Initial Commit

lib/html5lib/trie/__init__.py

@@ -0,0 +1,12 @@
+from __future__ import absolute_import, division, unicode_literals
+
+from .py import Trie as PyTrie
+
+Trie = PyTrie
+
+try:
+    from .datrie import Trie as DATrie
+except ImportError:
+    pass
+else:
+    Trie = DATrie

lib/html5lib/trie/_base.py

@@ -0,0 +1,37 @@
+from __future__ import absolute_import, division, unicode_literals
+
+from collections import Mapping
+
+
+class Trie(Mapping):
+    """Abstract base class for tries"""
+
+    def keys(self, prefix=None):
+        keys = super().keys()
+
+        if prefix is None:
+            return set(keys)
+
+        # Python 2.6: no set comprehensions
+        return set([x for x in keys if x.startswith(prefix)])
+
+    def has_keys_with_prefix(self, prefix):
+        for key in self.keys():
+            if key.startswith(prefix):
+                return True
+
+        return False
+
+    def longest_prefix(self, prefix):
+        if prefix in self:
+            return prefix
+
+        for i in range(1, len(prefix) + 1):
+            if prefix[:-i] in self:
+                return prefix[:-i]
+
+        raise KeyError(prefix)
+
+    def longest_prefix_item(self, prefix):
+        lprefix = self.longest_prefix(prefix)
+        return (lprefix, self[lprefix])

lib/html5lib/trie/datrie.py

@@ -0,0 +1,44 @@
+from __future__ import absolute_import, division, unicode_literals
+
+from datrie import Trie as DATrie
+from six import text_type
+
+from ._base import Trie as ABCTrie
+
+
+class Trie(ABCTrie):
+    def __init__(self, data):
+        chars = set()
+        for key in data.keys():
+            if not isinstance(key, text_type):
+                raise TypeError("All keys must be strings")
+            for char in key:
+                chars.add(char)
+
+        self._data = DATrie("".join(chars))
+        for key, value in data.items():
+            self._data[key] = value
+
+    def __contains__(self, key):
+        return key in self._data
+
+    def __len__(self):
+        return len(self._data)
+
+    def __iter__(self):
+        raise NotImplementedError()
+
+    def __getitem__(self, key):
+        return self._data[key]
+
+    def keys(self, prefix=None):
+        return self._data.keys(prefix)
+
+    def has_keys_with_prefix(self, prefix):
+        return self._data.has_keys_with_prefix(prefix)
+
+    def longest_prefix(self, prefix):
+        return self._data.longest_prefix(prefix)
+
+    def longest_prefix_item(self, prefix):
+        return self._data.longest_prefix_item(prefix)

lib/html5lib/trie/py.py

@@ -0,0 +1,67 @@
+from __future__ import absolute_import, division, unicode_literals
+from six import text_type
+
+from bisect import bisect_left
+
+from ._base import Trie as ABCTrie
+
+
+class Trie(ABCTrie):
+    def __init__(self, data):
+        if not all(isinstance(x, text_type) for x in data.keys()):
+            raise TypeError("All keys must be strings")
+
+        self._data = data
+        self._keys = sorted(data.keys())
+        self._cachestr = ""
+        self._cachepoints = (0, len(data))
+
+    def __contains__(self, key):
+        return key in self._data
+
+    def __len__(self):
+        return len(self._data)
+
+    def __iter__(self):
+        return iter(self._data)
+
+    def __getitem__(self, key):
+        return self._data[key]
+
+    def keys(self, prefix=None):
+        if prefix is None or prefix == "" or not self._keys:
+            return set(self._keys)
+
+        if prefix.startswith(self._cachestr):
+            lo, hi = self._cachepoints
+            start = i = bisect_left(self._keys, prefix, lo, hi)
+        else:
+            start = i = bisect_left(self._keys, prefix)
+
+        keys = set()
+        if start == len(self._keys):
+            return keys
+
+        while self._keys[i].startswith(prefix):
+            keys.add(self._keys[i])
+            i += 1
+
+        self._cachestr = prefix
+        self._cachepoints = (start, i)
+
+        return keys
+
+    def has_keys_with_prefix(self, prefix):
+        if prefix in self._data:
+            return True
+
+        if prefix.startswith(self._cachestr):
+            lo, hi = self._cachepoints
+            i = bisect_left(self._keys, prefix, lo, hi)
+        else:
+            i = bisect_left(self._keys, prefix)
+
+        if i == len(self._keys):
+            return False
+
+        return self._keys[i].startswith(prefix)