Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Improve sortKeys #556

Merged
merged 2 commits into from
Feb 15, 2024
Merged

Improve sortKeys #556

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e8744ed
.
lihaoyi Feb 15, 2024
7a1452c
.
lihaoyi Feb 15, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 16 additions & 1 deletion upickle/core/src-2.12/upickle/core/compat/SortInPlace.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,9 +1,24 @@
package upickle.core.compat

object SortInPlace {
def apply[T, B: Ordering](t: collection.mutable.ArrayBuffer[T])(f: T => B): Unit = {
def apply[T, B: Ordering](t: collection.mutable.ArrayBuffer[T])(f: PartialFunction[T, B]): Unit = {
val sorted = t.sortBy(f)
t.clear()
t.appendAll(sorted)
}
}

object DistinctBy{
def apply[T, V](items: collection.Seq[T])(f: T => V) = {
val output = collection.mutable.Buffer.empty[T]
val seen = collection.mutable.Set.empty[V]
for(item <- items){
val key = f(item)
if (!seen(key)) {
seen.add(key)
output.append(item)
}
}
output
}
}
8 changes: 7 additions & 1 deletion upickle/core/src-2.13+/upickle/core/compat/SortInPlace.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,13 @@
package upickle.core.compat

object SortInPlace {
def apply[T, B: scala.Ordering](t: collection.mutable.ArrayBuffer[T])(f: T => B): Unit = {
def apply[T, B: scala.Ordering](t: collection.mutable.ArrayBuffer[T])(f: PartialFunction[T, B]): Unit = {
t.sortInPlaceBy(f)
}
}

object DistinctBy{
def apply[T, V](items: collection.Seq[T])(f: T => V) = {
items.distinctBy(f)
}
}
57 changes: 39 additions & 18 deletions upickle/core/src/upickle/core/BufferedValue.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@ package upickle.core

import upickle.core.ParseUtils.reject
import scala.collection.mutable
import upickle.core.compat.{SortInPlace, DistinctBy}

/**
* A reified version of [[Visitor]], allowing visitor method calls to be buffered up,
Expand All @@ -13,35 +14,55 @@ sealed trait BufferedValue {

object BufferedValue extends Transformer[BufferedValue]{
def valueToSortKey(b: BufferedValue): String = b match{
case BufferedValue.Null(i) => "00"
case BufferedValue.True(i) => "01" + "true"
case BufferedValue.False(i) => "02" + "false"
case BufferedValue.Str(s, i) => "03" + s.toString
case BufferedValue.Num(s, _, _, i) => "04" + s.toString
case BufferedValue.Char(c, i) => "05" + c.toString
case BufferedValue.Binary(bytes, o, l, _) => "06" + new String(bytes, o, l)
case BufferedValue.Ext(tag, bytes, o, l, i) => "07" + tag.toString + new String(bytes, o, l)
case BufferedValue.Float32(f, i) => "08" + f.toString
case BufferedValue.Float64String(s, i) => "09" + s
case BufferedValue.Int32(n, i) => "10" + n.toString
case BufferedValue.Int64(n, i) => "11" + n.toString
case BufferedValue.NumRaw(d, i) => "12" + d.toString
case BufferedValue.UInt64(n, i) => "13" + n.toString
case BufferedValue.Arr(vs, i) => "14" + vs.map(valueToSortKey).mkString
case BufferedValue.Obj(kvs, _, i) => "15" + kvs.map{case (k, v) => valueToSortKey(k) + valueToSortKey(v)}.mkString
case BufferedValue.Null(i) => "null"
case BufferedValue.True(i) => "true"
case BufferedValue.False(i) => "false"
case BufferedValue.Str(s, i) => s.toString
case BufferedValue.Num(s, _, _, i) => s.toString
case BufferedValue.Char(c, i) => c.toString
case BufferedValue.Binary(bytes, o, l, _) => new String(bytes, o, l)
case BufferedValue.Ext(tag, bytes, o, l, i) => tag.toString + new String(bytes, o, l)
case BufferedValue.Float32(f, i) => f.toString
case BufferedValue.Float64String(s, i) => s
case BufferedValue.Int32(n, i) => n.toString
case BufferedValue.Int64(n, i) => n.toString
case BufferedValue.NumRaw(d, i) => d.toString
case BufferedValue.UInt64(n, i) => n.toString
case BufferedValue.Arr(vs, i) => vs.map(valueToSortKey).mkString
case BufferedValue.Obj(kvs, _, i) => kvs.map{case (k, v) => valueToSortKey(k) + valueToSortKey(v)}.mkString
}

