From 4626f9b32d8d45cb8e3f0d4891c53fa6ee1d0d70 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Thu, 27 Mar 2025 17:30:28 -0400
Subject: [PATCH 01/20] Temp pin to datafusion main

---
 Cargo.toml | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/Cargo.toml b/Cargo.toml
index b6164f89d31e8..c282d2283cfcb 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -206,3 +206,20 @@ used_underscore_binding = "warn"
 [workspace.lints.rust]
 unexpected_cfgs = { level = "warn", check-cfg = ["cfg(tarpaulin)"] }
 unused_qualifications = "deny"
+
+
+## Temporary arrow-rs patch until 55
+
+[patch.crates-io]
+arrow = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-array = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-buffer = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-cast = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-data = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-ipc = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-schema = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-select = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-string = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-ord = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow-flight = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+parquet = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }

From 98d62e0d0e448818de8a2bf01c35e5214e31534f Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 28 Mar 2025 18:41:06 -0400
Subject: [PATCH 02/20] Update cargo lock

---
 Cargo.lock | 64 ++++++++++++++++++++++++++++++++++++++++++++----------
 1 file changed, 53 insertions(+), 11 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 16fccdb2865b7..8a7181675972d 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -301,9 +301,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-buffer"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4e899dade2c3b7f5642eb8366cfd898958bcca099cde6dfea543c7e8d3ad88d4"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "bytes",
  "half",
@@ -349,9 +348,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-data"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0a329fb064477c9ec5f0870d2f5130966f91055c7c5bce2b3a084f116bc28c3b"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-buffer",
  "arrow-schema",
@@ -448,9 +446,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-schema"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "85934a9d0261e0fa5d4e2a5295107d743b543a6e0484a835d4b8db2da15306f9"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "bitflags 2.8.0",
  "serde",
@@ -3897,7 +3894,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "fc2f4eb4bc735547cfed7c0a4922cbd04a4655978c09b54f1f7b228750664c34"
 dependencies = [
  "cfg-if",
- "windows-targets 0.52.6",
+ "windows-targets 0.48.5",
 ]
 
 [[package]]
@@ -7046,7 +7043,7 @@ version = "0.1.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cf221c93e13a30d793f7645a0e7762c55d169dbb0a49671918a2319d289b10bb"
 dependencies = [
- "windows-sys 0.59.0",
+ "windows-sys 0.48.0",
 ]
 
 [[package]]
@@ -7501,3 +7498,48 @@ dependencies = [
  "cc",
  "pkg-config",
 ]
+
+[[patch.unused]]
+name = "arrow"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-array"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-cast"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-flight"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-ipc"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-ord"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-select"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "arrow-string"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+
+[[patch.unused]]
+name = "parquet"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"

From fc1f7a58a44a86d749913e71300cfe28e442479f Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 28 Mar 2025 18:45:10 -0400
Subject: [PATCH 03/20] update pyo3

---
 Cargo.lock                   | 167 +++++++++++++----------------------
 Cargo.toml                   |  14 +--
 datafusion/common/Cargo.toml |   2 +-
 3 files changed, 69 insertions(+), 114 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 8a7181675972d..1c59c62bbcd35 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -246,9 +246,8 @@ checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
 
 [[package]]
 name = "arrow"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dc208515aa0151028e464cc94a692156e945ce5126abd3537bb7fd6ba2143ed1"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -265,14 +264,13 @@ dependencies = [
  "arrow-string",
  "half",
  "pyo3",
- "rand 0.8.5",
+ "rand 0.9.0",
 ]
 
 [[package]]
 name = "arrow-arith"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e07e726e2b3f7816a85c6a45b6ec118eeeabf0b2a8c208122ad949437181f49a"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -284,9 +282,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-array"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a2262eba4f16c78496adfd559a29fe4b24df6088efc9985a873d58e92be022d5"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "ahash 0.8.11",
  "arrow-buffer",
@@ -311,9 +308,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-cast"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4103d88c5b441525ed4ac23153be7458494c2b0c9a11115848fdb9b81f6f886a"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -332,9 +328,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-csv"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "43d3cb0914486a3cae19a5cad2598e44e225d53157926d0ada03c20521191a65"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-cast",
@@ -359,9 +354,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-flight"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c7408f2bf3b978eddda272c7699f439760ebc4ac70feca25fefa82c5b8ce808d"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -386,9 +380,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-ipc"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ddecdeab02491b1ce88885986e25002a3da34dd349f682c7cfe67bab7cc17b86"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -400,9 +393,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-json"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d03b9340013413eb84868682ace00a1098c81a5ebc96d279f7ebf9a4cac3c0fd"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -413,16 +405,17 @@ dependencies = [
  "half",
  "indexmap 2.8.0",
  "lexical-core",
+ "memchr",
  "num",
  "serde",
  "serde_json",
+ "simdutf8",
 ]
 
 [[package]]
 name = "arrow-ord"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f841bfcc1997ef6ac48ee0305c4dfceb1f7c786fe31e67c1186edf775e1f1160"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -433,9 +426,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-row"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1eeb55b0a0a83851aa01f2ca5ee5648f607e8506ba6802577afdda9d75cdedcd"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -455,9 +447,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-select"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7e2932aece2d0c869dd2125feb9bd1709ef5c445daa3838ac4112dcfa0fda52c"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -469,9 +460,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-string"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "912e38bd6a7a7714c1d9b61df80315685553b7455e8a6045c27531d8ecd5b458"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -1344,9 +1334,9 @@ checksum = "613afe47fcd5fac7ccf1db93babcb082c5994d996f20b8b159f2ad1658eb5724"
 
 [[package]]
 name = "chrono"
-version = "0.4.39"
+version = "0.4.40"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7e36cc9d416881d2e24f9a963be5fb1cd90966419ac844274161d10488b3e825"
+checksum = "1a7964611d71df112cb1730f2ee67324fcf4d0fc6606acbbe9bfe06df124637c"
 dependencies = [
  "android-tzdata",
  "iana-time-zone",
@@ -1354,7 +1344,7 @@ dependencies = [
  "num-traits",
  "serde",
  "wasm-bindgen",
- "windows-targets 0.52.6",
+ "windows-link",
 ]
 
 [[package]]
@@ -2900,11 +2890,11 @@ checksum = "1d674e81391d1e1ab681a28d99df07927c6d4aa5b027d7da16ba32d1d21ecd99"
 
 [[package]]
 name = "flatbuffers"
-version = "24.12.23"
+version = "25.2.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4f1baf0dbf96932ec9a3038d57900329c015b0bfb7b63d904f3bc27e2b02a096"
+checksum = "1045398c1bfd89168b5fd3f1fc11f6e70b34f6f66300c87d44d3de849463abf1"
 dependencies = [
- "bitflags 1.3.2",
+ "bitflags 2.8.0",
  "rustc_version",
 ]
 
@@ -3982,7 +3972,7 @@ version = "0.11.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "75761162ae2b0e580d7e7c390558127e5f01b4194debd6221fd8c207fc80e3f5"
 dependencies = [
- "twox-hash",
+ "twox-hash 1.6.3",
 ]
 
 [[package]]
@@ -4343,9 +4333,8 @@ dependencies = [
 
 [[package]]
 name = "parquet"
-version = "54.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f88838dca3b84d41444a0341b19f347e8098a3898b0f21536654b8b799e11abd"
+version = "54.3.1"
+source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -4373,9 +4362,8 @@ dependencies = [
  "snap",
  "thrift",
  "tokio",
- "twox-hash",
+ "twox-hash 2.1.0",
  "zstd",
- "zstd-sys",
 ]
 
 [[package]]
@@ -4822,9 +4810,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3"
-version = "0.23.5"
+version = "0.24.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7778bffd85cf38175ac1f545509665d0b9b92a198ca7941f131f85f7a4f9a872"
+checksum = "7f1c6c3591120564d64db2261bec5f910ae454f01def849b9c22835a84695e86"
 dependencies = [
  "cfg-if",
  "indoc",
@@ -4840,9 +4828,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.23.5"
+version = "0.24.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "94f6cbe86ef3bf18998d9df6e0f3fc1050a8c5efa409bf712e661a4366e010fb"
+checksum = "e9b6c2b34cf71427ea37c7001aefbaeb85886a074795e35f161f5aecc7620a7a"
 dependencies = [
  "once_cell",
  "target-lexicon",
@@ -4850,9 +4838,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.23.5"
+version = "0.24.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e9f1b4c431c0bb1c8fb0a338709859eed0d030ff6daa34368d3b152a63dfdd8d"
+checksum = "5507651906a46432cdda02cd02dd0319f6064f1374c9147c45b978621d2c3a9c"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -4860,9 +4848,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.23.5"
+version = "0.24.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fbc2201328f63c4710f68abdf653c89d8dbc2858b88c5d88b0ff38a75288a9da"
+checksum = "b0d394b5b4fd8d97d48336bb0dd2aebabad39f1d294edd6bcd2cccf2eefe6f42"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
@@ -4872,9 +4860,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.23.5"
+version = "0.24.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fca6726ad0f3da9c9de093d6f116a93c1a38e417ed73bf138472cf4064f72028"
+checksum = "fd72da09cfa943b1080f621f024d2ef7e2773df7badd51aa30a2be1f8caa7c8e"
 dependencies = [
  "heck 0.5.0",
  "proc-macro2",
@@ -6110,9 +6098,9 @@ checksum = "55937e1799185b12863d447f42597ed69d9928686b8d88a1df17376a097d8369"
 
 [[package]]
 name = "target-lexicon"
-version = "0.12.16"
+version = "0.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "61c41af27dd6d1e27b1b16b489db798443478cef1f06a660c96db617ba5de3b1"
+checksum = "e502f78cdbb8ba4718f566c418c52bc729126ffd16baee5baa718cf25dd5a69a"
 
 [[package]]
 name = "tempfile"
@@ -6613,6 +6601,12 @@ dependencies = [
  "static_assertions",
 ]
 
+[[package]]
+name = "twox-hash"
+version = "2.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e7b17f197b3050ba473acf9181f7b1d3b66d1cf7356c6cc57886662276e65908"
+
 [[package]]
 name = "typed-arena"
 version = "2.0.2"
@@ -7105,6 +7099,12 @@ dependencies = [
  "syn 2.0.100",
 ]
 
+[[package]]
+name = "windows-link"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "76840935b766e1b0a05c0066835fb9ec80071d4c09a16f6bd5f7e655e3c14c38"
+
 [[package]]
 name = "windows-registry"
 version = "0.2.0"
@@ -7498,48 +7498,3 @@ dependencies = [
  "cc",
  "pkg-config",
 ]
-
-[[patch.unused]]
-name = "arrow"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-array"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-cast"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-flight"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-ipc"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-ord"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-select"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "arrow-string"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
-
-[[patch.unused]]
-name = "parquet"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
diff --git a/Cargo.toml b/Cargo.toml
index c282d2283cfcb..4c06b418820c9 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -87,19 +87,19 @@ ahash = { version = "0.8", default-features = false, features = [
     "runtime-rng",
 ] }
 apache-avro = { version = "0.17", default-features = false }
-arrow = { version = "54.2.1", features = [
+arrow = { version = "54.3.1", features = [
     "prettyprint",
     "chrono-tz",
 ] }
-arrow-buffer = { version = "54.1.0", default-features = false }
-arrow-flight = { version = "54.2.1", features = [
+arrow-buffer = { version = "54.3.1", default-features = false }
+arrow-flight = { version = "54.3.1", features = [
     "flight-sql-experimental",
 ] }
-arrow-ipc = { version = "54.2.0", default-features = false, features = [
+arrow-ipc = { version = "54.3.1", default-features = false, features = [
     "lz4",
 ] }
-arrow-ord = { version = "54.1.0", default-features = false }
-arrow-schema = { version = "54.1.0", default-features = false }
+arrow-ord = { version = "54.3.1", default-features = false }
+arrow-schema = { version = "54.3.1", default-features = false }
 async-trait = "0.1.88"
 bigdecimal = "0.4.7"
 bytes = "1.10"
@@ -149,7 +149,7 @@ itertools = "0.14"
 log = "^0.4"
 object_store = { version = "0.11.0", default-features = false }
 parking_lot = "0.12"
-parquet = { version = "54.2.1", default-features = false, features = [
+parquet = { version = "54.3.1", default-features = false, features = [
     "arrow",
     "async",
     "object_store",
diff --git a/datafusion/common/Cargo.toml b/datafusion/common/Cargo.toml
index 39b47a96bccf3..74e99163955e9 100644
--- a/datafusion/common/Cargo.toml
+++ b/datafusion/common/Cargo.toml
@@ -63,7 +63,7 @@ log = { workspace = true }
 object_store = { workspace = true, optional = true }
 parquet = { workspace = true, optional = true, default-features = true }
 paste = "1.0.15"
-pyo3 = { version = "0.23.5", optional = true }
+pyo3 = { version = "0.24.0", optional = true }
 recursive = { workspace = true, optional = true }
 sqlparser = { workspace = true }
 tokio = { workspace = true }

