1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
|
1.3.0 (2015-09-08):
- Added python 3 support.
- Added an example of parsing ISO 8601.
- Support rule arguments in parser protocols.
1.2.0 (2013-08-01):
- Added 'makeProtocol' for parsing stream protocols using Parsley
grammars and Twisted, with examples, docs, and a tutorial.
- Added 'stack' as support for 'makeProtocol'.
- Sped up parsing; started compiling python expressons only once.
(thanks to eevee!)
- Allowed tuple unpacking in assignment in rules. For example,
`rule1 = rule2:(a, b) -> a + b`. (thanks to introom!)
- Added a trace hook and an example of a trace visualizer using it.
- Fixed an example in the README. (thanks to ilmarga!)
- Added support for custom labels in grammars. For example,
`rule1 = rule2 ^(foo)` will make Parsley consider 'foo' the name for
'rule2' in error messages. (thanks to miebster!)
1.1.0 (2013-03-03):
- The 'spaces' rule has been provided under the alias 'ws'; 'spaces'
is now the deprecated name.
- Single quotes can now be used around multiple characters to match
them consecutively. For example, 'abc' is equivalent 'a' 'b' 'c'.
- The 'token' rule is thus deprecated as well, since "token('foo')"
is more typing than "ws 'foo'".
- Hex escapes can now be used in literals for matching bytes.
- 'makeGrammar' now takes an 'unwrap' argument, to get a grammar
class suitable for subclassing.
- 'makeGrammar' also takes an 'extends' argument that can be either a
regular python class or a wrapped Parsley grammar, for simple
subclassing.
- 'repeat' expressions now work for 0 repetitions properly.
- ometa.runtime.TIMING flag added for enabling debug output of
grammar load times.
- OMeta version 1 parser moved to ometa.compat.
- Twine removed.
- Various bugfixes around error reporting.
- Parsley no longer loads its internal grammars at import time. The
'stage' script in bin/ is used to regenerate the compiled versions
after modification.
- Experimental tree transformer grammars, with special syntax for
destructuring terms and creating string templates, have been added.
1.0.0 (2012-10-01):
- PyMeta is now Parsley.
- A new public API has been added in the 'parsley' module.
- Grammars now use the OMeta 2 syntax by default. Compatibility is
available via ometa.grammar.OMeta1.
- A generic AST-handling library has been added, 'terml'.
- Grammars now parse to term trees instead of nested lists. The code
generator has accordingly been greatly simplified.
- A grammar interpreter has been added, which parses by evaluating
the grammar term tree directly instead of generating a class.
- A push parser has been added which accepts data in arbitrary
chunks, maintaining state while waiting for more input.
0.4.0 (2010-05-15):
- Builders now work from a grammar AST, rather than being driven by
the grammar parser directly.
- All parse operations now carry error information, including what
was expected to be parsed and where.
* Consequently, ParseError now has a 'formatError' method that takes
the input string and returns a human-readable description the
parse failure.
- New example: TermL parser, in examples/terml.
- New script, bin/generate_parser. Takes a file containing a PyMeta
grammar as input and writes a Python module file.
- A couple bugfixes:
#248643 subclassed grammars don't inherit base namespace
#564135 makeGrammar error handling for nonsensical grammars is broken
0.3.0 (2008-07-12):
- Input stream now implemented like the Javascript version, with immutable
position objects.
0.2.0 (2008-04-17):
- Grammars generate Python source instead of ASTs.
- OMeta now has a "metagrammarClass" attribute, to ease extension of
the metagrammar.
0.1.1 (2008-03-27):
- Fixed a bug in 'super'.
0.1 (2008-03-26):
- Initial release.
|
