}, {c: cat}]
+`
+
+	var expectedJson = `a.b.0.c = dog
+a.b.1.c = cat`
+	var actualProps = yamlToProps(sampleYaml)
+	test.AssertResult(t, expectedJson, actualProps)
+}

pkg/yqlib/expression_parser_test.go

@@ -6,37 +6,66 @@ import (
 	"github.com/mikefarah/yq/v4/test"
 )
 
-func TestPathTreeNoArgsForTwoArgOp(t *testing.T) {
+func TestParserNoMatchingCloseCollect(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression("[1,2")
+	test.AssertResultComplex(t, "Bad expression, could not find matching `]`", err.Error())
+}
+func TestParserNoMatchingCloseObjectInCollect(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression(`[{"b": "c"]`)
+	test.AssertResultComplex(t, "Bad expression, could not find matching `}`", err.Error())
+}
+
+func TestParserNoMatchingCloseInCollect(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression(`[(.a]`)
+	test.AssertResultComplex(t, "Bad expression, could not find matching `)`", err.Error())
+}
+
+func TestParserNoMatchingCloseCollectObject(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression(`{"a": "b"`)
+	test.AssertResultComplex(t, "Bad expression, could not find matching `}`", err.Error())
+}
+
+func TestParserNoMatchingCloseCollectInCollectObject(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression(`{"b": [1}`)
+	test.AssertResultComplex(t, "Bad expression, could not find matching `]`", err.Error())
+}
+
+func TestParserNoMatchingCloseBracketInCollectObject(t *testing.T) {
+	_, err := NewExpressionParser().ParseExpression(`{"b": (1}`)
+	test.AssertResultComplex(t, "Bad expression, could not find matching `)`", err.Error())
+}
+
+func TestParserNoArgsForTwoArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression("=")
 	test.AssertResultComplex(t, "'=' expects 2 args but there is 0", err.Error())
 }
 
-func TestPathTreeOneLhsArgsForTwoArgOp(t *testing.T) {
+func TestParserOneLhsArgsForTwoArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression(".a =")
 	test.AssertResultComplex(t, "'=' expects 2 args but there is 1", err.Error())
 }
 
-func TestPathTreeOneRhsArgsForTwoArgOp(t *testing.T) {
+func TestParserOneRhsArgsForTwoArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression("= .a")
 	test.AssertResultComplex(t, "'=' expects 2 args but there is 1", err.Error())
 }
 
-func TestPathTreeTwoArgsForTwoArgOp(t *testing.T) {
+func TestParserTwoArgsForTwoArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression(".a = .b")
 	test.AssertResultComplex(t, nil, err)
 }
 
-func TestPathTreeNoArgsForOneArgOp(t *testing.T) {
+func TestParserNoArgsForOneArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression("explode")
 	test.AssertResultComplex(t, "'explode' expects 1 arg but received none", err.Error())
 }
 
-func TestPathTreeOneArgForOneArgOp(t *testing.T) {
+func TestParserOneArgForOneArgOp(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression("explode(.)")
 	test.AssertResultComplex(t, nil, err)
 }
 
-func TestPathTreeExtraArgs(t *testing.T) {
+func TestParserExtraArgs(t *testing.T) {
 	_, err := NewExpressionParser().ParseExpression("sortKeys(.) explode(.)")
 	test.AssertResultComplex(t, "Bad expression, please check expression syntax", err.Error())
 }

pkg/yqlib/expression_postfix.go

@@ -2,6 +2,7 @@ package yqlib
 
 import (
 	"errors"
+	"fmt"
 
 	logging "gopkg.in/op/go-logging.v1"
 )
@@ -24,6 +25,17 @@ func popOpToResult(opStack []*token, result []*Operation) ([]*token, []*Operatio
 	return opStack, append(result, newOp.Operation)
 }
 
+func validateNoOpenTokens(token *token) error {
+	if token.TokenType == openCollect {
+		return fmt.Errorf(("Bad expression, could not find matching `]`"))
+	} else if token.TokenType == openCollectObject {
+		return fmt.Errorf(("Bad expression, could not find matching `}`"))
+	} else if token.TokenType == openBracket {
+		return fmt.Errorf(("Bad expression, could not find matching `)`"))
+	}
+	return nil
+}
+
 func (p *expressionPostFixerImpl) ConvertToPostfix(infixTokens []*token) ([]*Operation, error) {
 	var result []*Operation
 	// surround the whole thing with brackets
@@ -45,6 +57,10 @@ func (p *expressionPostFixerImpl) ConvertToPostfix(infixTokens []*token) ([]*Ope
 			}
 
 			for len(opStack) > 0 && opStack[len(opStack)-1].TokenType != opener {
+				missingClosingTokenErr := validateNoOpenTokens(opStack[len(opStack)-1])
+				if missingClosingTokenErr != nil {
+					return nil, missingClosingTokenErr
+				}
 				opStack, result = popOpToResult(opStack, result)
 			}
 			if len(opStack) == 0 {
@@ -76,6 +92,11 @@ func (p *expressionPostFixerImpl) ConvertToPostfix(infixTokens []*token) ([]*Ope
 
 		case closeBracket:
 			for len(opStack) > 0 && opStack[len(opStack)-1].TokenType != openBracket {
+				missingClosingTokenErr := validateNoOpenTokens(opStack[len(opStack)-1])
+				if missingClosingTokenErr != nil {
+					return nil, missingClosingTokenErr
+				}
+
 				opStack, result = popOpToResult(opStack, result)
 			}
 			if len(opStack) == 0 {
@@ -98,6 +119,8 @@ func (p *expressionPostFixerImpl) ConvertToPostfix(infixTokens []*token) ([]*Ope
 		}
 	}
 
+	log.Debugf("opstackLen: %v", len(opStack))
+
 	if log.IsEnabledFor(logging.DEBUG) {
 		log.Debugf("PostFix Result:")
 		for _, currentToken := range result {

pkg/yqlib/expression_processing_test.go

@@ -15,6 +15,31 @@ var pathTests = []struct {
 	expectedTokens  []interface{}
 	expectedPostFix []interface{}
 }{
+	{
+		`.["cat"].["dog"]`,
+		append(make([]interface{}, 0), "SELF", "TRAVERSE_ARRAY", "[", "cat (string)", "]", "SHORT_PIPE", "SELF", "TRAVERSE_ARRAY", "[", "dog (string)", "]"),
+		append(make([]interface{}, 0), "SELF", "cat (string)", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SELF", "dog (string)", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE"),
+	},
+	{
+		`.["cat"]`,
+		append(make([]interface{}, 0), "SELF", "TRAVERSE_ARRAY", "[", "cat (string)", "]"),
+		append(make([]interface{}, 0), "SELF", "cat (string)", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY"),
+	},
+	{
+		"with(.a;.=3)",
+		append(make([]interface{}, 0), "WITH", "(", "a", "BLOCK", "SELF", "ASSIGN", "3 (int64)", ")"),
+		append(make([]interface{}, 0), "a", "SELF", "3 (int64)", "ASSIGN", "BLOCK", "WITH"),
+	},
+	{
+		"0x12",
+		append(make([]interface{}, 0), "18 (int64)"),
+		append(make([]interface{}, 0), "18 (int64)"),
+	},
+	{
+		"0X12",
+		append(make([]interface{}, 0), "18 (int64)"),
+		append(make([]interface{}, 0), "18 (int64)"),
+	},
 	{
 		".a\n",
 		append(make([]interface{}, 0), "a"),
@@ -112,8 +137,8 @@ var pathTests = []struct {
 	},
 	{
 		`.a.[]`,
-		append(make([]interface{}, 0), "a", "TRAVERSE_ARRAY", "[", "EMPTY", "]"),
-		append(make([]interface{}, 0), "a", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY"),
+		append(make([]interface{}, 0), "a", "SHORT_PIPE", "SELF", "TRAVERSE_ARRAY", "[", "EMPTY", "]"),
+		append(make([]interface{}, 0), "a", "SELF", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE"),
 	},
 	{
 		`.a[0]`,
@@ -127,8 +152,8 @@ var pathTests = []struct {
 	},
 	{
 		`.a.[0]`,
-		append(make([]interface{}, 0), "a", "TRAVERSE_ARRAY", "[", "0 (int64)", "]"),
-		append(make([]interface{}, 0), "a", "0 (int64)", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY"),
+		append(make([]interface{}, 0), "a", "SHORT_PIPE", "SELF", "TRAVERSE_ARRAY", "[", "0 (int64)", "]"),
+		append(make([]interface{}, 0), "a", "SELF", "0 (int64)", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE"),
 	},
 	{
 		`.a[].c`,
@@ -202,8 +227,8 @@ var pathTests = []struct {
 	},
 	{
 		`{.a: .c, .b.[]: .f.g[]}`,
-		append(make([]interface{}, 0), "{", "a", "CREATE_MAP", "c", "UNION", "b", "TRAVERSE_ARRAY", "[", "EMPTY", "]", "CREATE_MAP", "f", "SHORT_PIPE", "g", "TRAVERSE_ARRAY", "[", "EMPTY", "]", "}"),
-		append(make([]interface{}, 0), "a", "c", "CREATE_MAP", "b", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "f", "g", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE", "CREATE_MAP", "UNION", "COLLECT_OBJECT", "SHORT_PIPE"),
+		append(make([]interface{}, 0), "{", "a", "CREATE_MAP", "c", "UNION", "b", "SHORT_PIPE", "SELF", "TRAVERSE_ARRAY", "[", "EMPTY", "]", "CREATE_MAP", "f", "SHORT_PIPE", "g", "TRAVERSE_ARRAY", "[", "EMPTY", "]", "}"),
+		append(make([]interface{}, 0), "a", "c", "CREATE_MAP", "b", "SELF", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE", "f", "g", "EMPTY", "COLLECT", "SHORT_PIPE", "TRAVERSE_ARRAY", "SHORT_PIPE", "CREATE_MAP", "UNION", "COLLECT_OBJECT", "SHORT_PIPE"),
 	},
 	{
 		`explode(.a.b)`,

pkg/yqlib/expression_tokeniser.go

@@ -2,6 +2,7 @@ package yqlib
 
 import (
 	"fmt"
+	"regexp"
 	"strconv"
 	"strings"
 
@@ -111,6 +112,7 @@ func multiplyWithPrefs() lex.Action {
 		if strings.Contains(options, "d") {
 			prefs.DeepMergeArrays = true
 		}
+		prefs.TraversePrefs.DontFollowAlias = true
 		op := &Operation{OperationType: multiplyOpType, Value: multiplyOpType.Type, StringValue: options, Preferences: prefs}
 		return &token{TokenType: operationToken, Operation: op}, nil
 	}
@@ -129,6 +131,44 @@ func opTokenWithPrefs(op *operationType, assignOpType *operationType, preference
 	}
 }
 
+func extractNumberParamter(value string) (int, error) {
+	parameterParser := regexp.MustCompile(`.*\(([0-9]+)\)`)
+	matches := parameterParser.FindStringSubmatch(value)
+	var indent, errParsingInt = strconv.ParseInt(matches[1], 10, 32) // nolint
+	if errParsingInt != nil {
+		return 0, errParsingInt
+	}
+	return int(indent), nil
+}
+
+func flattenWithDepth() lex.Action {
+	return func(s *lex.Scanner, m *machines.Match) (interface{}, error) {
+		value := string(m.Bytes)
+		var depth, errParsingInt = extractNumberParamter(value)
+		if errParsingInt != nil {
+			return nil, errParsingInt
+		}
+
+		prefs := flattenPreferences{depth: depth}
+		op := &Operation{OperationType: flattenOpType, Value: flattenOpType.Type, StringValue: value, Preferences: prefs}
+		return &token{TokenType: operationToken, Operation: op}, nil
+	}
+}
+
+func encodeWithIndent(outputFormat PrinterOutputFormat) lex.Action {
+	return func(s *lex.Scanner, m *machines.Match) (interface{}, error) {
+		value := string(m.Bytes)
+		var indent, errParsingInt = extractNumberParamter(value)
+		if errParsingInt != nil {
+			return nil, errParsingInt
+		}
+
+		prefs := encoderPreferences{format: outputFormat, indent: indent}
+		op := &Operation{OperationType: encodeOpType, Value: encodeOpType.Type, StringValue: value, Preferences: prefs}
+		return &token{TokenType: operationToken, Operation: op}, nil
+	}
+}
+
 func assignAllCommentsOp(updateAssign bool) lex.Action {
 	return func(s *lex.Scanner, m *machines.Match) (interface{}, error) {
 		log.Debug("assignAllCommentsOp %v", string(m.Bytes))
@@ -166,6 +206,20 @@ func numberValue() lex.Action {
 	}
 }
 
+func hexValue() lex.Action {
+	return func(s *lex.Scanner, m *machines.Match) (interface{}, error) {
+		var originalString = string(m.Bytes)
+		var numberString = originalString[2:]
+		log.Debugf("numberString: %v", numberString)
+		var number, errParsingInt = strconv.ParseInt(numberString, 16, 64) // nolint
+		if errParsingInt != nil {
+			return nil, errParsingInt
+		}
+
+		return &token{TokenType: operationToken, Operation: createValueOperation(number, originalString)}, nil
+	}
+}
+
 func floatValue() lex.Action {
 	return func(s *lex.Scanner, m *machines.Match) (interface{}, error) {
 		var numberString = string(m.Bytes)
@@ -257,11 +311,36 @@ func initLexer() (*lex.Lexer, error) {
 	lexer.Add([]byte(`,`), opToken(unionOpType))
 	lexer.Add([]byte(`:\s*`), opToken(createMapOpType))
 	lexer.Add([]byte(`length`), opToken(lengthOpType))
+
+	lexer.Add([]byte(`flatten\([0-9]+\)`), flattenWithDepth())
+	lexer.Add([]byte(`flatten`), opTokenWithPrefs(flattenOpType, nil, flattenPreferences{depth: -1}))
+
+	lexer.Add([]byte(`toyaml\([0-9]+\)`), encodeWithIndent(YamlOutputFormat))
+	lexer.Add([]byte(`to_yaml\([0-9]+\)`), encodeWithIndent(YamlOutputFormat))
+
+	lexer.Add([]byte(`tojson\([0-9]+\)`), encodeWithIndent(JsonOutputFormat))
+	lexer.Add([]byte(`to_json\([0-9]+\)`), encodeWithIndent(JsonOutputFormat))
+
+	lexer.Add([]byte(`toyaml`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: YamlOutputFormat, indent: 2}))
+	lexer.Add([]byte(`tojson`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: JsonOutputFormat, indent: 2}))
+	lexer.Add([]byte(`toprops`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: PropsOutputFormat, indent: 2}))
+
+	lexer.Add([]byte(`to_yaml`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: YamlOutputFormat, indent: 2}))
+	lexer.Add([]byte(`to_json`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: JsonOutputFormat, indent: 2}))
+	lexer.Add([]byte(`to_props`), opTokenWithPrefs(encodeOpType, nil, encoderPreferences{format: PropsOutputFormat, indent: 2}))
+
+	lexer.Add([]byte(`fromyaml`), opToken(decodeOpType))
+	lexer.Add([]byte(`fromjson`), opToken(decodeOpType))
+
+	lexer.Add([]byte(`from_yaml`), opToken(decodeOpType))
+	lexer.Add([]byte(`from_json`), opToken(decodeOpType))
+
 	lexer.Add([]byte(`sortKeys`), opToken(sortKeysOpType))
 	lexer.Add([]byte(`select`), opToken(selectOpType))
 	lexer.Add([]byte(`has`), opToken(hasOpType))
 	lexer.Add([]byte(`unique`), opToken(uniqueOpType))
 	lexer.Add([]byte(`unique_by`), opToken(uniqueByOpType))
+	lexer.Add([]byte(`group_by`), opToken(groupByOpType))
 	lexer.Add([]byte(`explode`), opToken(explodeOpType))
 	lexer.Add([]byte(`or`), opToken(orOpType))
 	lexer.Add([]byte(`and`), opToken(andOpType))
@@ -276,11 +355,15 @@ func initLexer() (*lex.Lexer, error) {
 
 	lexer.Add([]byte(`join`), opToken(joinStringOpType))
 	lexer.Add([]byte(`sub`), opToken(subStringOpType))
+	lexer.Add([]byte(`match`), opToken(matchOpType))
+	lexer.Add([]byte(`capture`), opToken(captureOpType))
+	lexer.Add([]byte(`test`), opToken(testOpType))
 
 	lexer.Add([]byte(`any`), opToken(anyOpType))
 	lexer.Add([]byte(`any_c`), opToken(anyConditionOpType))
 	lexer.Add([]byte(`all`), opToken(allOpType))
 	lexer.Add([]byte(`all_c`), opToken(allConditionOpType))
+	lexer.Add([]byte(`contains`), opToken(containsOpType))
 
 	lexer.Add([]byte(`split`), opToken(splitStringOpType))
 	lexer.Add([]byte(`keys`), opToken(keysOpType))
@@ -298,6 +381,8 @@ func initLexer() (*lex.Lexer, error) {
 	lexer.Add([]byte(`from_entries`), opToken(fromEntriesOpType))
 	lexer.Add([]byte(`with_entries`), opToken(withEntriesOpType))
 
