This repo contains code to mirror other repos. It also contains the code that is getting mirrored.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

272 lines
10 KiB

  1. from .exceptions import GrammarError
  2. from .lexer import Token
  3. from .tree import Tree
  4. from .visitors import InlineTransformer # XXX Deprecated
  5. from .visitors import Transformer_InPlace
  6. from .visitors import _vargs_meta, _vargs_meta_inline
  7. ###{standalone
  8. from functools import partial, wraps
  9. from itertools import repeat, product
  10. class ExpandSingleChild:
  11. def __init__(self, node_builder):
  12. self.node_builder = node_builder
  13. def __call__(self, children):
  14. if len(children) == 1:
  15. return children[0]
  16. else:
  17. return self.node_builder(children)
  18. class PropagatePositions:
  19. def __init__(self, node_builder):
  20. self.node_builder = node_builder
  21. def __call__(self, children):
  22. res = self.node_builder(children)
  23. if isinstance(res, Tree):
  24. for c in children:
  25. if isinstance(c, Tree) and not c.meta.empty:
  26. res.meta.line = c.meta.line
  27. res.meta.column = c.meta.column
  28. res.meta.start_pos = c.meta.start_pos
  29. res.meta.empty = False
  30. break
  31. elif isinstance(c, Token):
  32. res.meta.line = c.line
  33. res.meta.column = c.column
  34. res.meta.start_pos = c.pos_in_stream
  35. res.meta.empty = False
  36. break
  37. for c in reversed(children):
  38. if isinstance(c, Tree) and not c.meta.empty:
  39. res.meta.end_line = c.meta.end_line
  40. res.meta.end_column = c.meta.end_column
  41. res.meta.end_pos = c.meta.end_pos
  42. res.meta.empty = False
  43. break
  44. elif isinstance(c, Token):
  45. res.meta.end_line = c.end_line
  46. res.meta.end_column = c.end_column
  47. res.meta.end_pos = c.end_pos
  48. res.meta.empty = False
  49. break
  50. return res
  51. class ChildFilter:
  52. def __init__(self, to_include, append_none, node_builder):
  53. self.node_builder = node_builder
  54. self.to_include = to_include
  55. self.append_none = append_none
  56. def __call__(self, children):
  57. filtered = []
  58. for i, to_expand, add_none in self.to_include:
  59. if add_none:
  60. filtered += [None] * add_none
  61. if to_expand:
  62. filtered += children[i].children
  63. else:
  64. filtered.append(children[i])
  65. if self.append_none:
  66. filtered += [None] * self.append_none
  67. return self.node_builder(filtered)
  68. class ChildFilterLALR(ChildFilter):
  69. "Optimized childfilter for LALR (assumes no duplication in parse tree, so it's safe to change it)"
  70. def __call__(self, children):
  71. filtered = []
  72. for i, to_expand, add_none in self.to_include:
  73. if add_none:
  74. filtered += [None] * add_none
  75. if to_expand:
  76. if filtered:
  77. filtered += children[i].children
  78. else: # Optimize for left-recursion
  79. filtered = children[i].children
  80. else:
  81. filtered.append(children[i])
  82. if self.append_none:
  83. filtered += [None] * self.append_none
  84. return self.node_builder(filtered)
  85. class ChildFilterLALR_NoPlaceholders(ChildFilter):
  86. "Optimized childfilter for LALR (assumes no duplication in parse tree, so it's safe to change it)"
  87. def __init__(self, to_include, node_builder):
  88. self.node_builder = node_builder
  89. self.to_include = to_include
  90. def __call__(self, children):
  91. filtered = []
  92. for i, to_expand in self.to_include:
  93. if to_expand:
  94. if filtered:
  95. filtered += children[i].children
  96. else: # Optimize for left-recursion
  97. filtered = children[i].children
  98. else:
  99. filtered.append(children[i])
  100. return self.node_builder(filtered)
  101. def _should_expand(sym):
  102. return not sym.is_term and sym.name.startswith('_')
  103. def maybe_create_child_filter(expansion, keep_all_tokens, ambiguous, _empty_indices):
  104. # Prepare empty_indices as: How many Nones to insert at each index?
  105. if _empty_indices:
  106. assert _empty_indices.count(False) == len(expansion)
  107. s = ''.join(str(int(b)) for b in _empty_indices)
  108. empty_indices = [len(ones) for ones in s.split('0')]
  109. assert len(empty_indices) == len(expansion)+1, (empty_indices, len(expansion))
  110. else:
  111. empty_indices = [0] * (len(expansion)+1)
  112. to_include = []
  113. nones_to_add = 0
  114. for i, sym in enumerate(expansion):
  115. nones_to_add += empty_indices[i]
  116. if keep_all_tokens or not (sym.is_term and sym.filter_out):
  117. to_include.append((i, _should_expand(sym), nones_to_add))
  118. nones_to_add = 0
  119. nones_to_add += empty_indices[len(expansion)]
  120. if _empty_indices or len(to_include) < len(expansion) or any(to_expand for i, to_expand,_ in to_include):
  121. if _empty_indices or ambiguous:
  122. return partial(ChildFilter if ambiguous else ChildFilterLALR, to_include, nones_to_add)
  123. else:
  124. # LALR without placeholders
  125. return partial(ChildFilterLALR_NoPlaceholders, [(i, x) for i,x,_ in to_include])
  126. class AmbiguousExpander:
  127. """Deal with the case where we're expanding children ('_rule') into a parent but the children
  128. are ambiguous. i.e. (parent->_ambig->_expand_this_rule). In this case, make the parent itself
  129. ambiguous with as many copies as their are ambiguous children, and then copy the ambiguous children
  130. into the right parents in the right places, essentially shifting the ambiguiuty up the tree."""
