repo stringlengths 5 53 | pr_number int32 1 321k | task_type stringclasses 2
values | issue_text stringlengths 0 81.2k | pr_title stringlengths 1 319 | pr_body stringlengths 0 105k | base_sha stringlengths 40 40 | head_sha stringlengths 40 40 | gold_diff stringlengths 0 202M | changed_files listlengths 0 100 | review_threads listlengths 0 100 | test_patch stringlengths 0 23.4M | merged bool 1
class |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
twitter/scalding | 1,942 | issue_to_patch | Adding Beam-Dataflow support for group functions in Scalding | This PR adds support for scalding group functions by handling:
- Grouped
- IdentityReduce
- UnsortedGrouped
- UnsortedIdentityReduce
- IteratorMappedReduce
- SortedGrouped
- IdentityValueSortedReduce
- ValueSortedReduce
via
- mapGroup
- sortedMapGroup
- sorted
methods in BeamOp.
- Opti... | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala
index 229474cab..62a2a0249 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala
+++ b/scalding-beam/sr... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backe... | [
{
"comment": "this is shame for this to be a no-op...\r\n\r\nCan you not do some kind of stateful map only operation that processes all the keys on the mapper and uses a cache of a fixed size to aggregate?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala",
"hun... | diff --git a/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala
index 7152bb229..451eb7ee3 100644
--- a/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala
+++ b/s... | true | |
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | none of this code is handling `OrderedSerialization` which I think will be a significant deoptimization relative to scalding.
You might want to pattern match on Ordering and if it is an OrderedSerialization, then serialize to bytes and use the binary comparator.
If nothing else, you can use the OrderedSerializati... | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "none of this code is handling `OrderedSerialization` which I think will be a significant deoptimization relative to scalding.\r\n\r\nYou might want to pattern match on Ordering and if it is an OrderedSerialization, then serialize to bytes and use the binary comparator.\r\n\r\nIf nothing else, you ... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | how does this work if ordV is being ignored? in what way is it a sort? | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "how does this work if ordV is being ignored? in what way is it a sort?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +55,101 @@ object BeamOp extends Serializable {\n }\n }\n \n- final case class TransformBeamOp[A, B](sour... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | so, here you can also match on the reduceFn:
```scala
reduceFn match {
case MapValueStream(SumAll(semigroup)) =>
// you can use beam combine here
}
```
there may be other functions you can optimize if you look in here:
https://github.com/twitter/scalding/blob/develop/scalding-core/src/main/scala/com/twitt... | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "so, here you can also match on the reduceFn:\r\n```scala\r\nreduceFn match {\r\n case MapValueStream(SumAll(semigroup)) =>\r\n // you can use beam combine here\r\n}\r\n```\r\nthere may be other functions you can optimize if you look in here:\r\nhttps://github.com/twitter/scalding/blob/develop/... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | we can pattern match on the `Ordering[V]` to see if it is a `OrderedSerialization` to recover a way to serialize. | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "we can pattern match on the `Ordering[V]` to see if it is a `OrderedSerialization` to recover a way to serialize.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +55,101 @@ object BeamOp extends Serializable {\n }\n }\n \n- ... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | this should probably take two coders, one for K and one for V since you can construct some of them using `Ordering[K]` or `Ordering[V]`
| 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "this should probably take two coders, one for K and one for V since you can construct some of them using `Ordering[K]` or `Ordering[V]`\r\n\r\n",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +55,101 @@ object BeamOp extends Serial... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | nit: using a while loop and `m.iterator` is faster (no closure):
```scala
val mit = m.iterator
while (mit.hasNext) {
val x = mit.next()
c.output(KV.of(x._1, x._2))
}
```
A bit of a pain to write things like that out, but avoiding closures can be measurable wins in some of the internals. | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "nit: using a while loop and `m.iterator` is faster (no closure):\r\n```scala\r\nval mit = m.iterator\r\nwhile (mit.hasNext) {\r\n val x = mit.next()\r\n c.output(KV.of(x._1, x._2))\r\n}\r\n```\r\nA bit of a pain to write things like that out, but avoiding closures can be measurable wins in some ... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | same comment about avoiding the closure. | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "same comment about avoiding the closure.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +25,31 @@ object BeamFunctions {\n def processElement(c: DoFn[A, B]#ProcessContext): Unit =\n c.output(f(c.element()))\n }\... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | this is going to pattern match on each value. I don't think we want this. I think we should do:
```scala
object OrderedSerializationCoder {
def apply[T](ord: Ordering[T])(implicit kryoCoder: KryoCoder): AtomicCoder[T] =
ord match {
case ordSer: OrderedSerialization[T] @unchecked => // make an orderedse... | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
index 27adfee52..1520328a0 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
+++ b/scalding-beam/src/main... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "this is going to pattern match on each value. I don't think we want this. I think we should do:\r\n\r\n```scala\r\nobject OrderedSerializationCoder {\r\n def apply[T](ord: Ordering[T])(implicit kryoCoder: KryoCoder): AtomicCoder[T] =\r\n ord match {\r\n case ordSer: OrderedSerialization[T] ... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | no need for `new` here. | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "no need for `new` here.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -101,45 +128,52 @@ object BeamOp extends Serializable {\n new PTransform[PCollection[(K, V)], PCollection[(K, V)]]() {\n override def expand(input:... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | we should probably override:
```scala
def hashCode = value.hashCode
def equals(that: Any) =
that match {
case WrappedComparable(thatValue, _) => value.equals(thatValue)
case _ => false
}
``` | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "we should probably override:\r\n```scala\r\ndef hashCode = value.hashCode\r\ndef equals(that: Any) =\r\n that match {\r\n case WrappedComparable(thatValue, _) => value.equals(thatValue)\r\n case _ => false\r\n }\r\n```",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_b... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Do we need this iterator thing here? Why not
```
m.foreach { case (k, v) =>
c.output(KV.of(k, v), BoundedWindow.TIMESTAMP_MIN_VALUE, GlobalWindow.INSTANCE)
}
```
? | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "Do we need this iterator thing here? Why not\r\n```\r\nm.foreach { case (k, v) => \r\n c.output(KV.of(k, v), BoundedWindow.TIMESTAMP_MIN_VALUE, GlobalWindow.INSTANCE)\r\n}\r\n```\r\n?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | nit: empty lines between methods | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "nit: empty lines between methods",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +25,40 @@ object BeamFunctions {\n def processElement(c: DoFn[A, B]#ProcessContext): Unit =\n c.output(f(c.element()))\n }\n+\n+ c... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Do we really want to keep it over `KV`, and not just `()`? | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "Do we really want to keep it over `KV`, and not just `()`?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +25,40 @@ object BeamFunctions {\n def processElement(c: DoFn[A, B]#ProcessContext): Unit =\n c.output(f(c.... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | nit: empty lines between methods | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "nit: empty lines between methods",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +78,131 @@ object BeamOp extends Serializable {\n }\n }\n \n- final case class TransformBeamOp[A, B](source: BeamOp[A], f: PTransform[PCollecti... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Why `kryoCoder` is being passed here? | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
index 27adfee52..1520328a0 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
+++ b/scalding-beam/src/main... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "Why `kryoCoder` is being passed here?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -27,3 +29,15 @@ object KryoCoder {\n implicit def castType[T](kryoCoder: KryoCoder): AtomicCoder[T] =\n kryoCoder.asInstanceOf[AtomicCoder[T]]... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | nit: not a `KryoCoder`, and better to call it fallback. Also no need to have it as implicit parameter here | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
index 27adfee52..1520328a0 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
+++ b/scalding-beam/src/main... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "nit: not a `KryoCoder`, and better to call it fallback. Also no need to have it as implicit parameter here",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -27,3 +29,15 @@ object KryoCoder {\n implicit def castType[T](kryoCoder: Kryo... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Why do we need this class? | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "Why do we need this class?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +78,131 @@ object BeamOp extends Serializable {\n }\n }\n \n- final case class TransformBeamOp[A, B](source: BeamOp[A], f: PTransform[PCollection[A],... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | I don't think `SortValues` cares about comparability of values, so unfortunately we need to implement it ourselves... | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "I don't think `SortValues` cares about comparability of values, so unfortunately we need to implement it ourselves...",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +78,131 @@ object BeamOp extends Serializable {\n }\n }\n \... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Add `assert` on `ord` being same as wrapped ordering | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "Add `assert` on `ord` being same as wrapped ordering",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +78,131 @@ object BeamOp extends Serializable {\n }\n }\n \n- final case class TransformBeamOp[A, B](source: BeamOp[A], f: ... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Nice idea, love it! | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "Nice idea, love it!",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -51,9 +78,131 @@ object BeamOp extends Serializable {\n }\n }\n \n- final case class TransformBeamOp[A, B](source: BeamOp[A], f: PTransform[PCollection[A], PColle... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | Making it `toArray` would be more performant I think, using raw java here would be even better. But I'm fine to keep any way, since I hope we'll rewrite this eventually.