From 9299b4bea974e1f09b1325338d88fe2cf240c31d Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 28 Mar 2025 18:52:21 -0400
Subject: [PATCH 04/20] vendor random generation

---
 .../functions-aggregate/benches/array_agg.rs  | 28 +++++++++++++++++--
 datafusion/functions/benches/chr.rs           | 10 +++++--
 2 files changed, 34 insertions(+), 4 deletions(-)

diff --git a/datafusion/functions-aggregate/benches/array_agg.rs b/datafusion/functions-aggregate/benches/array_agg.rs
index fb605e87ed0cc..e22be611d8d76 100644
--- a/datafusion/functions-aggregate/benches/array_agg.rs
+++ b/datafusion/functions-aggregate/benches/array_agg.rs
@@ -19,17 +19,23 @@ use std::sync::Arc;
 
 use arrow::array::{
     Array, ArrayRef, ArrowPrimitiveType, AsArray, ListArray, NullBufferBuilder,
+    PrimitiveArray,
 };
 use arrow::datatypes::{Field, Int64Type};
-use arrow::util::bench_util::create_primitive_array;
 use criterion::{black_box, criterion_group, criterion_main, Criterion};
 use datafusion_expr::Accumulator;
 use datafusion_functions_aggregate::array_agg::ArrayAggAccumulator;
 
 use arrow::buffer::OffsetBuffer;
-use arrow::util::test_util::seedable_rng;
 use rand::distributions::{Distribution, Standard};
+use rand::prelude::StdRng;
 use rand::Rng;
+use rand::SeedableRng;
+
+/// Returns fixed seedable RNG
+pub fn seedable_rng() -> StdRng {
+    StdRng::seed_from_u64(42)
+}
 
 fn merge_batch_bench(c: &mut Criterion, name: &str, values: ArrayRef) {
     let list_item_data_type = values.as_list::<i32>().values().data_type().clone();
@@ -46,6 +52,24 @@ fn merge_batch_bench(c: &mut Criterion, name: &str, values: ArrayRef) {
     });
 }
 
+pub fn create_primitive_array<T>(size: usize, null_density: f32) -> PrimitiveArray<T>
+where
+    T: ArrowPrimitiveType,
+    Standard: Distribution<T::Native>,
+{
+    let mut rng = seedable_rng();
+
+    (0..size)
+        .map(|_| {
+            if rng.gen::<f32>() < null_density {
+                None
+            } else {
+                Some(rng.gen())
+            }
+        })
+        .collect()
+}
+
 /// Create List array with the given item data type, null density, null locations and zero length lists density
 /// Creates an random (but fixed-seeded) array of a given size and null density
 pub fn create_list_array<T>(
diff --git a/datafusion/functions/benches/chr.rs b/datafusion/functions/benches/chr.rs
index 4750fb4666532..8575809c21c8b 100644
--- a/datafusion/functions/benches/chr.rs
+++ b/datafusion/functions/benches/chr.rs
@@ -17,15 +17,21 @@
 
 extern crate criterion;
 
-use arrow::{array::PrimitiveArray, datatypes::Int64Type, util::test_util::seedable_rng};
+use arrow::{array::PrimitiveArray, datatypes::Int64Type};
 use criterion::{black_box, criterion_group, criterion_main, Criterion};
 use datafusion_expr::{ColumnarValue, ScalarFunctionArgs};
 use datafusion_functions::string::chr;
-use rand::Rng;
+use rand::{Rng, SeedableRng};
 
 use arrow::datatypes::DataType;
+use rand::rngs::StdRng;
 use std::sync::Arc;
 
+/// Returns fixed seedable RNG
+pub fn seedable_rng() -> StdRng {
+    StdRng::seed_from_u64(42)
+}
+
 fn criterion_benchmark(c: &mut Criterion) {
     let cot_fn = chr();
     let size = 1024;

From 7b8320eee948396f7e76db694e769ed22481805a Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 28 Mar 2025 18:54:50 -0400
Subject: [PATCH 05/20] Update error message

---
 datafusion/sqllogictest/test_files/dates.slt      |  2 +-
 datafusion/sqllogictest/test_files/timestamps.slt | 10 +++++-----
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/datafusion/sqllogictest/test_files/dates.slt b/datafusion/sqllogictest/test_files/dates.slt
index 4425eee333735..148f0dfe64bb7 100644
--- a/datafusion/sqllogictest/test_files/dates.slt
+++ b/datafusion/sqllogictest/test_files/dates.slt
@@ -183,7 +183,7 @@ query error input contains invalid characters
 SELECT to_date('2020-09-08 12/00/00+00:00', '%c', '%+')
 
 # to_date with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_date('2020-09-08 12/00/00+00:00', '%q')
 
 statement ok
diff --git a/datafusion/sqllogictest/test_files/timestamps.slt b/datafusion/sqllogictest/test_files/timestamps.slt
index dcbcfbfa439d5..e3f8d2e4c8bb2 100644
--- a/datafusion/sqllogictest/test_files/timestamps.slt
+++ b/datafusion/sqllogictest/test_files/timestamps.slt
@@ -2241,23 +2241,23 @@ query error input contains invalid characters
 SELECT to_timestamp_seconds('2020-09-08 12/00/00+00:00', '%c', '%+')
 
 # to_timestamp with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_timestamp('2020-09-08 12/00/00+00:00', '%q')
 
 # to_timestamp_nanos with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_timestamp_nanos('2020-09-08 12/00/00+00:00', '%q')
 
 # to_timestamp_millis with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_timestamp_millis('2020-09-08 12/00/00+00:00', '%q')
 
 # to_timestamp_micros with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_timestamp_micros('2020-09-08 12/00/00+00:00', '%q')
 
 # to_timestamp_seconds with broken formatting
-query error bad or unsupported format string
+query error DataFusion error: Execution error: Error parsing timestamp from '2020\-09\-08 12/00/00\+00:00' using format '%q': trailing input
 SELECT to_timestamp_seconds('2020-09-08 12/00/00+00:00', '%q')
 
 # Create string timestamp table with different formats

From cdc55e5c30ee68dc7e1f4f450d6ad8c7af551ee8 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 28 Mar 2025 18:56:46 -0400
Subject: [PATCH 06/20] Update for extraction

---
 datafusion/sqllogictest/test_files/expr/date_part.slt | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/datafusion/sqllogictest/test_files/expr/date_part.slt b/datafusion/sqllogictest/test_files/expr/date_part.slt
index dec796aa59cb5..39c42cbe1e97f 100644
--- a/datafusion/sqllogictest/test_files/expr/date_part.slt
+++ b/datafusion/sqllogictest/test_files/expr/date_part.slt
@@ -884,7 +884,7 @@ SELECT extract(day from arrow_cast('14400 minutes', 'Interval(DayTime)'))
 query I
 SELECT extract(minute from arrow_cast('14400 minutes', 'Interval(DayTime)'))
 ----
-14400
+0
 
 query I
 SELECT extract(second from arrow_cast('5.1 seconds', 'Interval(DayTime)'))
@@ -894,7 +894,7 @@ SELECT extract(second from arrow_cast('5.1 seconds', 'Interval(DayTime)'))
 query I
 SELECT extract(second from arrow_cast('14400 minutes', 'Interval(DayTime)'))
 ----
-864000
+0
 
 query I
 SELECT extract(second from arrow_cast('2 months', 'Interval(MonthDayNano)'))
@@ -954,7 +954,7 @@ from t
 order by id;
 ----
 0 0 5
-1 0 15
+1 0 3
 2 0 0
 3 2 0
 4 0 8

From 53ec3530128546cf77006499cb3a9fb0d4111d3d Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 10:38:17 -0400
Subject: [PATCH 07/20] Update pin

---
 Cargo.lock | 165 ++++++++++++++++++++++++++++++++---------------------
 Cargo.toml |  39 +++++++------
 2 files changed, 121 insertions(+), 83 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 1dcca0dccb708..c778e8a7b846b 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -246,8 +246,8 @@ checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
 
 [[package]]
 name = "arrow"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -269,8 +269,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-arith"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -282,8 +282,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-array"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "ahash 0.8.11",
  "arrow-buffer",
@@ -298,8 +298,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-buffer"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "bytes",
  "half",
@@ -308,8 +308,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-cast"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -328,8 +328,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-csv"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-cast",
@@ -343,8 +343,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-data"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-buffer",
  "arrow-schema",
@@ -354,8 +354,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-flight"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -380,8 +380,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-ipc"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -393,8 +393,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-json"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -414,8 +414,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-ord"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -426,8 +426,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-row"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -438,8 +438,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-schema"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "bitflags 2.8.0",
  "serde",
@@ -447,8 +447,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-select"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -460,8 +460,8 @@ dependencies = [
 
 [[package]]
 name = "arrow-string"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -1841,7 +1841,7 @@ dependencies = [
  "itertools 0.14.0",
  "log",
  "nix",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "parquet",
  "paste",
@@ -1874,7 +1874,7 @@ dependencies = [
  "futures",
  "log",
  "mimalloc",
- "object_store",
+ "object_store 0.11.2",
  "parquet",
  "rand 0.8.5",
  "serde",
@@ -1905,7 +1905,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "tokio",
 ]
@@ -1927,7 +1927,7 @@ dependencies = [
  "datafusion-session",
  "futures",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "tempfile",
  "tokio",
 ]
@@ -1950,7 +1950,7 @@ dependencies = [
  "insta",
  "insta-cmd",
  "mimalloc",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "parquet",
  "predicates",
@@ -1977,7 +1977,7 @@ dependencies = [
  "insta",
  "libc",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parquet",
  "paste",
  "pyo3",
@@ -2020,7 +2020,7 @@ dependencies = [
  "glob",
  "itertools 0.14.0",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parquet",
  "rand 0.8.5",
  "tempfile",
@@ -2050,7 +2050,7 @@ dependencies = [
  "datafusion-session",
  "futures",
  "num-traits",
- "object_store",
+ "object_store 0.11.2",
  "rstest",
  "serde_json",
  "tokio",
@@ -2074,7 +2074,7 @@ dependencies = [
  "datafusion-physical-plan",
  "datafusion-session",
  "futures",
- "object_store",
+ "object_store 0.11.2",
  "regex",
  "tokio",
 ]
@@ -2097,7 +2097,7 @@ dependencies = [
  "datafusion-physical-plan",
  "datafusion-session",
  "futures",
- "object_store",
+ "object_store 0.11.2",
  "serde_json",
  "tokio",
 ]
@@ -2125,7 +2125,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "parquet",
  "rand 0.8.5",
@@ -2153,7 +2153,7 @@ dependencies = [
  "log",
  "mimalloc",
  "nix",
- "object_store",
+ "object_store 0.11.2",
  "prost",
  "tempfile",
  "test-utils",
@@ -2176,7 +2176,7 @@ dependencies = [
  "datafusion-expr",
  "futures",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "rand 0.8.5",
  "tempfile",
@@ -2493,7 +2493,7 @@ dependencies = [
  "datafusion-functions-window-common",
  "datafusion-proto-common",
  "doc-comment",
- "object_store",
+ "object_store 0.11.2",
  "pbjson",
  "prost",
  "serde",
@@ -2532,7 +2532,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "parking_lot",
  "tokio",
 ]
@@ -2578,7 +2578,7 @@ dependencies = [
  "indicatif",
  "itertools 0.14.0",
  "log",
- "object_store",
+ "object_store 0.11.2",
  "postgres-protocol",
  "postgres-types",
  "rust_decimal",
@@ -2603,7 +2603,7 @@ dependencies = [
  "datafusion-functions-aggregate",
  "insta",
  "itertools 0.14.0",
- "object_store",
+ "object_store 0.11.2",
  "pbjson-types",
  "prost",
  "serde_json",
@@ -2627,7 +2627,7 @@ dependencies = [
  "datafusion-sql",
  "getrandom 0.2.15",
  "insta",
- "object_store",
+ "object_store 0.11.2",
  "tokio",
  "url",
  "wasm-bindgen",
@@ -2922,6 +2922,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "11faaf5a5236997af9848be0bef4db95824b1d534ebc64d0f0c6cf3e67bd38dc"
 dependencies = [
  "crc32fast",
+ "libz-rs-sys",
  "miniz_oxide",
 ]
 
