Skip to content
Newer
Older
100644 55 lines (32 sloc) 1.72 KB
7d4e663 [svn r3933] r4727@delle: sbehnel | 2008-08-09 11:17:28 +0200
scoder authored Aug 9, 2008
1 Things to try out when life permits
2 ===================================
3
bf4d27e [svn r3938] r4739@delle: sbehnel | 2008-08-15 14:52:32 +0200
scoder authored Aug 19, 2008
4 * zlib-based parsing/serialising of compressed in-memory data
5
6 * requires a libxml2 I/O OutputBuffer with appropriate I/O functions
fc88bcb @scoder cleaned up TODO and IDEAS text files
scoder authored Oct 9, 2012
7 that call into the zlib compression routines
0a1aea1 [svn r4168] r5162@delle: sbehnel | 2009-06-20 17:59:17 +0200
scoder authored Jun 20, 2009
8
9 * lzma-based parsing/serialising of compressed in-memory data
10
11 * requires a libxml2 I/O OutputBuffer with appropriate I/O functions
12 that call into the lzma compression routines
13
14 * advantage over zlib: probably faster and better compression
15
16 * maybe embed the lzma C sources in the distro
17 http://www.7-zip.org/sdk.html
bf4d27e [svn r3938] r4739@delle: sbehnel | 2008-08-15 14:52:32 +0200
scoder authored Aug 19, 2008
18
7d4e663 [svn r3933] r4727@delle: sbehnel | 2008-08-09 11:17:28 +0200
scoder authored Aug 9, 2008
19 * generating XML using the ``with`` statement
20
21 http://comments.gmane.org/gmane.comp.python.general/579950?set_lines=100000
22
23 * parse-time validation against a user provided DTD
24
25 * currently only works for XML Schema
26
27 * somehow integrate RelaxNG compact notation (rnc versus rng)
28
29 * currently not supported by libxml2 (patch exists)
30
31 * support subclassing XSLTAccessControl to provide custom per-URL
32 access check methods
33
34 * maybe custom resolvers are enough, or can be combined with this?
bc40f3d [svn r3975] r4797@delle: sbehnel | 2008-09-17 15:20:13 +0200
scoder authored Sep 21, 2008
35
36 * reimplement iterparse() using the libxml2 xmlReader API
37
38 * Advantage: the implementation can be made safer than the current
39 SAX implementation, as the parser would not interact with the
40 Python-level tree.
41
42 * Disadvantage: the tree has to be built manually. In the current
43 SAX based implementation, libxml2 does it for us.
fc88bcb @scoder cleaned up TODO and IDEAS text files
scoder authored Oct 9, 2012
44
45 * rewrite iterparse() to accept a parser as argument instead of being
46 one
47
48 * disadvantage: iterparse() can't deal with all parser options
49
50 * provide an HTMLParser wrapper that handles broken encodings in broken
51 HTML better, e.g. using BeautifulSoup's "unicode dammit" analyser
52
53 * expose namespace prefixes through the QName class
54
Something went wrong with that request. Please try again.