1 _____________________________________________________________________________
5 foo.add(y.andnot(x)) ==> this is broken
9 - Get at least *some* sort of moderate improvement in the error messages
11 ..................................................
13 - evil problems with: (x y? z /ws)
14 - it gets even more evil than that
15 - basically, follow restrictions are not honored when the element
16 matches against the empty string
18 ______________________________________________________________________________
21 - precedes restrictions ("<-")
23 - MUST HAVE BETTER ERROR MESSAGES
24 - use for developing java15.g
27 - once this is ready, do big announcement
29 - topology no longer needed as an arg to parser?
31 - broader regression testing (for stuff like error messages, etc)
33 - More topology untangling [later]
34 - tib: use the lexer only for indentation increases/decreases
35 - grammar highlighting?
37 - Forest needs a "manual access" API
38 - the unwrap bit in Forest makes it really hard to expose an API for forests
42 ______________________________________________________________________________
45 - finalize metagrammar and rdp-op's
50 - RFC2822 (email message/headers)
51 - clean up the whole Walk situation (?)
54 ______________________________________________________________________________
57 - serialization of parse tables
59 - "ambiguity modulo dropped fragments"?
60 - can this be checked statically?
61 - eliminated statically?
63 - substring parsing for better error messages
66 - "Regular Right Part" grammars (NP Chapman, etc)
67 - Attribute unification
69 - inference of rejections for literals
70 - "prefer whitespace higher up" (?)
72 - Labeled edges on trees (associate a label with each slot in the
73 child array in Forest.Body? might make equality tough) --
74 equivalent to Feature Structures. Colon-labeling.
76 ______________________________________________________________________________
79 - Partly-Linear-PATR? (O(n^6) unification grammar)
81 - Implement a k-token peek buffer (for each state, see if it "dead
82 ends" during the next k Phases based solely on state -- ignoring
85 - Arrange for the SPPF corresponding to dropped subtrees to never be
86 generated (or merged, etc)
88 - Is there any way we can avoid creating a GSS.Node instance for
89 nodes which are transient in the sense that they have only one
92 - Re-read Rekers, particularly the stuff on optimal sharing
94 - Isolate the Element objects from Parse.Table/GSS so we can move
97 - consider allowing a Forest.Body to represent some other Tree whose
98 Body's should be [recursively] considered part of this Forest.
100 - perhaps not: right now we have a nice situation where
101 Forest.Ref instances become immutable once iterator()ed. This
102 also gives us a strong place to to culling with the certainty
103 that we won't throw out a Body which would later be salvaged
104 by some yet-to-be-added dependency.
106 - Figure out if there is a way to:
108 - allow unwrapping of children other than the very last one.
110 - fold repetitions into an array form in Forest, before
111 conversion to Tree. The major problem here is that multiple
112 tree-arrays are possible, all of different lengths. Worse,
113 even if they're all the same length, not all elements belong
114 in the same "possibility vector" as all others. You
115 essentially need a GSS to represent the array, which perhaps
116 is what the unfolded form was in the first place.
118 - Wikipedia grammar (needs to be both lexerless and boolean)
121 => Ordered Choice (";" operator)
123 - bring back in parse-table phase resolution of precedence (just
124 like associativity). This can be inferred from the use of ">"
125 when the rules are in one of these special forms:
136 where "_" is anything and "E" is the defining nonterminal.
137 Essentially what we're looking for is the situation where the
138 leftmost portion of one rule produces another rule, and the
139 rightmost portion of the latter produces the former.
141 I'm not 100% certain that this is as "strong" as the prefer/avoid
142 form (try to prove this, you probably can), but it's "what people
143 intend" most of the time.
145 - implement Johnstone's algorithm for "reduced, resolved LR
146 tables" to eliminate superfluous reductions on
149 ______________________________________________________________________________
152 - Rekers & Koorn note that GLR Substring Parsing can be used to do
153 really elegant and generalized "autocompletion".
156 ______________________________________________________________________________
159 - Incremental parse table construction
160 - "lazy GLR" and "lazy trees" -> language with first-class CF matching
161 - perhaps linear boolean grammars instead? (linear time, quad space)
162 - Forest parsing => chained parsers
163 - unification parsing, attributes, etc
165 - Take another stab at maximal-match? Nonterminal not-followed-by is
167 - Error recovery based on substring parsing