Extensible HTML/XML generator, cross-platform templating language, Oracle utilities and various other tools
Project description
XIST provides an extensible HTML and XML generator. XIST is also a XML parser with a very simple and pythonesque tree API. Every XML element type corresponds to a Python class and these Python classes provide a conversion method to transform the XML tree (e.g. into HTML). XIST can be considered ‘object oriented XSLT’.
XIST also includes the following modules and packages:
ll.ul4c is compiler for a cross-platform templating language with similar capabilities to Django’s templating language. UL4 templates are compiled to an internal format, which makes it possible to implement template renderers in other languages and makes the template code “secure” (i.e. template code can’t open or delete files).
There are implementations for Python, Java, Javascript and PHP.
ll.ul4on provides functions for encoding and decoding a lightweight machine-readable text-based format for serializing the object types supported by UL4. It is extensible to allow encoding/decoding arbitrary instances (i.e. it is basically a reimplementation of pickle, but with string input/output instead of bytes and with an eye towards cross-plattform support).
There are implementations for Python, Java, Javascript and PHP.
ll.orasql provides utilities for working with cx_Oracle:
It allows calling functions and procedures with keyword arguments.
Query results will be put into Record objects, where database fields are accessible as object attributes.
The Connection class provides methods for iterating through the database metadata.
Importing the modules adds support for URLs with the scheme oracle to ll.url.
ll.make is an object oriented make replacement. Like make it allows you to specify dependencies between files and actions to be executed when files don’t exist or are out of date with respect to one of their sources. But unlike make you can do this in a object oriented way and targets are not only limited to files.
ll.color provides classes and functions for handling RGB color values. This includes the ability to convert between different color models (RGB, HSV, HLS) as well as to and from CSS format, and several functions for modifying and mixing colors.
ll.sisyphus provides classes for running Python scripts as cron jobs.
ll.url provides classes for parsing and constructing RFC 2396 compliant URLs.
ll.nightshade can be used to serve the output of PL/SQL functions/procedures with CherryPy.
ll.misc provides several small utility functions and classes.
ll.astyle can be used for colored terminal output (via ANSI escape sequences).
ll.daemon can be used on UNIX to fork a daemon process.
ll.xml_codec contains a complete codec for encoding and decoding XML.
Changes in 5.0 (released 06/04/2013)
The HTML namespace (ll.xist.ns.html) has been updated to support the current HTML5 definition.
However old elements/attributes from the previous HTML namespace are still supported.
XIST now allows arbitrary elements and attributes. ll.xist.parse will parse any XML file, even if the pool object doesn’t contain an element for the element name, and even if an attribute name isn’t declared for an element.
Undeclared elements will be “plain” instances of ll.xist.xsc.Element (i.e. not instances of a subclass of ll.xist.xsc.Element) with the attributes xmlns and xmlname set accordingly and undeclared attributes will be “plain” instances of ll.xist.xsc.Attr (with proper xmlns and xmlname attributes).
This new feature requires several API changes which will be described below.
Validation is now off by default, to turn it on pass validate=True to parse.tree or parse.itertree for parsing, or the publisher object or the bytes, iterbytes, string or iterstring methods for publishing.
Accessing an attribute via __getattr__ (i.e. htmlelement.attrs.class_) only works for attributes that are declared for the class, all other attributes must be accessed via __getitem__ (i.e. htmlelement.attrs["class"]). __getitem__ always requires the XML name of the attribute. __getitem__ also allows an attribute name for a global attribute in Clark notation (i.e. htmlelement.attrs["{http://www.w3.org/XML/1998/namespace}lang"]). A global attribute can also be accessed via a (namespace name, attribute name) tuple (i.e. htmlelement.attrs[("http://www.w3.org/XML/1998/namespace", "lang")]). Using an attribute class or attribute object is also possible (i.e. htmlelement.attrs[xml.Attr.lang] or htmlelement.attrs[xml.Attr.lang('de')]).