Please put `TODO` here documenting this needs to be changed to external sorter like the one in beam | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
index 1307c2b8e..6d6898ce8 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
+++ b/scalding-beam/src/main/scala/co... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "Making it `toArray` would be more performant I think, using raw java here would be even better. But I'm fine to keep any way, since I hope we'll rewrite this eventually.\r\n\r\nPlease put `TODO` here documenting this needs to be changed to external sorter like the one in beam",
"path": "scaldi... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | I just noticed you are breaking the pair just to allocate again:
```scala
val pair = c.element()
val evicted = cache.put(Map(pair))
...
``` | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "I just noticed you are breaking the pair just to allocate again:\r\n```scala\r\nval pair = c.element()\r\nval evicted = cache.put(Map(pair))\r\n...\r\n```",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +24,42 @@ object Beam... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | also here:
```scala
c.output(mit.next())
``` | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "also here:\r\n```scala\r\nc.output(mit.next())\r\n```",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +24,42 @@ object BeamFunctions {\n def processElement(c: DoFn[A, B]#ProcessContext): Unit =\n c.output(f(c.eleme... | true | ||
twitter/scalding | 1,942 | comment_to_fix | Adding Beam-Dataflow support for group functions in Scalding | `c.output(mit.next(), ...)` | 18d2c9f28b4733b5bd79aca60250f86f96d61cbb | 17c08fd839d8fe9797ec18d484bf93a5c6f7e8ee | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
index 5aa86789c..b9da0ddd7 100644
--- a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala
+++ b/scalding-b... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala"
] | [
{
"comment": "`c.output(mit.next(), ...)`",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"hunk": "@@ -22,4 +24,42 @@ object BeamFunctions {\n def processElement(c: DoFn[A, B]#ProcessContext): Unit =\n c.output(f(c.element()))\n }\n+\n+ case c... | true | ||
twitter/scalding | 1,944 | issue_to_patch | Don't use deprecated sbt syntax | `property in Task` syntax was departed since sbt 0.13.x. I've changed it to modern `Task / property` syntax. | 155233a9e6575aa4aa41e7fe257dfbb18ce49723 | 168d79e79df6b5d7dda0c0d22c83e48113b00fa0 | diff --git a/build.sbt b/build.sbt
index 7f1177546..b6473ca5a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -54,9 +54,9 @@ val sharedSettings = scalariformSettings ++ Seq(
javacOptions ++= Seq("-source", "1.8", "-target", "1.8"),
- javacOptions in doc := Seq("-source", "1.8"),
+ doc / javacOptions := Seq("-source"... | [
"build.sbt",
"version.sbt"
] | [] | true | ||
twitter/scalding | 1,943 | issue_to_patch | Migrate scalding docs to mdoc | Tut was deprecated so I migrated scalding to mdoc. As part of this effort I discovered Jekyll being broken as part of our doc generation because of `-f` parameter, and I fixed that. | 132f9bd3767feab76f7284f73fc9deb0cec9116e | c42bdcbda468c35e77561949599ffa1ccef74d84 | diff --git a/.travis.yml b/.travis.yml
index ee1ebb0de..4ab8d832b 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -10,7 +10,7 @@ install:
- rvm use 2.3.0 --install --fuzzy
- gem update --system
- gem install sass
- - gem install jekyll -f -v 3.2.1
+ - gem install jekyll -v 3.2.1
cache:
directories:
diff ... | [
".travis.yml",
"build.sbt",
"project/plugins.sbt"
] | [] | true | ||
twitter/scalding | 1,941 | issue_to_patch | Upgrade scala and java versions | I'm using post java 13 on my laptop and turns out it's incompatible with scala 2.12.8 we use in scalding.
I've updated both scala version to the latest available, and java version to 1.8 because PermGen related code wasn't working properly in sbt runner we use, and java versions before 1.8 seem obsolete at this point ... | 09621d8aba1e376bf01bd0a371a1b7bb52756931 | 3bceec86622a2544c00ed848fd3ff26eac45dcd6 | diff --git a/.gitignore b/.gitignore
index 57fa41912d..6895853c85 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,3 +1,4 @@
+.bsp
.cache
.project
.settings
diff --git a/.travis.yml b/.travis.yml
index e71494b985..ee1ebb0de5 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -31,7 +31,7 @@ matrix:
env: BUILD="base" ... | [
".gitignore",
".travis.yml",
"build.sbt",
"project/plugins.sbt",
"sbt",
"scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/ByteBufferOrderedBuf.scala",
"scripts/scald.rb"
] | [] | true | ||
twitter/scalding | 1,940 | issue_to_patch | Beam-Dataflow support for Scalding | This is a first PR for beam support to scalding
It now extends support to basic operations in SimpleWrite only:
- Filter
- FilterKeys
- Mapped
- FlatMapped
- MapValues
- FlatMapValues
Added support for reading from Source path and Iterable
Added support for writing into Sink path
Added KryoCoder sharing... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/build.sbt b/build.sbt
index b8113c151..2fad14bba 100644
--- a/build.sbt
+++ b/build.sbt
@@ -35,6 +35,7 @@ val scalaCheckVersion = "1.13.4"
val scalaTestVersion = "3.0.1"
val scroogeVersion = "18.9.0"
val sparkVersion = "2.4.0"
+val beamVersion = "2.29.0"
val slf4jVersion = "1.6.6"
val thriftVersion = ... | [
"build.sbt",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamBackend.scala",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamFunctions.scala",
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"scalding-beam/src/main/scala/com/twitter/sc... | [
{
"comment": "I don't think you want to make a kryo for each item, that will be very slow. I think you want a private member variable.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,26 @@\n+package com.twitter.scalding.beam_backend\n+\n+im... | diff --git a/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala
new file mode 100644
index 000000000..7152bb229
--- /dev/null
+++ b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/Be... | true | |
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I don't think you want to make a kryo for each item, that will be very slow. I think you want a private member variable. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "I don't think you want to make a kryo for each item, that will be very slow. I think you want a private member variable.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,26 @@\n+package com.twitter.scalding.beam_backend\n+\n+im... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I would prefer this to be a dependency only in `test` scope | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/build.sbt b/build.sbt
index b8113c151..2fad14bba 100644
--- a/build.sbt
+++ b/build.sbt
@@ -35,6 +35,7 @@ val scalaCheckVersion = "1.13.4"
val scalaTestVersion = "3.0.1"
val scroogeVersion = "18.9.0"
val sparkVersion = "2.4.0"
+val beamVersion = "2.29.0"
val slf4jVersion = "1.6.6"
val thriftVersion = ... | [
"build.sbt"
] | [
{
"comment": "I would prefer this to be a dependency only in `test` scope",
"path": "build.sbt",
"hunk": "@@ -359,6 +361,18 @@ lazy val scaldingSpark = module(\"spark\").settings(\n )\n ).dependsOn(scaldingCore)\n \n+lazy val scaldingBeam = module(\"beam\").settings(\n+ libraryDependencies ++= Se... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | We should fix this before landing | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/build.sbt b/build.sbt
index b8113c151..2fad14bba 100644
--- a/build.sbt
+++ b/build.sbt
@@ -35,6 +35,7 @@ val scalaCheckVersion = "1.13.4"
val scalaTestVersion = "3.0.1"
val scroogeVersion = "18.9.0"
val sparkVersion = "2.4.0"
+val beamVersion = "2.29.0"
val slf4jVersion = "1.6.6"
val thriftVersion = ... | [
"build.sbt"
] | [
{
"comment": "We should fix this before landing",
"path": "build.sbt",
"hunk": "@@ -359,6 +361,18 @@ lazy val scaldingSpark = module(\"spark\").settings(\n )\n ).dependsOn(scaldingCore)\n \n+lazy val scaldingBeam = module(\"beam\").settings(\n+ libraryDependencies ++= Seq(\n+ \"org.apache.beam... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | This can't be correct, because this means you keep running same `pipeline` again and again with every `execute()` call | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "This can't be correct, because this means you keep running same `pipeline` again and again with every `execute()` call",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+impor... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | `i` name seems not very descriptive | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "`i` name seems not very descriptive",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Execution.Writer\n+import com.twitter.scalding.typed.{Resolv... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | You don't want to block `execute`, instead you want to create `Promise`, and resolve it when corresponding pipeline finishes | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "You don't want to block `execute`, instead you want to create `Promise`, and resolve it when corresponding pipeline finishes",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala",
"hunk": "@@ -0,0 +1,47 @@\n+package com.twitter.scalding.beam_backend\n+... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I wonder if we still need to do `.waitUntilFinish`. This will be needed when we support chaining of executions.