@@ -3943,6 +3944,15 @@ dependencies = [
  "escape8259",
 ]
 
+[[package]]
+name = "libz-rs-sys"
+version = "0.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "902bc563b5d65ad9bba616b490842ef0651066a1a1dc3ce1087113ffcb873c8d"
+dependencies = [
+ "zlib-rs",
+]
+
 [[package]]
 name = "linked-hash-map"
 version = "0.5.6"
@@ -4274,6 +4284,27 @@ dependencies = [
  "walkdir",
 ]
 
+[[package]]
+name = "object_store"
+version = "0.12.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e9ce831b09395f933addbc56d894d889e4b226eba304d4e7adbab591e26daf1e"
+dependencies = [
+ "async-trait",
+ "bytes",
+ "chrono",
+ "futures",
+ "http 1.2.0",
+ "humantime",
+ "itertools 0.14.0",
+ "parking_lot",
+ "percent-encoding",
+ "thiserror 2.0.12",
+ "tokio",
+ "tracing",
+ "url",
+]
+
 [[package]]
 name = "once_cell"
 version = "1.20.3"
@@ -4350,8 +4381,8 @@ dependencies = [
 
 [[package]]
 name = "parquet"
-version = "54.3.1"
-source = "git+https://github.com/apache/arrow-rs.git?rev=b30336da8c1318f3f45e22f0a377ca21830fddac#b30336da8c1318f3f45e22f0a377ca21830fddac"
+version = "55.0.0"
+source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -4372,7 +4403,7 @@ dependencies = [
  "lz4_flex",
  "num",
  "num-bigint",
- "object_store",
+ "object_store 0.12.0",
  "paste",
  "seq-macro",
  "simdutf8",
@@ -4752,7 +4783,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "be769465445e8c1474e9c5dac2018218498557af32d9ed057325ec9a41ae81bf"
 dependencies = [
  "heck 0.5.0",
- "itertools 0.14.0",
+ "itertools 0.13.0",
  "log",
  "multimap",
  "once_cell",
@@ -4772,7 +4803,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8a56d757972c98b346a9b766e3f02746cde6dd1cd1d1d563472929fdd74bec4d"
 dependencies = [
  "anyhow",
- "itertools 0.14.0",
+ "itertools 0.13.0",
  "proc-macro2",
  "quote",
  "syn 2.0.100",
@@ -4827,9 +4858,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7f1c6c3591120564d64db2261bec5f910ae454f01def849b9c22835a84695e86"
+checksum = "17da310086b068fbdcefbba30aeb3721d5bb9af8db4987d6735b2183ca567229"
 dependencies = [
  "cfg-if",
  "indoc",
@@ -4845,9 +4876,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e9b6c2b34cf71427ea37c7001aefbaeb85886a074795e35f161f5aecc7620a7a"
+checksum = "e27165889bd793000a098bb966adc4300c312497ea25cf7a690a9f0ac5aa5fc1"
 dependencies = [
  "once_cell",
  "target-lexicon",
@@ -4855,9 +4886,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5507651906a46432cdda02cd02dd0319f6064f1374c9147c45b978621d2c3a9c"
+checksum = "05280526e1dbf6b420062f3ef228b78c0c54ba94e157f5cb724a609d0f2faabc"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -4865,9 +4896,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b0d394b5b4fd8d97d48336bb0dd2aebabad39f1d294edd6bcd2cccf2eefe6f42"
+checksum = "5c3ce5686aa4d3f63359a5100c62a127c9f15e8398e5fdeb5deef1fed5cd5f44"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
@@ -4877,9 +4908,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.24.0"
+version = "0.24.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fd72da09cfa943b1080f621f024d2ef7e2773df7badd51aa30a2be1f8caa7c8e"
+checksum = "f4cf6faa0cbfb0ed08e89beb8103ae9724eb4750e3a78084ba4017cbe94f3855"
 dependencies = [
  "heck 0.5.0",
  "proc-macro2",
@@ -7054,7 +7085,7 @@ version = "0.1.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "cf221c93e13a30d793f7645a0e7762c55d169dbb0a49671918a2319d289b10bb"
 dependencies = [
- "windows-sys 0.48.0",
+ "windows-sys 0.59.0",
 ]
 
 [[package]]
@@ -7488,6 +7519,12 @@ dependencies = [
  "syn 2.0.100",
 ]
 
+[[package]]
+name = "zlib-rs"
+version = "0.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8b20717f0917c908dc63de2e44e97f1e6b126ca58d0e391cee86d504eb8fbd05"
+
 [[package]]
 name = "zstd"
 version = "0.13.3"
diff --git a/Cargo.toml b/Cargo.toml
index 7181d8f87ee36..593ff94ec9dd3 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -87,19 +87,19 @@ ahash = { version = "0.8", default-features = false, features = [
     "runtime-rng",
 ] }
 apache-avro = { version = "0.17", default-features = false }
-arrow = { version = "54.3.1", features = [
+arrow = { version = "55.0.0", features = [
     "prettyprint",
     "chrono-tz",
 ] }
-arrow-buffer = { version = "54.3.1", default-features = false }
-arrow-flight = { version = "54.3.1", features = [
+arrow-buffer = { version = "55.0.0", default-features = false }
+arrow-flight = { version = "55.0.0", features = [
     "flight-sql-experimental",
 ] }
-arrow-ipc = { version = "54.3.1", default-features = false, features = [
+arrow-ipc = { version = "55.0.0", default-features = false, features = [
     "lz4",
 ] }
-arrow-ord = { version = "54.3.1", default-features = false }
-arrow-schema = { version = "54.3.1", default-features = false }
+arrow-ord = { version = "55.0.0", default-features = false }
+arrow-schema = { version = "55.0.0", default-features = false }
 async-trait = "0.1.88"
 bigdecimal = "0.4.7"
 bytes = "1.10"
@@ -149,7 +149,7 @@ itertools = "0.14"
 log = "^0.4"
 object_store = { version = "0.11.0", default-features = false }
 parking_lot = "0.12"
-parquet = { version = "54.3.1", default-features = false, features = [
+parquet = { version = "55.0.0", default-features = false, features = [
     "arrow",
     "async",
     "object_store",
@@ -216,17 +216,18 @@ unused_qualifications = "deny"
 
 
 ## Temporary arrow-rs patch until 55
+## https://github.com/apache/arrow-rs/pull/7391
 
 [patch.crates-io]
-arrow = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-array = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-buffer = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-cast = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-data = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-ipc = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-schema = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-select = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-string = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-ord = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-arrow-flight = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
-parquet = { git = "https://github.com/apache/arrow-rs.git", rev = "b30336da8c1318f3f45e22f0a377ca21830fddac" }
+arrow = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-array = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-buffer = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-cast = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-data = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-ipc = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-schema = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-select = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-string = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-ord = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow-flight = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+parquet = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }

From c103a03e2d6a5d1e6f424709d7fb3357a3838174 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 10:46:14 -0400
Subject: [PATCH 08/20] Upgrade object_store

---
 Cargo.lock | 91 ++++++++++++++++--------------------------------------
 Cargo.toml |  2 +-
 2 files changed, 27 insertions(+), 66 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 9f74e6f0a7a14..177d779c43f60 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1841,7 +1841,7 @@ dependencies = [
  "itertools 0.14.0",
  "log",
  "nix",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "parquet",
  "paste",
@@ -1874,7 +1874,7 @@ dependencies = [
  "futures",
  "log",
  "mimalloc",
- "object_store 0.11.2",
+ "object_store",
  "parquet",
  "rand 0.8.5",
  "serde",
@@ -1905,7 +1905,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "tokio",
 ]
@@ -1927,7 +1927,7 @@ dependencies = [
  "datafusion-session",
  "futures",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "tempfile",
  "tokio",
 ]
@@ -1950,7 +1950,7 @@ dependencies = [
  "insta",
  "insta-cmd",
  "mimalloc",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "parquet",
  "predicates",
@@ -1977,7 +1977,7 @@ dependencies = [
  "insta",
  "libc",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parquet",
  "paste",
  "pyo3",
@@ -2021,7 +2021,7 @@ dependencies = [
  "glob",
  "itertools 0.14.0",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parquet",
  "rand 0.8.5",
  "tempfile",
@@ -2051,7 +2051,7 @@ dependencies = [
  "datafusion-session",
  "futures",
  "num-traits",
- "object_store 0.11.2",
+ "object_store",
  "rstest",
  "serde_json",
  "tokio",
@@ -2075,7 +2075,7 @@ dependencies = [
  "datafusion-physical-plan",
  "datafusion-session",
  "futures",
- "object_store 0.11.2",
+ "object_store",
  "regex",
  "tokio",
 ]
@@ -2098,7 +2098,7 @@ dependencies = [
  "datafusion-physical-plan",
  "datafusion-session",
  "futures",
- "object_store 0.11.2",
+ "object_store",
  "serde_json",
  "tokio",
 ]
@@ -2126,7 +2126,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "parquet",
  "rand 0.8.5",
@@ -2154,7 +2154,7 @@ dependencies = [
  "log",
  "mimalloc",
  "nix",
- "object_store 0.11.2",
+ "object_store",
  "prost",
  "tempfile",
  "test-utils",
@@ -2177,7 +2177,7 @@ dependencies = [
  "datafusion-expr",
  "futures",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "rand 0.8.5",
  "tempfile",
@@ -2495,7 +2495,7 @@ dependencies = [
  "datafusion-functions-window-common",
  "datafusion-proto-common",
  "doc-comment",
- "object_store 0.11.2",
+ "object_store",
  "pbjson",
  "prost",
  "serde",
@@ -2534,7 +2534,7 @@ dependencies = [
  "futures",
  "itertools 0.14.0",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "parking_lot",
  "tokio",
 ]
@@ -2580,7 +2580,7 @@ dependencies = [
  "indicatif",
  "itertools 0.14.0",
  "log",
- "object_store 0.11.2",
+ "object_store",
  "postgres-protocol",
  "postgres-types",
  "rust_decimal",
@@ -2605,7 +2605,7 @@ dependencies = [
  "datafusion-functions-aggregate",
  "insta",
  "itertools 0.14.0",
- "object_store 0.11.2",
+ "object_store",
  "pbjson-types",
  "prost",
  "serde_json",
@@ -2629,7 +2629,7 @@ dependencies = [
  "datafusion-sql",
  "getrandom 0.2.15",
  "insta",
- "object_store 0.11.2",
+ "object_store",
  "tokio",
  "url",
  "wasm-bindgen",
@@ -4257,18 +4257,21 @@ dependencies = [
 
