Accepting request 1064743 from devel:languages:python:numeric
OBS-URL: https://build.opensuse.org/request/show/1064743 OBS-URL: https://build.opensuse.org/package/show/openSUSE:Factory/python-fastparquet?expand=0&rev=26
This commit is contained in:
commit
208da3606a
@ -1,3 +0,0 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:8e098493e2e87f8e004c3fc9b4e94ef57df9d7656fd8aa1a5626b6e6f3ff2b8e
|
||||
size 28903475
|
3
fastparquet-2023.2.0.tar.gz
Normal file
3
fastparquet-2023.2.0.tar.gz
Normal file
@ -0,0 +1,3 @@
|
||||
version https://git-lfs.github.com/spec/v1
|
||||
oid sha256:6224c7fda7d900e4334b339c4376b58ad2c27892fe7faa67ae28c380100fbe24
|
||||
size 28902717
|
@ -1,183 +0,0 @@
|
||||
From f035a2d7b37e2cbdef1a99bc6130be7e4afcf35f Mon Sep 17 00:00:00 2001
|
||||
From: Martin Durant <martin.durant@alumni.utoronto.ca>
|
||||
Date: Thu, 22 Dec 2022 12:57:32 -0500
|
||||
Subject: [PATCH 1/5] more print
|
||||
|
||||
---
|
||||
fastparquet/cencoding.pyx | 5 +++++
|
||||
1 file changed, 5 insertions(+)
|
||||
|
||||
diff --git a/fastparquet/cencoding.pyx b/fastparquet/cencoding.pyx
|
||||
index 4ab48be6..c151aff9 100644
|
||||
--- a/fastparquet/cencoding.pyx
|
||||
+++ b/fastparquet/cencoding.pyx
|
||||
@@ -225,6 +225,7 @@ cdef void delta_read_bitpacked(NumpyIO file_obj, uint8_t bitwidth,
|
||||
stop += 8
|
||||
else:
|
||||
o.write_int((data >> stop) & mask)
|
||||
+ print("bitpack value", (data >> stop) & mask, data, stop, mask)
|
||||
stop -= bitwidth
|
||||
count -= 1
|
||||
|
||||
@@ -239,11 +240,13 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
const uint8_t[:] bitwidths
|
||||
uint8_t bitwidth
|
||||
values_per_miniblock = block_size // miniblock_per_block
|
||||
+ print("\nstart", count, value, values_per_miniblock)
|
||||
while True:
|
||||
min_delta = zigzag_long(read_unsigned_var_int(file_obj))
|
||||
bitwidths = file_obj.read(miniblock_per_block)
|
||||
for i in range(miniblock_per_block):
|
||||
bitwidth = bitwidths[i]
|
||||
+ print("\n miniblock", i, "width", bitwidth)
|
||||
if bitwidth:
|
||||
temp = o.loc
|
||||
if count > 1:
|
||||
@@ -253,6 +256,7 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
for j in range(values_per_miniblock):
|
||||
temp = o.read_int()
|
||||
o.loc -= 4
|
||||
+ print("miniblock value", value)
|
||||
o.write_int(value)
|
||||
value += min_delta + temp
|
||||
count -= 1
|
||||
@@ -260,6 +264,7 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
return
|
||||
else:
|
||||
for j in range(values_per_miniblock):
|
||||
+ print("miniblock value", value)
|
||||
o.write_int(value)
|
||||
value += min_delta
|
||||
count -= 1
|
||||
|
||||
From c453e140355055be1077f2c99b24785444b4ab20 Mon Sep 17 00:00:00 2001
|
||||
From: Martin Durant <martin.durant@alumni.utoronto.ca>
|
||||
Date: Thu, 22 Dec 2022 13:02:27 -0500
|
||||
Subject: [PATCH 2/5] more
|
||||
|
||||
---
|
||||
fastparquet/cencoding.pyx | 2 +-
|
||||
1 file changed, 1 insertion(+), 1 deletion(-)
|
||||
|
||||
diff --git a/fastparquet/cencoding.pyx b/fastparquet/cencoding.pyx
|
||||
index c151aff9..6c2051c6 100644
|
||||
--- a/fastparquet/cencoding.pyx
|
||||
+++ b/fastparquet/cencoding.pyx
|
||||
@@ -256,7 +256,7 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
for j in range(values_per_miniblock):