Same for returning 0 as the unique id. We would need to keep some state which can increment this counter. For now we can keep it like this and address this in the follow up patch. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "I wonder if we still need to do `.waitUntilFinish`. This will be needed when we support chaining of executions.\r\n\r\nSame for returning 0 as the unique id. We would need to keep some state which can increment this counter. For now we can keep it like this and address this in the follow up patch.... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | nit: extra new lines, should be only one | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/build.sbt b/build.sbt
index b8113c151..2fad14bba 100644
--- a/build.sbt
+++ b/build.sbt
@@ -35,6 +35,7 @@ val scalaCheckVersion = "1.13.4"
val scalaTestVersion = "3.0.1"
val scroogeVersion = "18.9.0"
val sparkVersion = "2.4.0"
+val beamVersion = "2.29.0"
val slf4jVersion = "1.6.6"
val thriftVersion = ... | [
"build.sbt"
] | [
{
"comment": "nit: extra new lines, should be only one",
"path": "build.sbt",
"hunk": "@@ -359,6 +361,18 @@ lazy val scaldingSpark = module(\"spark\").settings(\n )\n ).dependsOn(scaldingCore)\n \n+lazy val scaldingBeam = module(\"beam\").settings(\n+ libraryDependencies ++= Seq(\n+ \"org.apac... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Add comment indicating we need that because scalding config depends on hadoop | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/build.sbt b/build.sbt
index b8113c151..2fad14bba 100644
--- a/build.sbt
+++ b/build.sbt
@@ -35,6 +35,7 @@ val scalaCheckVersion = "1.13.4"
val scalaTestVersion = "3.0.1"
val scroogeVersion = "18.9.0"
val sparkVersion = "2.4.0"
+val beamVersion = "2.29.0"
val slf4jVersion = "1.6.6"
val thriftVersion = ... | [
"build.sbt"
] | [
{
"comment": "Add comment indicating we need that because scalding config depends on hadoop",
"path": "build.sbt",
"hunk": "@@ -359,6 +361,18 @@ lazy val scaldingSpark = module(\"spark\").settings(\n )\n ).dependsOn(scaldingCore)\n \n+lazy val scaldingBeam = module(\"beam\").settings(\n+ libraryD... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | `.head` doesn't seem correct here, would be great to fix it or add `todo` | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "`.head` doesn't seem correct here, would be great to fix it or add `todo`",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Execution.Writer\n+imp... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I don't think you really need `ExecutionContext` here | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "I don't think you really need `ExecutionContext` here",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Execution.Writer\n+import com.twitter.scal... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Same as `Source` - no need for execution context | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "Same as `Source` - no need for execution context",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Execution.Writer\n+import com.twitter.scalding.... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | `ExecutionContext` doesn't seem necessary | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "`ExecutionContext` doesn't seem necessary",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Config\n+import com.twitter.scalding.typed.TypedSource\n... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | The fact we use `KryoCoder` here which we might want to have customizable indicates we want to have access to config in `run()` method, would be great to add it | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "The fact we use `KryoCoder` here which we might want to have customizable indicates we want to have access to config in `run()` method, would be great to add it",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Given `cancel` method on `PipelineResult` object I don't think this future should be uncancellable. You might want to add `todo` or fix it. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "Given `cancel` method on `PipelineResult` object I don't think this future should be uncancellable. You might want to add `todo` or fix it.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala",
"hunk": "@@ -0,0 +1,50 @@\n+package com.twitter.scalding.... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Given `cancel` method on `PipelineResult` object I don't think this future should be uncancellable. You might want to add `todo` or fix it.
You don't need to do `map` separately instead you can return it as is as part of your Future block:
```
Future {
pipeline.run().waitUntilFinish()
(0L, ExecutionCounters.... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "Given `cancel` method on `PipelineResult` object I don't think this future should be uncancellable. You might want to add `todo` or fix it.\r\n\r\nYou don't need to do `map` separately instead you can return it as is as part of your Future block:\r\n```\r\nFuture {\r\n pipeline.run().waitUntilFin... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Put comments explaining what we need this config | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "Put comments explaining what we need this config",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,42 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.chill.Kryo\n+import com.twitter.chill.config.ScalaMapCo... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I'm not sure that's good default, I think we permit customers to pass nulls around | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "I'm not sure that's good default, I think we permit customers to pass nulls around",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,42 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.chill.Kryo\n+import c... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | todo: on caching `bos` and `bosOutput` in thread local | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "todo: on caching `bos` and `bosOutput` in thread local",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,42 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.chill.Kryo\n+import com.twitter.chill.config.Scal... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | nit: typo | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "nit: typo",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,42 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.chill.Kryo\n+import com.twitter.chill.config.ScalaMapConfig\n+import com.twitter.scalding.seri... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | nit: remote | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala
new file mode 100644
index 000000000..7152bb229
--- /dev/null
+++ b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/Be... | [
"scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala"
] | [
{
"comment": "nit: remote",
"path": "scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala",
"hunk": "@@ -0,0 +1,62 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.{Config, Execution, TextLine, TypedPipe}\n+import java.io.File\n+import java... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | here you probably want to make an error if the localPaths have more than one item in them. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
new file mode 100644
index 000000000..424cb29da
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala
@@ ... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala"
] | [
{
"comment": "here you probably want to make an error if the localPaths have more than one item in them.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamMode.scala",
"hunk": "@@ -0,0 +1,79 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Execu... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I think you could write this as `parDo(new MapFn(f))` without the `this`. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "I think you could write this as `parDo(new MapFn(f))` without the `this`.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Config\n+import com.twit... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | no need for `this`
can we remove the cast? | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "no need for `this`\r\n\r\ncan we remove the cast?",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Config\n+import com.twitter.scalding.typed.Typed... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | case class... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "case class...",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Config\n+import com.twitter.scalding.typed.TypedSource\n+import org.apache.beam.sdk.... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | case class. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "case class.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,80 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.twitter.scalding.Config\n+import com.twitter.scalding.typed.TypedSource\n+import org.apache.beam.sdk.Pi... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | you might want to add a TODO to allow overriding these optimization rules based on the conf. This is done in cascading mode. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "you might want to add a TODO to allow overriding these optimization rules based on the conf. This is done in cascading mode.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala",
"hunk": "@@ -0,0 +1,50 @@\n+package com.twitter.scalding.beam_backend\n+... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | are there no counters in Beam?
you don't want to return `0L` for each item. The contract is to return a new unique Long for each call to this function. You can keep an atomicLong and increment on each call, or something like that. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "are there no counters in Beam?\r\n\r\nyou don't want to return `0L` for each item. The contract is to return a new unique Long for each call to this function. You can keep an atomicLong and increment on each call, or something like that.",
"path": "scalding-beam/src/main/scala/com/twitter/scal... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I am not a fan of this.
1. note that when you are building the `BeamOp` you have access to the `Config`.
2. we have traditionally configured kryo by using the `Config` to get a `KryoHadoop` instance.
What I would prefer is in the BeamPlanner you build the `KryoInstantiator` and then pass that as needed.