+	lexer.Add([]byte(`with`), opToken(withOpType))
+
 	lexer.Add([]byte(`lineComment`), opTokenWithPrefs(getCommentOpType, assignCommentOpType, commentOpPreferences{LineComment: true}))
 
 	lexer.Add([]byte(`headComment`), opTokenWithPrefs(getCommentOpType, assignCommentOpType, commentOpPreferences{HeadComment: true}))
@@ -320,11 +405,12 @@ func initLexer() (*lex.Lexer, error) {
 	lexer.Add([]byte("( |\t|\n|\r)+"), skip)
 
 	lexer.Add([]byte(`\."[^ "]+"\??`), pathToken(true))
-	lexer.Add([]byte(`\.[^ \}\{\:\[\],\|\.\[\(\)=\n]+\??`), pathToken(false))
+	lexer.Add([]byte(`\.[^ ;\}\{\:\[\],\|\.\[\(\)=\n]+\??`), pathToken(false))
 	lexer.Add([]byte(`\.`), selfToken())
 
 	lexer.Add([]byte(`\|`), opToken(pipeOpType))
 
+	lexer.Add([]byte(`0[xX][0-9A-Fa-f]+`), hexValue())
 	lexer.Add([]byte(`-?\d+(\.\d+)`), floatValue())
 	lexer.Add([]byte(`-?[1-9](\.\d+)?[Ee][-+]?\d+`), floatValue())
 	lexer.Add([]byte(`-?\d+`), numberValue())
@@ -349,7 +435,8 @@ func initLexer() (*lex.Lexer, error) {
 	lexer.Add([]byte(`\-`), opToken(subtractOpType))
 	lexer.Add([]byte(`\-=`), opToken(subtractAssignOpType))
 	lexer.Add([]byte(`\$[a-zA-Z_-0-9]+`), getVariableOpToken())
-	lexer.Add([]byte(`as`), opToken(assignVariableOpType))
+	lexer.Add([]byte(`as`), opTokenWithPrefs(assignVariableOpType, nil, assignVarPreferences{}))
+	lexer.Add([]byte(`ref`), opTokenWithPrefs(assignVariableOpType, nil, assignVarPreferences{IsReference: true}))
 
 	err := lexer.CompileNFA()
 	if err != nil {
@@ -414,17 +501,16 @@ func (p *expressionTokeniserImpl) handleToken(tokens []*token, index int, postPr
 	log.Debug("processing %v", currentToken.toString(true))
 
 	if currentToken.TokenType == traverseArrayCollect {
+		// `.[exp]`` works by creating a traversal array of [self, exp] and piping that into the traverse array operator
 		//need to put a traverse array then a collect currentToken
 		// do this by adding traverse then converting currentToken to collect
 
-		if index == 0 || tokens[index-1].TokenType != operationToken ||
-			tokens[index-1].Operation.OperationType != traversePathOpType {
-			log.Debug("  adding self")
-			op := &Operation{OperationType: selfReferenceOpType, StringValue: "SELF"}
-			postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})
-		}
+		log.Debug("  adding self")
+		op := &Operation{OperationType: selfReferenceOpType, StringValue: "SELF"}
+		postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})
+
 		log.Debug("  adding traverse array")
-		op := &Operation{OperationType: traverseArrayOpType, StringValue: "TRAVERSE_ARRAY"}
+		op = &Operation{OperationType: traverseArrayOpType, StringValue: "TRAVERSE_ARRAY"}
 		postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})
 
 		currentToken = &token{TokenType: openCollect}
@@ -452,20 +538,15 @@ func (p *expressionTokeniserImpl) handleToken(tokens []*token, index int, postPr
 	}
 
 	if index != len(tokens)-1 && currentToken.CheckForPostTraverse &&
-		tokens[index+1].TokenType == operationToken &&
-		tokens[index+1].Operation.OperationType == traversePathOpType {
+		((tokens[index+1].TokenType == operationToken && (tokens[index+1].Operation.OperationType == traversePathOpType)) ||
+			(tokens[index+1].TokenType == traverseArrayCollect)) {
 		log.Debug("  adding pipe because the next thing is traverse")
-		op := &Operation{OperationType: shortPipeOpType, Value: "PIPE"}
+		op := &Operation{OperationType: shortPipeOpType, Value: "PIPE", StringValue: "."}
 		postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})
 	}
 	if index != len(tokens)-1 && currentToken.CheckForPostTraverse &&
 		tokens[index+1].TokenType == openCollect {
 
-		// if tokens[index].TokenType == closeCollect {
-		// 	log.Debug("  adding pipe because next is opencollect")
-		// 	op := &Operation{OperationType: shortPipeOpType, Value: "PIPE"}
-		// 	postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})
-		// }
 		log.Debug("  adding traverArray because next is opencollect")
 		op := &Operation{OperationType: traverseArrayOpType}
 		postProcessedTokens = append(postProcessedTokens, &token{TokenType: operationToken, Operation: op})

pkg/yqlib/file_utils.go

@@ -2,6 +2,7 @@ package yqlib
 
 import (
 	"io"
+	"io/ioutil"
 	"os"
 )
 
@@ -23,6 +24,14 @@ func safelyRenameFile(from string, to string) {
 	}
 }
 
+func tryRemoveFile(filename string) {
+	log.Debug("Removing temp file: %v", filename)
+	removeErr := os.Remove(filename)
+	if removeErr != nil {
+		log.Errorf("Failed to remove temp file: %v", filename)
+	}
+}
+
 // thanks https://stackoverflow.com/questions/21060945/simple-way-to-copy-a-file-in-golang
 func copyFileContents(src, dst string) (err error) {
 	// ignore CWE-22 gosec issue - that's more targetted for http based apps that run in a public directory,
@@ -44,6 +53,13 @@ func copyFileContents(src, dst string) (err error) {
 	return out.Sync()
 }
 
+func SafelyCloseReader(reader io.Reader) {
+	switch reader := reader.(type) {
+	case *os.File:
+		safelyCloseFile(reader)
+	}
+}
+
 func safelyCloseFile(file *os.File) {
 	err := file.Close()
 	if err != nil {
@@ -51,3 +67,22 @@ func safelyCloseFile(file *os.File) {
 		log.Error(err.Error())
 	}
 }
+
+func createTempFile() (*os.File, error) {
+	_, err := os.Stat(os.TempDir())
+	if os.IsNotExist(err) {
+		err = os.Mkdir(os.TempDir(), 0700)
+		if err != nil {
+			return nil, err
+		}
+	} else if err != nil {
+		return nil, err
+	}
+
+	file, err := ioutil.TempFile("", "temp")
+	if err != nil {
+		return nil, err
+	}
+
+	return file, err
+}

pkg/yqlib/front_matter.go

@@ -0,0 +1,93 @@
+package yqlib
+
+import (
+	"bufio"
+	"io"
+	"os"
+)
+
+type frontMatterHandler interface {
+	Split() error
+	GetYamlFrontMatterFilename() string
+	GetContentReader() io.Reader
+	CleanUp()
+}
+
+type frontMatterHandlerImpl struct {
+	originalFilename        string
+	yamlFrontMatterFilename string
+	contentReader           io.Reader
+}
+
+func NewFrontMatterHandler(originalFilename string) frontMatterHandler {
+	return &frontMatterHandlerImpl{originalFilename, "", nil}
+}
+
+func (f *frontMatterHandlerImpl) GetYamlFrontMatterFilename() string {
+	return f.yamlFrontMatterFilename
+}
+
+func (f *frontMatterHandlerImpl) GetContentReader() io.Reader {
+	return f.contentReader
+}
+
+func (f *frontMatterHandlerImpl) CleanUp() {
+	tryRemoveFile(f.yamlFrontMatterFilename)
+}
+
+// Splits the given file by yaml front matter
+// yaml content will be saved to first temporary file
+// remaining content will be saved to second temporary file
+func (f *frontMatterHandlerImpl) Split() error {
+	var reader *bufio.Reader
+	var err error
+	if f.originalFilename == "-" {
+		reader = bufio.NewReader(os.Stdin)
+	} else {
+		file, err := os.Open(f.originalFilename) // #nosec
+		if err != nil {
+			return err
+		}
+		reader = bufio.NewReader(file)
+	}
+	f.contentReader = reader
+
+	yamlTempFile, err := createTempFile()
+	if err != nil {
+		return err
+	}
+	f.yamlFrontMatterFilename = yamlTempFile.Name()
+	log.Debug("yamlTempFile: %v", yamlTempFile.Name())
+
+	lineCount := 0
+
+	for {
+		peekBytes, err := reader.Peek(3)
+		if err == io.EOF {
+			// we've finished reading the yaml content..I guess
+			break
+		} else if err != nil {
+			return err
+		}
+		if lineCount > 0 && string(peekBytes) == "---" {
+			// we've finished reading the yaml content..
+			break
+		}
+		line, errReading := reader.ReadString('\n')
+		lineCount = lineCount + 1
+		if errReading != nil && errReading != io.EOF {
+			return errReading
+		}
+
+		_, errWriting := yamlTempFile.Write([]byte(line))
+
+		if errWriting != nil {
+			return errWriting
+		}
+	}
+
+	safelyCloseFile(yamlTempFile)
+
+	return nil
+
+}

pkg/yqlib/front_matter_test.go

@@ -0,0 +1,142 @@
+package yqlib
+
+import (
+	"io/ioutil"
+	"testing"
+
+	"github.com/mikefarah/yq/v4/test"
+)
+
+func createTestFile(content string) string {
+	tempFile, err := createTempFile()
+	if err != nil {
+		panic(err)
+	}
+
+	_, err = tempFile.Write([]byte(content))
+	if err != nil {
+		panic(err)
+	}
+
+	safelyCloseFile(tempFile)
+
+	return tempFile.Name()
+}
+
+func readFile(filename string) string {
+	bytes, err := ioutil.ReadFile(filename)
+	if err != nil {
+		panic(err)
+	}
+	return string(bytes)
+}
+
+func TestFrontMatterSplitWithLeadingSep(t *testing.T) {
+	file := createTestFile(`---
+a: apple
+b: banana
+---
+not a 
+yaml: doc
+`)
+
+	expectedYamlFm := `---
+a: apple
+b: banana
+`
+
+	expectedContent := `---
+not a 
+yaml: doc
+`
+
+	fmHandler := NewFrontMatterHandler(file)
+	err := fmHandler.Split()
+	if err != nil {
+		panic(err)
+	}
+
+	yamlFm := readFile(fmHandler.GetYamlFrontMatterFilename())
+
+	test.AssertResult(t, expectedYamlFm, yamlFm)
+
+	contentBytes, err := ioutil.ReadAll(fmHandler.GetContentReader())
+	if err != nil {
+		panic(err)
+	}
+	test.AssertResult(t, expectedContent, string(contentBytes))
+
+	tryRemoveFile(file)
+	fmHandler.CleanUp()
+}
+
+func TestFrontMatterSplitWithNoLeadingSep(t *testing.T) {
+	file := createTestFile(`a: apple
+b: banana
+---
+not a 
+yaml: doc
+`)
+
+	expectedYamlFm := `a: apple
+b: banana
+`
+
+	expectedContent := `---
+not a 
+yaml: doc
+`
+
+	fmHandler := NewFrontMatterHandler(file)
+	err := fmHandler.Split()
+	if err != nil {
+		panic(err)
+	}
+
+	yamlFm := readFile(fmHandler.GetYamlFrontMatterFilename())
+
+	test.AssertResult(t, expectedYamlFm, yamlFm)
+
+	contentBytes, err := ioutil.ReadAll(fmHandler.GetContentReader())
+	if err != nil {
+		panic(err)
+	}
+	test.AssertResult(t, expectedContent, string(contentBytes))
+
+	tryRemoveFile(file)
+	fmHandler.CleanUp()
+}
+
+func TestFrontMatterSplitWithArray(t *testing.T) {
+	file := createTestFile(`[1,2,3]
+---
+not a 
+yaml: doc
+`)
+
+	expectedYamlFm := "[1,2,3]\n"
+
+	expectedContent := `---
+not a 
+yaml: doc
+`
+
+	fmHandler := NewFrontMatterHandler(file)
+	err := fmHandler.Split()
+	if err != nil {
+		panic(err)
+	}
+
+	yamlFm := readFile(fmHandler.GetYamlFrontMatterFilename())
+
+	test.AssertResult(t, expectedYamlFm, yamlFm)
+
+	contentBytes, err := ioutil.ReadAll(fmHandler.GetContentReader())
+	if err != nil {
+		panic(err)
+	}
+	test.AssertResult(t, expectedContent, string(contentBytes))
+
+	tryRemoveFile(file)
+	fmHandler.CleanUp()
+}

pkg/yqlib/lib.go

@@ -6,6 +6,8 @@ import (
 	"bytes"
 	"container/list"
 	"fmt"
+	"strconv"
+	"strings"
 
 	logging "gopkg.in/op/go-logging.v1"
 	yaml "gopkg.in/yaml.v3"
@@ -57,9 +59,12 @@ var shortPipeOpType = &operationType{Type: "SHORT_PIPE", NumArgs: 2, Precedence:
 
 var lengthOpType = &operationType{Type: "LENGTH", NumArgs: 0, Precedence: 50, Handler: lengthOperator}
 var collectOpType = &operationType{Type: "COLLECT", NumArgs: 0, Precedence: 50, Handler: collectOperator}
+var encodeOpType = &operationType{Type: "ENCODE", NumArgs: 0, Precedence: 50, Handler: encodeOperator}
+var decodeOpType = &operationType{Type: "DECODE", NumArgs: 0, Precedence: 50, Handler: decodeOperator}
 
 var anyOpType = &operationType{Type: "ANY", NumArgs: 0, Precedence: 50, Handler: anyOperator}
 var allOpType = &operationType{Type: "ALL", NumArgs: 0, Precedence: 50, Handler: allOperator}
+var containsOpType = &operationType{Type: "CONTAINS", NumArgs: 1, Precedence: 50, Handler: containsOperator}
 var anyConditionOpType = &operationType{Type: "ANY_CONDITION", NumArgs: 1, Precedence: 50, Handler: anyOperator}
 var allConditionOpType = &operationType{Type: "ALL_CONDITION", NumArgs: 1, Precedence: 50, Handler: allOperator}
 