  131. def __init__(self, to_expand, tree_class, node_builder):
  132. self.node_builder = node_builder
  133. self.tree_class = tree_class
  134. self.to_expand = to_expand
  135. def __call__(self, children):
  136. def _is_ambig_tree(child):
  137. return hasattr(child, 'data') and child.data == '_ambig'
  138. #### When we're repeatedly expanding ambiguities we can end up with nested ambiguities.
  139. # All children of an _ambig node should be a derivation of that ambig node, hence
  140. # it is safe to assume that if we see an _ambig node nested within an ambig node
  141. # it is safe to simply expand it into the parent _ambig node as an alternative derivation.
  142. ambiguous = []
  143. for i, child in enumerate(children):
  144. if _is_ambig_tree(child):
  145. if i in self.to_expand:
  146. ambiguous.append(i)
  147. to_expand = [j for j, grandchild in enumerate(child.children) if _is_ambig_tree(grandchild)]
  148. child.expand_kids_by_index(*to_expand)
  149. if not ambiguous:
  150. return self.node_builder(children)
  151. expand = [ iter(child.children) if i in ambiguous else repeat(child) for i, child in enumerate(children) ]
  152. return self.tree_class('_ambig', [self.node_builder(list(f[0])) for f in product(zip(*expand))])
  153. def maybe_create_ambiguous_expander(tree_class, expansion, keep_all_tokens):
  154. to_expand = [i for i, sym in enumerate(expansion)
  155. if keep_all_tokens or ((not (sym.is_term and sym.filter_out)) and _should_expand(sym))]
  156. if to_expand:
  157. return partial(AmbiguousExpander, to_expand, tree_class)
  158. def ptb_inline_args(func):
  159. @wraps(func)
  160. def f(children):
  161. return func(*children)
  162. return f
  163. def inplace_transformer(func):
  164. @wraps(func)
  165. def f(children):
  166. # function name in a Transformer is a rule name.
  167. tree = Tree(func.__name__, children)
  168. return func(tree)
  169. return f
  170. def apply_visit_wrapper(func, name, wrapper):
  171. if wrapper is _vargs_meta or wrapper is _vargs_meta_inline:
  172. raise NotImplementedError("Meta args not supported for internal transformer")
  173. @wraps(func)
  174. def f(children):
  175. return wrapper(func, name, children, None)
  176. return f
  177. class ParseTreeBuilder:
  178. def __init__(self, rules, tree_class, propagate_positions=False, keep_all_tokens=False, ambiguous=False, maybe_placeholders=False):
  179. self.tree_class = tree_class
  180. self.propagate_positions = propagate_positions
  181. self.always_keep_all_tokens = keep_all_tokens
  182. self.ambiguous = ambiguous
  183. self.maybe_placeholders = maybe_placeholders
  184. self.rule_builders = list(self._init_builders(rules))
  185. def _init_builders(self, rules):
  186. for rule in rules:
  187. options = rule.options
  188. keep_all_tokens = self.always_keep_all_tokens or options.keep_all_tokens
  189. expand_single_child = options.expand1
  190. wrapper_chain = list(filter(None, [
  191. (expand_single_child and not rule.alias) and ExpandSingleChild,
  192. maybe_create_child_filter(rule.expansion, keep_all_tokens, self.ambiguous, options.empty_indices if self.maybe_placeholders else None),
  193. self.propagate_positions and PropagatePositions,
  194. self.ambiguous and maybe_create_ambiguous_expander(self.tree_class, rule.expansion, keep_all_tokens),
  195. ]))
  196. yield rule, wrapper_chain
  197. def create_callback(self, transformer=None):
  198. callbacks = {}
  199. for rule, wrapper_chain in self.rule_builders:
  200. user_callback_name = rule.alias or rule.options.template_source or rule.origin.name
  201. try:
  202. f = getattr(transformer, user_callback_name)
  203. # XXX InlineTransformer is deprecated!
  204. wrapper = getattr(f, 'visit_wrapper', None)
  205. if wrapper is not None:
  206. f = apply_visit_wrapper(f, user_callback_name, wrapper)
  207. else:
  208. if isinstance(transformer, InlineTransformer):
  209. f = ptb_inline_args(f)
  210. elif isinstance(transformer, Transformer_InPlace):
  211. f = inplace_transformer(f)
  212. except AttributeError:
  213. f = partial(self.tree_class, user_callback_name)
  214. for w in wrapper_chain:
  215. f = w(f)
  216. if rule in callbacks:
  217. raise GrammarError("Rule '%s' already exists" % (rule,))
  218. callbacks[rule] = f
  219. return callbacks
  220. ###}