def maybeSortKeysTransform[T, V](tr: Transformer[T],
t: T,
sortKeys: Boolean,
f: Visitor[_, V]): V = {

def rec(x: BufferedValue): Unit = {
x match {
case BufferedValue.Arr(items, i) => items.map(rec)
case BufferedValue.Obj(items, jsonableKeys, i) =>
upickle.core.compat.SortInPlace[(BufferedValue, BufferedValue), String](items) {
case (k, v) => valueToSortKey(k)

// Special case handling for objects whose keys are all numbers
DistinctBy(items)(_._1.getClass) match{
case collection.Seq((_: BufferedValue.Num, _)) =>
SortInPlace(items) { case (k: BufferedValue.Num, v) => k.s.toString.toDouble}
case collection.Seq((_: BufferedValue.Float32, _)) =>
SortInPlace(items) { case (k: BufferedValue.Float32, v) => k.d }
case collection.Seq((_: BufferedValue.Float64String, _)) =>
SortInPlace(items) { case (k: BufferedValue.Float64String, v) => k.s.toDouble }
case collection.Seq((_: BufferedValue.Int32, _)) =>
SortInPlace(items) { case (k: BufferedValue.Int32, v) => k.i }
case collection.Seq((_: BufferedValue.Int64, _)) =>
SortInPlace(items) { case (k: BufferedValue.Int64, v) => k.i }
case collection.Seq((_: BufferedValue.NumRaw, _)) =>
SortInPlace(items) { case (k: BufferedValue.NumRaw, v) => k.d }
case collection.Seq((_: BufferedValue.UInt64, _)) =>
SortInPlace(items) { case (k: BufferedValue.UInt64, v) => k.i }
case _ =>
// Fall back to generic string-based sorting routine
SortInPlace(items) { case (k, v) => valueToSortKey(k)}
}

items.foreach { case (c, v) => (c, rec(v)) }
case v =>
}
Expand Down
141 changes: 113 additions & 28 deletions upickle/test/src/upickle/StructTests.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -610,34 +610,119 @@ object StructTests extends TestSuite {
test("null") - rw(ujson.Null, """null""")
}
test("sortKeys") {
val raw = """{"d": [{"c": 0, "b": 1}], "a": []}"""
val sorted =
"""{
| "a": [],
| "d": [
| {
| "b": 1,
| "c": 0
| }
| ]
|}""".stripMargin
val struct = upickle.default.read[Map[String, Seq[Map[String, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted

val baos = new java.io.ByteArrayOutputStream
upickle.default.writeToOutputStream(struct, baos, indent = 4, sortKeys = true)
baos.toString ==> sorted

val writer = new java.io.StringWriter
upickle.default.writeTo(struct, writer, indent = 4, sortKeys = true)
writer.toString ==> sorted

new String(upickle.default.writeToByteArray(struct, indent = 4, sortKeys = true)) ==> sorted

val baos2 = new java.io.ByteArrayOutputStream
upickle.default.stream(struct, indent = 4, sortKeys = true).writeBytesTo(baos2)
baos2.toString() ==> sorted
test("streaming") {
val raw = """{"d": [{"c": 0, "b": 1}], "a": []}"""
val sorted =
"""{
| "a": [],
| "d": [
| {
| "b": 1,
| "c": 0
| }
| ]
|}""".stripMargin
val struct = upickle.default.read[Map[String, Seq[Map[String, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted

val baos = new java.io.ByteArrayOutputStream
upickle.default.writeToOutputStream(struct, baos, indent = 4, sortKeys = true)
baos.toString ==> sorted