@@ -67,6 +72,8 @@ var toEntriesOpType = &operationType{Type: "TO_ENTRIES", NumArgs: 0, Precedence:
 var fromEntriesOpType = &operationType{Type: "FROM_ENTRIES", NumArgs: 0, Precedence: 50, Handler: fromEntriesOperator}
 var withEntriesOpType = &operationType{Type: "WITH_ENTRIES", NumArgs: 1, Precedence: 50, Handler: withEntriesOperator}
 
+var withOpType = &operationType{Type: "WITH", NumArgs: 1, Precedence: 50, Handler: withOperator}
+
 var splitDocumentOpType = &operationType{Type: "SPLIT_DOC", NumArgs: 0, Precedence: 50, Handler: splitDocumentOperator}
 var getVariableOpType = &operationType{Type: "GET_VARIABLE", NumArgs: 0, Precedence: 55, Handler: getVariableOperator}
 var getStyleOpType = &operationType{Type: "GET_STYLE", NumArgs: 0, Precedence: 50, Handler: getStyleOperator}
@@ -83,6 +90,9 @@ var explodeOpType = &operationType{Type: "EXPLODE", NumArgs: 1, Precedence: 50,
 var sortKeysOpType = &operationType{Type: "SORT_KEYS", NumArgs: 1, Precedence: 50, Handler: sortKeysOperator}
 var joinStringOpType = &operationType{Type: "JOIN", NumArgs: 1, Precedence: 50, Handler: joinStringOperator}
 var subStringOpType = &operationType{Type: "SUBSTR", NumArgs: 1, Precedence: 50, Handler: substituteStringOperator}
+var matchOpType = &operationType{Type: "MATCH", NumArgs: 1, Precedence: 50, Handler: matchOperator}
+var captureOpType = &operationType{Type: "CAPTURE", NumArgs: 1, Precedence: 50, Handler: captureOperator}
+var testOpType = &operationType{Type: "TEST", NumArgs: 1, Precedence: 50, Handler: testOperator}
 var splitStringOpType = &operationType{Type: "SPLIT", NumArgs: 1, Precedence: 50, Handler: splitStringOperator}
 
 var keysOpType = &operationType{Type: "KEYS", NumArgs: 0, Precedence: 50, Handler: keysOperator}
@@ -103,6 +113,8 @@ var selectOpType = &operationType{Type: "SELECT", NumArgs: 1, Precedence: 50, Ha
 var hasOpType = &operationType{Type: "HAS", NumArgs: 1, Precedence: 50, Handler: hasOperator}
 var uniqueOpType = &operationType{Type: "UNIQUE", NumArgs: 0, Precedence: 50, Handler: unique}
 var uniqueByOpType = &operationType{Type: "UNIQUE_BY", NumArgs: 1, Precedence: 50, Handler: uniqueBy}
+var groupByOpType = &operationType{Type: "GROUP_BY", NumArgs: 1, Precedence: 50, Handler: groupBy}
+var flattenOpType = &operationType{Type: "FLATTEN_BY", NumArgs: 0, Precedence: 50, Handler: flattenOp}
 var deleteChildOpType = &operationType{Type: "DELETE", NumArgs: 1, Precedence: 40, Handler: deleteChildOperator}
 
 type Operation struct {
@@ -114,8 +126,77 @@ type Operation struct {
 	UpdateAssign  bool // used for assign ops, when true it means we evaluate the rhs given the lhs
 }
 
-func createValueOperation(value interface{}, stringValue string) *Operation {
-	var node yaml.Node = yaml.Node{Kind: yaml.ScalarNode}
+func recurseNodeArrayEqual(lhs *yaml.Node, rhs *yaml.Node) bool {
+	if len(lhs.Content) != len(rhs.Content) {
+		return false
+	}
+
+	for index := 0; index < len(lhs.Content); index = index + 1 {
+		if !recursiveNodeEqual(lhs.Content[index], rhs.Content[index]) {
+			return false
+		}
+	}
+	return true
+}
+
+func findInArray(array *yaml.Node, item *yaml.Node) int {
+
+	for index := 0; index < len(array.Content); index = index + 1 {
+		if recursiveNodeEqual(array.Content[index], item) {
+			return index
+		}
+	}
+	return -1
+}
+
+func recurseNodeObjectEqual(lhs *yaml.Node, rhs *yaml.Node) bool {
+	if len(lhs.Content) != len(rhs.Content) {
+		return false
+	}
+
+	for index := 0; index < len(lhs.Content); index = index + 2 {
+		key := lhs.Content[index]
+		value := lhs.Content[index+1]
+
+		indexInRhs := findInArray(rhs, key)
+
+		if indexInRhs == -1 || !recursiveNodeEqual(value, rhs.Content[indexInRhs+1]) {
+			return false
+		}
+	}
+	return true
+}
+
+func recursiveNodeEqual(lhs *yaml.Node, rhs *yaml.Node) bool {
+	if lhs.Kind != rhs.Kind || lhs.Tag != rhs.Tag {
+		return false
+	} else if lhs.Tag == "!!null" {
+		return true
+
+	} else if lhs.Kind == yaml.ScalarNode {
+		return lhs.Value == rhs.Value
+	} else if lhs.Kind == yaml.SequenceNode {
+		return recurseNodeArrayEqual(lhs, rhs)
+	} else if lhs.Kind == yaml.MappingNode {
+		return recurseNodeObjectEqual(lhs, rhs)
+	}
+	return false
+
+}
+
+// yaml numbers can be hex encoded...
+func parseInt(numberString string) (string, int64, error) {
+	if strings.HasPrefix(numberString, "0x") ||
+		strings.HasPrefix(numberString, "0X") {
+		num, err := strconv.ParseInt(numberString[2:], 16, 64) // nolint
+		return "0x%X", num, err
+	}
+	num, err := strconv.ParseInt(numberString, 10, 64) // nolint
+	return "%v", num, err
+}
+
+func createScalarNode(value interface{}, stringValue string) *yaml.Node {
+	var node = &yaml.Node{Kind: yaml.ScalarNode}
 	node.Value = stringValue
 
 	switch value.(type) {
@@ -130,17 +211,25 @@ func createValueOperation(value interface{}, stringValue string) *Operation {
 	case nil:
 		node.Tag = "!!null"
 	}
+	return node
+}
+
+func createValueOperation(value interface{}, stringValue string) *Operation {
+	var node *yaml.Node = createScalarNode(value, stringValue)
 
 	return &Operation{
 		OperationType: valueOpType,
 		Value:         value,
 		StringValue:   stringValue,
-		CandidateNode: &CandidateNode{Node: &node},
+		CandidateNode: &CandidateNode{Node: node},
 	}
 }
 
 // debugging purposes only
 func (p *Operation) toString() string {
+	if p == nil {
+		return "OP IS NIL"
+	}
 	if p.OperationType == traversePathOpType {
 		return fmt.Sprintf("%v", p.Value)
 	} else if p.OperationType == selfReferenceOpType {
@@ -158,7 +247,7 @@ func NodesToString(collection *list.List) string {
 		return ""
 	}
 
-	result := ""
+	result := fmt.Sprintf("%v results\n", collection.Len())
 	for el := collection.Front(); el != nil; el = el.Next() {
 		result = result + "\n" + NodeToString(el.Value.(*CandidateNode))
 	}

pkg/yqlib/matchKeyString.go

@@ -9,26 +9,50 @@ func matchKey(name string, pattern string) (matched bool) {
 		log.Debug("wild!")
 		return true
 	}
-	return deepMatch([]rune(name), []rune(pattern))
+	return deepMatch(name, pattern)
 }
 
-func deepMatch(str, pattern []rune) bool {
-	for len(pattern) > 0 {
-		switch pattern[0] {
-		default:
-			if len(str) == 0 || str[0] != pattern[0] {
-				return false
+// deepMatch reports whether the name matches the pattern in linear time.
+// Source https://research.swtch.com/glob
+func deepMatch(name, pattern string) bool {
+	px := 0
+	nx := 0
+	nextPx := 0
+	nextNx := 0
+	for px < len(pattern) || nx < len(name) {
+		if px < len(pattern) {
+			c := pattern[px]
+			switch c {
+			default: // ordinary character
+				if nx < len(name) && name[nx] == c {
+					px++
+					nx++
+					continue
+				}
+			case '?': // single-character wildcard
+				if nx < len(name) {
+					px++
+					nx++
+					continue
+				}
+			case '*': // zero-or-more-character wildcard
+				// Try to match at nx.
+				// If that doesn't work out,
+				// restart at nx+1 next.
+				nextPx = px
+				nextNx = nx + 1
+				px++
+				continue
 			}
-		case '?':
-			if len(str) == 0 {
-				return false
-			}
-		case '*':
-			return deepMatch(str, pattern[1:]) ||
-				(len(str) > 0 && deepMatch(str[1:], pattern))
 		}
-		str = str[1:]
-		pattern = pattern[1:]
+		// Mismatch. Maybe restart.
+		if 0 < nextNx && nextNx <= len(name) {
+			px = nextPx
+			nx = nextNx
+			continue
+		}
+		return false
 	}
-	return len(str) == 0 && len(pattern) == 0
+	// Matched all of pattern to all of name. Success.
+	return true
 }

pkg/yqlib/matchKeyString_test.go

@@ -0,0 +1,40 @@
+package yqlib
+
+import (
+	"strings"
+	"testing"
+)
+
+func TestDeepMatch(t *testing.T) {
+	var tests = []struct {
+		name    string
+		pattern string
+		ok      bool
+	}{
+		{"", "", true},
+		{"", "x", false},
+		{"x", "", false},
+		{"abc", "abc", true},
+		{"abc", "*", true},
+		{"abc", "*c", true},
+		{"abc", "*b", false},
+		{"abc", "a*", true},
+		{"abc", "b*", false},
+		{"a", "a*", true},
+		{"a", "*a", true},
+		{"axbxcxdxe", "a*b*c*d*e*", true},
+		{"axbxcxdxexxx", "a*b*c*d*e*", true},
+		{"abxbbxdbxebxczzx", "a*b?c*x", true},
+		{"abxbbxdbxebxczzy", "a*b?c*x", false},
+		{strings.Repeat("a", 100), "a*a*a*a*b", false},
+		{"xxx", "*x", true},
+	}
+
+	for _, tt := range tests {
+		t.Run(tt.name+" "+tt.pattern, func(t *testing.T) {
+			if want, got := tt.ok, deepMatch(tt.name, tt.pattern); want != got {
+				t.Errorf("Expected %v got %v", want, got)
+			}
+		})
+	}
+}

pkg/yqlib/operator_add.go

@@ -76,17 +76,17 @@ func addScalars(target *CandidateNode, lhs *yaml.Node, rhs *yaml.Node) (*Candida
 		target.Node.Tag = "!!str"
 		target.Node.Value = lhs.Value + rhs.Value
 	} else if lhs.Tag == "!!int" && rhs.Tag == "!!int" {
-		lhsNum, err := strconv.Atoi(lhs.Value)
+		format, lhsNum, err := parseInt(lhs.Value)
 		if err != nil {
 			return nil, err
 		}
-		rhsNum, err := strconv.Atoi(rhs.Value)
+		_, rhsNum, err := parseInt(rhs.Value)
 		if err != nil {
 			return nil, err
 		}
 		sum := lhsNum + rhsNum
 		target.Node.Tag = "!!int"
-		target.Node.Value = fmt.Sprintf("%v", sum)
+		target.Node.Value = fmt.Sprintf(format, sum)
 	} else if (lhs.Tag == "!!int" || lhs.Tag == "!!float") && (rhs.Tag == "!!int" || rhs.Tag == "!!float") {
 		lhsNum, err := strconv.ParseFloat(lhs.Value, 64)
 		if err != nil {

pkg/yqlib/operator_alternative.go

@@ -2,7 +2,7 @@ package yqlib
 
 func alternativeOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
 	log.Debugf("-- alternative")
-	return crossFunction(d, context.ReadOnlyClone(), expressionNode, alternativeFunc, true)
+	return crossFunction(d, context, expressionNode, alternativeFunc, true)
 }
 
 func alternativeFunc(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {

pkg/yqlib/operator_alternative_test.go

@@ -6,10 +6,13 @@ import (
 
 var alternativeOperatorScenarios = []expressionScenario{
 	{
+		// to match jq - we do not use a readonly clone context on the LHS.
 		skipDoc:    true,
 		expression: `.b // .c`,
 		document:   `a: bridge`,
-		expected:   []string{},
+		expected: []string{
+			"D0, P[c], (!!null)::null\n",
+		},
 	},
 	{
 		skipDoc:    true,

pkg/yqlib/operator_anchors_aliases.go

@@ -141,6 +141,48 @@ func explodeOperator(d *dataTreeNavigator, context Context, expressionNode *Expr
 	return context, nil
 }
 