|
||||
temp = o.read_int()
|
||||
o.loc -= 4
|
||||
- print("miniblock value", value)
|
||||
+ print("miniblock value (bw)", value)
|
||||
o.write_int(value)
|
||||
value += min_delta + temp
|
||||
count -= 1
|
||||
|
||||
From 76b8c4fefe703d980b72ad37841bb2e76b9da590 Mon Sep 17 00:00:00 2001
|
||||
From: Martin Durant <martin.durant@alumni.utoronto.ca>
|
||||
Date: Thu, 22 Dec 2022 13:32:27 -0500
|
||||
Subject: [PATCH 3/5] more
|
||||
|
||||
---
|
||||
fastparquet/cencoding.pyx | 1 +
|
||||
1 file changed, 1 insertion(+)
|
||||
|
||||
diff --git a/fastparquet/cencoding.pyx b/fastparquet/cencoding.pyx
|
||||
index 6c2051c6..9335b52a 100644
|
||||
--- a/fastparquet/cencoding.pyx
|
||||
+++ b/fastparquet/cencoding.pyx
|
||||
@@ -223,6 +223,7 @@ cdef void delta_read_bitpacked(NumpyIO file_obj, uint8_t bitwidth,
|
||||
if stop < 0:
|
||||
data = ((data & 0X00FFFFFFFFFFFFFF) << 8) | file_obj.read_byte()
|
||||
stop += 8
|
||||
+ print("bin stop", bin(stop), bin(data))
|
||||
else:
|
||||
o.write_int((data >> stop) & mask)
|
||||
print("bitpack value", (data >> stop) & mask, data, stop, mask)
|
||||
|
||||
From bbb32d8ea7f2cb97a817a22c0357f1a023f79b42 Mon Sep 17 00:00:00 2001
|
||||
From: Martin Durant <martin.durant@alumni.utoronto.ca>
|
||||
Date: Thu, 22 Dec 2022 13:39:14 -0500
|
||||
Subject: [PATCH 4/5] trial
|
||||
|
||||
---
|
||||
fastparquet/cencoding.pyx | 4 ++--
|
||||
1 file changed, 2 insertions(+), 2 deletions(-)
|
||||
|
||||
diff --git a/fastparquet/cencoding.pyx b/fastparquet/cencoding.pyx
|
||||
index 9335b52a..1be0a001 100644
|
||||
--- a/fastparquet/cencoding.pyx
|
||||
+++ b/fastparquet/cencoding.pyx
|
||||
@@ -18,7 +18,7 @@ cdef extern from "string.h":
|
||||
from cpython cimport (
|
||||
PyBytes_FromStringAndSize, PyBytes_GET_SIZE, PyUnicode_DecodeUTF8,
|
||||
)
|
||||
-from libc.stdint cimport uint8_t, uint32_t, int32_t, uint64_t, int64_t
|
||||
+from libc.stdint cimport int8_t, uint8_t, uint32_t, int32_t, uint64_t, int64_t
|
||||
|
||||
|
||||
cpdef void read_rle(NumpyIO file_obj, int32_t header, int32_t bit_width, NumpyIO o, int32_t itemsize=4):
|
||||
@@ -217,7 +217,7 @@ cdef void delta_read_bitpacked(NumpyIO file_obj, uint8_t bitwidth,
|
||||
NumpyIO o, uint64_t count, uint8_t itemsize=4):
|
||||
cdef:
|
||||
uint64_t data = 0
|
||||
- char stop = -bitwidth
|
||||
+ int8_t stop = -bitwidth
|
||||
uint64_t mask = 0XFFFFFFFFFFFFFFFF >> (64 - bitwidth)
|
||||
while count > 0:
|
||||
if stop < 0:
|
||||
|
||||
From 43e34e28b2e108f178b05bba8c109e2b131f5fc2 Mon Sep 17 00:00:00 2001
|
||||
From: Martin Durant <martin.durant@alumni.utoronto.ca>
|
||||
Date: Thu, 22 Dec 2022 13:43:11 -0500
|
||||
Subject: [PATCH 5/5] fix
|
||||
|
||||
---
|
||||
fastparquet/cencoding.pyx | 6 ------
|
||||
1 file changed, 6 deletions(-)
|
||||
|
||||
diff --git a/fastparquet/cencoding.pyx b/fastparquet/cencoding.pyx
|
||||
index 1be0a001..90ba15db 100644
|
||||
--- a/fastparquet/cencoding.pyx
|
||||
+++ b/fastparquet/cencoding.pyx
|
||||
@@ -223,10 +223,8 @@ cdef void delta_read_bitpacked(NumpyIO file_obj, uint8_t bitwidth,
|
||||
if stop < 0:
|
||||
data = ((data & 0X00FFFFFFFFFFFFFF) << 8) | file_obj.read_byte()
|
||||
stop += 8
|
||||
- print("bin stop", bin(stop), bin(data))
|
||||