val writer = new java.io.StringWriter
upickle.default.writeTo(struct, writer, indent = 4, sortKeys = true)
writer.toString ==> sorted

new String(upickle.default.writeToByteArray(struct, indent = 4, sortKeys = true)) ==> sorted

val baos2 = new java.io.ByteArrayOutputStream
upickle.default.stream(struct, indent = 4, sortKeys = true).writeBytesTo(baos2)
baos2.toString() ==> sorted
}

test("ints") {
val raw = """{"27": [{"10": 0, "2": 1}], "3": []}"""
val sorted =
"""{
| "3": [],
| "27": [
| {
| "2": 1,
| "10": 0
| }
| ]
|}""".stripMargin
val struct = upickle.default.read[Map[Int, Seq[Map[Int, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted
}
test("longs") {
val raw = """{"27": [{"10": 0, "2": 1}], "300": []}"""
val sorted =
"""{
| "27": [
| {
| "2": 1,
| "10": 0
| }
| ],
| "300": []
|}""".stripMargin
val struct = upickle.default.read[Map[Long, Seq[Map[Long, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted
}
test("floats") {
val raw = """{"27.5": [{"10.5": 0, "2.5": 1}], "3.5": []}"""
val sorted =
"""{
| "3.5": [],
| "27.5": [
| {
| "2.5": 1,
| "10.5": 0
| }
| ]
|}""".stripMargin
val struct = upickle.default.read[Map[Float, Seq[Map[Float, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted
}
test("doubles") {
val raw = """{"27.5": [{"10.5": 0, "2.5": 1}], "3.5": []}"""
val sorted =
"""{
| "3.5": [],
| "27.5": [
| {
| "2.5": 1,
| "10.5": 0
| }
| ]
|}""".stripMargin
val struct = upickle.default.read[Map[Double, Seq[Map[Double, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted
}
test("strings") {
// Make sure that when we treat things as Strings, they are sorted
// as strings, unlike the above cases where they are treated as numbers
val raw = """{"27.5": [{"10.5": 0, "2.5": 1}], "3.5": []}"""
val sorted =
"""{
| "27.5": [
| {
| "10.5": 0,
| "2.5": 1
| }
| ],
| "3.5": []
|}""".stripMargin
val struct = upickle.default.read[Map[String, Seq[Map[String, Int]]]](raw)

upickle.default.write(struct, indent = 4, sortKeys = true) ==> sorted
}
}
}
}
Expand Down
17 changes: 12 additions & 5 deletions upickleReadme/Readme.scalatex
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,7 @@
)
)

@sect("uPickle 3.1.4")
@sect("uPickle 3.1.5")
@div(display.flex, alignItems.center, flexDirection.column)
@div
@a(href := "https://gitter.im/lihaoyi/upickle")(
Expand Down Expand Up @@ -74,8 +74,8 @@

@sect{Getting Started}
@hl.scala
"com.lihaoyi" %% "upickle" % "3.1.4" // SBT
ivy"com.lihaoyi::upickle:3.1.4" // Mill
"com.lihaoyi" %% "upickle" % "3.1.5" // SBT
ivy"com.lihaoyi::upickle:3.1.5" // Mill

@p
And then you can immediately start writing and reading common Scala
Expand All @@ -93,8 +93,8 @@
@p
For ScalaJS applications, use this dependencies instead:
@hl.scala
"com.lihaoyi" %%% "upickle" % "3.1.4" // SBT
ivy"com.lihaoyi::upickle::3.1.4" // Mill
"com.lihaoyi" %%% "upickle" % "3.1.5" // SBT
ivy"com.lihaoyi::upickle::3.1.5" // Mill

@sect{Scala Versions}
@p
Expand Down Expand Up @@ -886,6 +886,13 @@
JSON library, and inherits a lot of it's performance from Erik's work.

@sect{Version History}
@sect{3.1.5}
@ul
@li
Add the @code{sortKeys = true} flag that can be passed to @code{upickle.default.write}
or @code{ujson.write}, allowing you to ensure the generated JSON has object keys in sorted
order

@sect{3.1.4}
@ul
@li
Expand Down
Loading