 [[package]]
 name = "object_store"
-version = "0.11.2"
+version = "0.12.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3cfccb68961a56facde1163f9319e0d15743352344e7808a11795fb99698dcaf"
+checksum = "e9ce831b09395f933addbc56d894d889e4b226eba304d4e7adbab591e26daf1e"
 dependencies = [
  "async-trait",
  "base64 0.22.1",
  "bytes",
  "chrono",
+ "form_urlencoded",
  "futures",
+ "http 1.2.0",
+ "http-body-util",
  "humantime",
  "hyper",
- "itertools 0.13.0",
+ "itertools 0.14.0",
  "md-5",
  "parking_lot",
  "percent-encoding",
@@ -4279,28 +4282,7 @@ dependencies = [
  "rustls-pemfile",
  "serde",
  "serde_json",
- "snafu",
- "tokio",
- "tracing",
- "url",
- "walkdir",
-]
-
-[[package]]
-name = "object_store"
-version = "0.12.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e9ce831b09395f933addbc56d894d889e4b226eba304d4e7adbab591e26daf1e"
-dependencies = [
- "async-trait",
- "bytes",
- "chrono",
- "futures",
- "http 1.2.0",
- "humantime",
- "itertools 0.14.0",
- "parking_lot",
- "percent-encoding",
+ "serde_urlencoded",
  "thiserror 2.0.12",
  "tokio",
  "tracing",
@@ -4405,7 +4387,7 @@ dependencies = [
  "lz4_flex",
  "num",
  "num-bigint",
- "object_store 0.12.0",
+ "object_store",
  "paste",
  "seq-macro",
  "simdutf8",
@@ -5810,27 +5792,6 @@ version = "1.14.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "7fcf8323ef1faaee30a44a340193b1ac6814fd9b7b4e88e9d4519a3e4abe1cfd"
 
-[[package]]
-name = "snafu"
-version = "0.8.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "223891c85e2a29c3fe8fb900c1fae5e69c2e42415e3177752e8718475efa5019"
-dependencies = [
- "snafu-derive",
-]
-
-[[package]]
-name = "snafu-derive"
-version = "0.8.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "03c3c6b7927ffe7ecaa769ee0e3994da3b8cafc8f444578982c83ecb161af917"
-dependencies = [
- "heck 0.5.0",
- "proc-macro2",
- "quote",
- "syn 2.0.100",
-]
-
 [[package]]
 name = "snap"
 version = "1.1.1"
diff --git a/Cargo.toml b/Cargo.toml
index 593ff94ec9dd3..f65d23169ad5e 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -147,7 +147,7 @@ hashbrown = { version = "0.14.5", features = ["raw"] }
 indexmap = "2.8.0"
 itertools = "0.14"
 log = "^0.4"
-object_store = { version = "0.11.0", default-features = false }
+object_store = { version = "0.12.0", default-features = false }
 parking_lot = "0.12"
 parquet = { version = "55.0.0", default-features = false, features = [
     "arrow",

From 586851bcee0f5daed336c3815fa5093a84d3c5ba Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 10:49:05 -0400
Subject: [PATCH 09/20] fix feature

---
 Cargo.lock                      | 1 +
 datafusion/execution/Cargo.toml | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/Cargo.lock b/Cargo.lock
index 177d779c43f60..357d3a922456c 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -4287,6 +4287,7 @@ dependencies = [
  "tokio",
  "tracing",
  "url",
+ "walkdir",
 ]
 
 [[package]]
diff --git a/datafusion/execution/Cargo.toml b/datafusion/execution/Cargo.toml
index 8f642f3384d2e..20e507e98b68a 100644
--- a/datafusion/execution/Cargo.toml
+++ b/datafusion/execution/Cargo.toml
@@ -44,7 +44,7 @@ datafusion-common = { workspace = true, default-features = true }
 datafusion-expr = { workspace = true }
 futures = { workspace = true }
 log = { workspace = true }
-object_store = { workspace = true }
+object_store = { workspace = true, features = ["fs"] }
 parking_lot = { workspace = true }
 rand = { workspace = true }
 tempfile = { workspace = true }

From cabfb589f52b40f3de6d2946025f85f99baf58bf Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 10:58:10 -0400
Subject: [PATCH 10/20] Update file size handling

---
 datafusion/datasource/src/file_groups.rs | 13 +++++++------
 datafusion/datasource/src/mod.rs         | 24 +++++++++++++++---------
 2 files changed, 22 insertions(+), 15 deletions(-)

diff --git a/datafusion/datasource/src/file_groups.rs b/datafusion/datasource/src/file_groups.rs
index a1f966c22f35f..15c86427ed00a 100644
--- a/datafusion/datasource/src/file_groups.rs
+++ b/datafusion/datasource/src/file_groups.rs
@@ -224,10 +224,11 @@ impl FileGroupPartitioner {
             return None;
         }
 
-        let target_partition_size = (total_size as usize).div_ceil(target_partitions);
+        let target_partition_size =
+            (total_size as u64).div_ceil(target_partitions as u64);
 
         let current_partition_index: usize = 0;
-        let current_partition_size: usize = 0;
+        let current_partition_size: u64 = 0;
 
         // Partition byte range evenly for all `PartitionedFile`s
         let repartitioned_files = flattened_files
@@ -497,15 +498,15 @@ struct ToRepartition {
     /// the index from which the original file will be taken
     source_index: usize,
     /// the size of the original file
-    file_size: usize,
+    file_size: u64,
     /// indexes of which group(s) will this be distributed to (including `source_index`)
     new_groups: Vec<usize>,
 }
 
 impl ToRepartition {
-    // how big will each file range be when this file is read in its new groups?
-    fn range_size(&self) -> usize {
-        self.file_size / self.new_groups.len()
+    /// How big will each file range be when this file is read in its new groups?
+    fn range_size(&self) -> u64 {
+        self.file_size / (self.new_groups.len() as u64)
     }
 }
 
diff --git a/datafusion/datasource/src/mod.rs b/datafusion/datasource/src/mod.rs
index c02f84c74d64b..1322fefdda505 100644
--- a/datafusion/datasource/src/mod.rs
+++ b/datafusion/datasource/src/mod.rs
@@ -52,7 +52,7 @@ pub use self::url::ListingTableUrl;
 use crate::file_groups::FileGroup;
 use chrono::TimeZone;
 use datafusion_common::stats::Precision;
-use datafusion_common::{ColumnStatistics, Result};
+use datafusion_common::{exec_datafusion_err, ColumnStatistics, Result};
 use datafusion_common::{ScalarValue, Statistics};
 use file_meta::FileMeta;
 use futures::{Stream, StreamExt};
@@ -123,7 +123,7 @@ impl PartitionedFile {
             object_meta: ObjectMeta {
                 location: Path::from(path.into()),
                 last_modified: chrono::Utc.timestamp_nanos(0),
-                size: size as usize,
+                size,
                 e_tag: None,
                 version: None,
             },
@@ -141,7 +141,7 @@ impl PartitionedFile {
             object_meta: ObjectMeta {
                 location: Path::from(path),
                 last_modified: chrono::Utc.timestamp_nanos(0),
-                size: size as usize,
+                size,
                 e_tag: None,
                 version: None,
             },
@@ -224,7 +224,7 @@ impl From<ObjectMeta> for PartitionedFile {
 ///   Indicates that the range calculation determined no further action is
 ///   necessary, possibly because the calculated range is empty or invalid.
 pub enum RangeCalculation {
-    Range(Option<Range<usize>>),
+    Range(Option<Range<u64>>),
     TerminateEarly,
 }
 
@@ -250,7 +250,12 @@ pub async fn calculate_range(
     match file_meta.range {
         None => Ok(RangeCalculation::Range(None)),
         Some(FileRange { start, end }) => {
-            let (start, end) = (start as usize, end as usize);
+            let start: u64 = start.try_into().map_err(|_| {
+                exec_datafusion_err!("Expect start range to fit in u64, got {start}")
+            })?;
+            let end: u64 = end.try_into().map_err(|_| {
+                exec_datafusion_err!("Expect end range to fit in u64, got {end}")
+            })?;
 
             let start_delta = if start != 0 {
                 find_first_newline(store, location, start - 1, file_size, newline).await?
@@ -289,10 +294,10 @@ pub async fn calculate_range(
 async fn find_first_newline(
     object_store: &Arc<dyn ObjectStore>,
     location: &Path,
-    start: usize,
-    end: usize,
+    start: u64,
+    end: u64,
     newline: u8,
-) -> Result<usize> {
+) -> Result<u64> {
     let options = GetOptions {
         range: Some(GetRange::Bounded(start..end)),
         ..Default::default()
@@ -305,10 +310,11 @@ async fn find_first_newline(
 
     while let Some(chunk) = result_stream.next().await.transpose()? {
         if let Some(position) = chunk.iter().position(|&byte| byte == newline) {
+            let position = position as u64;
             return Ok(index + position);
         }
 
-        index += chunk.len();
+        index += chunk.len() as u64;
     }
 
     Ok(index)

From d980c009a65420e5221b5ba1b95ef3a3ae7dba0f Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 11:39:06 -0400
Subject: [PATCH 11/20] bash for object store API changes

---
 .../datasource/physical_plan/arrow_file.rs    | 20 +++++++++----------
 datafusion/core/src/test/object_store.rs      |  4 ++--
 .../datasource-parquet/src/file_format.rs     |  3 ++-
 datafusion/datasource-parquet/src/reader.rs   | 16 ++++++++-------
 4 files changed, 23 insertions(+), 20 deletions(-)

diff --git a/datafusion/core/src/datasource/physical_plan/arrow_file.rs b/datafusion/core/src/datasource/physical_plan/arrow_file.rs
index 5dcf4df73f57a..21d962e6d11fb 100644
--- a/datafusion/core/src/datasource/physical_plan/arrow_file.rs
+++ b/datafusion/core/src/datasource/physical_plan/arrow_file.rs
@@ -305,7 +305,7 @@ impl FileOpener for ArrowOpener {
                     )?;
                     // read footer according to footer_len
                     let get_option = GetOptions {
-                        range: Some(GetRange::Suffix(10 + footer_len)),
+                        range: Some(GetRange::Suffix(10 + (footer_len as u64))),
                         ..Default::default()
                     };
                     let get_result = object_store
@@ -332,9 +332,9 @@ impl FileOpener for ArrowOpener {
                         .iter()
                         .flatten()
                         .map(|block| {
-                            let block_len = block.bodyLength() as usize
-                                + block.metaDataLength() as usize;
-                            let block_offset = block.offset() as usize;
+                            let block_len =
+                                block.bodyLength() as u64 + block.metaDataLength() as u64;
+                            let block_offset = block.offset() as u64;
                             block_offset..block_offset + block_len
                         })
                         .collect_vec();
@@ -354,9 +354,9 @@ impl FileOpener for ArrowOpener {
                         .iter()
                         .flatten()
                         .filter(|block| {
-                            let block_offset = block.offset() as usize;
-                            block_offset >= range.start as usize
-                                && block_offset < range.end as usize
+                            let block_offset = block.offset() as u64;
+                            block_offset >= range.start as u64
+                                && block_offset < range.end as u64
                         })
                         .copied()
                         .collect_vec();
@@ -364,9 +364,9 @@ impl FileOpener for ArrowOpener {
                     let recordbatch_ranges = recordbatches
                         .iter()
                         .map(|block| {
-                            let block_len = block.bodyLength() as usize
-                                + block.metaDataLength() as usize;
-                            let block_offset = block.offset() as usize;
+                            let block_len =
+                                block.bodyLength() as u64 + block.metaDataLength() as u64;
+                            let block_offset = block.offset() as u64;
                             block_offset..block_offset + block_len
                         })
                         .collect_vec();
diff --git a/datafusion/core/src/test/object_store.rs b/datafusion/core/src/test/object_store.rs
index e1328770cabdd..8b19658bb1473 100644
--- a/datafusion/core/src/test/object_store.rs
+++ b/datafusion/core/src/test/object_store.rs
@@ -66,7 +66,7 @@ pub fn local_unpartitioned_file(path: impl AsRef<std::path::Path>) -> ObjectMeta
     ObjectMeta {
         location,
         last_modified: metadata.modified().map(chrono::DateTime::from).unwrap(),
-        size: metadata.len() as usize,
+        size: metadata.len(),
         e_tag: None,
         version: None,
     }
@@ -166,7 +166,7 @@ impl ObjectStore for BlockingObjectStore {
     fn list(
         &self,
         prefix: Option<&Path>,
-    ) -> BoxStream<'_, object_store::Result<ObjectMeta>> {
+    ) -> BoxStream<'static, object_store::Result<ObjectMeta>> {
         self.inner.list(prefix)
     }
 