Note:... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "I am not a fan of this.\r\n\r\n1. note that when you are building the `BeamOp` you have access to the `Config`.\r\n2. we have traditionally configured kryo by using the `Config` to get a `KryoHadoop` instance.\r\n\r\nWhat I would prefer is in the BeamPlanner you build the `KryoInstantiator` and th... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | this is the only place ProcessFn is used, Why not make:
```scala
case class ProcessPredicate[A](fn: A => Boolean) extends ProcessFunction[A, java.lang.Boolean] {
def apply(a: A) = java.lang.Boolean.valueOf(fn(a))
}
```
This reduces unused abstraction and also reduces slightly the indirection at runtime (you avo... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "this is the only place ProcessFn is used, Why not make:\r\n```scala\r\ncase class ProcessPredicate[A](fn: A => Boolean) extends ProcessFunction[A, java.lang.Boolean] {\r\n def apply(a: A) = java.lang.Boolean.valueOf(fn(a))\r\n}\r\n```\r\nThis reduces unused abstraction and also reduces slightly t... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I am still pretty sure: `def map[B](f: A => B)...` works here because `Function1` is contravariant in the first parameter. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "I am still pretty sure: `def map[B](f: A => B)...` works here because `Function1` is contravariant in the first parameter.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,61 @@\n+package com.twitter.scalding.beam_backend\n+\n+imp... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | same comment here: `def flatMap[B](f: A => ...)` should work because `Function1` is contravariant in the first parameter. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
new file mode 100644
index 000000000..1307c2b8e
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala
@@ -0,0 +... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala"
] | [
{
"comment": "same comment here: `def flatMap[B](f: A => ...)` should work because `Function1` is contravariant in the first parameter.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamOp.scala",
"hunk": "@@ -0,0 +1,61 @@\n+package com.twitter.scalding.beam_backend\n+\n+impo... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | `Unit` is the type and companion object for the type, `()` is the value. This is working because scala is casting the companion object into the value unit.
It is better to write: `case Nil => ()` here.
that is not a shortcut, is actually the correct thing that works without relying on Any to Unit casting.
It's... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "`Unit` is the type and companion object for the type, `()` is the value. This is working because scala is casting the companion object into the value unit.\r\n\r\nIt is better to write: `case Nil => ()` here.\r\n\r\nthat is not a shortcut, is actually the correct thing that works without relying o... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | I think you should at least log or something if you are going to ignore items.
Honestly, I would prefer: `???` so you don't forget to implement it later. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "I think you should at least log or something if you are going to ignore items.\r\n\r\nHonestly, I would prefer: `???` so you don't forget to implement it later.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala",
"hunk": "@@ -0,0 +1,60 @@\n+package ... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | we should get the runId *before* we start the job, not after:
```scala
val runId = state.getAndIncrement() // do this before returning
Future {
result.waitUntilFinish()
(runId, ...)
}
```
honestly the return type from the method should probably be `(Long, Future[ExecutionCounters])` but that it wasn't imple... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala
new file mode 100644
index 000000000..cb379e890
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.sca... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/BeamWriter.scala"
] | [
{
"comment": "we should get the runId *before* we start the job, not after:\r\n```scala\r\nval runId = state.getAndIncrement() // do this before returning\r\nFuture {\r\n result.waitUntilFinish()\r\n (runId, ...)\r\n}\r\n```\r\nhonestly the return type from the method should probably be `(Long, Future[Executi... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | maybe make this `private[this]` as well. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "maybe make this `private[this]` as well.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,28 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.esotericsoftware.kryo.io.Input\n+import com.twitter.chill.{ KryoInstant... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | this isn't safe... this can only write 1 byte (even though it takes an int). so if you have more than 255 bytes of data it will fail.
See:
https://github.com/twitter/chill/blob/19daefccbcda7d50f3130985de6e016be6eafa01/chill-hadoop/src/main/java/com/twitter/chill/hadoop/KryoDeserializer.java#L49
you might want to... | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "this isn't safe... this can only write 1 byte (even though it takes an int). so if you have more than 255 bytes of data it will fail.\r\n\r\nSee:\r\nhttps://github.com/twitter/chill/blob/19daefccbcda7d50f3130985de6e016be6eafa01/chill-hadoop/src/main/java/com/twitter/chill/hadoop/KryoDeserializer.j... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | same comment about this not actually being safe. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "same comment about this not actually being safe.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala",
"hunk": "@@ -0,0 +1,28 @@\n+package com.twitter.scalding.beam_backend\n+\n+import com.esotericsoftware.kryo.io.Input\n+import com.twitter.chill.{ Kry... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | 1. can we not call this `inferType` but perhaps `castType`
2. can we accept a `KryoCoder` as the argument not an `AtomicCoder`? It isn't safe in general, it is safe(ish) for kryo because it can handle any type at runtime. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
new file mode 100644
index 000000000..27adfee52
--- /dev/null
+++ b/scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala
... | [
"scalding-beam/src/main/scala/com/twitter/scalding/beam_backend/KryoCoder.scala"
] | [
{
"comment": "1. can we not call this `inferType` but perhaps `castType`\r\n2. can we accept a `KryoCoder` as the argument not an `AtomicCoder`? It isn't safe in general, it is safe(ish) for kryo because it can handle any type at runtime.",
"path": "scalding-beam/src/main/scala/com/twitter/scalding/beam_bac... | true | ||
twitter/scalding | 1,940 | comment_to_fix | Beam-Dataflow support for Scalding | Can we open the files using normal Java IO and make some assertion about them?
Like read all the numbers and put them in a set and assert it matches the expected Set?
I'd like a bit more assurance than it just didn't crash. | b6d465f6fc308d66e5a53c62662d5860846d8dc4 | 5a915abc2d7bf40d40de27d60b3135537da1612e | diff --git a/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala
new file mode 100644
index 000000000..7152bb229
--- /dev/null
+++ b/scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/Be... | [
"scalding-beam/src/test/scala/com/twitter/scalding/beam_backend/BeamBackendTests.scala"
] | [
{
"comment": "Can we open the files using normal Java IO and make some assertion about them?\n\nLike read all the numbers and put them in a set and assert it matches the expected Set?\n\nI'd like a bit more assurance than it just didn't crash.",
"path": "scalding-beam/src/test/scala/com/twitter/scalding/bea... | true | ||
twitter/scalding | 1,939 | issue_to_patch | Update sbt to 1.5.6 | This PR updates sbt version to latest available one: `1.5.6`.
I've updated sbt plugins versions to be compatible with newer version of sbt, as well as updates `build.sbt` to work successfully.
I'll follow up with two PRs: to move to scalafmt and mdoc plugins | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/.travis.yml b/.travis.yml
index 2a5218f60d..e71494b985 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -31,7 +31,7 @@ matrix:
env: BUILD="base" TEST_TARGET="scalding-args scalding-date maple scalding-quotation"
script: "scripts/run_test.sh"
- - scala: 2.12.4
+ - scala: 2.12.8
env... | [
".travis.yml",
"build.sbt",
"project/build.properties",
"project/plugins.sbt",
"sbt",
"scripts/scald.rb",
"scripts/test_repl_tutorial.sh"
] | [
{
"comment": "can we just remove scalariform? scalafmt is the winner here.",
"path": "build.sbt",
"hunk": "@@ -44,14 +40,14 @@ val jlineVersion = \"2.14.3\"\n \n val printDependencyClasspath = taskKey[Unit](\"Prints location of the dependencies\")\n \n-val sharedSettings = assemblySettings ++ scalarifor... | diff --git a/scripts/test_repl_tutorial.sh b/scripts/test_repl_tutorial.sh
index bae32ca466..d82c02384b 100755
--- a/scripts/test_repl_tutorial.sh
+++ b/scripts/test_repl_tutorial.sh
@@ -23,7 +23,8 @@ System.exit(1)
# Test running repl from sbt.
cd $BASE_DIR
-echo "System.exit(0)" | ./sbt ++$TRAVIS_SCALA_VERSION sc... | true | |
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | can we just remove scalariform? scalafmt is the winner here. | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "can we just remove scalariform? scalafmt is the winner here.",
"path": "build.sbt",
"hunk": "@@ -44,14 +40,14 @@ val jlineVersion = \"2.14.3\"\n \n val printDependencyClasspath = taskKey[Unit](\"Prints location of the dependencies\")\n \n-val sharedSettings = assemblySettings ++ scalarifor... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | can we make a val holding this version constant for this?
Also, fixing to a particular scala version here is really scary. That sounds like a great way for runtime errors if someone tries `2.11` for this artifact.