+func reconstructAliasedMap(node *yaml.Node, context Context) error {
+	var newContent = list.New()
+	// can I short cut here by prechecking if there's an anchor in the map?
+	// no it needs to recurse in overrideEntry.
+
+	for index := 0; index < len(node.Content); index = index + 2 {
+		keyNode := node.Content[index]
+		valueNode := node.Content[index+1]
+		log.Debugf("traversing %v", keyNode.Value)
+		if keyNode.Value != "<<" {
+			err := overrideEntry(node, keyNode, valueNode, index, context.ChildContext(newContent))
+			if err != nil {
+				return err
+			}
+		} else {
+			if valueNode.Kind == yaml.SequenceNode {
+				log.Debugf("an alias merge list!")
+				for index := len(valueNode.Content) - 1; index >= 0; index = index - 1 {
+					aliasNode := valueNode.Content[index]
+					err := applyAlias(node, aliasNode.Alias, index, context.ChildContext(newContent))
+					if err != nil {
+						return err
+					}
+				}
+			} else {
+				log.Debugf("an alias merge!")
+				err := applyAlias(node, valueNode.Alias, index, context.ChildContext(newContent))
+				if err != nil {
+					return err
+				}
+			}
+		}
+	}
+	node.Content = make([]*yaml.Node, newContent.Len())
+	index := 0
+	for newEl := newContent.Front(); newEl != nil; newEl = newEl.Next() {
+		node.Content[index] = newEl.Value.(*yaml.Node)
+		index++
+	}
+	return nil
+}
+
 func explodeNode(node *yaml.Node, context Context) error {
 	node.Anchor = ""
 	switch node.Kind {
@@ -165,43 +207,35 @@ func explodeNode(node *yaml.Node, context Context) error {
 		}
 		return nil
 	case yaml.MappingNode:
-		var newContent = list.New()
+		// //check the map has an alias in it
+		hasAlias := false
 		for index := 0; index < len(node.Content); index = index + 2 {
 			keyNode := node.Content[index]
-			valueNode := node.Content[index+1]
-			log.Debugf("traversing %v", keyNode.Value)
-			if keyNode.Value != "<<" {
-				err := overrideEntry(node, keyNode, valueNode, index, context.ChildContext(newContent))
+			if keyNode.Value == "<<" {
+				hasAlias = true
+				break
+			}
+		}
+
+		if hasAlias {
+			// this is a slow op, which is why we want to check before running it.
+			return reconstructAliasedMap(node, context)
+		} else {
+			// this map has no aliases, but it's kids might
+			for index := 0; index < len(node.Content); index = index + 2 {
+				keyNode := node.Content[index]
+				valueNode := node.Content[index+1]
+				err := explodeNode(keyNode, context)
 				if err != nil {
 					return err
 				}
-			} else {
-				if valueNode.Kind == yaml.SequenceNode {
-					log.Debugf("an alias merge list!")
-					for index := len(valueNode.Content) - 1; index >= 0; index = index - 1 {
-						aliasNode := valueNode.Content[index]
-						err := applyAlias(node, aliasNode.Alias, index, context.ChildContext(newContent))
-						if err != nil {
-							return err
-						}
-					}
-				} else {
-					log.Debugf("an alias merge!")
-					err := applyAlias(node, valueNode.Alias, index, context.ChildContext(newContent))
-					if err != nil {
-						return err
-					}
+				err = explodeNode(valueNode, context)
+				if err != nil {
+					return err
 				}
 			}
+			return nil
 		}
-		node.Content = make([]*yaml.Node, newContent.Len())
-		index := 0
-		for newEl := newContent.Front(); newEl != nil; newEl = newEl.Next() {
-			node.Content[index] = newEl.Value.(*yaml.Node)
-			index++
-		}
-
-		return nil
 	default:
 		return nil
 	}

pkg/yqlib/operator_anchors_aliases_test.go

@@ -12,6 +12,29 @@ var specDocument = `- &CENTER { x: 1, y: 2 }
 
 var expectedSpecResult = "D0, P[4], (!!map)::x: 1\ny: 2\nr: 10\n"
 
+var simpleArrayRef = `
+item_value: &item_value
+  value: true
+
+thingOne:
+  name: item_1
+  <<: *item_value
+
+thingTwo:
+  name: item_2
+  <<: *item_value
+`
+
+var expectedUpdatedArrayRef = `D0, P[], (doc)::item_value: &item_value
+    value: true
+thingOne:
+    name: item_1
+    value: false
+thingTwo:
+    name: item_2
+    !!merge <<: *item_value
+`
+
 var anchorOperatorScenarios = []expressionScenario{
 	{
 		description:    "Merge one map",
@@ -197,6 +220,13 @@ foobar:
 			"D0, P[], (doc)::{f: {a: cat, b: {f: cat}, cat: {f: cat}}}\n",
 		},
 	},
+	{
+		description:    "Dereference and update a field",
+		subdescription: "`Use explode with multiply to dereference an object",
+		document:       simpleArrayRef,
+		expression:     `.thingOne |= explode(.) * {"value": false}`,
+		expected:       []string{expectedUpdatedArrayRef},
+	},
 }
 
 func TestAnchorAliasOperatorScenarios(t *testing.T) {

pkg/yqlib/operator_assign_test.go

@@ -112,9 +112,10 @@ var assignOperatorScenarios = []expressionScenario{
 		},
 	},
 	{
-		description: "Update selected results",
-		document:    `{a: {b: apple, c: cactus}}`,
-		expression:  `(.a[] | select(. == "apple")) = "frog"`,
+		description:    "Update deeply selected results",
+		subdescription: "Note that the LHS is wrapped in brackets! This is to ensure we dont first filter out the yaml and then update the snippet.",
+		document:       `{a: {b: apple, c: cactus}}`,
+		expression:     `(.a[] | select(. == "apple")) = "frog"`,
 		expected: []string{
 			"D0, P[], (doc)::{a: {b: frog, c: cactus}}\n",
 		},

pkg/yqlib/operator_collect_object.go

@@ -20,8 +20,7 @@ import (
 func collectObjectOperator(d *dataTreeNavigator, originalContext Context, expressionNode *ExpressionNode) (Context, error) {
 	log.Debugf("-- collectObjectOperation")
 
-	context := originalContext.Clone()
-	context.DontAutoCreate = false
+	context := originalContext.WritableClone()
 
 	if context.MatchingNodes.Len() == 0 {
 		node := &yaml.Node{Kind: yaml.MappingNode, Tag: "!!map", Value: "{}"}

pkg/yqlib/operator_contains.go

@@ -0,0 +1,111 @@
+package yqlib
+
+import (
+	"fmt"
+	"strings"
+
+	yaml "gopkg.in/yaml.v3"
+)
+
+func containsOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	return crossFunction(d, context.ReadOnlyClone(), expressionNode, containsWithNodes, false)
+}
+
+func containsArrayElement(array *yaml.Node, item *yaml.Node) (bool, error) {
+	for index := 0; index < len(array.Content); index = index + 1 {
+		containedInArray, err := contains(array.Content[index], item)
+		if err != nil {
+			return false, err
+		}
+		if containedInArray {
+			return true, nil
+		}
+	}
+	return false, nil
+}
+
+func containsArray(lhs *yaml.Node, rhs *yaml.Node) (bool, error) {
+	if rhs.Kind != yaml.SequenceNode {
+		return containsArrayElement(lhs, rhs)
+	}
+	for index := 0; index < len(rhs.Content); index = index + 1 {
+		itemInArray, err := containsArrayElement(lhs, rhs.Content[index])
+		if err != nil {
+			return false, err
+		}
+		if !itemInArray {
+			return false, nil
+		}
+	}
+	return true, nil
+}
+
+func containsObject(lhs *yaml.Node, rhs *yaml.Node) (bool, error) {
+	if rhs.Kind != yaml.MappingNode {
+		return false, nil
+	}
+	for index := 0; index < len(rhs.Content); index = index + 2 {
+		rhsKey := rhs.Content[index]
+		rhsValue := rhs.Content[index+1]
+		log.Debugf("Looking for %v in the lhs", rhsKey.Value)
+		lhsKeyIndex := findInArray(lhs, rhsKey)
+		log.Debugf("index is %v", lhsKeyIndex)
+		if lhsKeyIndex < 0 || lhsKeyIndex%2 != 0 {
+			return false, nil
+		}
+		lhsValue := lhs.Content[lhsKeyIndex+1]
+		log.Debugf("lhsValue is %v", lhsValue.Value)
+
+		itemInArray, err := contains(lhsValue, rhsValue)
+		log.Debugf("rhsValue is %v", rhsValue.Value)
+		if err != nil {
+			return false, err
+		}
+		if !itemInArray {
+			return false, nil
+		}
+	}
+	return true, nil
+}
+
+func containsScalars(lhs *yaml.Node, rhs *yaml.Node) (bool, error) {
+	if lhs.Tag == "!!str" {
+		return strings.Contains(lhs.Value, rhs.Value), nil
+	}
+	return lhs.Value == rhs.Value, nil
+}
+
+func contains(lhs *yaml.Node, rhs *yaml.Node) (bool, error) {
+	switch lhs.Kind {
+	case yaml.MappingNode:
+		return containsObject(lhs, rhs)
+	case yaml.SequenceNode:
+		return containsArray(lhs, rhs)
+	case yaml.ScalarNode:
+		if rhs.Kind != yaml.ScalarNode || lhs.Tag != rhs.Tag {
+			return false, nil
+		}
+		if lhs.Tag == "!!null" {
+			return rhs.Tag == "!!null", nil
+		}
+		return containsScalars(lhs, rhs)
+	}
+
+	return false, fmt.Errorf("%v not yet supported for contains", lhs.Tag)
+}
+
+func containsWithNodes(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {
+	lhs.Node = unwrapDoc(lhs.Node)
+	rhs.Node = unwrapDoc(rhs.Node)
+
+	if lhs.Node.Kind != rhs.Node.Kind {
+		return nil, fmt.Errorf("%v cannot check contained in %v", rhs.Node.Tag, lhs.Node.Tag)
+	}
+
+	result, err := contains(lhs.Node, rhs.Node)
+	if err != nil {
+		return nil, err
+	}
+
+	return createBooleanCandidate(lhs, result), nil
+}

pkg/yqlib/operator_contains_test.go

@@ -0,0 +1,82 @@
+package yqlib
+
+import "testing"
+
+var containsOperatorScenarios = []expressionScenario{
+	{
+		skipDoc:    true,
+		expression: `null | contains(~)`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `3 | contains(3)`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `3 | contains(32)`,
+		expected: []string{
+			"D0, P[], (!!bool)::false\n",
+		},
+	},
+	{
+		description:    "Array contains array",
+		subdescription: "Array is equal or subset of",
+		document:       `["foobar", "foobaz", "blarp"]`,
+		expression:     `contains(["baz", "bar"])`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `["dog", "cat", "giraffe"] | contains(["camel"])`,
+		expected: []string{
+			"D0, P[], (!!bool)::false\n",
+		},
+	},
+	{
+		description: "Object included in array",
+		document:    `{"foo": 12, "bar":[1,2,{"barp":12, "blip":13}]}`,
+		expression:  `contains({"bar": [{"barp": 12}]})`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+	{
+		description: "Object not included in array",
+		document:    `{"foo": 12, "bar":[1,2,{"barp":12, "blip":13}]}`,
+		expression:  `contains({"foo": 12, "bar": [{"barp": 15}]})`,
+		expected: []string{
+			"D0, P[], (!!bool)::false\n",
+		},
+	},
+	{
+		description: "String contains substring",
+		document:    `"foobar"`,
+		expression:  `contains("bar")`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+	{
+		description: "String equals string",
+		document:    `"meow"`,
+		expression:  `contains("meow")`,
+		expected: []string{
+			"D0, P[], (!!bool)::true\n",
+		},
+	},
+}
+
+func TestContainsOperatorScenarios(t *testing.T) {
+	for _, tt := range containsOperatorScenarios {
+		testScenario(t, &tt)
+	}
+	documentScenarios(t, "Contains", containsOperatorScenarios)
+}

pkg/yqlib/operator_encoder_decoder.go

@@ -0,0 +1,91 @@
+package yqlib
+
+import (
+	"bufio"
+	"bytes"
+	"container/list"
+	"regexp"
+	"strings"
+
+	"gopkg.in/yaml.v3"
+)
+
+func yamlToString(candidate *CandidateNode, prefs encoderPreferences) (string, error) {
+	var output bytes.Buffer
+	log.Debug("printing with indent: %v", prefs.indent)
+
+	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), prefs.format, true, false, prefs.indent, true)
+	err := printer.PrintResults(candidate.AsList())
+	return output.String(), err
+}
+
+type encoderPreferences struct {
+	format PrinterOutputFormat
+	indent int
+}
+
+/* encodes object as yaml string */
+
+func encodeOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	preferences := expressionNode.Operation.Preferences.(encoderPreferences)
+	var results = list.New()
+
+	hasOnlyOneNewLine := regexp.MustCompile("[^\n].*\n$")
+	endWithNewLine := regexp.MustCompile(".*\n$")
+	chomper := regexp.MustCompile("\n+$")
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		stringValue, err := yamlToString(candidate, preferences)
+
+		if err != nil {
+			return Context{}, err
+		}
+
+		// remove trailing newlines if needed.
+		// check if we originally decoded this path, and the original thing had a single line.
+		originalList := context.GetVariable("decoded: " + candidate.GetKey())
+		if originalList != nil && originalList.Len() > 0 && hasOnlyOneNewLine.MatchString(stringValue) {
+
+			original := originalList.Front().Value.(*CandidateNode)
+			originalNode := unwrapDoc(original.Node)
+			// original block did not have a new line at the end, get rid of this one too
+			if !endWithNewLine.MatchString(originalNode.Value) {
+				stringValue = chomper.ReplaceAllString(stringValue, "")
+			}
+		}
+
+		// dont print a new line when printing json on a single line.
+		if preferences.format == JsonOutputFormat && preferences.indent == 0 {
+			stringValue = chomper.ReplaceAllString(stringValue, "")
+		}
+
+		stringContentNode := &yaml.Node{Kind: yaml.ScalarNode, Tag: "!!str", Value: stringValue}
+		results.PushBack(candidate.CreateChild(nil, stringContentNode))
+	}
+	return context.ChildContext(results), nil
+}
+
+/* takes a string and decodes it back into an object */
+func decodeOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+
+	var results = list.New()
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+
+		context.SetVariable("decoded: "+candidate.GetKey(), candidate.AsList())
+
+		var dataBucket yaml.Node
+		log.Debugf("got: [%v]", candidate.Node.Value)
+		decoder := yaml.NewDecoder(strings.NewReader(unwrapDoc(candidate.Node).Value))
+		errorReading := decoder.Decode(&dataBucket)
+		if errorReading != nil {
+			return Context{}, errorReading
+		}
+		//first node is a doc
+		node := unwrapDoc(&dataBucket)
+
+		results.PushBack(candidate.CreateChild(nil, node))
+	}
+	return context.ChildContext(results), nil
+}