diff --git a/datafusion/datasource-parquet/src/file_format.rs b/datafusion/datasource-parquet/src/file_format.rs
index 1d9a67fd2eb6d..f534d6769b9f1 100644
--- a/datafusion/datasource-parquet/src/file_format.rs
+++ b/datafusion/datasource-parquet/src/file_format.rs
@@ -739,6 +739,7 @@ impl MetadataFetch for ObjectStoreFetch<'_> {
         &mut self,
         range: Range<usize>,
     ) -> BoxFuture<'_, Result<Bytes, ParquetError>> {
+        let range = range.start as u64..range.end as u64;
         async {
             self.store
                 .get_range(&self.meta.location, range)
@@ -765,7 +766,7 @@ pub async fn fetch_parquet_metadata(
 
     ParquetMetaDataReader::new()
         .with_prefetch_hint(size_hint)
-        .load_and_finish(fetch, file_size)
+        .load_and_finish(fetch, file_size as usize)
         .await
         .map_err(DataFusionError::from)
 }
diff --git a/datafusion/datasource-parquet/src/reader.rs b/datafusion/datasource-parquet/src/reader.rs
index 5924a5b5038fc..141b30d431379 100644
--- a/datafusion/datasource-parquet/src/reader.rs
+++ b/datafusion/datasource-parquet/src/reader.rs
@@ -18,19 +18,19 @@
 //! [`ParquetFileReaderFactory`] and [`DefaultParquetFileReaderFactory`] for
 //! low level control of parquet file readers
 
+use crate::ParquetFileMetrics;
 use bytes::Bytes;
 use datafusion_datasource::file_meta::FileMeta;
 use datafusion_physical_plan::metrics::ExecutionPlanMetricsSet;
 use futures::future::BoxFuture;
 use object_store::ObjectStore;
+use parquet::arrow::arrow_reader::ArrowReaderOptions;
 use parquet::arrow::async_reader::{AsyncFileReader, ParquetObjectReader};
 use parquet::file::metadata::ParquetMetaData;
 use std::fmt::Debug;
 use std::ops::Range;
 use std::sync::Arc;
 
-use crate::ParquetFileMetrics;
-
 /// Interface for reading parquet files.
 ///
 /// The combined implementations of [`ParquetFileReaderFactory`] and
@@ -114,10 +114,11 @@ impl AsyncFileReader for ParquetFileReader {
         self.inner.get_byte_ranges(ranges)
     }
 
-    fn get_metadata(
-        &mut self,
-    ) -> BoxFuture<'_, parquet::errors::Result<Arc<ParquetMetaData>>> {
-        self.inner.get_metadata()
+    fn get_metadata<'a>(
+        &'a mut self,
+        options: Option<&'a ArrowReaderOptions>,
+    ) -> BoxFuture<'a, parquet::errors::Result<Arc<ParquetMetaData>>> {
+        self.inner.get_metadata(options)
     }
 }
 
@@ -135,7 +136,8 @@ impl ParquetFileReaderFactory for DefaultParquetFileReaderFactory {
             metrics,
         );
         let store = Arc::clone(&self.store);
-        let mut inner = ParquetObjectReader::new(store, file_meta.object_meta);
+        let mut inner = ParquetObjectReader::new(store, file_meta.object_meta.location)
+            .with_file_size(file_meta.object_meta.size as usize);
 
         if let Some(hint) = metadata_size_hint {
             inner = inner.with_footer_size_hint(hint)

From 2dd1827bff47eaa7210356e1bbb6622986ee124b Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 11:41:05 -0400
Subject: [PATCH 12/20] few more

---
 datafusion/core/src/test_util/parquet.rs         | 2 +-
 datafusion/proto/src/physical_plan/from_proto.rs | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/datafusion/core/src/test_util/parquet.rs b/datafusion/core/src/test_util/parquet.rs
index 084554eecbdb0..f5753af64d93f 100644
--- a/datafusion/core/src/test_util/parquet.rs
+++ b/datafusion/core/src/test_util/parquet.rs
@@ -102,7 +102,7 @@ impl TestParquetFile {
 
         println!("Generated test dataset with {num_rows} rows");
 
-        let size = std::fs::metadata(&path)?.len() as usize;
+        let size = std::fs::metadata(&path)?.len();
 
         let mut canonical_path = path.canonicalize()?;
 
diff --git a/datafusion/proto/src/physical_plan/from_proto.rs b/datafusion/proto/src/physical_plan/from_proto.rs
index c949e3c9f8cb1..88963ed6bba52 100644
--- a/datafusion/proto/src/physical_plan/from_proto.rs
+++ b/datafusion/proto/src/physical_plan/from_proto.rs
@@ -555,7 +555,7 @@ impl TryFrom<&protobuf::PartitionedFile> for PartitionedFile {
             object_meta: ObjectMeta {
                 location: Path::from(val.path.as_str()),
                 last_modified: Utc.timestamp_nanos(val.last_modified_ns as i64),
-                size: val.size as usize,
+                size: val.size,
                 e_tag: None,
                 version: None,
             },

From dff9490ee720ae5f9139a3813bdef1b192e8a9bc Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 15:06:41 -0400
Subject: [PATCH 13/20] Update APIs more

---
 .../examples/advanced_parquet_index.rs         |  5 ++++-
 .../core/src/datasource/file_format/arrow.rs   |  4 ++--
 .../core/src/datasource/file_format/csv.rs     | 17 +++++++++--------
 .../core/src/datasource/file_format/parquet.rs |  6 +++---
 datafusion/core/src/datasource/mod.rs          |  2 +-
 .../src/datasource/physical_plan/parquet.rs    |  4 ++--
 datafusion/core/tests/parquet/custom_reader.rs |  5 ++++-
 datafusion/core/tests/parquet/page_pruning.rs  |  2 +-
 datafusion/core/tests/sql/path_partition.rs    | 18 ++++++++++--------
 .../datasource-parquet/src/row_group_filter.rs |  7 +++++--
 datafusion/proto/src/physical_plan/to_proto.rs |  2 +-
 .../substrait/src/physical_plan/producer.rs    |  2 +-
 12 files changed, 43 insertions(+), 31 deletions(-)

diff --git a/datafusion-examples/examples/advanced_parquet_index.rs b/datafusion-examples/examples/advanced_parquet_index.rs
index b8c303e221618..113c43dc13236 100644
--- a/datafusion-examples/examples/advanced_parquet_index.rs
+++ b/datafusion-examples/examples/advanced_parquet_index.rs
@@ -571,7 +571,9 @@ impl ParquetFileReaderFactory for CachedParquetFileReaderFactory {
             .to_string();
 
         let object_store = Arc::clone(&self.object_store);
-        let mut inner = ParquetObjectReader::new(object_store, file_meta.object_meta);
+        let mut inner =
+            ParquetObjectReader::new(object_store, file_meta.object_meta.location)
+                .with_file_size(file_meta.object_meta.size as usize);
 
         if let Some(hint) = metadata_size_hint {
             inner = inner.with_footer_size_hint(hint)
@@ -618,6 +620,7 @@ impl AsyncFileReader for ParquetReaderWithCache {
 
     fn get_metadata(
         &mut self,
+        _options: Option<&ArrowReaderOptions>,
     ) -> BoxFuture<'_, datafusion::parquet::errors::Result<Arc<ParquetMetaData>>> {
         println!("get_metadata: {} returning cached metadata", self.filename);
 
diff --git a/datafusion/core/src/datasource/file_format/arrow.rs b/datafusion/core/src/datasource/file_format/arrow.rs
index 6c7c9463cf3b7..ea027ae91a061 100644
--- a/datafusion/core/src/datasource/file_format/arrow.rs
+++ b/datafusion/core/src/datasource/file_format/arrow.rs
@@ -442,7 +442,7 @@ mod tests {
         let object_meta = ObjectMeta {
             location,
             last_modified: DateTime::default(),
-            size: usize::MAX,
+            size: u64::MAX,
             e_tag: None,
             version: None,
         };
@@ -485,7 +485,7 @@ mod tests {
         let object_meta = ObjectMeta {
             location,
             last_modified: DateTime::default(),
-            size: usize::MAX,
+            size: u64::MAX,
             e_tag: None,
             version: None,
         };
diff --git a/datafusion/core/src/datasource/file_format/csv.rs b/datafusion/core/src/datasource/file_format/csv.rs
index 309458975ab6c..9fa4c00e6af25 100644
--- a/datafusion/core/src/datasource/file_format/csv.rs
+++ b/datafusion/core/src/datasource/file_format/csv.rs
@@ -72,7 +72,7 @@ mod tests {
     #[derive(Debug)]
     struct VariableStream {
         bytes_to_repeat: Bytes,
-        max_iterations: usize,
+        max_iterations: u64,
         iterations_detected: Arc<Mutex<usize>>,
     }
 
@@ -103,14 +103,15 @@ mod tests {
 
         async fn get(&self, location: &Path) -> object_store::Result<GetResult> {
             let bytes = self.bytes_to_repeat.clone();
-            let range = 0..bytes.len() * self.max_iterations;
+            let len = bytes.len() as u64;
+            let range = 0..len * self.max_iterations;
             let arc = self.iterations_detected.clone();
             let stream = futures::stream::repeat_with(move || {
                 let arc_inner = arc.clone();
                 *arc_inner.lock().unwrap() += 1;
                 Ok(bytes.clone())
             })
-            .take(self.max_iterations)
+            .take(self.max_iterations as usize)
             .boxed();
 
             Ok(GetResult {
@@ -138,7 +139,7 @@ mod tests {
         async fn get_ranges(
             &self,
             _location: &Path,
-            _ranges: &[Range<usize>],
+            _ranges: &[Range<u64>],
         ) -> object_store::Result<Vec<Bytes>> {
             unimplemented!()
         }
@@ -154,7 +155,7 @@ mod tests {
         fn list(
             &self,
             _prefix: Option<&Path>,
-        ) -> BoxStream<'_, object_store::Result<ObjectMeta>> {
+        ) -> BoxStream<'static, object_store::Result<ObjectMeta>> {
             unimplemented!()
         }
 