If it only runs for 2.10, we should have the build filter this out when scalaVersion is not 2.10. | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "can we make a val holding this version constant for this?\r\n\r\nAlso, fixing to a particular scala version here is really scary. That sounds like a great way for runtime errors if someone tries `2.11` for this artifact.\r\n\r\nIf it only runs for 2.10, we should have the build filter this out whe... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | side note: we should migrate to mdoc, tut is EOL I believe.
also, I though `Tut / scalacOption` is the new syntax? | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "side note: we should migrate to mdoc, tut is EOL I believe.\r\n\r\nalso, I though `Tut / scalacOption` is the new syntax?",
"path": "build.sbt",
"hunk": "@@ -720,15 +722,17 @@ lazy val docSettings = Seq(\n )\n \n \n+/*\n // Documentation is generated for projects defined in\n // `docsSou... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | Looking at the rest of the code here I think the pattern would be to capture the version in a variable, like hbaseVersion and hadoopVersion.
Why are these Scala 2.10?
| b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "Looking at the rest of the code here I think the pattern would be to capture the version in a variable, like hbaseVersion and hadoopVersion.\r\n\r\nWhy are these Scala 2.10?\r\n\r\n",
"path": "build.sbt",
"hunk": "@@ -476,13 +470,21 @@ lazy val scaldingHRaven = module(\"hraven\").settings(... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | think we can just nuke this. | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/project/plugins.sbt b/project/plugins.sbt
index 2294d9e17b..4e251c1ded 100644
--- a/project/plugins.sbt
+++ b/project/plugins.sbt
@@ -1,21 +1,19 @@
-resolvers += Resolver.url("artifactory", url("https://scalasbt.artifactoryonline.com/scalasbt/sbt-plugin-releases"))(Resolver.ivyStylePatterns)
-
resolvers +... | [
"project/plugins.sbt"
] | [
{
"comment": "think we can just nuke this.",
"path": "project/plugins.sbt",
"hunk": "@@ -1,21 +1,21 @@\n-resolvers += Resolver.url(\"artifactory\", url(\"https://scalasbt.artifactoryonline.com/scalasbt/sbt-plugin-releases\"))(Resolver.ivyStylePatterns)\n+resolvers += Resolver.sbtPluginRepo(\"releases\")... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | New line at EOF | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "New line at EOF",
"path": "build.sbt",
"hunk": "@@ -720,15 +722,17 @@ lazy val docSettings = Seq(\n )\n \n \n+/*\n // Documentation is generated for projects defined in\n // `docsSourcesAndProjects`.\n lazy val docs = project\n .enablePlugins(MicrositesPlugin)\n+ .enablePlugins(ScalaU... | true | ||
twitter/scalding | 1,939 | comment_to_fix | Update sbt to 1.5.6 | Why not keeping the import? | b9ef1183072dda2b6974c7296221e18e840de63e | a7d2ef2c2855f1dce13a44ce19d69f9a2aa27c44 | diff --git a/build.sbt b/build.sbt
index d42b61c24c..b8113c151a 100644
--- a/build.sbt
+++ b/build.sbt
@@ -1,12 +1,10 @@
-import AssemblyKeys._
import ReleaseTransformations._
-import com.typesafe.sbt.SbtGhPages.GhPagesKeys._
-import com.typesafe.sbt.SbtScalariform._
+import com.typesafe.sbt.SbtScalariform.scalariform... | [
"build.sbt"
] | [
{
"comment": "Why not keeping the import?",
"path": "build.sbt",
"hunk": "@@ -44,14 +40,14 @@ val jlineVersion = \"2.14.3\"\n \n val printDependencyClasspath = taskKey[Unit](\"Prints location of the dependencies\")\n \n-val sharedSettings = assemblySettings ++ scalariformSettings ++ Seq(\n+val sharedSet... | true | ||
twitter/scalding | 1,938 | issue_to_patch | Fix race between jvm shutdown and `writer.finished` | Currently `writer.finished` happens in `onComplete` callback on `Future` result in `Exucution`. However since `onComplete` isn't being called before future being resolved and called asynchroniously after future being resolved, it leads to a race and runtime error:
- User's code as last operation in `main` executes `Ex... | f5c55b580b757a76cf8441235802b9a2d33e2fcc | 45ca0e9ce4bd1fbf729423415530819c8dd845c5 | diff --git a/scalding-core/src/main/scala/com/twitter/scalding/Execution.scala b/scalding-core/src/main/scala/com/twitter/scalding/Execution.scala
index df767fc90..ef58ed34f 100644
--- a/scalding-core/src/main/scala/com/twitter/scalding/Execution.scala
+++ b/scalding-core/src/main/scala/com/twitter/scalding/Execution.s... | [
"scalding-core/src/main/scala/com/twitter/scalding/Execution.scala",
"scalding-parquet/src/test/java/com/twitter/scalding/parquet/example/TestInputOutputFormat.java"
] | [
{
"comment": "I think we are technically supposed to do:\r\n```scala\r\nblocking {\r\n Await.ready(cancelHandler.stop(), ...)\r\n```\r\nhere:\r\n\r\nhttps://docs.scala-lang.org/overviews/core/futures.html#blocking-inside-a-future\r\n\r\nOther than that, it looks good.\r\n\r\nSmall nit, we should probably impor... | diff --git a/scalding-parquet/src/test/java/com/twitter/scalding/parquet/example/TestInputOutputFormat.java b/scalding-parquet/src/test/java/com/twitter/scalding/parquet/example/TestInputOutputFormat.java
index 05ee467ea..dde364701 100644
--- a/scalding-parquet/src/test/java/com/twitter/scalding/parquet/example/TestInp... | true | |
twitter/scalding | 1,937 | issue_to_patch | Alternative implementation to DeprecatedParquetInputFormat with fix | When combining `N` Parquet files, the first record of files `2 to N` gets skipped while the last record from the previous file is returned instead. This means losing some records while others get duplicated, quite bad.
This was fixed a month ago in https://github.com/apache/parquet-mr/pull/844 but we would need to ... | 20aee268c156841e8bdf77bb94354ffde74ec3b9 | 1c6862cbc156745b3ca100cb92137a587c42600a | diff --git a/scalding-parquet-scrooge/src/main/java/com/twitter/scalding/parquet/scrooge/ParquetScroogeScheme.java b/scalding-parquet-scrooge/src/main/java/com/twitter/scalding/parquet/scrooge/ParquetScroogeScheme.java
index cff692b347..3b41696ae6 100644
--- a/scalding-parquet-scrooge/src/main/java/com/twitter/scalding... | [
"scalding-parquet-scrooge/src/main/java/com/twitter/scalding/parquet/scrooge/ParquetScroogeScheme.java",
"scalding-parquet/src/main/java/com/twitter/scalding/parquet/ScaldingDeprecatedParquetInputFormat.java",
"scalding-parquet/src/main/java/com/twitter/scalding/parquet/thrift/ParquetTBaseScheme.java",
"scald... | [] | diff --git a/scalding-parquet/src/test/java/com/twitter/scalding/parquet/ScaldingDeprecatedInputFormatTest.java b/scalding-parquet/src/test/java/com/twitter/scalding/parquet/ScaldingDeprecatedInputFormatTest.java
new file mode 100644
index 0000000000..ed86a7813e
--- /dev/null
+++ b/scalding-parquet/src/test/java/com/tw... | true | |
twitter/scalding | 1,936 | issue_to_patch | Remove mapreduce.input.fileinputformat.inputdir setting in memory source | Memory source sets the mapreduce.input.fileinputformat.inputdir property to
a random UUID value. Often in clusters with HDFS federation, paths like that are
not valid namespaces.
While this path is not usually checked since this is a memory source, in clusters
where Kerberos is enabled, Hadoop lists the input dirs ... | 11eac58a875096d9060d697da8d4783622c27687 | 2893db5483665c6e765eccb8bc46a51fb8f7a879 | diff --git a/maple/src/main/java/com/twitter/maple/tap/MemorySourceTap.java b/maple/src/main/java/com/twitter/maple/tap/MemorySourceTap.java
index 1d07de3a23..d3ef58e624 100644
--- a/maple/src/main/java/com/twitter/maple/tap/MemorySourceTap.java
+++ b/maple/src/main/java/com/twitter/maple/tap/MemorySourceTap.java
@@ -4... | [
"maple/src/main/java/com/twitter/maple/tap/MemorySourceTap.java"
] | [] | true | ||
twitter/scalding | 1,934 | issue_to_patch | TypedPipe.groupRandomly() use Ordering with EquivSerialization | Seen `"Scalding's ordered serialization logic exhausted the finite supply of boxed classes."` error when creating many parallel Executions from the same job. We traced it to groupRandomly() not using an EquivSerialization in its groupBy(), taking up new slots of boxed classes for every Execution.
