From f2c8b1e3629ba23f1db62352868cb95374d42cdb Mon Sep 17 00:00:00 2001
From: Ines Montani <ines@ines.io>
Date: Wed, 18 Sep 2019 20:24:41 +0200
Subject: [PATCH] Simplify lookup hashing

Just use get_string_id, which already does everything ensure_hash was supposed to do
---
 spacy/lookups.py                          | 15 ++++-----------
 spacy/tests/vocab_vectors/test_lookups.py | 13 +++++++------
 2 files changed, 11 insertions(+), 17 deletions(-)

diff --git a/spacy/lookups.py b/spacy/lookups.py
index c1f351fe5..05a60f289 100644
--- a/spacy/lookups.py
+++ b/spacy/lookups.py
@@ -7,16 +7,9 @@ from preshed.bloom import BloomFilter
 
 from .errors import Errors
 from .util import SimpleFrozenDict, ensure_path
-from .compat import basestring_
 from .strings import get_string_id
 
 
-def ensure_hash(key):
-    if isinstance(key, basestring_):
-        return get_string_id(key)
-    return key
-
-
 class Lookups(object):
     """Container for large lookup tables and dictionaries, e.g. lemmatization
     data or tokenizer exception lists. Lookups are available via vocab.lookups,
@@ -202,7 +195,7 @@ class Table(OrderedDict):
         key (unicode / int): The key to set.
         value: The value to set.
         """
-        key = ensure_hash(key)
+        key = get_string_id(key)
         OrderedDict.__setitem__(self, key, value)
         self.bloom.add(key)
 
@@ -221,7 +214,7 @@ class Table(OrderedDict):
         key (unicode / int): The key to get.
         RETURNS: The value.
         """
-        key = ensure_hash(key)
+        key = get_string_id(key)
         return OrderedDict.__getitem__(self, key)
 
     def get(self, key, default=None):
@@ -231,7 +224,7 @@ class Table(OrderedDict):
         default: The default value to return.
         RETURNS: The value.
         """
-        key = ensure_hash(key)
+        key = get_string_id(key)
         return OrderedDict.get(self, key, default)
 
     def __contains__(self, key):
@@ -240,7 +233,7 @@ class Table(OrderedDict):
         key (unicode / int): The key to check.
         RETURNS (bool): Whether the key is in the table.
         """
-        key = ensure_hash(key)
+        key = get_string_id(key)
         # This can give a false positive, so we need to check it after
         if key not in self.bloom:
             return False
diff --git a/spacy/tests/vocab_vectors/test_lookups.py b/spacy/tests/vocab_vectors/test_lookups.py
index 02f25532a..f78dd33c4 100644
--- a/spacy/tests/vocab_vectors/test_lookups.py
+++ b/spacy/tests/vocab_vectors/test_lookups.py
@@ -2,7 +2,8 @@
 from __future__ import unicode_literals
 
 import pytest
-from spacy.lookups import Lookups, Table, ensure_hash
+from spacy.lookups import Lookups, Table
+from spacy.strings import get_string_id
 from spacy.vocab import Vocab
 
 from ..util import make_tempdir
@@ -45,17 +46,17 @@ def test_table_api():
     table = Table(name="table", data=data)
     assert len(table) == len(data)
     assert "foo" in table
-    assert ensure_hash("foo") in table
+    assert get_string_id("foo") in table
     assert table["foo"] == "bar"
-    assert table[ensure_hash("foo")] == "bar"
+    assert table[get_string_id("foo")] == "bar"
     assert table.get("foo") == "bar"
     assert table.get("abc") is None
     table["abc"] = 123
     assert table["abc"] == 123
-    assert table[ensure_hash("abc")] == 123
+    assert table[get_string_id("abc")] == 123
     table.set("def", 456)
     assert table["def"] == 456
-    assert table[ensure_hash("def")] == 456
+    assert table[get_string_id("def")] == 456
 
 
 def test_table_api_to_from_bytes():
@@ -66,7 +67,7 @@ def test_table_api_to_from_bytes():
     assert new_table.name == "table"
     assert len(new_table) == 3
     assert new_table["foo"] == "bar"
-    assert new_table[ensure_hash("foo")] == "bar"
+    assert new_table[get_string_id("foo")] == "bar"
     new_table2 = Table(data={"def": 456})
     new_table2.from_bytes(table_bytes)
     assert len(new_table2) == 3