else:
|
||||
o.write_int((data >> stop) & mask)
|
||||
- print("bitpack value", (data >> stop) & mask, data, stop, mask)
|
||||
stop -= bitwidth
|
||||
count -= 1
|
||||
|
||||
@@ -241,13 +239,11 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
const uint8_t[:] bitwidths
|
||||
uint8_t bitwidth
|
||||
values_per_miniblock = block_size // miniblock_per_block
|
||||
- print("\nstart", count, value, values_per_miniblock)
|
||||
while True:
|
||||
min_delta = zigzag_long(read_unsigned_var_int(file_obj))
|
||||
bitwidths = file_obj.read(miniblock_per_block)
|
||||
for i in range(miniblock_per_block):
|
||||
bitwidth = bitwidths[i]
|
||||
- print("\n miniblock", i, "width", bitwidth)
|
||||
if bitwidth:
|
||||
temp = o.loc
|
||||
if count > 1:
|
||||
@@ -257,7 +253,6 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
for j in range(values_per_miniblock):
|
||||
temp = o.read_int()
|
||||
o.loc -= 4
|
||||
- print("miniblock value (bw)", value)
|
||||
o.write_int(value)
|
||||
value += min_delta + temp
|
||||
count -= 1
|
||||
@@ -265,7 +260,6 @@ cpdef void delta_binary_unpack(NumpyIO file_obj, NumpyIO o):
|
||||
return
|
||||
else:
|
||||
for j in range(values_per_miniblock):
|
||||
- print("miniblock value", value)
|
||||
o.write_int(value)
|
||||
value += min_delta
|
||||
count -= 1
|
@ -1,3 +1,24 @@
|
||||
-------------------------------------------------------------------
|
||||
Thu Feb 9 15:55:08 UTC 2023 - Arun Persaud <arun@gmx.de>
|
||||
|
||||
- update to version 2023.2.0:
|
||||
* revert one-level set of filters (#852)
|
||||
* full size dict for decoding V2 pages (#850)
|
||||
* infer_object_encoding fix (#847)
|
||||
* row filtering with V2 pages (#845)
|
||||
|
||||
-------------------------------------------------------------------
|
||||
Wed Feb 8 18:25:03 UTC 2023 - Arun Persaud <arun@gmx.de>
|
||||
|
||||
- specfile:
|
||||
* remove fastparquet-pr835.patch, implemented upstream
|
||||
|
||||
- update to version 2023.1.0:
|
||||
* big improvement to write speed
|
||||
* paging support for bigger row-groups
|
||||
* pandas 2.0 support
|
||||
* delta for big-endian architecture
|
||||
|
||||
-------------------------------------------------------------------
|
||||
Mon Jan 2 20:38:49 UTC 2023 - Ben Greiner <code@bnavigator.de>
|
||||
|
||||
|
@ -17,16 +17,14 @@
|
||||
|
||||
|
||||
Name: python-fastparquet
|
||||
Version: 2022.12.0
|
||||
Version: 2023.2.0
|
||||
Release: 0
|
||||
Summary: Python support for Parquet file format
|
||||
License: Apache-2.0
|
||||
URL: https://github.com/dask/fastparquet/
|
||||
# Use GitHub archive, because it containts the test modules and data, requires setting version manuall for setuptools_scm
|
||||
Source: https://github.com/dask/fastparquet/archive/%{version}.tar.gz#/fastparquet-%{version}.tar.gz
|
||||
# PATCH-FIX-UPSTREAM fastparquet-pr835.patch gh#dask/fastparquet#835
|
||||
Patch2: fastparquet-pr835.patch
|
||||
BuildRequires: %{python_module Cython}
|
||||
BuildRequires: %{python_module Cython >= 0.29.23}
|
||||
BuildRequires: %{python_module base >= 3.8}
|
||||
BuildRequires: %{python_module cramjam >= 2.3.0}
|
||||
# version requirement not declared for runtime, but necessary for tests.
|
||||
|
Loading…
Reference in New Issue
Block a user