Converted `identity... | 97bd99babc3e340e38324c4e747731f5c971b449 | 1058f861a1a819835b3c9cc8237bbba7543ac364 | diff --git a/scalding-core/src/main/scala/com/twitter/scalding/typed/TypedPipe.scala b/scalding-core/src/main/scala/com/twitter/scalding/typed/TypedPipe.scala
index afe23cc625..a49269f0d4 100644
--- a/scalding-core/src/main/scala/com/twitter/scalding/typed/TypedPipe.scala
+++ b/scalding-core/src/main/scala/com/twitter/... | [
"scalding-core/src/main/scala/com/twitter/scalding/typed/TypedPipe.scala"
] | [
{
"comment": "small note: why don't we do `delegate.hash(x)` here? seems weird that we delegate through the others. Maybe a good reason, we have had weird bugs related to non-uniform partitioning on this method before, and this may be related to fixing that, but it would be nice if we had commented that.",
... | true | ||
twitter/scalding | 1,931 | issue_to_patch | Add additional guards in `ReferencedClassFinder` | During upgrade to Scala 2.12.10 in Twitter we discovered `illegal cyclic reference` exception failing from `scalaType.member(universe.stringToTermName(field.getName)).typeSignature` code in `ReferencedClassFinder`.
In this PR I've guarded this code with the same check as we do for `getClassType`.
Unfortunately I ... | 943c57bbc343a31692c0daa5e6e928a8630f7f24 | 437595d327f6340ffe1b7158382e389ef751f87a | diff --git a/scalding-core/src/main/scala/com/twitter/scalding/ReferencedClassFinder.scala b/scalding-core/src/main/scala/com/twitter/scalding/ReferencedClassFinder.scala
index 53353b4554..b05ce273eb 100644
--- a/scalding-core/src/main/scala/com/twitter/scalding/ReferencedClassFinder.scala
+++ b/scalding-core/src/main/... | [
"scalding-core/src/main/scala/com/twitter/scalding/ReferencedClassFinder.scala"
] | [] | true | ||
twitter/scalding | 1,926 | issue_to_patch | Add type ascriptions to serialization code | Scalding macros expand into a large amount of code, most of which
contained no or very little type ascriptions, leaving a lot of
unnecessary work to the compiler. By explicitly adding these type
ascriptions in the generated code, we can reduce compilation times. | 5b820212ec3723349e270c59da0bf94a57fb53c3 | 87ebd364c650b76dde7f0c7a5b3b9ddaa96c30f8 | diff --git a/scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/ByteBufferOrderedBuf.scala b/scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/ByteBufferOrderedBuf.scala
index d5b356c267.... | [
"scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/ByteBufferOrderedBuf.scala",
"scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/CaseClassOrderedBuf.scala",
"scalding-serializa... | [
{
"comment": "I wonder if we even need this binding. It looks unused doesn’t it? Same with `$b`",
"path": "scalding-serialization/src/main/scala/com/twitter/scalding/serialization/macros/impl/ordered_serialization/providers/TraversablesOrderedBuf.scala",
"hunk": "@@ -136,8 +136,8 @@ object TraversablesO... | true | ||
twitter/scalding | 1,930 | issue_to_patch | Fix scalding travis CI | Scalding CI is broken because of unsecured connection to maven. I fixed it by setting `sbt.repository.secure` to `true`. | 5b820212ec3723349e270c59da0bf94a57fb53c3 | 2350560557f98ebff0ad405730d6f4317c7e1c82 | [
"scripts/run_test.sh"
] | [] | diff --git a/scripts/run_test.sh b/scripts/run_test.sh
index a986828d4c..fbd414abbd 100755
--- a/scripts/run_test.sh
+++ b/scripts/run_test.sh
@@ -19,27 +19,27 @@ PROGRESS_REPORTER_PID=$!
echo "running..."
-echo time ./sbt -Dhttp.keepAlive=false -Dsbt.repository.secure=false ++$TRAVIS_SCALA_VERSION "$(withCmd "tes... | true | ||
twitter/scalding | 1,198 | issue_to_patch | Typed Parquet Tuple | - Used as sink to write tuple with primitive fields and also collection types(List, Set, Map)
- Create Parquet schema generation macro
| 16d62c431f4aefdb2e1f7c3e54f79c7a1ad0c589 | 536bd0c43981475d48922f303df22df8618de1ea | diff --git a/project/Build.scala b/project/Build.scala
index ffa4197dd9..ec0f6c2ec5 100644
--- a/project/Build.scala
+++ b/project/Build.scala
@@ -33,8 +33,10 @@ object ScaldingBuild extends Build {
val hravenVersion = "0.9.13"
val jacksonVersion = "2.4.2"
val json4SVersion = "3.2.11"
+ val paradiseVersion = ... | [
"project/Build.scala",
"scalding-parquet/src/main/scala/com/twitter/scalding/parquet/tuple/TypedParquet.scala",
"scalding-parquet/src/main/scala/com/twitter/scalding/parquet/tuple/macros/Macros.scala",
"scalding-parquet/src/main/scala/com/twitter/scalding/parquet/tuple/macros/impl/ParquetSchemaProvider.scala"... | [
{
"comment": "I'm curious, why do you use \".\" to capture nesting instead of actually nesting groups?\n",
"path": "scalding-parquet/src/test/scala/com/twitter/scalding/parquet/tuple/macros/MacroUnitTests.scala",
"hunk": "@@ -0,0 +1,86 @@\n+package com.twitter.scalding.parquet.tuple.macros\n+\n+import o... | diff --git a/scalding-parquet/src/test/scala/com/twitter/scalding/parquet/tuple/TypedParquetTupleTest.scala b/scalding-parquet/src/test/scala/com/twitter/scalding/parquet/tuple/TypedParquetTupleTest.scala
new file mode 100644
index 0000000000..0b7483818f
--- /dev/null
+++ b/scalding-parquet/src/test/scala/com/twitter/s... | true | |
twitter/scalding | 1,923 | issue_to_patch | Fix typos | Should be non-semantic.
Uses https://en.wikipedia.org/wiki/Wikipedia:Lists_of_common_misspellings/For_machines to find likely typos, with https://github.com/bwignall/typochecker to help automate the checking. | 73ce124f3c0156bd41f5394d2b609a5cd74f370e | a2826072ac8b02c04246bd673f45e09ef5964bfd | diff --git a/CHANGES.md b/CHANGES.md
index d1ddc64c01..eeb38c973e 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -96,7 +96,7 @@ This is the first Scalding release that publishes artifacts for Scala 2.12!
* Update Build.scala : #1361
* Allow overriding of hadoop configuration options for a single source/sink : #1362
*... | [
"CHANGES.md",
"scalding-cats/src/main/scala/com/twitter/scalding/hellcats/HellCats.scala",
"scalding-core/src/main/scala/com/twitter/scalding/CPromise.scala",
"scalding-core/src/main/scala/com/twitter/scalding/Job.scala",
"scalding-core/src/main/scala/com/twitter/scalding/JobTest.scala",
"scalding-core/sr... | [] | diff --git a/scalding-core/src/test/scala/com/twitter/scalding/FileSourceTest.scala b/scalding-core/src/test/scala/com/twitter/scalding/FileSourceTest.scala
index 7b5e94616c..a8ea468d10 100644
--- a/scalding-core/src/test/scala/com/twitter/scalding/FileSourceTest.scala
+++ b/scalding-core/src/test/scala/com/twitter/sca... | true | |
twitter/scalding | 1,920 | issue_to_patch | Cancellable Executions | This change tries to enable cancellation of executions by stopping the underlying flow.
