\r by itself is mapped to \n which causes things like AWS S3 Key names to have information loss when parsed by SweetXxml

[cmarkle]

Can SweetXml be configured / motivated to not map single \r characters to \n? This causes issues when parsing XML form AWS S3 storage service when object names contain \r (which although not smart to use, is legal to use).

Working with this XML returned from Amazon AWS S3 storage service:

<ListBucketResult xmlns="http://s3.amazonaws.com/doc/2006-03-01/">
  <Name>obfuscated_bucket_name</Name>
  <Prefix/>
  <NextContinuationToken>1ILUW_obfuscated_continuation_token_1yMUM</NextContinuationToken>
  <KeyCount>1000</KeyCount>
  <MaxKeys>1000</MaxKeys>
  <Delimiter/>
  <IsTruncated>true</IsTruncated>
  <Contents>
    <Key>workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -&gt; MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon&#13;</Key>
    <LastModified>2022-03-25T23:52:22.122Z</LastModified>
    <ETag>"d41d8cd98f00b204e9800998ecf8427e"</ETag>
    <Size>0</Size>
    <StorageClass>STANDARD</StorageClass>
  </Contents>
  <Contents>
    <Key>workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -&gt; MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon&#13;inmiddle</Key>
    <LastModified>2022-03-26T23:52:22.122Z</LastModified>
    <ETag>"c41d8cd98f00b204e9800998ecf8427e"</ETag>
    <Size>0</Size>
    <StorageClass>STANDARD</StorageClass>
  </Contents>
</ListBucketResult>

I used the SweetXml xpath snippet from lib/ex_aws/s3/parsers.ex in the ex_aws/ex_aws_s3 project, as-is:

iex(19)>  xml
"<?xml version=\"1.0\" encoding=\"UTF-8\" standalone=\"yes\"?>\n<ListBucketResult xmlns=\"http://s3.amazonaws.com/doc/2006-03-01/\">\n  <Name>obfuscated_bucket_name</Name>\n  <Prefix></Prefix>\n  <NextContinuationToken>1ILUW_obfuscated_continuation_token_1yMUM</NextContinuationToken>\n  <KeyCount>1000</KeyCount>\n  <MaxKeys>1000</MaxKeys>\n  <Delimiter></Delimiter>\n  <IsTruncated>true</IsTruncated>\n  <Contents>\n    <Key>workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -&gt; MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon&#13;</Key>\n    <LastModified>2022-03-25T23:52:22.122Z</LastModified>\n    <ETag>&quot;d41d8cd98f00b204e9800998ecf8427e&quot;</ETag>\n    <Size>0</Size>\n    <StorageClass>STANDARD</StorageClass>\n  </Contents>\n  <Contents>\n    <Key>workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -&gt; MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon&#13;inmiddle</Key>\n    <LastModified>2022-03-26T23:52:22.122Z</LastModified>\n    <ETag>&quot;c41d8cd98f00b204e9800998ecf8427e&quot;</ETag>\n    <Size>0</Size>\n    <StorageClass>STANDARD</StorageClass>\n  </Contents>\n</ListBucketResult>"

iex(22)> parsed_body =
...(22)>         xml |> SweetXml.xpath(~x"//ListBucketResult",
...(22)>           name: ~x"./Name/text()"s,
...(22)>           is_truncated: ~x"./IsTruncated/text()"s,
...(22)>           prefix: ~x"./Prefix/text()"s,
...(22)>           marker: ~x"./Marker/text()"s,
...(22)>           next_continuation_token: ~x"./NextContinuationToken/text()"s,
...(22)>           key_count: ~x"./KeyCount/text()"s,
...(22)>           max_keys: ~x"./MaxKeys/text()"s,
...(22)>           next_marker: ~x"./NextMarker/text()"s,
...(22)>           contents: [
...(22)>             ~x"./Contents"l,
...(22)>             key: ~x"./Key/text()"s,
...(22)>             last_modified: ~x"./LastModified/text()"s,
...(22)>             e_tag: ~x"./ETag/text()"s,
...(22)>             size: ~x"./Size/text()"s,
...(22)>             storage_class: ~x"./StorageClass/text()"s,
...(22)>             owner: [
...(22)>               ~x"./Owner"o,
...(22)>               id: ~x"./ID/text()"s,
...(22)>               display_name: ~x"./DisplayName/text()"s
...(22)>             ]
...(22)>           ],
...(22)>           common_prefixes: [
...(22)>             ~x"./CommonPrefixes"l,
...(22)>             prefix: ~x"./Prefix/text()"s
...(22)>           ]
...(22)>         )
%{
  common_prefixes: [],
  contents: [
    %{
      e_tag: "\"d41d8cd98f00b204e9800998ecf8427e\"",
      key: "workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -> MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon\n",
      last_modified: "2022-03-25T23:52:22.122Z",
      owner: nil,
      size: "0",
      storage_class: "STANDARD"
    },
    %{
      e_tag: "\"c41d8cd98f00b204e9800998ecf8427e\"",
      key: "workspaces/51109/packages/ND-7H46rSQ.asp-package/contents/WHALE -> MH/Whale_TTB_DOCUMENTS/Aspera Inboxes/Icon\ninmiddle",
      last_modified: "2022-03-26T23:52:22.122Z",
      owner: nil,
      size: "0",
      storage_class: "STANDARD"
    }
  ],
  is_truncated: "true",
  key_count: "1000",
  marker: "",
  max_keys: "1000",
  name: "obfuscated_bucket_name",
  next_continuation_token: "1ILUW_obfuscated_continuation_token_1yMUM",
  next_marker: "",
  prefix: ""
}

Note that the S3 Key's with \r in them have their \r characters mapped to \n. In general this is probably "right" and in conformance with the XML specs wrt end of line handling, but I think we should ideally be able to use SweetXml in some way against this type of input and have the \r and other similar characters mapped to their XML escape equivalents (e.g., 
 in the \r case).

Can SweetXml be used in a way in which this mapping of \r to \n is not done?