@@ -179,7 +180,7 @@ mod tests {
     }
 
     impl VariableStream {
-        pub fn new(bytes_to_repeat: Bytes, max_iterations: usize) -> Self {
+        pub fn new(bytes_to_repeat: Bytes, max_iterations: u64) -> Self {
             Self {
                 bytes_to_repeat,
                 max_iterations,
@@ -371,7 +372,7 @@ mod tests {
         let object_meta = ObjectMeta {
             location: Path::parse("/")?,
             last_modified: DateTime::default(),
-            size: usize::MAX,
+            size: u64::MAX,
             e_tag: None,
             version: None,
         };
@@ -429,7 +430,7 @@ mod tests {
         let object_meta = ObjectMeta {
             location: Path::parse("/")?,
             last_modified: DateTime::default(),
-            size: usize::MAX,
+            size: u64::MAX,
             e_tag: None,
             version: None,
         };
diff --git a/datafusion/core/src/datasource/file_format/parquet.rs b/datafusion/core/src/datasource/file_format/parquet.rs
index 67a7ba8dc776e..76009ccd80b05 100644
--- a/datafusion/core/src/datasource/file_format/parquet.rs
+++ b/datafusion/core/src/datasource/file_format/parquet.rs
@@ -331,7 +331,7 @@ mod tests {
         fn list(
             &self,
             _prefix: Option<&Path>,
-        ) -> BoxStream<'_, object_store::Result<ObjectMeta>> {
+        ) -> BoxStream<'static, object_store::Result<ObjectMeta>> {
             Box::pin(futures::stream::once(async {
                 Err(object_store::Error::NotImplemented)
             }))
@@ -408,7 +408,7 @@ mod tests {
         )));
 
         // Use the file size as the hint so we can get the full metadata from the first fetch
-        let size_hint = meta[0].size;
+        let size_hint = meta[0].size as usize;
 
         fetch_parquet_metadata(store.upcast().as_ref(), &meta[0], Some(size_hint))
             .await
@@ -443,7 +443,7 @@ mod tests {
         )));
 
         // Use the a size hint larger than the file size to make sure we don't panic
-        let size_hint = meta[0].size + 100;
+        let size_hint = (meta[0].size + 100) as usize;
 
         fetch_parquet_metadata(store.upcast().as_ref(), &meta[0], Some(size_hint))
             .await
diff --git a/datafusion/core/src/datasource/mod.rs b/datafusion/core/src/datasource/mod.rs
index a15b2b6ffe137..25a89644cd2a4 100644
--- a/datafusion/core/src/datasource/mod.rs
+++ b/datafusion/core/src/datasource/mod.rs
@@ -106,7 +106,7 @@ mod tests {
         let meta = ObjectMeta {
             location,
             last_modified: metadata.modified().map(chrono::DateTime::from).unwrap(),
-            size: metadata.len() as usize,
+            size: metadata.len(),
             e_tag: None,
             version: None,
         };
diff --git a/datafusion/core/src/datasource/physical_plan/parquet.rs b/datafusion/core/src/datasource/physical_plan/parquet.rs
index 5c06c3902c1c8..5986460cb539a 100644
--- a/datafusion/core/src/datasource/physical_plan/parquet.rs
+++ b/datafusion/core/src/datasource/physical_plan/parquet.rs
@@ -1786,13 +1786,13 @@ mod tests {
         path: &str,
         store: Arc<dyn ObjectStore>,
         batch: RecordBatch,
-    ) -> usize {
+    ) -> u64 {
         let mut writer =
             ArrowWriter::try_new(BytesMut::new().writer(), batch.schema(), None).unwrap();
         writer.write(&batch).unwrap();
         writer.flush().unwrap();
         let bytes = writer.into_inner().unwrap().into_inner().freeze();
-        let total_size = bytes.len();
+        let total_size = bytes.len() as u64;
         let path = Path::from(path);
         let payload = object_store::PutPayload::from_bytes(bytes);
         store
diff --git a/datafusion/core/tests/parquet/custom_reader.rs b/datafusion/core/tests/parquet/custom_reader.rs
index ce5c0d720174d..c06da2e75a461 100644
--- a/datafusion/core/tests/parquet/custom_reader.rs
+++ b/datafusion/core/tests/parquet/custom_reader.rs
@@ -44,6 +44,7 @@ use insta::assert_snapshot;
 use object_store::memory::InMemory;
 use object_store::path::Path;
 use object_store::{ObjectMeta, ObjectStore};
+use parquet::arrow::arrow_reader::ArrowReaderOptions;
 use parquet::arrow::async_reader::AsyncFileReader;
 use parquet::arrow::ArrowWriter;
 use parquet::errors::ParquetError;
@@ -186,7 +187,7 @@ async fn store_parquet_in_memory(
                 location: Path::parse(format!("file-{offset}.parquet"))
                     .expect("creating path"),
                 last_modified: chrono::DateTime::from(SystemTime::now()),
-                size: buf.len(),
+                size: buf.len() as u64,
                 e_tag: None,
                 version: None,
             };
@@ -222,6 +223,7 @@ impl AsyncFileReader for ParquetFileReader {
     ) -> BoxFuture<'_, parquet::errors::Result<Bytes>> {
         self.metrics.bytes_scanned.add(range.end - range.start);
 
+        let range = range.start as u64..range.end as u64;
         self.store
             .get_range(&self.meta.location, range)
             .map_err(|e| {
@@ -232,6 +234,7 @@ impl AsyncFileReader for ParquetFileReader {
 
     fn get_metadata(
         &mut self,
+        _options: Option<&ArrowReaderOptions>,
     ) -> BoxFuture<'_, parquet::errors::Result<Arc<ParquetMetaData>>> {
         Box::pin(async move {
             let metadata = fetch_parquet_metadata(
diff --git a/datafusion/core/tests/parquet/page_pruning.rs b/datafusion/core/tests/parquet/page_pruning.rs
index 7006bf083eeed..f693485cbe018 100644
--- a/datafusion/core/tests/parquet/page_pruning.rs
+++ b/datafusion/core/tests/parquet/page_pruning.rs
@@ -52,7 +52,7 @@ async fn get_parquet_exec(state: &SessionState, filter: Expr) -> DataSourceExec
     let meta = ObjectMeta {
         location,
         last_modified: metadata.modified().map(chrono::DateTime::from).unwrap(),
-        size: metadata.len() as usize,
+        size: metadata.len(),
         e_tag: None,
         version: None,
     };
diff --git a/datafusion/core/tests/sql/path_partition.rs b/datafusion/core/tests/sql/path_partition.rs
index bf8466d849f25..fa6c7432413f1 100644
--- a/datafusion/core/tests/sql/path_partition.rs
+++ b/datafusion/core/tests/sql/path_partition.rs
@@ -712,7 +712,7 @@ impl ObjectStore for MirroringObjectStore {
         let meta = ObjectMeta {
             location: location.clone(),
             last_modified: metadata.modified().map(chrono::DateTime::from).unwrap(),
-            size: metadata.len() as usize,
+            size: metadata.len(),
             e_tag: None,
             version: None,
         };
@@ -728,14 +728,15 @@ impl ObjectStore for MirroringObjectStore {
     async fn get_range(
         &self,
         location: &Path,
-        range: Range<usize>,
+        range: Range<u64>,
     ) -> object_store::Result<Bytes> {
         self.files.iter().find(|x| *x == location).unwrap();
         let path = std::path::PathBuf::from(&self.mirrored_file);
         let mut file = File::open(path).unwrap();
-        file.seek(SeekFrom::Start(range.start as u64)).unwrap();
+        file.seek(SeekFrom::Start(range.start)).unwrap();
 
         let to_read = range.end - range.start;
+        let to_read: usize = to_read.try_into().unwrap();
         let mut data = Vec::with_capacity(to_read);
         let read = file.take(to_read as u64).read_to_end(&mut data).unwrap();
         assert_eq!(read, to_read);
@@ -750,9 +751,10 @@ impl ObjectStore for MirroringObjectStore {
     fn list(
         &self,
         prefix: Option<&Path>,
-    ) -> BoxStream<'_, object_store::Result<ObjectMeta>> {
+    ) -> BoxStream<'static, object_store::Result<ObjectMeta>> {
         let prefix = prefix.cloned().unwrap_or_default();
-        Box::pin(stream::iter(self.files.iter().filter_map(
+        let size = self.file_size;
+        Box::pin(stream::iter(self.files.clone().into_iter().filter_map(
             move |location| {
                 // Don't return for exact prefix match
                 let filter = location
@@ -762,9 +764,9 @@ impl ObjectStore for MirroringObjectStore {
 
                 filter.then(|| {
                     Ok(ObjectMeta {
-                        location: location.clone(),
+                        location,
                         last_modified: Utc.timestamp_nanos(0),
-                        size: self.file_size as usize,
+                        size,
                         e_tag: None,
                         version: None,
                     })
@@ -802,7 +804,7 @@ impl ObjectStore for MirroringObjectStore {
                 let object = ObjectMeta {
                     location: k.clone(),
                     last_modified: Utc.timestamp_nanos(0),
-                    size: self.file_size as usize,
+                    size: self.file_size,
                     e_tag: None,
                     version: None,
                 };
diff --git a/datafusion/datasource-parquet/src/row_group_filter.rs b/datafusion/datasource-parquet/src/row_group_filter.rs
index 9d5f9fa16b6eb..96b21c8703e73 100644
--- a/datafusion/datasource-parquet/src/row_group_filter.rs
+++ b/datafusion/datasource-parquet/src/row_group_filter.rs
@@ -1513,7 +1513,7 @@ mod tests {
         let object_meta = ObjectMeta {
             location: object_store::path::Path::parse(file_name).expect("creating path"),
             last_modified: chrono::DateTime::from(std::time::SystemTime::now()),
-            size: data.len(),
+            size: data.len() as u64,
             e_tag: None,
             version: None,
         };
@@ -1526,8 +1526,11 @@ mod tests {
         let metrics = ExecutionPlanMetricsSet::new();
         let file_metrics =
             ParquetFileMetrics::new(0, object_meta.location.as_ref(), &metrics);
+        let inner = ParquetObjectReader::new(Arc::new(in_memory), object_meta.location)
+            .with_file_size(object_meta.size as usize);
+
         let reader = ParquetFileReader {
-            inner: ParquetObjectReader::new(Arc::new(in_memory), object_meta),
+            inner,
             file_metrics: file_metrics.clone(),
         };
         let mut builder = ParquetRecordBatchStreamBuilder::new(reader).await.unwrap();
diff --git a/datafusion/proto/src/physical_plan/to_proto.rs b/datafusion/proto/src/physical_plan/to_proto.rs
index f6546ff3f2a64..bcd67bac59aeb 100644
--- a/datafusion/proto/src/physical_plan/to_proto.rs
+++ b/datafusion/proto/src/physical_plan/to_proto.rs
@@ -441,7 +441,7 @@ impl TryFrom<&PartitionedFile> for protobuf::PartitionedFile {
         })? as u64;
         Ok(protobuf::PartitionedFile {
             path: pf.object_meta.location.as_ref().to_owned(),
-            size: pf.object_meta.size as u64,
+            size: pf.object_meta.size,
             last_modified_ns,
             partition_values: pf
                 .partition_values
diff --git a/datafusion/substrait/src/physical_plan/producer.rs b/datafusion/substrait/src/physical_plan/producer.rs
index 9ba0e0c964e9e..cb725a7277fd3 100644
--- a/datafusion/substrait/src/physical_plan/producer.rs
+++ b/datafusion/substrait/src/physical_plan/producer.rs
@@ -61,7 +61,7 @@ pub fn to_substrait_rel(
                     substrait_files.push(FileOrFiles {
                         partition_index: partition_index.try_into().unwrap(),
                         start: 0,
-                        length: file.object_meta.size as u64,
+                        length: file.object_meta.size,
                         path_type: Some(PathType::UriPath(
                             file.object_meta.location.as_ref().to_string(),
                         )),

From a8b3c4dc03edc7a4b1811a3fe79e1ac9385fe015 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Mon, 7 Apr 2025 15:35:30 -0400
Subject: [PATCH 14/20] update expected message