Using __setattr__ to set attributes only works for declared attributes too. Using __setitem__ to set attributes supports the same kind of arguments as __getitem__ does. For declared attributes the resulting attribute object will always be an instance of the declared attribute class. For all other attributes it will be an instance of ll.xist.xsc.Attr except when an attribute class or instance is used as the key. In this case the attribute will be an instance of that class.
The methods convert, clone, __copy__, __deepcopy__, compacted, withsep, reversed, filtered, shuffled, mapped and normalized make sure that plain nodes are copied properly, i.e. they retain their custom xmlns and xmlname attributes.
The keys in an attribute dictionary (i.e. an ll.xist.xsc.Attrs object) are no longer the attribute classes, but the (namespace name, attribute name) tuples:
>>> node = html.div({xml.Attrs.lang: 'de'}, id='id42', class_='foo') >>> list(node.attrs.keys()) [('http://www.w3.org/XML/1998/namespace', 'lang'), (None, 'class'), (None, 'id')]For all methods that existed in Python/XML pairs (e.g. withnames and withnames_xml in xsc.Attrs or elementclass and elementclass_xml in xsc.Pool etc.) there is only one version now: A method without the _xml suffix that accepts the XML version of the name.
The method checkvalid has been renamed to validate. It no longer calls warnings.warn itself, but is a generator that returns the warning objects. Furthermore the model objects now get passed the complete path instead of only the target node (this is used to implement HTML5’s transparent content model).
Validating whether an attribute is allowed is now done in Attrs.validateattr. The default implementation yields warnings about undeclared local attributes. The HTML5 namespace extends this to also accept any attribute whose name starts with data- or aria-.
Node comparison now ignores the classes for elements, entities and processsing instructions, so that plain nodes compare equal to instances of Element, Entity or ProcInst subclasses as long as the name and content of the node matches.
ll.xist.parse.Tidy no longer has a skipbad argument.
Converter contexts now support string as keys (which must be hierarchical dot-separated names similar to Java package names (e.g. "org.example.project.handler") to avoid name collisions).
The docbook module has been updated to support DocBook 5.0.
URL objects are pickable now.
When whitespace is removed in the literal text of UL4 templates (via the keepws parameter), any initial spaces (before the first line feed) are now no longer removed.
If you have Cython installed and the environment variable LL_USE_CYTHON set, several modules will now be compiled into extension modules.
It’s now possible to expose attributes and methods of objects to UL4 templates. Exposing attributes can be done by setting a class or instance attribute ul4attrs to a sequence of attribute names. Exposing methods can be done with the decorators ul4c.expose_method and ul4c.expose_generatormethod.
A new UL4 function list has been added. This function works like the Python function list, creating a copy of a sequence or materialzing an iterator.
A new UL4 function slice has been added. It works like itertools.slice, i.e. returning a slice from an iterator.
The function html.astext that converts an XIST tree containing HTML to plain text is now implemented in plain Python so it no longer requires a text mode browser. The function also got more configurable.
The objects available to db2ul4 scripts have been changed: oracle, sqlite and mysql are now objects with a connect method that returns a connection object. A connection object now has a method query that executes the query and returns an iterator over the results. Furthermore query supports keyword arguments for parameterized queries, i.e. you can now do:
<?code db = oracle.connect("user/pwd@db")?> <?for row in db.query("select * from foo where bar = :bar", bar=42)?> <?print row?> <?end for?>The system object now has an execute method that executes the system command.
Fixed a bug in orasql.OracleFileResource.close that surfaced when writing to on Oracle object.
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
Built Distribution
Hashes for ll_xist-5.0-py3.3-macosx-10.7-x86_64.egg
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 91ebc0a9c022f8c526088cc0d25eef724952c00211404c7cc46369e82a907240 |
|
| MD5 | 654056b308da36fec8775007c091027e |
|
| BLAKE2b-256 | f4661060629cb38c6b45dcf91be061608ba8dfb76d1a400dce604077144f2924 |