pkg/yqlib/operator_encoder_decoder_test.go

@@ -0,0 +1,117 @@
+package yqlib
+
+import (
+	"testing"
+)
+
+var prefix = "D0, P[], (doc)::a:\n    cool:\n        bob: dylan\n"
+
+var encoderDecoderOperatorScenarios = []expressionScenario{
+	{
+		description:    "Encode value as yaml string",
+		subdescription: "Indent defaults to 2",
+		document:       "a:\n  cool:\n    bob: dylan",
+		expression:     `.b = (.a | to_yaml)`,
+		expected: []string{
+			prefix + "b: |\n    cool:\n      bob: dylan\n",
+		},
+	},
+	{
+		description:    "Encode value as yaml string, with custom indentation",
+		subdescription: "You can specify the indentation level as the first parameter.",
+		document:       "a:\n  cool:\n    bob: dylan",
+		expression:     `.b = (.a | to_yaml(8))`,
+		expected: []string{
+			prefix + "b: |\n    cool:\n            bob: dylan\n",
+		},
+	},
+	{
+		description:    "Encode value as yaml string, using toyaml",
+		subdescription: "Does the same thing as to_yaml, matching jq naming convention.",
+		document:       `{a: {cool: "thing"}}`,
+		expression:     `.b = (.a | to_yaml)`,
+		expected: []string{
+			`D0, P[], (doc)::{a: {cool: "thing"}, b: "{cool: \"thing\"}\n"}
+`,
+		},
+	},
+	{
+		description: "Encode value as json string",
+		document:    `{a: {cool: "thing"}}`,
+		expression:  `.b = (.a | to_json)`,
+		expected: []string{
+			`D0, P[], (doc)::{a: {cool: "thing"}, b: "{\n  \"cool\": \"thing\"\n}\n"}
+`,
+		},
+	},
+	{
+		description:    "Encode value as json string, on one line",
+		subdescription: "Pass in a 0 indent to print json on a single line.",
+		document:       `{a: {cool: "thing"}}`,
+		expression:     `.b = (.a | to_json(0))`,
+		expected: []string{
+			`D0, P[], (doc)::{a: {cool: "thing"}, b: '{"cool":"thing"}'}
+`,
+		},
+	},
+	{
+		description: "Encode value as props string",
+		document:    `{a: {cool: "thing"}}`,
+		expression:  `.b = (.a | to_props)`,
+		expected: []string{
+			`D0, P[], (doc)::{a: {cool: "thing"}, b: "cool = thing\n"}
+`,
+		},
+	},
+	{
+		description: "Decode a yaml encoded string",
+		document:    `a: "foo: bar"`,
+		expression:  `.b = (.a | from_yaml)`,
+		expected: []string{
+			"D0, P[], (doc)::a: \"foo: bar\"\nb:\n    foo: bar\n",
+		},
+	},
+	{
+		description:           "Update a multiline encoded yaml string",
+		dontFormatInputForDoc: true,
+		document:              "a: |\n  foo: bar\n  baz: dog\n",
+		expression:            `.a |= (from_yaml | .foo = "cat" | to_yaml)`,
+		expected: []string{
+			"D0, P[], (doc)::a: |\n    foo: cat\n    baz: dog\n",
+		},
+	},
+	{
+		skipDoc:               true,
+		dontFormatInputForDoc: true,
+		document:              "a: |-\n  foo: bar\n  baz: dog\n",
+		expression:            `.a |= (from_yaml | .foo = "cat" | to_yaml)`,
+		expected: []string{
+			"D0, P[], (doc)::a: |-\n    foo: cat\n    baz: dog\n",
+		},
+	},
+	{
+		description:           "Update a single line encoded yaml string",
+		dontFormatInputForDoc: true,
+		document:              "a: 'foo: bar'",
+		expression:            `.a |= (from_yaml | .foo = "cat" | to_yaml)`,
+		expected: []string{
+			"D0, P[], (doc)::a: 'foo: cat'\n",
+		},
+	},
+	{
+		skipDoc:               true,
+		dontFormatInputForDoc: true,
+		document:              "a: \"foo: bar\"",
+		expression:            `.a |= (from_yaml | .foo = {"a": "frog"} | to_yaml)`,
+		expected: []string{
+			"D0, P[], (doc)::a: \"foo:\\n  a: frog\"\n",
+		},
+	},
+}
+
+func TestEncoderDecoderOperatorScenarios(t *testing.T) {
+	for _, tt := range encoderDecoderOperatorScenarios {
+		testScenario(t, &tt)
+	}
+	documentScenarios(t, "Encoder and Decoder", encoderDecoderOperatorScenarios)
+}

pkg/yqlib/operator_flatten.go

@@ -0,0 +1,54 @@
+package yqlib
+
+import (
+	"fmt"
+
+	yaml "gopkg.in/yaml.v3"
+)
+
+type flattenPreferences struct {
+	depth int
+}
+
+func flatten(node *yaml.Node, depth int) {
+	if depth == 0 {
+		return
+	}
+	if node.Kind != yaml.SequenceNode {
+		return
+	}
+	content := node.Content
+	newSeq := make([]*yaml.Node, 0)
+
+	for i := 0; i < len(content); i++ {
+		if content[i].Kind == yaml.SequenceNode {
+			flatten(content[i], depth-1)
+			for j := 0; j < len(content[i].Content); j++ {
+				newSeq = append(newSeq, content[i].Content[j])
+			}
+		} else {
+			newSeq = append(newSeq, content[i])
+		}
+	}
+	node.Content = newSeq
+}
+
+func flattenOp(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+
+	log.Debugf("-- flatten Operator")
+	depth := expressionNode.Operation.Preferences.(flattenPreferences).depth
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		candidateNode := unwrapDoc(candidate.Node)
+		if candidateNode.Kind != yaml.SequenceNode {
+			return Context{}, fmt.Errorf("Only arrays are supported for flatten")
+		}
+
+		flatten(candidateNode, depth)
+
+	}
+
+	return context, nil
+
+}

pkg/yqlib/operator_flatten_test.go

@@ -0,0 +1,48 @@
+package yqlib
+
+import (
+	"testing"
+)
+
+var flattenOperatorScenarios = []expressionScenario{
+	{
+		description:    "Flatten",
+		subdescription: "Recursively flattens all arrays",
+		document:       `[1, [2], [[3]]]`,
+		expression:     `flatten`,
+		expected: []string{
+			"D0, P[], (doc)::[1, 2, 3]\n",
+		},
+	},
+	{
+		description: "Flatten with depth of one",
+		document:    `[1, [2], [[3]]]`,
+		expression:  `flatten(1)`,
+		expected: []string{
+			"D0, P[], (doc)::[1, 2, [3]]\n",
+		},
+	},
+	{
+		description: "Flatten empty array",
+		document:    `[[]]`,
+		expression:  `flatten`,
+		expected: []string{
+			"D0, P[], (doc)::[]\n",
+		},
+	},
+	{
+		description: "Flatten array of objects",
+		document:    `[{foo: bar}, [{foo: baz}]]`,
+		expression:  `flatten`,
+		expected: []string{
+			"D0, P[], (doc)::[{foo: bar}, {foo: baz}]\n",
+		},
+	},
+}
+
+func TestFlattenOperatorScenarios(t *testing.T) {
+	for _, tt := range flattenOperatorScenarios {
+		testScenario(t, &tt)
+	}
+	documentScenarios(t, "Flatten", flattenOperatorScenarios)
+}

pkg/yqlib/operator_group_by.go

@@ -0,0 +1,76 @@
+package yqlib
+
+import (
+	"container/list"
+	"fmt"
+
+	"github.com/elliotchance/orderedmap"
+	yaml "gopkg.in/yaml.v3"
+)
+
+func processIntoGroups(d *dataTreeNavigator, context Context, rhsExp *ExpressionNode, node *yaml.Node) (*orderedmap.OrderedMap, error) {
+	var newMatches = orderedmap.NewOrderedMap()
+	for _, node := range node.Content {
+		child := &CandidateNode{Node: node}
+		rhs, err := d.GetMatchingNodes(context.SingleReadonlyChildContext(child), rhsExp)
+
+		if err != nil {
+			return nil, err
+		}
+
+		keyValue := "null"
+
+		if rhs.MatchingNodes.Len() > 0 {
+			first := rhs.MatchingNodes.Front()
+			keyCandidate := first.Value.(*CandidateNode)
+			keyValue = keyCandidate.Node.Value
+		}
+
+		groupList, exists := newMatches.Get(keyValue)
+
+		if !exists {
+			groupList = list.New()
+			newMatches.Set(keyValue, groupList)
+		}
+		groupList.(*list.List).PushBack(node)
+	}
+	return newMatches, nil
+}
+
+func groupBy(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+
+	log.Debugf("-- groupBy Operator")
+	var results = list.New()
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		candidateNode := unwrapDoc(candidate.Node)
+
+		if candidateNode.Kind != yaml.SequenceNode {
+			return Context{}, fmt.Errorf("Only arrays are supported for group by")
+		}
+
+		newMatches, err := processIntoGroups(d, context, expressionNode.Rhs, candidateNode)
+
+		if err != nil {
+			return Context{}, err
+		}
+
+		resultNode := &yaml.Node{Kind: yaml.SequenceNode, Tag: "!!seq"}
+		for groupEl := newMatches.Front(); groupEl != nil; groupEl = groupEl.Next() {
+			groupResultNode := &yaml.Node{Kind: yaml.SequenceNode, Tag: "!!seq"}
+			groupList := groupEl.Value.(*list.List)
+			for groupItem := groupList.Front(); groupItem != nil; groupItem = groupItem.Next() {
+				groupResultNode.Content = append(groupResultNode.Content, groupItem.Value.(*yaml.Node))
+			}
+
+			resultNode.Content = append(resultNode.Content, groupResultNode)
+		}
+
+		results.PushBack(candidate.CreateChild(nil, resultNode))
+
+	}
+
+	return context.ChildContext(results), nil
+
+}

pkg/yqlib/operator_group_by_test.go

@@ -0,0 +1,31 @@
+package yqlib
+
+import (
+	"testing"
+)
+
+var groupByOperatorScenarios = []expressionScenario{
+	{
+		description: "Group by field",
+		document:    `[{foo: 1, bar: 10}, {foo: 3, bar: 100}, {foo: 1, bar: 1}]`,
+		expression:  `group_by(.foo)`,
+		expected: []string{
+			"D0, P[], (!!seq)::- - {foo: 1, bar: 10}\n  - {foo: 1, bar: 1}\n- - {foo: 3, bar: 100}\n",
+		},
+	},
+	{
+		description: "Group by field, with nuls",
+		document:    `[{cat: dog}, {foo: 1, bar: 10}, {foo: 3, bar: 100}, {no: foo for you}, {foo: 1, bar: 1}]`,
+		expression:  `group_by(.foo)`,
+		expected: []string{
+			"D0, P[], (!!seq)::- - {cat: dog}\n  - {no: foo for you}\n- - {foo: 1, bar: 10}\n  - {foo: 1, bar: 1}\n- - {foo: 3, bar: 100}\n",
+		},
+	},
+}
+
+func TestGroupByOperatorScenarios(t *testing.T) {
+	for _, tt := range groupByOperatorScenarios {
+		testScenario(t, &tt)
+	}
+	documentScenarios(t, "Group By", groupByOperatorScenarios)
+}

pkg/yqlib/operator_multiply.go

@@ -6,6 +6,7 @@ import (
 
 	"container/list"
 
+	"github.com/jinzhu/copier"
 	yaml "gopkg.in/yaml.v3"
 )
 
@@ -20,8 +21,29 @@ func multiplyOperator(d *dataTreeNavigator, context Context, expressionNode *Exp
 	return crossFunction(d, context, expressionNode, multiply(expressionNode.Operation.Preferences.(multiplyPreferences)), false)
 }
 
+func getNewBlankNode(lhs *yaml.Node, rhs *yaml.Node) *yaml.Node {
+
+	blankNode := &yaml.Node{}
+
+	if lhs.HeadComment != "" {
+		blankNode.HeadComment = lhs.HeadComment
+	} else if rhs.HeadComment != "" {
+		blankNode.HeadComment = rhs.HeadComment
+	}
+
+	if lhs.FootComment != "" {
+		blankNode.FootComment = lhs.FootComment
+	} else if rhs.FootComment != "" {
+		blankNode.FootComment = rhs.FootComment
+	}
+
+	return blankNode
+}
+
 func multiply(preferences multiplyPreferences) func(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {
 	return func(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {
+		// need to do this before unWrapping the potential document node
+		newBlankNode := getNewBlankNode(lhs.Node, rhs.Node)
 		lhs.Node = unwrapDoc(lhs.Node)
 		rhs.Node = unwrapDoc(rhs.Node)
 		log.Debugf("Multipling LHS: %v", lhs.Node.Tag)
@@ -29,14 +51,21 @@ func multiply(preferences multiplyPreferences) func(d *dataTreeNavigator, contex
 
 		if lhs.Node.Kind == yaml.MappingNode && rhs.Node.Kind == yaml.MappingNode ||
 			(lhs.Node.Kind == yaml.SequenceNode && rhs.Node.Kind == yaml.SequenceNode) {
-
-			var newBlank = lhs.CreateChild(nil, &yaml.Node{})
-
-			var newThing, err = mergeObjects(d, context.WritableClone(), newBlank, lhs, multiplyPreferences{})
+			var newBlank = CandidateNode{}
+			err := copier.CopyWithOption(&newBlank, lhs, copier.Option{IgnoreEmpty: true, DeepCopy: true})
 			if err != nil {
 				return nil, err
 			}
-			return mergeObjects(d, context.WritableClone(), newThing, rhs, preferences)
+			newBlank.Node.HeadComment = newBlankNode.HeadComment
+			newBlank.Node.FootComment = newBlankNode.FootComment
+
+			// var newBlank = lhs.CreateChild(nil, newBlankNode)
+			// log.Debugf("merge - merge lhs into blank")
+			// var newThing, err = mergeObjects(d, context.WritableClone(), newBlank, lhs, multiplyPreferences{})
+			// if err != nil {
+			// 	return nil, err
+			// }
+			return mergeObjects(d, context.WritableClone(), &newBlank, rhs, preferences)
 		} else if lhs.Node.Tag == "!!int" && rhs.Node.Tag == "!!int" {
 			return multiplyIntegers(lhs, rhs)
 		} else if (lhs.Node.Tag == "!!int" || lhs.Node.Tag == "!!float") && (rhs.Node.Tag == "!!int" || rhs.Node.Tag == "!!float") {
@@ -70,25 +99,26 @@ func multiplyIntegers(lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, e
 	target.Node.Style = lhs.Node.Style
 	target.Node.Tag = "!!int"
 
-	lhsNum, err := strconv.Atoi(lhs.Node.Value)
+	format, lhsNum, err := parseInt(lhs.Node.Value)
 	if err != nil {
 		return nil, err
 	}
-	rhsNum, err := strconv.Atoi(rhs.Node.Value)
+	_, rhsNum, err := parseInt(rhs.Node.Value)
 	if err != nil {
 		return nil, err
 	}
-	target.Node.Value = fmt.Sprintf("%v", lhsNum*rhsNum)
+	target.Node.Value = fmt.Sprintf(format, lhsNum*rhsNum)
 	return target, nil
 }
 
 func mergeObjects(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode, preferences multiplyPreferences) (*CandidateNode, error) {
-	shouldAppendArrays := preferences.AppendArrays
 	var results = list.New()
 