---
 datafusion/core/src/datasource/file_format/mod.rs | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/datafusion/core/src/datasource/file_format/mod.rs b/datafusion/core/src/datasource/file_format/mod.rs
index e921f0158e540..ad8c0bdb5680e 100644
--- a/datafusion/core/src/datasource/file_format/mod.rs
+++ b/datafusion/core/src/datasource/file_format/mod.rs
@@ -127,7 +127,7 @@ mod tests {
             .write_parquet(out_dir_url, DataFrameWriteOptions::new(), None)
             .await
             .expect_err("should fail because input file does not match inferred schema");
-        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value d for column 0 at line 4");
+        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value 'd' as type 'Int64' for column 0 at line 4. Row data: '[d,4]'");
         Ok(())
     }
 }

From 84725ff50982f1165fc24ac7faa3ccce3c3237b2 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 8 Apr 2025 09:24:08 -0400
Subject: [PATCH 15/20] update error messages

---
 datafusion/core/src/datasource/physical_plan/csv.rs  | 2 +-
 datafusion/core/src/datasource/physical_plan/json.rs | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/datafusion/core/src/datasource/physical_plan/csv.rs b/datafusion/core/src/datasource/physical_plan/csv.rs
index 5914924797dce..3ef4030134520 100644
--- a/datafusion/core/src/datasource/physical_plan/csv.rs
+++ b/datafusion/core/src/datasource/physical_plan/csv.rs
@@ -658,7 +658,7 @@ mod tests {
             )
             .await
             .expect_err("should fail because input file does not match inferred schema");
-        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value d for column 0 at line 4");
+        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value 'd' as type 'Int64' for column 0 at line 4. Row data: '[d,4]'");
         Ok(())
     }
 
diff --git a/datafusion/core/src/datasource/physical_plan/json.rs b/datafusion/core/src/datasource/physical_plan/json.rs
index 910c4316d9734..736248fbd95df 100644
--- a/datafusion/core/src/datasource/physical_plan/json.rs
+++ b/datafusion/core/src/datasource/physical_plan/json.rs
@@ -495,7 +495,7 @@ mod tests {
             .write_json(out_dir_url, DataFrameWriteOptions::new(), None)
             .await
             .expect_err("should fail because input file does not match inferred schema");
-        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value d for column 0 at line 4");
+        assert_eq!(e.strip_backtrace(), "Arrow error: Parser error: Error while parsing value 'd' as type 'Int64' for column 0 at line 4. Row data: '[d,4]'");
         Ok(())
     }
 

From 9bfc8a324a571083a3d722b15f8d76d67e8eee10 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 8 Apr 2025 11:58:41 -0400
Subject: [PATCH 16/20] Update to apache

---
 Cargo.lock | 32 ++++++++++++++++----------------
 Cargo.toml | 27 +++++++++++++--------------
 2 files changed, 29 insertions(+), 30 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 3b6d474b57e3d..51cb3411ce97a 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -247,7 +247,7 @@ checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
 [[package]]
 name = "arrow"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -270,7 +270,7 @@ dependencies = [
 [[package]]
 name = "arrow-arith"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -283,7 +283,7 @@ dependencies = [
 [[package]]
 name = "arrow-array"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "ahash 0.8.11",
  "arrow-buffer",
@@ -299,7 +299,7 @@ dependencies = [
 [[package]]
 name = "arrow-buffer"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "bytes",
  "half",
@@ -309,7 +309,7 @@ dependencies = [
 [[package]]
 name = "arrow-cast"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -329,7 +329,7 @@ dependencies = [
 [[package]]
 name = "arrow-csv"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-cast",
@@ -344,7 +344,7 @@ dependencies = [
 [[package]]
 name = "arrow-data"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-buffer",
  "arrow-schema",
@@ -355,7 +355,7 @@ dependencies = [
 [[package]]
 name = "arrow-flight"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -381,7 +381,7 @@ dependencies = [
 [[package]]
 name = "arrow-ipc"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -394,7 +394,7 @@ dependencies = [
 [[package]]
 name = "arrow-json"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -415,7 +415,7 @@ dependencies = [
 [[package]]
 name = "arrow-ord"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -427,7 +427,7 @@ dependencies = [
 [[package]]
 name = "arrow-row"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -439,7 +439,7 @@ dependencies = [
 [[package]]
 name = "arrow-schema"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "bitflags 2.8.0",
  "serde",
@@ -448,7 +448,7 @@ dependencies = [
 [[package]]
 name = "arrow-select"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -461,7 +461,7 @@ dependencies = [
 [[package]]
 name = "arrow-string"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4367,7 +4367,7 @@ dependencies = [
 [[package]]
 name = "parquet"
 version = "55.0.0"
-source = "git+https://github.com/alamb/arrow-rs.git?branch=alamb/prepare_for_55.0.0#8c911537c385c60bd6870206f5267735d6223007"
+source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
diff --git a/Cargo.toml b/Cargo.toml
index 86f81f379ea7d..05bf187e5de72 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -214,20 +214,19 @@ used_underscore_binding = "warn"
 unexpected_cfgs = { level = "warn", check-cfg = ["cfg(tarpaulin)"] }
 unused_qualifications = "deny"
 
-
-## Temporary arrow-rs patch until 55
+## Temporary arrow-rs patch until 55 is released
 ## https://github.com/apache/arrow-rs/pull/7391
 
 [patch.crates-io]
-arrow = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-array = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-buffer = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-cast = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-data = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-ipc = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-schema = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-select = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-string = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-ord = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-arrow-flight = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
-parquet = { git = "https://github.com/alamb/arrow-rs.git", branch = "alamb/prepare_for_55.0.0" }
+arrow = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-array = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-buffer = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-cast = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-data = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-ipc = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-schema = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-select = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-string = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-ord = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+arrow-flight = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
+parquet = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }

From 3d646d7a4784663e6de1ddd022f2b1cc7fac1442 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 8 Apr 2025 10:41:45 -0400
Subject: [PATCH 17/20] Update API for nicer parquet u64s

---
 .../examples/advanced_parquet_index.rs              |  6 +++---
 datafusion/core/tests/parquet/custom_reader.rs      |  6 +++---
 datafusion/datasource-parquet/src/file_format.rs    |  8 ++------
 datafusion/datasource-parquet/src/reader.rs         | 13 +++++++------
 .../datasource-parquet/src/row_group_filter.rs      |  2 +-
 5 files changed, 16 insertions(+), 19 deletions(-)

diff --git a/datafusion-examples/examples/advanced_parquet_index.rs b/datafusion-examples/examples/advanced_parquet_index.rs
index 113c43dc13236..03ef3d66f9d71 100644
--- a/datafusion-examples/examples/advanced_parquet_index.rs
+++ b/datafusion-examples/examples/advanced_parquet_index.rs
@@ -573,7 +573,7 @@ impl ParquetFileReaderFactory for CachedParquetFileReaderFactory {
         let object_store = Arc::clone(&self.object_store);
         let mut inner =
             ParquetObjectReader::new(object_store, file_meta.object_meta.location)
-                .with_file_size(file_meta.object_meta.size as usize);
+                .with_file_size(file_meta.object_meta.size);
 
         if let Some(hint) = metadata_size_hint {
             inner = inner.with_footer_size_hint(hint)
@@ -601,7 +601,7 @@ struct ParquetReaderWithCache {
 impl AsyncFileReader for ParquetReaderWithCache {
     fn get_bytes(
         &mut self,
-        range: Range<usize>,
+        range: Range<u64>,
     ) -> BoxFuture<'_, datafusion::parquet::errors::Result<Bytes>> {
         println!("get_bytes: {} Reading range {:?}", self.filename, range);
         self.inner.get_bytes(range)
@@ -609,7 +609,7 @@ impl AsyncFileReader for ParquetReaderWithCache {
 
     fn get_byte_ranges(
         &mut self,
-        ranges: Vec<Range<usize>>,
+        ranges: Vec<Range<u64>>,
     ) -> BoxFuture<'_, datafusion::parquet::errors::Result<Vec<Bytes>>> {
         println!(
             "get_byte_ranges: {} Reading ranges {:?}",
diff --git a/datafusion/core/tests/parquet/custom_reader.rs b/datafusion/core/tests/parquet/custom_reader.rs
index c06da2e75a461..761a78a29fd3a 100644
--- a/datafusion/core/tests/parquet/custom_reader.rs
+++ b/datafusion/core/tests/parquet/custom_reader.rs
@@ -219,11 +219,11 @@ struct ParquetFileReader {
 impl AsyncFileReader for ParquetFileReader {
     fn get_bytes(
         &mut self,
-        range: Range<usize>,
+        range: Range<u64>,
     ) -> BoxFuture<'_, parquet::errors::Result<Bytes>> {
-        self.metrics.bytes_scanned.add(range.end - range.start);
+        let bytes_scanned = range.end - range.start;
+        self.metrics.bytes_scanned.add(bytes_scanned as usize);
 
-        let range = range.start as u64..range.end as u64;
         self.store
             .get_range(&self.meta.location, range)
             .map_err(|e| {
diff --git a/datafusion/datasource-parquet/src/file_format.rs b/datafusion/datasource-parquet/src/file_format.rs
index f534d6769b9f1..7617d4d70ceed 100644
--- a/datafusion/datasource-parquet/src/file_format.rs
+++ b/datafusion/datasource-parquet/src/file_format.rs
@@ -735,11 +735,7 @@ impl<'a> ObjectStoreFetch<'a> {
 }
 
 impl MetadataFetch for ObjectStoreFetch<'_> {
-    fn fetch(
-        &mut self,
-        range: Range<usize>,
-    ) -> BoxFuture<'_, Result<Bytes, ParquetError>> {
-        let range = range.start as u64..range.end as u64;
+    fn fetch(&mut self, range: Range<u64>) -> BoxFuture<'_, Result<Bytes, ParquetError>> {
         async {
             self.store
                 .get_range(&self.meta.location, range)
@@ -766,7 +762,7 @@ pub async fn fetch_parquet_metadata(
 
     ParquetMetaDataReader::new()
         .with_prefetch_hint(size_hint)
-        .load_and_finish(fetch, file_size as usize)
+        .load_and_finish(fetch, file_size)
         .await
         .map_err(DataFusionError::from)
 }
diff --git a/datafusion/datasource-parquet/src/reader.rs b/datafusion/datasource-parquet/src/reader.rs
index 141b30d431379..27ec843c1991d 100644
--- a/datafusion/datasource-parquet/src/reader.rs
+++ b/datafusion/datasource-parquet/src/reader.rs
@@ -96,21 +96,22 @@ pub(crate) struct ParquetFileReader {
 impl AsyncFileReader for ParquetFileReader {
     fn get_bytes(
         &mut self,
-        range: Range<usize>,
+        range: Range<u64>,
     ) -> BoxFuture<'_, parquet::errors::Result<Bytes>> {
-        self.file_metrics.bytes_scanned.add(range.end - range.start);
+        let bytes_scanned = range.end - range.start;
+        self.file_metrics.bytes_scanned.add(bytes_scanned as usize);
         self.inner.get_bytes(range)
     }
 
     fn get_byte_ranges(
         &mut self,
-        ranges: Vec<Range<usize>>,
+        ranges: Vec<Range<u64>>,
     ) -> BoxFuture<'_, parquet::errors::Result<Vec<Bytes>>>
     where
         Self: Send,
     {
-        let total = ranges.iter().map(|r| r.end - r.start).sum();
-        self.file_metrics.bytes_scanned.add(total);
+        let total: u64 = ranges.iter().map(|r| r.end - r.start).sum();
+        self.file_metrics.bytes_scanned.add(total as usize);
         self.inner.get_byte_ranges(ranges)
     }
 
