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.

278 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. # local reference to Tree.meta reduces number of presence checks
  24. if isinstance(res, Tree):
  25. res_meta = res.meta
  26. for c in children:
  27. if isinstance(c, Tree):
  28. child_meta = c.meta
  29. if not child_meta.empty:
  30. res_meta.line = child_meta.line
  31. res_meta.column = child_meta.column
  32. res_meta.start_pos = child_meta.start_pos
  33. res_meta.empty = False
  34. break
  35. elif isinstance(c, Token):
  36. res_meta.line = c.line
  37. res_meta.column = c.column
  38. res_meta.start_pos = c.pos_in_stream
  39. res_meta.empty = False
  40. break
  41. for c in reversed(children):
  42. if isinstance(c, Tree):
  43. child_meta = c.meta
  44. if not child_meta.empty:
  45. res_meta.end_line = child_meta.end_line
  46. res_meta.end_column = child_meta.end_column
  47. res_meta.end_pos = child_meta.end_pos
  48. res_meta.empty = False
  49. break
  50. elif isinstance(c, Token):
  51. res_meta.end_line = c.end_line
  52. res_meta.end_column = c.end_column
  53. res_meta.end_pos = c.end_pos
  54. res_meta.empty = False
  55. break
  56. return res
  57. class ChildFilter:
  58. def __init__(self, to_include, append_none, node_builder):
  59. self.node_builder = node_builder
  60. self.to_include = to_include
  61. self.append_none = append_none
  62. def __call__(self, children):
  63. filtered = []
  64. for i, to_expand, add_none in self.to_include:
  65. if add_none:
  66. filtered += [None] * add_none
  67. if to_expand:
  68. filtered += children[i].children
  69. else:
  70. filtered.append(children[i])
  71. if self.append_none:
  72. filtered += [None] * self.append_none
  73. return self.node_builder(filtered)
  74. class ChildFilterLALR(ChildFilter):
  75. "Optimized childfilter for LALR (assumes no duplication in parse tree, so it's safe to change it)"
  76. def __call__(self, children):
  77. filtered = []
  78. for i, to_expand, add_none in self.to_include:
  79. if add_none:
  80. filtered += [None] * add_none
  81. if to_expand:
  82. if filtered:
  83. filtered += children[i].children
  84. else: # Optimize for left-recursion
  85. filtered = children[i].children
  86. else:
  87. filtered.append(children[i])
  88. if self.append_none:
  89. filtered += [None] * self.append_none
  90. return self.node_builder(filtered)
  91. class ChildFilterLALR_NoPlaceholders(ChildFilter):
  92. "Optimized childfilter for LALR (assumes no duplication in parse tree, so it's safe to change it)"
  93. def __init__(self, to_include, node_builder):
  94. self.node_builder = node_builder
  95. self.to_include = to_include
  96. def __call__(self, children):
  97. filtered = []
  98. for i, to_expand in self.to_include:
  99. if to_expand:
  100. if filtered:
  101. filtered += children[i].children
  102. else: # Optimize for left-recursion
  103. filtered = children[i].children
  104. else:
  105. filtered.append(children[i])
  106. return self.node_builder(filtered)
  107. def _should_expand(sym):
  108. return not sym.is_term and sym.name.startswith('_')
  109. def maybe_create_child_filter(expansion, keep_all_tokens, ambiguous, _empty_indices):
  110. # Prepare empty_indices as: How many Nones to insert at each index?
  111. if _empty_indices:
  112. assert _empty_indices.count(False) == len(expansion)
  113. s = ''.join(str(int(b)) for b in _empty_indices)
  114. empty_indices = [len(ones) for ones in s.split('0')]
  115. assert len(empty_indices) == len(expansion)+1, (empty_indices, len(expansion))
  116. else:
  117. empty_indices = [0] * (len(expansion)+1)
  118. to_include = []
  119. nones_to_add = 0
  120. for i, sym in enumerate(expansion):
  121. nones_to_add += empty_indices[i]
  122. if keep_all_tokens or not (sym.is_term and sym.filter_out):
  123. to_include.append((i, _should_expand(sym), nones_to_add))
  124. nones_to_add = 0
  125. nones_to_add += empty_indices[len(expansion)]
  126. if _empty_indices or len(to_include) < len(expansion) or any(to_expand for i, to_expand,_ in to_include):
  127. if _empty_indices or ambiguous:
  128. return partial(ChildFilter if ambiguous else ChildFilterLALR, to_include, nones_to_add)
  129. else:
  130. # LALR without placeholders
  131. return partial(ChildFilterLALR_NoPlaceholders, [(i, x) for i,x,_ in to_include])
  132. class AmbiguousExpander:
  133. """Deal with the case where we're expanding children ('_rule') into a parent but the children
  134. are ambiguous. i.e. (parent->_ambig->_expand_this_rule). In this case, make the parent itself
  135. ambiguous with as many copies as their are ambiguous children, and then copy the ambiguous children
  136. into the right parents in the right places, essentially shifting the ambiguiuty up the tree."""