Prior to this change, parallel executions that have started running could not be cancelled, even if one execution failed. This is problematic because if one parallel execution fails, others will continue running and the overall j... | 59d932731f7396eaaf6024624d1ce6660534ca77 | 94cb720084664bb360ec290bfc1324063cec304d | diff --git a/.travis.yml b/.travis.yml
index 42137d4c4f..2a5218f60d 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -1,5 +1,5 @@
language: scala
-jdk: oraclejdk8
+jdk: openjdk8
sudo: false
before_install:
diff --git a/scalding-core/src/main/java/com/twitter/scalding/cascading_interop/FlowListenerPromise.java b/scald... | [
".travis.yml",
"scalding-core/src/main/java/com/twitter/scalding/cascading_interop/FlowListenerPromise.java",
"scalding-core/src/main/scala/com/twitter/scalding/CFuture.scala",
"scalding-core/src/main/scala/com/twitter/scalding/CPromise.scala",
"scalding-core/src/main/scala/com/twitter/scalding/Cancellation... | [
{
"comment": "I think we can do `Future.zip(other.stop(), outer.stop())` I don't think we need to cancel in order.",
"path": "scalding-core/src/main/scala/com/twitter/scalding/CancellationHandler.scala",
"hunk": "@@ -0,0 +1,37 @@\n+package com.twitter.scalding\n+\n+import scala.concurrent.{ Future, Exec... | diff --git a/scalding-core/src/test/scala/com/twitter/scalding/ExecutionTest.scala b/scalding-core/src/test/scala/com/twitter/scalding/ExecutionTest.scala
index 7d64554861..bddda3f9c9 100644
--- a/scalding-core/src/test/scala/com/twitter/scalding/ExecutionTest.scala
+++ b/scalding-core/src/test/scala/com/twitter/scaldi... | true | |
twitter/twemproxy | 210 | issue_to_patch | MGET improve | hi, @manjuraj, I'm trying to import mget performance by rewrite it like this::
orig:
```
mget k1 k2 k3 k4 k5 k6 k7 k8 k9
```
after rewrite:
```
mget k1 k3 k6
mget k2 k4 k7
mget k5 k8 k9
```
this will cost less time on `req_done()` function calls,
I use this redis-benchmark::
```
if (test_is_selected("mget... | 97ced21b0405cc0a46b002c9074b34abee15acb7 | 257b564a6037524348364ea478373d0f8ea3aaaf | diff --git a/.gitignore b/.gitignore
index 014e7e3e..2556202e 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,3 +1,6 @@
+# pyc
+*.pyc
+
# Compiled Object files
*.lo
*.o
diff --git a/notes/redis.md b/notes/redis.md
index f969eb83..50df07e4 100644
--- a/notes/redis.md
+++ b/notes/redis.md
@@ -79,7 +79,7 @@
+------... | [
".gitignore",
"notes/redis.md",
"scripts/benchmark-mget.py",
"src/nc_core.c",
"src/nc_core.h",
"src/nc_log.c",
"src/nc_message.c",
"src/nc_message.h",
"src/nc_request.c",
"src/nc_server.c",
"src/nc_server.h",
"src/nc_string.h",
"src/proto/nc_memcache.c",
"src/proto/nc_proto.h",
"src/prot... | [
{
"comment": "what is this for?\n",
"path": "src/nc_log.c",
"hunk": "@@ -250,5 +250,12 @@ _log_hexdump(const char *file, int line, char *data, int datalen,\n l->nerror++;\n }\n \n+ if (len >= size - 1) {",
"resolving_sha": "257b564a6037524348364ea478373d0f8ea3aaaf",
"resolving_dif... | true | ||
twitter/twemproxy | 396 | issue_to_patch | Add SO_REUSEPORT support to socket | Add SO_REUSEPORT support to socket. Introduce new feature for Twemproxy running on newer kernels. With this feature you are able to do upgrades, restarts without any downtime. Kernel does load balancing between processes with the same host:port pairs.
| c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_conf.c b/src/nc_conf.c
index fa720796..bcb9910f 100644
--- a/src/nc_conf.c
+++ b/src/nc_conf.c
@@ -78,6 +78,10 @@ static struct command conf_commands[] = {
conf_set_bool,
offsetof(struct conf_pool, tcpkeepalive) },
+ { string("reuseport"),
+ conf_set_bool,
+ offsetof(str... | [
"src/nc_conf.c",
"src/nc_conf.h",
"src/nc_proxy.c",
"src/nc_server.h",
"src/nc_stats.c",
"src/nc_util.c",
"src/nc_util.h"
] | [
{
"comment": "@ton31337 twemproxy can run on mac or freebsd also. so I think you should check other os too not only linux :)\n",
"path": "src/nc_stats.c",
"hunk": "@@ -827,7 +828,9 @@ stats_listen(struct stats *st)\n log_error(\"socket failed: %s\", strerror(errno));\n return NC_ERROR;\n... | true | ||
twitter/twemproxy | 396 | comment_to_fix | Add SO_REUSEPORT support to socket | I think this should be replaced with `#ifdef SO_REUSEPORT`, which would cover all of the OS cases | c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_util.c b/src/nc_util.c
index e2768089..d9f1a6b1 100644
--- a/src/nc_util.c
+++ b/src/nc_util.c
@@ -75,6 +75,20 @@ nc_set_reuseaddr(int sd)
return setsockopt(sd, SOL_SOCKET, SO_REUSEADDR, &reuse, len);
}
+int
+nc_set_reuseport(int sd)
+{
+ int reuse;
+ socklen_t len;
+
+ reuse = 1;
+... | [
"src/nc_util.c"
] | [
{
"comment": "I think this should be replaced with `#ifdef SO_REUSEPORT`, which would cover all of the OS cases",
"path": "src/nc_util.c",
"hunk": "@@ -75,6 +78,24 @@ nc_set_reuseaddr(int sd)\n return setsockopt(sd, SOL_SOCKET, SO_REUSEADDR, &reuse, len);\n }\n \n+int\n+nc_set_reuseport(int sd)\n+{\... | true | ||
twitter/twemproxy | 396 | comment_to_fix | Add SO_REUSEPORT support to socket | The `__linux` specific code can be removed if `#ifdef SO_REUSEPORT` is used instead. | c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_util.c b/src/nc_util.c
index e2768089..d9f1a6b1 100644
--- a/src/nc_util.c
+++ b/src/nc_util.c
@@ -75,6 +75,20 @@ nc_set_reuseaddr(int sd)
return setsockopt(sd, SOL_SOCKET, SO_REUSEADDR, &reuse, len);
}
+int
+nc_set_reuseport(int sd)
+{
+ int reuse;
+ socklen_t len;
+
+ reuse = 1;
+... | [
"src/nc_util.c"
] | [
{
"comment": "The `__linux` specific code can be removed if `#ifdef SO_REUSEPORT` is used instead.",
"path": "src/nc_util.c",
"hunk": "@@ -22,6 +22,9 @@\n #include <unistd.h>\n #include <fcntl.h>\n #include <netdb.h>\n+#ifdef __linux\n+#include <linux/version.h>",
"resolving_sha": "3c9cec1d44491daf2... | true | ||
twitter/twemproxy | 396 | comment_to_fix | Add SO_REUSEPORT support to socket | 1. This conditional can be moved into `proxy_reuse` - if reuseport config is set and this fails, then NC_ERROR should be returned
2. This only seems to have a documented effect for AF_INET/AF_INET6 https://man7.org/linux/man-pages/man7/socket.7.html
```
switch (p->family) {
case AF_INET:
case AF_INET... | c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_proxy.c b/src/nc_proxy.c
index 38cdd9e6..b70be97a 100644
--- a/src/nc_proxy.c
+++ b/src/nc_proxy.c
@@ -142,6 +142,16 @@ proxy_listen(struct context *ctx, struct conn *p)
return NC_ERROR;
}
+ if (pool->reuseport) {
+ status = nc_set_reuseport(p->sd);
+ if (status < 0)... | [
"src/nc_proxy.c"
] | [
{
"comment": "1. This conditional can be moved into `proxy_reuse` - if reuseport config is set and this fails, then NC_ERROR should be returned\r\n2. This only seems to have a documented effect for AF_INET/AF_INET6 https://man7.org/linux/man-pages/man7/socket.7.html\r\n\r\n```\r\n switch (p->family) {\r\n ... | true | ||
twitter/twemproxy | 396 | comment_to_fix | Add SO_REUSEPORT support to socket | Looks good, SO_REUSEPORT would definitely help with errors I've encountered during startup when stats port is listening on all interfaces.
| c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_stats.c b/src/nc_stats.c
index fa4d92c7..363a9e88 100644
--- a/src/nc_stats.c
+++ b/src/nc_stats.c
@@ -827,7 +827,7 @@ stats_listen(struct stats *st)
log_error("socket failed: %s", strerror(errno));
return NC_ERROR;
}
-
+ nc_set_reuseport(st->sd);
status = nc_set_reuse... | [
"src/nc_stats.c"
] | [
{
"comment": "Looks good, SO_REUSEPORT would definitely help with errors I've encountered during startup when stats port is listening on all interfaces.\r\n",
"path": "src/nc_stats.c",
"hunk": "@@ -827,7 +827,7 @@ stats_listen(struct stats *st)\n log_error(\"socket failed: %s\", strerror(errno))... | true | ||
twitter/twemproxy | 396 | comment_to_fix | Add SO_REUSEPORT support to socket | ```suggestion
```
I meant remove the change to includes entirely.
