What’s new in PyTables 1.2

Author:

Francesc Altet

Contact:

faltet@carabos.com

Author:

Ivan Vilata i Balaguer

Contact:

ivilata@carabos.com

This document details the modifications to PyTables since version 1.1. Its main purpose is help you ensure that your programs will be runnable when you switch from PyTables 1.1 to PyTables 1.2.

API additions

  • The user is now allowed to set arbitrary Python (non-persistent) attributes on any instance of Node. If the name matches that of a child node, the later will no longer be accessible via natural naming, but it will still be available via File.getNode(), Group._f_getChild() and the group children dictionaries.

    Of course, this allows the user to overwrite internal (^_[cfgv]_) PyTables variables, but this is the way most Python packages work.

  • The new Group._f_getChild() method allows to get a child node (be it visible or not) by its name. This should be more intuitive that using getattr() or using the group children dictionaries.

  • The new File.isVisibleNode(), Node._f_isVisible() and Leaf.isVisible() methods tell whether a node is visible or not, i.e. if the node will appear in listing operations such as Group._f_listNodes().

Backward-incompatible changes

  • File.objects, File.groups and File.leaves can no longer be used to iterate over all the nodes in the file. However, they still may be used to access any node by its path.

  • File.__contains__() returns a true value when it is asked for an existent node, be it visible or not. This is more consistent with Group.__contains__().

  • Using Group.__delattr__() to remove a child is no longer supported. Please use Group._f_remove() instead.

  • The indexprops attribute is now present on all Table instances, be they indexed or not. In the last case, it is None.

  • Table.getWhereList() now has flavor parameter equal to “NumArray” by default, which is more consistent with other methods. Before, flavor defaulted to “List”.

  • The extVersion variable does no longer exist. It did not make much sense either, since the canonical version of the whole PyTables package is that of __version__.

  • The Row.nrow() has been converted into a property, so you have to replace any call to Row.nrow() into Row.nrow.

Deprecated features

  • The objects, groups and leaves mappings in File are retained only for compatibility purposes. Using File.getNode() is recommended to access nodes, File.__contains__() to check for node existence, and File.walkNodes() for iteration purposes. Using isinstance() and *isVisible*() methods is the preferred way of checking node type and visibility.

    Please note that the aforementioned mappings use the named methods internally, so the former have no special performance gains over the later.

API refinements

  • The isHDF5File() and isPyTablesFile() functions know how to handle nonexistent or unreadable files. An IOError is raised in those cases.

Bug fixes (affecting API)

  • None


Enjoy data!

—The PyTables Team