@@ -137,7 +138,7 @@ impl ParquetFileReaderFactory for DefaultParquetFileReaderFactory {
         );
         let store = Arc::clone(&self.store);
         let mut inner = ParquetObjectReader::new(store, file_meta.object_meta.location)
-            .with_file_size(file_meta.object_meta.size as usize);
+            .with_file_size(file_meta.object_meta.size);
 
         if let Some(hint) = metadata_size_hint {
             inner = inner.with_footer_size_hint(hint)
diff --git a/datafusion/datasource-parquet/src/row_group_filter.rs b/datafusion/datasource-parquet/src/row_group_filter.rs
index 96b21c8703e73..13418cdeee223 100644
--- a/datafusion/datasource-parquet/src/row_group_filter.rs
+++ b/datafusion/datasource-parquet/src/row_group_filter.rs
@@ -1527,7 +1527,7 @@ mod tests {
         let file_metrics =
             ParquetFileMetrics::new(0, object_meta.location.as_ref(), &metrics);
         let inner = ParquetObjectReader::new(Arc::new(in_memory), object_meta.location)
-            .with_file_size(object_meta.size as usize);
+            .with_file_size(object_meta.size);
 
         let reader = ParquetFileReader {
             inner,

From 86aab05cf3b13a1eaf741c6a245447c15a541e11 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 8 Apr 2025 12:56:06 -0400
Subject: [PATCH 18/20] Fix wasm build

---
 datafusion/core/src/datasource/file_format/arrow.rs        | 1 +
 datafusion/core/src/datasource/physical_plan/arrow_file.rs | 1 +
 datafusion/datasource-csv/src/source.rs                    | 1 +
 datafusion/datasource-json/src/file_format.rs              | 1 +
 datafusion/datasource-json/src/source.rs                   | 1 +
 5 files changed, 5 insertions(+)

diff --git a/datafusion/core/src/datasource/file_format/arrow.rs b/datafusion/core/src/datasource/file_format/arrow.rs
index ea027ae91a061..7fc27453d1ad5 100644
--- a/datafusion/core/src/datasource/file_format/arrow.rs
+++ b/datafusion/core/src/datasource/file_format/arrow.rs
@@ -144,6 +144,7 @@ impl FileFormat for ArrowFormat {
         for object in objects {
             let r = store.as_ref().get(&object.location).await?;
             let schema = match r.payload {
+                #[cfg(not(target_arch = "wasm32"))]
                 GetResultPayload::File(mut file, _) => {
                     let reader = FileReader::try_new(&mut file, None)?;
                     reader.schema()
diff --git a/datafusion/core/src/datasource/physical_plan/arrow_file.rs b/datafusion/core/src/datasource/physical_plan/arrow_file.rs
index 21d962e6d11fb..f0a1f94d87e1f 100644
--- a/datafusion/core/src/datasource/physical_plan/arrow_file.rs
+++ b/datafusion/core/src/datasource/physical_plan/arrow_file.rs
@@ -273,6 +273,7 @@ impl FileOpener for ArrowOpener {
                 None => {
                     let r = object_store.get(file_meta.location()).await?;
                     match r.payload {
+                        #[cfg(not(target_arch = "wasm32"))]
                         GetResultPayload::File(file, _) => {
                             let arrow_reader = arrow::ipc::reader::FileReader::try_new(
                                 file, projection,
diff --git a/datafusion/datasource-csv/src/source.rs b/datafusion/datasource-csv/src/source.rs
index 6db4d18703204..f5d45cd3fc881 100644
--- a/datafusion/datasource-csv/src/source.rs
+++ b/datafusion/datasource-csv/src/source.rs
@@ -704,6 +704,7 @@ impl FileOpener for CsvOpener {
             let result = store.get_opts(file_meta.location(), options).await?;
 
             match result.payload {
+                #[cfg(not(target_arch = "wasm32"))]
                 GetResultPayload::File(mut file, _) => {
                     let is_whole_file_scanned = file_meta.range.is_none();
                     let decoder = if is_whole_file_scanned {
diff --git a/datafusion/datasource-json/src/file_format.rs b/datafusion/datasource-json/src/file_format.rs
index a6c52312e4127..8d0515804fc7b 100644
--- a/datafusion/datasource-json/src/file_format.rs
+++ b/datafusion/datasource-json/src/file_format.rs
@@ -209,6 +209,7 @@ impl FileFormat for JsonFormat {
 
             let r = store.as_ref().get(&object.location).await?;
             let schema = match r.payload {
+                #[cfg(not(target_arch = "wasm32"))]
                 GetResultPayload::File(file, _) => {
                     let decoder = file_compression_type.convert_read(file)?;
                     let mut reader = BufReader::new(decoder);
diff --git a/datafusion/datasource-json/src/source.rs b/datafusion/datasource-json/src/source.rs
index f1adccf9ded7d..ee96d050966d6 100644
--- a/datafusion/datasource-json/src/source.rs
+++ b/datafusion/datasource-json/src/source.rs
@@ -355,6 +355,7 @@ impl FileOpener for JsonOpener {
             let result = store.get_opts(file_meta.location(), options).await?;
 
             match result.payload {
+                #[cfg(not(target_arch = "wasm32"))]
                 GetResultPayload::File(mut file, _) => {
                     let bytes = match file_meta.range {
                         None => file_compression_type.convert_read(file)?,

From 2a30ca3bbdffecc1cb828a7cda05adff84da2c85 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 11 Apr 2025 13:27:17 -0400
Subject: [PATCH 19/20] Remove pin

---
 Cargo.lock | 52 ++++++++++++++++++++++++++++++++++------------------
 Cargo.toml | 17 -----------------
 2 files changed, 34 insertions(+), 35 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 1444ea39e24aa..753bb53f2c352 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -247,7 +247,8 @@ checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
 [[package]]
 name = "arrow"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3095aaf545942ff5abd46654534f15b03a90fba78299d661e045e5d587222f0d"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -270,7 +271,8 @@ dependencies = [
 [[package]]
 name = "arrow-arith"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "00752064ff47cee746e816ddb8450520c3a52cbad1e256f6fa861a35f86c45e7"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -283,7 +285,8 @@ dependencies = [
 [[package]]
 name = "arrow-array"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cebfe926794fbc1f49ddd0cdaf898956ca9f6e79541efce62dabccfd81380472"
 dependencies = [
  "ahash 0.8.11",
  "arrow-buffer",
@@ -299,7 +302,8 @@ dependencies = [
 [[package]]
 name = "arrow-buffer"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0303c7ec4cf1a2c60310fc4d6bbc3350cd051a17bf9e9c0a8e47b4db79277824"
 dependencies = [
  "bytes",
  "half",
@@ -309,7 +313,8 @@ dependencies = [
 [[package]]
 name = "arrow-cast"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "335f769c5a218ea823d3760a743feba1ef7857cba114c01399a891c2fff34285"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -329,7 +334,8 @@ dependencies = [
 [[package]]
 name = "arrow-csv"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "510db7dfbb4d5761826516cc611d97b3a68835d0ece95b034a052601109c0b1b"
 dependencies = [
  "arrow-array",
  "arrow-cast",
@@ -344,7 +350,8 @@ dependencies = [
 [[package]]
 name = "arrow-data"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e8affacf3351a24039ea24adab06f316ded523b6f8c3dbe28fbac5f18743451b"
 dependencies = [
  "arrow-buffer",
  "arrow-schema",
@@ -355,7 +362,8 @@ dependencies = [
 [[package]]
 name = "arrow-flight"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e2e0fad280f41a918d53ba48288a246ff04202d463b3b380fbc0edecdcb52cfd"
 dependencies = [
  "arrow-arith",
  "arrow-array",
@@ -381,7 +389,8 @@ dependencies = [
 [[package]]
 name = "arrow-ipc"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "69880a9e6934d9cba2b8630dd08a3463a91db8693b16b499d54026b6137af284"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -394,7 +403,8 @@ dependencies = [
 [[package]]
 name = "arrow-json"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d8dafd17a05449e31e0114d740530e0ada7379d7cb9c338fd65b09a8130960b0"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -415,7 +425,8 @@ dependencies = [
 [[package]]
 name = "arrow-ord"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "895644523af4e17502d42c3cb6b27cb820f0cb77954c22d75c23a85247c849e1"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -427,7 +438,8 @@ dependencies = [
 [[package]]
 name = "arrow-row"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9be8a2a4e5e7d9c822b2b8095ecd77010576d824f654d347817640acfc97d229"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -439,7 +451,8 @@ dependencies = [
 [[package]]
 name = "arrow-schema"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7450c76ab7c5a6805be3440dc2e2096010da58f7cab301fdc996a4ee3ee74e49"
 dependencies = [
  "bitflags 2.8.0",
  "serde",
@@ -448,7 +461,8 @@ dependencies = [
 [[package]]
 name = "arrow-select"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "aa5f5a93c75f46ef48e4001535e7b6c922eeb0aa20b73cf58d09e13d057490d8"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -461,7 +475,8 @@ dependencies = [
 [[package]]
 name = "arrow-string"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6e7005d858d84b56428ba2a98a107fe88c0132c61793cf6b8232a1f9bfc0452b"
 dependencies = [
  "arrow-array",
  "arrow-buffer",
@@ -4368,7 +4383,8 @@ dependencies = [
 [[package]]
 name = "parquet"
 version = "55.0.0"
-source = "git+https://github.com/apache/arrow-rs.git?rev=9322547590ab32efeff8c0486e4a3a2cb5887a26#9322547590ab32efeff8c0486e4a3a2cb5887a26"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cd31a8290ac5b19f09ad77ee7a1e6a541f1be7674ad410547d5f1eef6eef4a9c"
 dependencies = [
  "ahash 0.8.11",
  "arrow-array",
@@ -4769,7 +4785,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "be769465445e8c1474e9c5dac2018218498557af32d9ed057325ec9a41ae81bf"
 dependencies = [
  "heck 0.5.0",
- "itertools 0.14.0",
+ "itertools 0.13.0",
  "log",
  "multimap",
  "once_cell",
@@ -4789,7 +4805,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8a56d757972c98b346a9b766e3f02746cde6dd1cd1d1d563472929fdd74bec4d"
 dependencies = [
  "anyhow",
- "itertools 0.14.0",
+ "itertools 0.13.0",
  "proc-macro2",
  "quote",
  "syn 2.0.100",
diff --git a/Cargo.toml b/Cargo.toml
index 05bf187e5de72..de53b7df50d98 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -213,20 +213,3 @@ used_underscore_binding = "warn"
 [workspace.lints.rust]
 unexpected_cfgs = { level = "warn", check-cfg = ["cfg(tarpaulin)"] }
 unused_qualifications = "deny"
-
-## Temporary arrow-rs patch until 55 is released
-## https://github.com/apache/arrow-rs/pull/7391
-
-[patch.crates-io]
-arrow = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-array = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-buffer = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-cast = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-data = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-ipc = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-schema = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-select = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-string = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-ord = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-arrow-flight = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }
-parquet = { git = "https://github.com/apache/arrow-rs.git", rev = "9322547590ab32efeff8c0486e4a3a2cb5887a26" }

From 1f0711e6f33903226d835808e6b178032f38e178 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Fri, 11 Apr 2025 13:33:08 -0400
Subject: [PATCH 20/20] Fix signature

---
 datafusion/core/tests/tracing/traceable_object_store.rs | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/datafusion/core/tests/tracing/traceable_object_store.rs b/datafusion/core/tests/tracing/traceable_object_store.rs
index e979200c8d9bf..dfcafc3a63da1 100644
--- a/datafusion/core/tests/tracing/traceable_object_store.rs
+++ b/datafusion/core/tests/tracing/traceable_object_store.rs
@@ -96,7 +96,7 @@ impl ObjectStore for TraceableObjectStore {
     fn list(
         &self,
         prefix: Option<&Path>,
-    ) -> BoxStream<'_, object_store::Result<ObjectMeta>> {
+    ) -> BoxStream<'static, object_store::Result<ObjectMeta>> {
         futures::executor::block_on(assert_traceability());
         self.inner.list(prefix)
     }