[h5py] 251/455: Update announce file
Ghislain Vaillant
ghisvail-guest at moszumanska.debian.org
Thu Jul 2 18:19:38 UTC 2015
This is an automated email from the git hooks/post-receive script.
ghisvail-guest pushed a commit to annotated tag 1.3.0
in repository h5py.
commit ecaa1d7713c28d1e716c8c6bb76a141b679be2e8
Author: andrewcollette <andrew.collette at gmail.com>
Date: Fri May 29 00:06:13 2009 +0000
Update announce file
---
ANN.txt | 41 ++++++++++++++++++++++++++---------------
1 file changed, 26 insertions(+), 15 deletions(-)
diff --git a/ANN.txt b/ANN.txt
index 3cc98fd..63363a8 100644
--- a/ANN.txt
+++ b/ANN.txt
@@ -1,6 +1,6 @@
-=====================================
-Announcing HDF5 for Python (h5py) 1.1
-=====================================
+============================================
+Announcing HDF5 for Python (h5py) 1.2 *BETA*
+============================================
What is h5py?
-------------
@@ -21,23 +21,33 @@ In addition to providing interoperability with existing HDF5 datasets
and platforms, h5py is a convienient way to store and retrieve
arbitrary NumPy data and metadata.
+Version 1.2 is a significant update to h5py, introducing several new features.
-New features in 1.1
--------------------
- - A new compression filter based on the LZF library, which provides
- transparent compression many times faster than the standard HDF5
- GZIP filter.
+What's new in 1.2
+-----------------
+
+ - Variable-length strings are now supported! They are mapped to native
+ Python strings via the NumPy "object" type. VL strings may be read,
+ written and created from h5py, and are allowed in all HDF5 contexts,
+ even as members of compound or array types.
+
+ - Enumerated types are now fully supported; they can be used in NumPy
+ anywhere integer types are allowed, and are stored as native HDF5
+ enums. Conversion between integers and enums is supported.
- - Efficient broadcasting using HDF5 hyperslab selections; for example,
- you can write to a (2000 x 100 x 50) selection from a (100 x 50)
- source array.
+ - The NumPy "array" dtype is now allowed as a top-level type when
+ creating a dataset, not just as a member of a compound type.
- - Now supports the NumPy boolean type
+ - Many different low-level HDF5 drivers can be now be used when creating
+ a file, which allows purely in-memory ("core") files, multi-volume
+ ("family") files, and files which use low-level buffered I/O.
- - Auto-completion for IPython 0.9.X (contributed by Darren Dale)
+ - HDF5 exceptions now inherit from common Python built-ins like TypeError
+ and ValueError (in addition to current HDF5 error hierarchy), freeing
+ the user from knowledge of the HDF5 error system.
- - Installable via easy_install
+ - More Pythonic dictionary-style naming system for groups and attributes
Standard features
@@ -51,7 +61,8 @@ Standard features
* Compound ("recarray")
* Strings
* Boolean
- * Array (as members of a compound type only)
+ * Array
+ * Enumeration (integer)
* Void
- Random access to datasets using the standard NumPy slicing syntax,
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-science/packages/h5py.git
More information about the debian-science-commits
mailing list