-	// shouldn't recurse arrays if appending
-	prefs := recursiveDescentPreferences{RecurseArray: !shouldAppendArrays,
+	// only need to recurse the array if we are doing a deep merge
+	prefs := recursiveDescentPreferences{RecurseArray: preferences.DeepMergeArrays,
 		TraversePreferences: traversePreferences{DontFollowAlias: true, IncludeMapKeys: true}}
+	log.Debugf("merge - preferences.DeepMergeArrays %v", preferences.DeepMergeArrays)
+	log.Debugf("merge - preferences.AppendArrays %v", preferences.AppendArrays)
 	err := recursiveDecent(d, results, context.SingleChildContext(rhs), prefs)
 	if err != nil {
 		return nil, err
@@ -118,14 +148,22 @@ func applyAssignment(d *dataTreeNavigator, context Context, pathIndexToStartFrom
 	log.Debugf("merge - applyAssignment lhs %v, rhs: %v", lhs.GetKey(), rhs.GetKey())
 
 	lhsPath := rhs.Path[pathIndexToStartFrom:]
+	log.Debugf("merge - lhsPath %v", lhsPath)
 
 	assignmentOp := &Operation{OperationType: assignAttributesOpType}
 	if shouldAppendArrays && rhs.Node.Kind == yaml.SequenceNode {
 		assignmentOp.OperationType = addAssignOpType
+		log.Debugf("merge - assignmentOp.OperationType = addAssignOpType")
 	} else if !preferences.DeepMergeArrays && rhs.Node.Kind == yaml.SequenceNode ||
 		(rhs.Node.Kind == yaml.ScalarNode || rhs.Node.Kind == yaml.AliasNode) {
 		assignmentOp.OperationType = assignOpType
 		assignmentOp.UpdateAssign = false
+		log.Debugf("merge - rhs.Node.Kind == yaml.SequenceNode: %v", rhs.Node.Kind == yaml.SequenceNode)
+		log.Debugf("merge - rhs.Node.Kind == yaml.ScalarNode: %v", rhs.Node.Kind == yaml.ScalarNode)
+		log.Debugf("merge - rhs.Node.Kind == yaml.AliasNode: %v", rhs.Node.Kind == yaml.AliasNode)
+		log.Debugf("merge - assignmentOp.OperationType = assignOpType, no updateassign")
+	} else {
+		log.Debugf("merge - assignmentOp := &Operation{OperationType: assignAttributesOpType}")
 	}
 	rhsOp := &Operation{OperationType: valueOpType, CandidateNode: rhs}
 

pkg/yqlib/operator_multiply_test.go

@@ -24,13 +24,121 @@ list2:
     - "123"
 `
 
+var mergeArrayWithAnchors = `sample:
+- &a
+- <<: *a
+`
+
 var mergeArraysObjectKeysText = `It's a complex command, the trickyness comes from needing to have the right context in the expressions.
 First we save the second array into a variable '$two' which lets us reference it later.
 We then need to update the first array. We will use the relative update (|=) because we need to update relative to the current element of the array in the LHS in the RHS expression. 
 We set the current element of the first array as $cur. Now we multiply (merge) $cur with the matching entry in $two, by passing $two through a select filter.
 `
 
+var docWithHeader = `
+# here
+
+a: apple
+`
+
+var nodeWithHeader = `
+# here
+a: apple
+`
+
+var docNoComments = `
+b: banana
+`
+
+var docWithFooter = `
+a: apple
+
+# footer
+`
+
+var nodeWithFooter = `
+a: apple
+# footer`
+
+var document = `
+a: &cat {name: cat}
+b: {name: dog}
+c: 
+  <<: *cat
+`
+
 var multiplyOperatorScenarios = []expressionScenario{
+	{
+		skipDoc:    true,
+		document:   mergeArrayWithAnchors,
+		expression: `. * .`,
+		expected: []string{
+			"D0, P[], (!!map)::sample:\n    - &a\n    - !!merge <<: *a\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   docWithHeader,
+		document2:  docNoComments,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::# here\na: apple\nb: banana\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   nodeWithHeader,
+		document2:  docNoComments,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::# here\na: apple\nb: banana\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   docNoComments,
+		document2:  docWithHeader,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::# here\nb: banana\na: apple\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   docNoComments,
+		document2:  nodeWithHeader,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::b: banana\n# here\na: apple\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   docWithFooter,
+		document2:  docNoComments,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::a: apple\nb: banana\n\n# footer\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   nodeWithFooter,
+		document2:  docNoComments,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{ // not sure why there's an extra newline *shrug*
+			"D0, P[], (!!map)::a: apple\n# footer\n\nb: banana\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   docNoComments,
+		document2:  docWithFooter,
+		expression: `select(fi == 0) * select(fi == 1)`,
+		expected: []string{
+			"D0, P[], (!!map)::b: banana\na: apple\n\n# footer\n",
+		},
+	},
 	{
 		description: "Multiply integers",
 		expression:  `3 * 4`,
@@ -304,7 +412,15 @@ var multiplyOperatorScenarios = []expressionScenario{
 		document:    mergeDocSample,
 		expression:  `.foobar * .foobarList`,
 		expected: []string{
-			"D0, P[foobar], (!!map)::c: foobarList_c\n<<: [*foo, *bar]\nthing: foobar_thing\nb: foobarList_b\n",
+			"D0, P[foobar], (!!map)::c: foobarList_c\n!!merge <<: [*foo, *bar]\nthing: foobar_thing\nb: foobarList_b\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   document,
+		expression: `.b * .c`,
+		expected: []string{
+			"D0, P[b], (!!map)::{name: dog, <<: *cat}\n",
 		},
 	},
 }

pkg/yqlib/operator_select.go

@@ -17,22 +17,25 @@ func selectOperator(d *dataTreeNavigator, context Context, expressionNode *Expre
 			return Context{}, err
 		}
 
-		// grab the first value
-		first := rhs.MatchingNodes.Front()
+		// find any truthy node
+		var errDecoding error
+		includeResult := false
 
-		if first != nil {
-			result := first.Value.(*CandidateNode)
-			log.Debugf("result %v", NodeToString(result))
-			includeResult, errDecoding := isTruthy(result)
+		for resultEl := rhs.MatchingNodes.Front(); resultEl != nil; resultEl = resultEl.Next() {
+			result := resultEl.Value.(*CandidateNode)
+			includeResult, errDecoding = isTruthy(result)
 			log.Debugf("isTruthy %v", includeResult)
 			if errDecoding != nil {
 				return Context{}, errDecoding
 			}
-
 			if includeResult {
-				results.PushBack(candidate)
+				break
 			}
 		}
+
+		if includeResult {
+			results.PushBack(candidate)
+		}
 	}
 	return context.ChildContext(results), nil
 }

pkg/yqlib/operator_select_test.go

@@ -5,6 +5,28 @@ import (
 )
 
 var selectOperatorScenarios = []expressionScenario{
+	{
+		skipDoc:    true,
+		document:   `cat`,
+		expression: `select(false, true)`,
+		expected: []string{
+			"D0, P[], (doc)::cat\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   `cat`,
+		expression: `select(true, false)`,
+		expected: []string{
+			"D0, P[], (doc)::cat\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   `cat`,
+		expression: `select(false)`,
+		expected:   []string{},
+	},
 	{
 		description: "Select elements from array",
 		document:    `[cat,goat,dog]`,

pkg/yqlib/operator_strings.go

@@ -74,6 +74,232 @@ func substituteStringOperator(d *dataTreeNavigator, context Context, expressionN
 
 }
 
+func addMatch(original []*yaml.Node, match string, offset int, name string) []*yaml.Node {
+
+	newContent := append(original,
+		createScalarNode("string", "string"))
+
+	if offset < 0 {
+		// offset of -1 means there was no match, force a null value like jq
+		newContent = append(newContent,
+			createScalarNode(nil, "null"),
+		)
+	} else {
+		newContent = append(newContent,
+			createScalarNode(match, match),
+		)
+	}
+
+	newContent = append(newContent,
+		createScalarNode("offset", "offset"),
+		createScalarNode(offset, fmt.Sprintf("%v", offset)),
+		createScalarNode("length", "length"),
+		createScalarNode(len(match), fmt.Sprintf("%v", len(match))))
+
+	if name != "" {
+		newContent = append(newContent,
+			createScalarNode("name", "name"),
+			createScalarNode(name, name),
+		)
+	}
+	return newContent
+}
+
+type matchPreferences struct {
+	Global bool
+}
+
+func getMatches(matchPrefs matchPreferences, regEx *regexp.Regexp, value string) ([][]string, [][]int) {
+	var allMatches [][]string
+	var allIndices [][]int
+
+	if matchPrefs.Global {
+		allMatches = regEx.FindAllStringSubmatch(value, -1)
+		allIndices = regEx.FindAllStringSubmatchIndex(value, -1)
+	} else {
+		allMatches = [][]string{regEx.FindStringSubmatch(value)}
+		allIndices = [][]int{regEx.FindStringSubmatchIndex(value)}
+	}
+
+	log.Debug("allMatches, %v", allMatches)
+	return allMatches, allIndices
+}
+
+func match(matchPrefs matchPreferences, regEx *regexp.Regexp, candidate *CandidateNode, value string, results *list.List) {
+	subNames := regEx.SubexpNames()
+	allMatches, allIndices := getMatches(matchPrefs, regEx, value)
+
+	// if all matches just has an empty array in it,
+	// then nothing matched
+	if len(allMatches) > 0 && len(allMatches[0]) == 0 {
+		return
+	}
+
+	for i, matches := range allMatches {
+		capturesNode := &yaml.Node{Kind: yaml.SequenceNode}
+		match, submatches := matches[0], matches[1:]
+		for j, submatch := range submatches {
+			captureNode := &yaml.Node{Kind: yaml.MappingNode}
+			captureNode.Content = addMatch(capturesNode.Content, submatch, allIndices[i][2+j*2], subNames[j+1])
+			capturesNode.Content = append(capturesNode.Content, captureNode)
+		}
+
+		node := &yaml.Node{Kind: yaml.MappingNode}
+		node.Content = addMatch(node.Content, match, allIndices[i][0], "")
+		node.Content = append(node.Content,
+			createScalarNode("captures", "captures"),
+			capturesNode,
+		)
+		results.PushBack(candidate.CreateChild(nil, node))
+
+	}
+
+}
+
+func capture(matchPrefs matchPreferences, regEx *regexp.Regexp, candidate *CandidateNode, value string, results *list.List) {
+	subNames := regEx.SubexpNames()
+	allMatches, allIndices := getMatches(matchPrefs, regEx, value)
+
+	// if all matches just has an empty array in it,
+	// then nothing matched
+	if len(allMatches) > 0 && len(allMatches[0]) == 0 {
+		return
+	}
+
+	for i, matches := range allMatches {
+		capturesNode := &yaml.Node{Kind: yaml.MappingNode}
+
+		_, submatches := matches[0], matches[1:]
+		for j, submatch := range submatches {
+			capturesNode.Content = append(capturesNode.Content,
+				createScalarNode(subNames[j+1], subNames[j+1]))
+
+			offset := allIndices[i][2+j*2]
+			// offset of -1 means there was no match, force a null value like jq
+			if offset < 0 {
+				capturesNode.Content = append(capturesNode.Content,
+					createScalarNode(nil, "null"),
+				)
+			} else {
+				capturesNode.Content = append(capturesNode.Content,
+					createScalarNode(submatch, submatch),
+				)
+			}
+		}
+
+		results.PushBack(candidate.CreateChild(nil, capturesNode))
+
+	}
+
+}
+
+func extractMatchArguments(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (*regexp.Regexp, matchPreferences, error) {
+	regExExpNode := expressionNode.Rhs
+
+	matchPrefs := matchPreferences{}
+
+	// we got given parameters e.g. match(exp; params)
+	if expressionNode.Rhs.Operation.OperationType == blockOpType {
+		block := expressionNode.Rhs
+		regExExpNode = block.Lhs
+		replacementNodes, err := d.GetMatchingNodes(context, block.Rhs)
+		if err != nil {
+			return nil, matchPrefs, err
+		}
+		paramText := ""
+		if replacementNodes.MatchingNodes.Front() != nil {
+			paramText = replacementNodes.MatchingNodes.Front().Value.(*CandidateNode).Node.Value
+		}
+		if strings.Contains(paramText, "g") {
+			paramText = strings.ReplaceAll(paramText, "g", "")
+			matchPrefs.Global = true
+		}
+		if strings.Contains(paramText, "i") {
+			return nil, matchPrefs, fmt.Errorf(`'i' is not a valid option for match. To ignore case, use an expression like match("(?i)cat")`)
+		}
+		if len(paramText) > 0 {
+			return nil, matchPrefs, fmt.Errorf(`Unrecognised match params '%v', please see docs at https://mikefarah.gitbook.io/yq/operators/string-operators`, paramText)
+		}
+	}
+
+	regExNodes, err := d.GetMatchingNodes(context.ReadOnlyClone(), regExExpNode)
+	if err != nil {
+		return nil, matchPrefs, err
+	}
+	log.Debug(NodesToString(regExNodes.MatchingNodes))
+	regExStr := ""
+	if regExNodes.MatchingNodes.Front() != nil {
+		regExStr = regExNodes.MatchingNodes.Front().Value.(*CandidateNode).Node.Value
+	}
+	log.Debug("regEx %v", regExStr)
+	regEx, err := regexp.Compile(regExStr)
+	return regEx, matchPrefs, err
+}
+
+func matchOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	regEx, matchPrefs, err := extractMatchArguments(d, context, expressionNode)
+	if err != nil {
+		return Context{}, err
+	}
+
+	var results = list.New()
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		node := unwrapDoc(candidate.Node)
+		if node.Tag != "!!str" {
+			return Context{}, fmt.Errorf("cannot match with %v, can only match strings. Hint: Most often you'll want to use '|=' over '=' for this operation", node.Tag)
+		}
+
+		match(matchPrefs, regEx, candidate, node.Value, results)
+	}
+
+	return context.ChildContext(results), nil
+}
+
+func captureOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	regEx, matchPrefs, err := extractMatchArguments(d, context, expressionNode)
+	if err != nil {
+		return Context{}, err
+	}
+
+	var results = list.New()
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		node := unwrapDoc(candidate.Node)
+		if node.Tag != "!!str" {
+			return Context{}, fmt.Errorf("cannot match with %v, can only match strings. Hint: Most often you'll want to use '|=' over '=' for this operation", node.Tag)
+		}
+		capture(matchPrefs, regEx, candidate, node.Value, results)
+
+	}
+
+	return context.ChildContext(results), nil
+}
+
+func testOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	regEx, _, err := extractMatchArguments(d, context, expressionNode)
+	if err != nil {
+		return Context{}, err
+	}
+
+	var results = list.New()
+
+	for el := context.MatchingNodes.Front(); el != nil; el = el.Next() {
+		candidate := el.Value.(*CandidateNode)
+		node := unwrapDoc(candidate.Node)
+		if node.Tag != "!!str" {
+			return Context{}, fmt.Errorf("cannot match with %v, can only match strings. Hint: Most often you'll want to use '|=' over '=' for this operation", node.Tag)
+		}
+		matches := regEx.FindStringSubmatch(node.Value)
+		results.PushBack(createBooleanCandidate(candidate, len(matches) > 0))
+
+	}
+
+	return context.ChildContext(results), nil
+}
+
 func joinStringOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
 	log.Debugf("-- joinStringOperator")
 	joinStr := ""