  137. def __init__(self, to_expand, tree_class, node_builder):
  138. self.node_builder = node_builder
  139. self.tree_class = tree_class
  140. self.to_expand = to_expand
  141. def __call__(self, children):
  142. def _is_ambig_tree(child):
  143. return hasattr(child, 'data') and child.data == '_ambig'
  144. #### When we're repeatedly expanding ambiguities we can end up with nested ambiguities.
  145. # All children of an _ambig node should be a derivation of that ambig node, hence
  146. # it is safe to assume that if we see an _ambig node nested within an ambig node
  147. # it is safe to simply expand it into the parent _ambig node as an alternative derivation.
  148. ambiguous = []
  149. for i, child in enumerate(children):
  150. if _is_ambig_tree(child):
  151. if i in self.to_expand:
  152. ambiguous.append(i)
  153. to_expand = [j for j, grandchild in enumerate(child.children) if _is_ambig_tree(grandchild)]
  154. child.expand_kids_by_index(*to_expand)
  155. if not ambiguous:
  156. return self.node_builder(children)
  157. expand = [ iter(child.children) if i in ambiguous else repeat(child) for i, child in enumerate(children) ]
  158. return self.tree_class('_ambig', [self.node_builder(list(f[0])) for f in product(zip(*expand))])
  159. def maybe_create_ambiguous_expander(tree_class, expansion, keep_all_tokens):
  160. to_expand = [i for i, sym in enumerate(expansion)
  161. if keep_all_tokens or ((not (sym.is_term and sym.filter_out)) and _should_expand(sym))]
  162. if to_expand:
  163. return partial(AmbiguousExpander, to_expand, tree_class)
  164. def ptb_inline_args(func):
  165. @wraps(func)
  166. def f(children):
  167. return func(*children)
  168. return f
  169. def inplace_transformer(func):
  170. @wraps(func)
  171. def f(children):
  172. # function name in a Transformer is a rule name.
  173. tree = Tree(func.__name__, children)
  174. return func(tree)
  175. return f
  176. def apply_visit_wrapper(func, name, wrapper):
  177. if wrapper is _vargs_meta or wrapper is _vargs_meta_inline:
  178. raise NotImplementedError("Meta args not supported for internal transformer")
  179. @wraps(func)
  180. def f(children):
  181. return wrapper(func, name, children, None)
  182. return f
  183. class ParseTreeBuilder:
  184. def __init__(self, rules, tree_class, propagate_positions=False, keep_all_tokens=False, ambiguous=False, maybe_placeholders=False):
  185. self.tree_class = tree_class
  186. self.propagate_positions = propagate_positions
  187. self.always_keep_all_tokens = keep_all_tokens
  188. self.ambiguous = ambiguous
  189. self.maybe_placeholders = maybe_placeholders
  190. self.rule_builders = list(self._init_builders(rules))
  191. def _init_builders(self, rules):
  192. for rule in rules:
  193. options = rule.options
  194. keep_all_tokens = self.always_keep_all_tokens or options.keep_all_tokens
  195. expand_single_child = options.expand1
  196. wrapper_chain = list(filter(None, [
  197. (expand_single_child and not rule.alias) and ExpandSingleChild,
  198. maybe_create_child_filter(rule.expansion, keep_all_tokens, self.ambiguous, options.empty_indices if self.maybe_placeholders else None),
  199. self.propagate_positions and PropagatePositions,
  200. self.ambiguous and maybe_create_ambiguous_expander(self.tree_class, rule.expansion, keep_all_tokens),
  201. ]))
  202. yield rule, wrapper_chain
  203. def create_callback(self, transformer=None):
  204. callbacks = {}
  205. for rule, wrapper_chain in self.rule_builders:
  206. user_callback_name = rule.alias or rule.options.template_source or rule.origin.name
  207. try:
  208. f = getattr(transformer, user_callback_name)
  209. # XXX InlineTransformer is deprecated!
  210. wrapper = getattr(f, 'visit_wrapper', None)
  211. if wrapper is not None:
  212. f = apply_visit_wrapper(f, user_callback_name, wrapper)
  213. else:
  214. if isinstance(transformer, InlineTransformer):
  215. f = ptb_inline_args(f)
  216. elif isinstance(transformer, Transformer_InPlace):
  217. f = inplace_transformer(f)
  218. except AttributeError:
  219. f = partial(self.tree_class, user_callback_name)
  220. for w in wrapper_chain:
  221. f = w(f)
  222. if rule in callbacks:
  223. raise GrammarError("Rule '%s' already exists" % (rule,))
  224. callbacks[rule] = f
  225. return callbacks
  226. ###}