This will also fail to build on non-linux platforms, which is why there was originally the `#ifdef __linux` check in a previous revision of your PR | c3f433b86bb0a90e96a1f50d356e3e77c2e26f3b | 3c9cec1d44491daf2abcc860a1efe9666e9fe6fb | diff --git a/src/nc_util.c b/src/nc_util.c
index e2768089..d9f1a6b1 100644
--- a/src/nc_util.c
+++ b/src/nc_util.c
@@ -75,6 +75,20 @@ nc_set_reuseaddr(int sd)
return setsockopt(sd, SOL_SOCKET, SO_REUSEADDR, &reuse, len);
}
+int
+nc_set_reuseport(int sd)
+{
+ int reuse;
+ socklen_t len;
+
+ reuse = 1;
+... | [
"src/nc_util.c"
] | [
{
"comment": "```suggestion\r\n```\r\n\r\nI meant remove the change to includes entirely.\r\n\r\nThis will also fail to build on non-linux platforms, which is why there was originally the `#ifdef __linux` check in a previous revision of your PR",
"path": "src/nc_util.c",
"hunk": "@@ -22,6 +22,9 @@\n #in... | true | ||
twitter/twemproxy | 654 | issue_to_patch | fix parse eval on SW_ARG2 caclulate r->rlen error | Problem
This bug is very hidden and must be reproduced using carefully constructed commands.
The key point is：
* the command is eval
* first enter parse SW_ARG2 (m = p + r->rlen) >= b->last
eg(mbuf is 512):
eval "return &nb... | 0cf49ad3f2f848c20be1710f4ee358a7abca1680 | 724941030ab561dc5afb3ab55b9efe305c36673b | diff --git a/src/proto/nc_redis.c b/src/proto/nc_redis.c
index a5ea210f..a80f84a3 100644
--- a/src/proto/nc_redis.c
+++ b/src/proto/nc_redis.c
@@ -1633,7 +1633,19 @@ redis_parse_req(struct msg *r)
m = p + r->rlen;
if (m >= b->last) {
- r->rlen -= (uint32_t)(b->last - p);
+ ... | [
"src/proto/nc_redis.c"
] | [
{
"comment": "```suggestion\r\n /* \r\n * For EVAL/EVALHASH, the r->token has been assigned a value. When \r\n * m >= b->last happens we will need to repair mbuf. So, when r->token == NULL, \r\n * we will need to calculate r->rlen again.\r\n ... | true | ||
twitter/twemproxy | 654 | comment_to_fix | fix parse eval on SW_ARG2 caclulate r->rlen error | ```suggestion
/*
* For EVAL/EVALHASH, the r->token has been assigned a value. When
* m >= b->last happens we will need to repair mbuf. So, when r->token == NULL,
* we will need to calculate r->rlen again.
*
* A... | 0cf49ad3f2f848c20be1710f4ee358a7abca1680 | 724941030ab561dc5afb3ab55b9efe305c36673b | diff --git a/src/proto/nc_redis.c b/src/proto/nc_redis.c
index a5ea210f..a80f84a3 100644
--- a/src/proto/nc_redis.c
+++ b/src/proto/nc_redis.c
@@ -1633,7 +1633,19 @@ redis_parse_req(struct msg *r)
m = p + r->rlen;
if (m >= b->last) {
- r->rlen -= (uint32_t)(b->last - p);
+ ... | [
"src/proto/nc_redis.c"
] | [
{
"comment": "```suggestion\r\n /* \r\n * For EVAL/EVALHASH, the r->token has been assigned a value. When \r\n * m >= b->last happens we will need to repair mbuf. So, when r->token == NULL, \r\n * we will need to calculate r->rlen again.\r\n ... | true | ||
twitter/twemproxy | 643 | issue_to_patch | Release twemproxy 0.5.0 | 7fb53c512dfde1385ade98a59188043c78768ce7 | ca45646e4e976fa42231fa7114f2cab504e60706 | diff --git a/ChangeLog b/ChangeLog
index 116588be..d7e15161 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+ 2021-13-07 Tyson Andre <tysonandre775@hotmail.com>
+ * twemproxy: version 0.5.0 release
+ Same as 0.5.0-RC1
+
2021-06-07 Tyson Andre <tysonandre775@hotmail.com>
* twemproxy: version 0.5.0... | [
"ChangeLog",
"configure.ac",
"scripts/nutcracker.spec"
] | [] | true | |||
twitter/twemproxy | 640 | issue_to_patch | Release twemproxy 0.5.0-RC1 | To be merged **Tuesday, July 6th or later** if no blockers are discovered.
- Distribution tarballs will be included with the release (To be generated with`autoreconf -fvi` on a fresh checkout)
- The release will be marked as pre-release
- Dates will be updated
- At least a week after 0.5.0-RC1 is out, create a re... | 329e1ab30df19dc81e5eec2752c4928a9cc7345e | f852ffa9c041116b04db8508d84d40bca7df784c | diff --git a/ChangeLog b/ChangeLog
index 72bb3e69..116588be 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,5 @@
- 2021-??-?? Tyson Andre
- * twemproxy: version 0.5.0 release (dev)
+ 2021-06-07 Tyson Andre <tysonandre775@hotmail.com>
+ * twemproxy: version 0.5.0-RC1 release
Add 'tcpkeepalive' pool boolea... | [
"ChangeLog",
"README.md",
"configure.ac",
"scripts/nutcracker.spec"
] | [] | true | ||
twitter/twemproxy | 641 | issue_to_patch | Add miscellaneous tests, fix compilation warnings | Warn about `strdup` being missing during `configure` for libyaml.
(I was checking compatibility with older gcc and should have
used `-std=gnu99`, not `-std=c99`)
Avoid warning about pointer type casts and fallthrough.
Add tests of remaining supported hash functions being compatible with
libmemcached results. | f46d1b85fb5ff670c6015ac8ab8ab519a4a40461 | cfeb446c7cd9e2ca12d80e87c6f39a4aaff859b8 | diff --git a/configure.ac b/configure.ac
index b501391c..12cc1b45 100644
--- a/configure.ac
+++ b/configure.ac
@@ -84,6 +84,7 @@ AC_CHECK_FUNCS([dup2 gethostname gettimeofday strerror])
AC_CHECK_FUNCS([socket])
AC_CHECK_FUNCS([memchr memmove memset])
AC_CHECK_FUNCS([strchr strndup strtoul])
+AC_CHECK_FUNCS([strdup])... | [
"configure.ac",
"src/hashkit/nc_crc32.c",
"src/test_all.c"
] | [] | diff --git a/src/test_all.c b/src/test_all.c
index 22d147cc..c22ec7c6 100644
--- a/src/test_all.c
+++ b/src/test_all.c
@@ -42,6 +42,19 @@ static void test_hash_algorithms(void) {
/* refer to libmemcached tests/hash_results.h */
expect_same_uint32_t(2297466611U, hash_one_at_a_time("apple", 5), "should have exp... | true | |
twitter/twemproxy | 639 | issue_to_patch | Update unit test for redis/memcache request testing | - Always clear errno before testing parsing.
- Update bug reporting address for `make check` to be the github issues url
- Migrate checks of parsing a lot of redis<= 3.2 commands to the C unit test | b649777f3e32a8ec47b2ee59595ca1e49f6b6602 | b4f000bd0cb0fb7e50341742411fb056f84badda | diff --git a/ci/Dockerfile b/ci/Dockerfile
index c360264c..6586490d 100644
--- a/ci/Dockerfile
+++ b/ci/Dockerfile
@@ -1,5 +1,6 @@
-# Dockerfile to build nutcracker
-# ARGS: none
+# Dockerfile to create a slower debug build of nutcracker with assertions enabled
+# for continuous integration checks.
+# ARGS: REDIS_VER
... | [
"ci/Dockerfile",
"configure.ac",
"src/test_all.c"
] | [] | diff --git a/src/test_all.c b/src/test_all.c
index 33743596..22d147cc 100644
--- a/src/test_all.c
+++ b/src/test_all.c
@@ -61,6 +61,7 @@ static void test_config_parsing(void) {
}
static void test_redis_parse_req_success_case(const char* data, int expected_type) {
+ const int original_failures = failures;
st... | true |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.