pkg/yqlib/operator_strings_test.go

@@ -13,6 +13,110 @@ var stringsOperatorScenarios = []expressionScenario{
 			"D0, P[], (!!str)::cat; meow; 1; ; true\n",
 		},
 	},
+	{
+		description: "Match string",
+		document:    `foo bar foo`,
+		expression:  `match("foo")`,
+		expected: []string{
+			"D0, P[], ()::string: foo\noffset: 0\nlength: 3\ncaptures: []\n",
+		},
+	},
+	{
+		description: "Match string, case insensitive",
+		document:    `foo bar FOO`,
+		expression:  `[match("(?i)foo"; "g")]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- string: foo\n  offset: 0\n  length: 3\n  captures: []\n- string: FOO\n  offset: 8\n  length: 3\n  captures: []\n",
+		},
+	},
+	{
+		description: "Match with capture groups",
+		document:    `abc abc`,
+		expression:  `[match("(abc)+"; "g")]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- string: abc\n  offset: 0\n  length: 3\n  captures:\n    - string: abc\n      offset: 0\n      length: 3\n- string: abc\n  offset: 4\n  length: 3\n  captures:\n    - string: abc\n      offset: 4\n      length: 3\n",
+		},
+	},
+	{
+		description: "Match with named capture groups",
+		document:    `foo bar foo foo  foo`,
+		expression:  `[match("foo (?P<bar123>bar)? foo"; "g")]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- string: foo bar foo\n  offset: 0\n  length: 11\n  captures:\n    - string: bar\n      offset: 4\n      length: 3\n      name: bar123\n- string: foo  foo\n  offset: 12\n  length: 8\n  captures:\n    - string: null\n      offset: -1\n      length: 0\n      name: bar123\n",
+		},
+	},
+	{
+		description: "Capture named groups into a map",
+		document:    `xyzzy-14`,
+		expression:  `capture("(?P<a>[a-z]+)-(?P<n>[0-9]+)")`,
+		expected: []string{
+			"D0, P[], ()::a: xyzzy\nn: \"14\"\n",
+		},
+	},
+	{
+		skipDoc:     true,
+		description: "Capture named groups into a map, with null",
+		document:    `xyzzy-14`,
+		expression:  `capture("(?P<a>[a-z]+)-(?P<n>[0-9]+)(?P<bar123>bar)?")`,
+		expected: []string{
+			"D0, P[], ()::a: xyzzy\nn: \"14\"\nbar123: null\n",
+		},
+	},
+	{
+		description: "Match without global flag",
+		document:    `cat cat`,
+		expression:  `match("cat")`,
+		expected: []string{
+			"D0, P[], ()::string: cat\noffset: 0\nlength: 3\ncaptures: []\n",
+		},
+	},
+	{
+		description: "Match with global flag",
+		document:    `cat cat`,
+		expression:  `[match("cat"; "g")]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- string: cat\n  offset: 0\n  length: 3\n  captures: []\n- string: cat\n  offset: 4\n  length: 3\n  captures: []\n",
+		},
+	},
+	{
+		skipDoc:     true,
+		description: "No match",
+		document:    `dog`,
+		expression:  `match("cat"; "g")`,
+		expected:    []string{},
+	},
+	{
+		skipDoc:     true,
+		description: "No match",
+		expression:  `"dog" | match("cat", "g")`,
+		expected:    []string{},
+	},
+	{
+		skipDoc:     true,
+		description: "No match",
+		expression:  `"dog" | match("cat")`,
+		expected:    []string{},
+	},
+	{
+		description:    "Test using regex",
+		subdescription: "Like jq'q equivalant, this works like match but only returns true/false instead of full match details",
+		document:       `["cat", "dog"]`,
+		expression:     `.[] | test("at")`,
+		expected: []string{
+			"D0, P[0], (!!bool)::true\n",
+			"D0, P[1], (!!bool)::false\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   `["cat*", "cat*", "cat"]`,
+		expression: `.[] | test("cat\*")`,
+		expected: []string{
+			"D0, P[0], (!!bool)::true\n",
+			"D0, P[1], (!!bool)::true\n",
+			"D0, P[2], (!!bool)::false\n",
+		},
+	},
 	{
 		description:    "Substitute / Replace string",
 		subdescription: "This uses golang regex, described [here](https://github.com/google/re2/wiki/Syntax)\nNote the use of `|=` to run in context of the current string value.",

pkg/yqlib/operator_style_test.go

@@ -14,10 +14,9 @@ var styleOperatorScenarios = []expressionScenario{
 		},
 	},
 	{
-		description:    "Update and set style of a particular node using path variables",
-		subdescription: "You can use a variable to re-use a path",
-		document:       `a: {b: thing, c: something}`,
-		expression:     `.a.b as $x | $x = "new" | $x style="double"`,
+		description: "Update and set style of a particular node using path variables",
+		document:    `a: {b: thing, c: something}`,
+		expression:  `with(.a.b ; . = "new" | . style="double")`,
 		expected: []string{
 			"D0, P[], (doc)::a: {b: \"new\", c: something}\n",
 		},

pkg/yqlib/operator_subtract.go

@@ -5,7 +5,7 @@ import (
 
 	"strconv"
 
-	yaml "gopkg.in/yaml.v3"
+	"gopkg.in/yaml.v3"
 )
 
 func createSubtractOp(lhs *ExpressionNode, rhs *ExpressionNode) *ExpressionNode {
@@ -24,6 +24,24 @@ func subtractOperator(d *dataTreeNavigator, context Context, expressionNode *Exp
 	return crossFunction(d, context.ReadOnlyClone(), expressionNode, subtract, false)
 }
 
+func subtractArray(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {
+	newLhsArray := make([]*yaml.Node, 0)
+
+	for lindex := 0; lindex < len(lhs.Node.Content); lindex = lindex + 1 {
+		shouldInclude := true
+		for rindex := 0; rindex < len(rhs.Node.Content) && shouldInclude; rindex = rindex + 1 {
+			if recursiveNodeEqual(lhs.Node.Content[lindex], rhs.Node.Content[rindex]) {
+				shouldInclude = false
+			}
+		}
+		if shouldInclude {
+			newLhsArray = append(newLhsArray, lhs.Node.Content[lindex])
+		}
+	}
+	lhs.Node.Content = newLhsArray
+	return lhs, nil
+}
+
 func subtract(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *CandidateNode) (*CandidateNode, error) {
 	lhs.Node = unwrapDoc(lhs.Node)
 	rhs.Node = unwrapDoc(rhs.Node)
@@ -40,14 +58,13 @@ func subtract(d *dataTreeNavigator, context Context, lhs *CandidateNode, rhs *Ca
 	case yaml.MappingNode:
 		return nil, fmt.Errorf("Maps not yet supported for subtraction")
 	case yaml.SequenceNode:
-		return nil, fmt.Errorf("Sequences not yet supported for subtraction")
-		// target.Node.Kind = yaml.SequenceNode
-		// target.Node.Style = lhsNode.Style
-		// target.Node.Tag = "!!seq"
-		// target.Node.Content = append(lhsNode.Content, toNodes(rhs)...)
+		if rhs.Node.Kind != yaml.SequenceNode {
+			return nil, fmt.Errorf("%v (%v) cannot be subtracted from %v", rhs.Node.Tag, rhs.Path, lhsNode.Tag)
+		}
+		return subtractArray(d, context, lhs, rhs)
 	case yaml.ScalarNode:
 		if rhs.Node.Kind != yaml.ScalarNode {
-			return nil, fmt.Errorf("%v (%v) cannot be added to a %v", rhs.Node.Tag, rhs.Path, lhsNode.Tag)
+			return nil, fmt.Errorf("%v (%v) cannot be subtracted from %v", rhs.Node.Tag, rhs.Path, lhsNode.Tag)
 		}
 		target.Node.Kind = yaml.ScalarNode
 		target.Node.Style = lhsNode.Style
@@ -62,17 +79,17 @@ func subtractScalars(target *CandidateNode, lhs *yaml.Node, rhs *yaml.Node) (*Ca
 	if lhs.Tag == "!!str" {
 		return nil, fmt.Errorf("strings cannot be subtracted")
 	} else if lhs.Tag == "!!int" && rhs.Tag == "!!int" {
-		lhsNum, err := strconv.Atoi(lhs.Value)
+		format, lhsNum, err := parseInt(lhs.Value)
 		if err != nil {
 			return nil, err
 		}
-		rhsNum, err := strconv.Atoi(rhs.Value)
+		_, rhsNum, err := parseInt(rhs.Value)
 		if err != nil {
 			return nil, err
 		}
 		result := lhsNum - rhsNum
 		target.Node.Tag = "!!int"
-		target.Node.Value = fmt.Sprintf("%v", result)
+		target.Node.Value = fmt.Sprintf(format, result)
 	} else if (lhs.Tag == "!!int" || lhs.Tag == "!!float") && (rhs.Tag == "!!int" || rhs.Tag == "!!float") {
 		lhsNum, err := strconv.ParseFloat(lhs.Value, 64)
 		if err != nil {

pkg/yqlib/operator_subtract_test.go

@@ -13,6 +13,51 @@ var subtractOperatorScenarios = []expressionScenario{
 			"D0, P[], (doc)::{}\n",
 		},
 	},
+	{
+		description: "Array subtraction",
+		expression:  `[1,2] - [2,3]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- 1\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `[2,1,2,2] - [2,3]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- 1\n",
+		},
+	},
+	{
+		description: "Array subtraction with nested array",
+		expression:  `[[1], 1, 2] - [[1], 3]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- 1\n- 2\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `[[1], 1, [[[2]]]] - [[1], [[[3]]]]`,
+		expected: []string{
+			"D0, P[], (!!seq)::- 1\n- - - - 2\n",
+		},
+	},
+	{
+		description:    "Array subtraction with nested object",
+		subdescription: `Note that order of the keys does not matter`,
+		document:       `[{a: b, c: d}, {a: b}]`,
+		expression:     `. - [{"c": "d", "a": "b"}]`,
+		expected: []string{
+			"D0, P[], (!!seq)::[{a: b}]\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		document:   `[{a: [1], c: d}, {a: [2], c: d}, {a: b}]`,
+		expression: `. - [{"c": "d", "a": [1]}]`,
+		expected: []string{
+			"D0, P[], (!!seq)::[{a: [2], c: d}, {a: b}]\n",
+		},
+	},
 	{
 		description:    "Number subtraction - float",
 		subdescription: "If the lhs or rhs are floats then the expression will be calculated with floats.",

pkg/yqlib/operator_union.go

@@ -1,17 +1,36 @@
 package yqlib
 
 func unionOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	log.Debug("unionOperator")
+	log.Debug("context: %v", NodesToString(context.MatchingNodes))
 	lhs, err := d.GetMatchingNodes(context, expressionNode.Lhs)
 	if err != nil {
 		return Context{}, err
 	}
+	log.Debug("lhs: %v", NodesToString(lhs.MatchingNodes))
+	log.Debug("rhs input: %v", NodesToString(context.MatchingNodes))
+	log.Debug("rhs: %v", expressionNode.Rhs.Operation.toString())
 	rhs, err := d.GetMatchingNodes(context, expressionNode.Rhs)
+
 	if err != nil {
 		return Context{}, err
 	}
-	for el := rhs.MatchingNodes.Front(); el != nil; el = el.Next() {
-		node := el.Value.(*CandidateNode)
-		lhs.MatchingNodes.PushBack(node)
+	log.Debug("lhs: %v", lhs.ToString())
+	log.Debug("rhs: %v", rhs.ToString())
+
+	// this can happen when both expressions modify the context
+	// instead of creating their own.
+	/// (.foo = "bar"), (.thing = "cat")
+
+	if rhs.MatchingNodes != lhs.MatchingNodes {
+
+		for el := rhs.MatchingNodes.Front(); el != nil; el = el.Next() {
+			node := el.Value.(*CandidateNode)
+			log.Debug("processing %v", NodeToString(node))
+
+			lhs.MatchingNodes.PushBack(node)
+		}
 	}
+	log.Debug("all together: %v", lhs.ToString())
 	return lhs, nil
 }

pkg/yqlib/operator_union_test.go

@@ -5,20 +5,21 @@ import (
 )
 
 var unionOperatorScenarios = []expressionScenario{
-	// {
-	// 	skipDoc:    true,
-	// 	document:   "{}",
-	// 	expression: `(.a, .b.c) as $x`,
-	// 	expected: []string{
-	// 		"D0, P[], (doc)::{}\n",
-	// 	},
-	// },
-	// {
-	// 	skipDoc:    true,
-	// 	document:   "{}",
-	// 	expression: `(.a, .b.c)`,
-	// 	expected:   []string{},
-	// },
+	{
+		skipDoc:    true,
+		document:   "{}",
+		expression: `(.a, .b.c) as $x`,
+		expected: []string{
+			"D0, P[], (doc)::{}\n",
+		},
+	},
+	{
+		skipDoc:    true,
+		expression: `(.foo = "bar"), (.toe = "jam")`,
+		expected: []string{
+			"D0, P[], ()::foo: bar\ntoe: jam\n",
+		},
+	},
 	{
 		description: "Combine scalars",
 		expression:  `1, true, "cat"`,

pkg/yqlib/operator_value_test.go

@@ -12,6 +12,76 @@ var valueOperatorScenarios = []expressionScenario{
 			"D0, P[], (!!int)::1\n",
 		},
 	},
+	{
+		document:   ``,
+		expression: `0x9f`,
+		expected: []string{
+			"D0, P[], (!!int)::0x9f\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `0x1A`,
+		expected: []string{
+			"D0, P[], (!!int)::0x1A\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `0x1A + 2`,
+		expected: []string{
+			"D0, P[], (!!int)::0x1C\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `0x12 * 2`,
+		expected: []string{
+			"D0, P[], (!!int)::0x24\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `0xF - 1`,
+		expected: []string{
+			"D0, P[], (!!int)::0xE\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `12`,
+		expected: []string{
+			"D0, P[], (!!int)::12\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `12 + 2`,
+		expected: []string{
+			"D0, P[], (!!int)::14\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `12 * 2`,
+		expected: []string{
+			"D0, P[], (!!int)::24\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `12 - 2`,
+		expected: []string{
+			"D0, P[], (!!int)::10\n",
+		},
+	},
+	{
+		document:   ``,
+		expression: `0X12`,
+		expected: []string{
+			"D0, P[], (!!int)::0X12\n",
+		},
+	},
 	{
 		document:   ``,
 		expression: `-1`,

pkg/yqlib/operator_variables.go

@@ -15,6 +15,10 @@ func getVariableOperator(d *dataTreeNavigator, context Context, expressionNode *
 	return context.ChildContext(result), nil
 }
 
+type assignVarPreferences struct {
+	IsReference bool
+}
+
 func assignVariableOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
 	lhs, err := d.GetMatchingNodes(context.ReadOnlyClone(), expressionNode.Lhs)
 	if err != nil {
@@ -24,6 +28,15 @@ func assignVariableOperator(d *dataTreeNavigator, context Context, expressionNod
 		return Context{}, fmt.Errorf("RHS of 'as' operator must be a variable name e.g. $foo")
 	}
 	variableName := expressionNode.Rhs.Operation.StringValue
-	context.SetVariable(variableName, lhs.MatchingNodes)
+
+	prefs := expressionNode.Operation.Preferences.(assignVarPreferences)
+
+	var variableValue *list.List
+	if prefs.IsReference {
+		variableValue = lhs.MatchingNodes
+	} else {
+		variableValue = lhs.DeepClone().MatchingNodes
+	}
+	context.SetVariable(variableName, variableValue)
 	return context, nil
 }

pkg/yqlib/operator_variables_test.go

@@ -34,15 +34,32 @@ var variableOperatorScenarios = []expressionScenario{
 		description:    "Using variables as a lookup",
 		subdescription: "Example taken from [jq](https://stedolan.github.io/jq/manual/#Variable/SymbolicBindingOperator:...as$identifier|...)",
 		document: `{"posts": [{"title": "Frist psot", "author": "anon"},
-		{"title": "A well-written article", "author": "person1"}],
-"realnames": {"anon": "Anonymous Coward",
-				"person1": "Person McPherson"}}`,
+			{"title": "A well-written article", "author": "person1"}],
+	"realnames": {"anon": "Anonymous Coward",
+					"person1": "Person McPherson"}}`,
 		expression: `.realnames as $names | .posts[] | {"title":.title, "author": $names[.author]}`,
 		expected: []string{
 			"D0, P[], (!!map)::title: \"Frist psot\"\nauthor: \"Anonymous Coward\"\n",
 			"D0, P[], (!!map)::title: \"A well-written article\"\nauthor: \"Person McPherson\"\n",
 		},
 	},
+	{
+		description: "Using variables to swap values",
+		document:    "a: a_value\nb: b_value",
+		expression:  `.a as $x  | .b as $y | .b = $x | .a = $y`,
+		expected: []string{
+			"D0, P[], (doc)::a: b_value\nb: a_value\n",
+		},
+	},
+	{
+		description:    "Use ref to reference a path repeatedly",
+		subdescription: "Note: You may find the `with` operator more useful.",
+		document:       `a: {b: thing, c: something}`,
+		expression:     `.a.b ref $x | $x = "new" | $x style="double"`,
+		expected: []string{
+			"D0, P[], (doc)::a: {b: \"new\", c: something}\n",
+		},
+	},
 }
 
 func TestVariableOperatorScenarios(t *testing.T) {

pkg/yqlib/operator_with.go

@@ -0,0 +1,30 @@
+package yqlib
+
+import "fmt"
+
+func withOperator(d *dataTreeNavigator, context Context, expressionNode *ExpressionNode) (Context, error) {
+	log.Debugf("-- withOperator")
+	// with(path, exp)
+
+	if expressionNode.Rhs.Operation.OperationType != blockOpType {
+		return Context{}, fmt.Errorf("with must be given a block, got %v instead", expressionNode.Rhs.Operation.OperationType.Type)
+	}
+
+	pathExp := expressionNode.Rhs.Lhs
+
+	updateContext, err := d.GetMatchingNodes(context, pathExp)
+
+	if err != nil {
+		return Context{}, err
+	}
+
+	updateExp := expressionNode.Rhs.Rhs
+
+	_, err = d.GetMatchingNodes(updateContext, updateExp)
+	if err != nil {
+		return Context{}, err
+	}
+
+	return context, nil
+
+}

pkg/yqlib/operator_with_test.go

@@ -0,0 +1,38 @@
+package yqlib
+
+import "testing"
+
+var withOperatorScenarios = []expressionScenario{
+	{
+		description: "Update and style",
+		document:    `a: {deeply: {nested: value}}`,
+		expression:  `with(.a.deeply.nested; . = "newValue" | . style="single")`,
+		expected: []string{
+			"D0, P[], (doc)::a: {deeply: {nested: 'newValue'}}\n",
+		},
+	},
+	{
+		description: "Update multiple deeply nested properties",
+		document:    `a: {deeply: {nested: value, other: thing}}`,
+		expression:  `with(.a.deeply; .nested = "newValue" | .other= "newThing")`,
+		expected: []string{
+			"D0, P[], (doc)::a: {deeply: {nested: newValue, other: newThing}}\n",
+		},
+	},
+	{
+		description:    "Update array elements relatively",
+		subdescription: "The second expression runs with each element of the array as it's contextual root. This allows you to make updates relative to the element.",
+		document:       `myArray: [{a: apple},{a: banana}]`,
+		expression:     `with(.myArray[]; .b = .a + " yum")`,
+		expected: []string{
+			"D0, P[], (doc)::myArray: [{a: apple, b: apple yum}, {a: banana, b: banana yum}]\n",
+		},
+	},
+}
+
+func TestWithOperatorScenarios(t *testing.T) {
+	for _, tt := range withOperatorScenarios {
+		testScenario(t, &tt)
+	}
+	documentScenarios(t, "With", withOperatorScenarios)
+}

pkg/yqlib/operators_test.go

@@ -107,14 +107,14 @@ func copyFromHeader(title string, out *os.File) error {
 
 func formatYaml(yaml string, filename string) string {
 	var output bytes.Buffer
-	printer := NewPrinter(bufio.NewWriter(&output), false, true, false, 2, true)
+	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
 
 	node, err := NewExpressionParser().ParseExpression(".. style= \"\"")
 	if err != nil {
 		panic(err)
 	}
 	streamEvaluator := NewStreamEvaluator()
-	err = streamEvaluator.Evaluate(filename, strings.NewReader(yaml), node, printer)
+	_, err = streamEvaluator.Evaluate(filename, strings.NewReader(yaml), node, printer, "")
 	if err != nil {
 		panic(err)
 	}
@@ -216,7 +216,7 @@ func documentInput(w *bufio.Writer, s expressionScenario) (string, string) {
 func documentOutput(t *testing.T, w *bufio.Writer, s expressionScenario, formattedDoc string, formattedDoc2 string) {
 	var output bytes.Buffer
 	var err error
-	printer := NewPrinter(bufio.NewWriter(&output), false, true, false, 2, true)
+	printer := NewPrinterWithSingleWriter(bufio.NewWriter(&output), YamlOutputFormat, true, false, 2, true)
 
 	node, err := NewExpressionParser().ParseExpression(s.expression)
 	if err != nil {

pkg/yqlib/printer.go

@@ -3,7 +3,9 @@ package yqlib
 import (
 	"bufio"
 	"container/list"
+	"fmt"
 	"io"
+	"strings"
 
 	yaml "gopkg.in/yaml.v3"
 )
@@ -11,35 +13,67 @@ import (
 type Printer interface {
 	PrintResults(matchingNodes *list.List) error
 	PrintedAnything() bool
+	//e.g. when given a front-matter doc, like jekyll
+	SetAppendix(reader io.Reader)
+}
+
+type PrinterOutputFormat uint32
+
+const (
+	YamlOutputFormat = 1 << iota
+	JsonOutputFormat
+	PropsOutputFormat
+)
+
+func OutputFormatFromString(format string) (PrinterOutputFormat, error) {
+	switch format {
+	case "yaml", "y":
+		return YamlOutputFormat, nil
+	case "json", "j":
+		return JsonOutputFormat, nil
+	case "props", "p":
+		return PropsOutputFormat, nil
+	default:
+		return 0, fmt.Errorf("Unknown fromat '%v' please use [yaml|json|props]", format)
+	}
 }
 
 type resultsPrinter struct {
-	outputToJSON       bool
+	outputFormat       PrinterOutputFormat
 	unwrapScalar       bool
 	colorsEnabled      bool
 	indent             int
 	printDocSeparators bool
-	writer             io.Writer
+	printerWriter      PrinterWriter
 	firstTimePrinting  bool
 	previousDocIndex   uint
 	previousFileIndex  int
 	printedMatches     bool
 	treeNavigator      DataTreeNavigator
+	appendixReader     io.Reader
 }
 
-func NewPrinter(writer io.Writer, outputToJSON bool, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
+func NewPrinterWithSingleWriter(writer io.Writer, outputFormat PrinterOutputFormat, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
+	return NewPrinter(NewSinglePrinterWriter(writer), outputFormat, unwrapScalar, colorsEnabled, indent, printDocSeparators)
+}
+
+func NewPrinter(printerWriter PrinterWriter, outputFormat PrinterOutputFormat, unwrapScalar bool, colorsEnabled bool, indent int, printDocSeparators bool) Printer {
 	return &resultsPrinter{
-		writer:             writer,
-		outputToJSON:       outputToJSON,
+		printerWriter:      printerWriter,
+		outputFormat:       outputFormat,
 		unwrapScalar:       unwrapScalar,
 		colorsEnabled:      colorsEnabled,
 		indent:             indent,
-		printDocSeparators: !outputToJSON && printDocSeparators,
+		printDocSeparators: outputFormat == YamlOutputFormat && printDocSeparators,
 		firstTimePrinting:  true,
 		treeNavigator:      NewDataTreeNavigator(),
 	}
 }
 
+func (p *resultsPrinter) SetAppendix(reader io.Reader) {
+	p.appendixReader = reader
+}
+
 func (p *resultsPrinter) PrintedAnything() bool {
 	return p.printedMatches
 }
@@ -49,11 +83,14 @@ func (p *resultsPrinter) printNode(node *yaml.Node, writer io.Writer) error {
 		(node.Tag != "!!bool" || node.Value != "false"))
 
 	var encoder Encoder
-	if node.Kind == yaml.ScalarNode && p.unwrapScalar && !p.outputToJSON {
+	if node.Kind == yaml.ScalarNode && p.unwrapScalar && p.outputFormat == YamlOutputFormat {
 		return p.writeString(writer, node.Value+"\n")
 	}
-	if p.outputToJSON {
+
+	if p.outputFormat == JsonOutputFormat {
 		encoder = NewJsonEncoder(writer, p.indent)
+	} else if p.outputFormat == PropsOutputFormat {
+		encoder = NewPropertiesEncoder(writer)
 	} else {
 		encoder = NewYamlEncoder(writer, p.indent, p.colorsEnabled)
 	}
@@ -65,16 +102,58 @@ func (p *resultsPrinter) writeString(writer io.Writer, txt string) error {
 	return errorWriting
 }
 
-func (p *resultsPrinter) safelyFlush(writer *bufio.Writer) {
-	if err := writer.Flush(); err != nil {
-		log.Error("Error flushing writer!")
-		log.Error(err.Error())
+func (p *resultsPrinter) processLeadingContent(mappedDoc *CandidateNode, writer io.Writer) error {
+	if strings.Contains(mappedDoc.Node.HeadComment, "$yqLeadingContent$") {
+		log.Debug("headcommentwas %v", mappedDoc.Node.HeadComment)
+		log.Debug("finished headcomment")
+		reader := bufio.NewReader(strings.NewReader(mappedDoc.Node.HeadComment))
+		mappedDoc.Node.HeadComment = ""
+
+		for {
+
+			readline, errReading := reader.ReadString('\n')
+			if errReading != nil && errReading != io.EOF {
+				return errReading
+			}
+			if strings.Contains(readline, "$yqLeadingContent$") {
+				// skip this
+
+			} else if strings.Contains(readline, "$yqDocSeperator$") {
+				if p.printDocSeparators {
+					if err := p.writeString(writer, "---\n"); err != nil {
+						return err
+					}
+				}
+			} else if p.outputFormat == YamlOutputFormat {
+				if err := p.writeString(writer, readline); err != nil {
+					return err
+				}
+			}
+
+			if errReading == io.EOF {
+				if readline != "" {
+					// the last comment we read didn't have a new line, put one in
+					if err := p.writeString(writer, "\n"); err != nil {
+						return err
+					}
+				}
+				break
+			}
+		}
+
 	}
+	return nil
 }
 
 func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 	log.Debug("PrintResults for %v matches", matchingNodes.Len())
-	if p.outputToJSON {
+
+	if matchingNodes.Len() == 0 {
+		log.Debug("no matching results, nothing to print")
+		return nil
+	}
+
+	if p.outputFormat != YamlOutputFormat {
 		explodeOp := Operation{OperationType: explodeOpType}
 		explodeNode := ExpressionNode{Operation: &explodeOp}
 		context, err := p.treeNavigator.GetMatchingNodes(Context{MatchingNodes: matchingNodes}, &explodeNode)
@@ -84,13 +163,6 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 		matchingNodes = context.MatchingNodes
 	}
 
-	bufferedWriter := bufio.NewWriter(p.writer)
-	defer p.safelyFlush(bufferedWriter)
-
-	if matchingNodes.Len() == 0 {
-		log.Debug("no matching results, nothing to print")
-		return nil
-	}
 	if p.firstTimePrinting {
 		node := matchingNodes.Front().Value.(*CandidateNode)
 		p.previousDocIndex = node.Document
@@ -99,20 +171,53 @@ func (p *resultsPrinter) PrintResults(matchingNodes *list.List) error {
 	}
 
 	for el := matchingNodes.Front(); el != nil; el = el.Next() {
+
 		mappedDoc := el.Value.(*CandidateNode)
 		log.Debug("-- print sep logic: p.firstTimePrinting: %v, previousDocIndex: %v, mappedDoc.Document: %v, printDocSeparators: %v", p.firstTimePrinting, p.previousDocIndex, mappedDoc.Document, p.printDocSeparators)
-		if (p.previousDocIndex != mappedDoc.Document || p.previousFileIndex != mappedDoc.FileIndex) && p.printDocSeparators {
+
+		writer, errorWriting := p.printerWriter.GetWriter(mappedDoc)
+		if errorWriting != nil {
+			return errorWriting
+		}
+
+		commentStartsWithSeparator := strings.Contains(mappedDoc.Node.HeadComment, "$yqLeadingContent$\n$yqDocSeperator$")
+
+		if (p.previousDocIndex != mappedDoc.Document || p.previousFileIndex != mappedDoc.FileIndex) && p.printDocSeparators && !commentStartsWithSeparator {
 			log.Debug("-- writing doc sep")
-			if err := p.writeString(bufferedWriter, "---\n"); err != nil {
+			if err := p.writeString(writer, "---\n"); err != nil {
 				return err
 			}
 		}
 
-		if err := p.printNode(mappedDoc.Node, bufferedWriter); err != nil {
+		if err := p.processLeadingContent(mappedDoc, writer); err != nil {
+			return err
+		}
+
+		if err := p.printNode(mappedDoc.Node, writer); err != nil {
 			return err
 		}
 
 		p.previousDocIndex = mappedDoc.Document
+		if err := writer.Flush(); err != nil {
+			return err
+		}
+	}
+
+	if p.appendixReader != nil && p.outputFormat == YamlOutputFormat {
+		writer, err := p.printerWriter.GetWriter(nil)
+		if err != nil {
+			return err
+		}
+
+		log.Debug("Piping appendix reader...")
+		betterReader := bufio.NewReader(p.appendixReader)
+		_, err = io.Copy(writer, betterReader)
+		if err != nil {
+			return err
+		}
+		if err := writer.Flush(); err != nil {
+			return err
+		}
 	}
 
 	return nil