Intermediate state

spacy/errors.py

@@ -955,7 +955,7 @@ class Errors(metaclass=ErrorsWithCodes):
     E1046 = ("{cls_name} is an abstract class and cannot be instantiated. If you are looking for spaCy's default "
              "knowledge base, use `InMemoryLookupKB`.")
     E1047 = ("Invalid rich group config '{label}'.")
-    E1048 = ("Length > 31 in rich group config '{label}.")
+    E1048 = ("Length > 63 in rich group config '{label}.")
     E1049 = ("Rich group config {label} specifies lengths that are not in ascending order.")
     E1050 = ("Error splitting UTF-8 byte string into separate characters.")
 

spacy/ml/models/tok2vec.py

@@ -199,6 +199,8 @@ def _verify_rich_config_group(
     if lengths is not None or rows is not None:
         if is_search_char_group and (search_chars is None or len(search_chars) == 0):
             raise ValueError(Errors.E1047.format(label=label))
+        if len(search_chars) > 63:
+            raise ValueError(Errors.E1048.format(label=label))
         if lengths is None or rows is None:
             raise ValueError(Errors.E1047.format(label=label))
         if len(lengths) != len(rows):
@@ -208,7 +210,7 @@ def _verify_rich_config_group(
     elif search_chars is not None:
         raise ValueError(Errors.E1047.format(label=label))
     if lengths is not None:
-        if lengths[-1] > 31:
+        if lengths[-1] > 63:
             raise ValueError(Errors.E1048.format(label=label))
         if len(lengths) != len(set(lengths)) or lengths != sorted(lengths):
             raise ValueError(Errors.E1049.format(label=label))

spacy/tests/doc/test_doc_api.py

@@ -3,7 +3,7 @@ from pickle import EMPTY_DICT
 import weakref
 
 import numpy
-import ctypes
+from time import time
 from numpy.testing import assert_array_equal
 from murmurhash.mrmr import hash
 import pytest
@@ -1438,8 +1438,13 @@ def _encode_and_hash(input: str, *, reverse: bool = False) -> int:
 @pytest.mark.parametrize("case_sensitive", [True, False])
 def test_get_character_combination_hashes_good_case(en_tokenizer, case_sensitive):
     doc = en_tokenizer("spaCy✨ and Prodigy")
-    ps1, ps2, ps3, ps4 = get_search_char_byte_arrays("Rp", case_sensitive)
+
-    ss1, ss2, ss3, ss4 = get_search_char_byte_arrays("xx✨rp", case_sensitive)
+    ps_search_chars, ps_width_offsets = get_search_char_byte_arrays(
+        "Rp", case_sensitive
+    )
+    ss_search_chars, ss_width_offsets = get_search_char_byte_arrays(
+        "xx✨rp", case_sensitive
+    )
     hashes = doc.get_character_combination_hashes(
         cs=case_sensitive,
         p_lengths=bytes(
@@ -1449,7 +1454,6 @@ def test_get_character_combination_hashes_good_case(en_tokenizer, case_sensitive
                 4,
             )
         ),
-        p_max_l=4,
         s_lengths=bytes(
             (
                 2,
@@ -1458,35 +1462,18 @@ def test_get_character_combination_hashes_good_case(en_tokenizer, case_sensitive
                 5,
             )
         ),
-        s_max_l=5,
+        ps_search_chars=ps_search_chars,
-        ps_1byte_ch=ps1,
+        ps_width_offsets=ps_width_offsets,
-        ps_1byte_ch_l=len(ps1),
-        ps_2byte_ch=ps2,
-        ps_2byte_ch_l=len(ps2),
-        ps_3byte_ch=ps3,
-        ps_3byte_ch_l=len(ps3),
-        ps_4byte_ch=ps4,
-        ps_4byte_ch_l=len(ps4),
         ps_lengths=bytes((2,)),
-        ps_max_l=2,
+        ss_search_chars=ss_search_chars,
-        ss_1byte_ch=ss1,
+        ss_width_offsets=ss_width_offsets,
-        ss_1byte_ch_l=len(ss1),
-        ss_2byte_ch=ss2,
-        ss_2byte_ch_l=len(ss2),
-        ss_3byte_ch=ss3,
-        ss_3byte_ch_l=len(ss3),
-        ss_4byte_ch=ss4,
-        ss_4byte_ch_l=len(ss4),
         ss_lengths=bytes(
             (
                 1,
                 2,
             )
         ),
-        ss_max_l=2,
-        hashes_per_tok=10,
     )
-
     assert hashes[0][0] == _encode_and_hash("s")
     assert hashes[0][1] == _encode_and_hash("spa")
     assert hashes[0][2] == _encode_and_hash("spaC" if case_sensitive else "spac")
@@ -1494,7 +1481,6 @@ def test_get_character_combination_hashes_good_case(en_tokenizer, case_sensitive
     assert hashes[0][4] == _encode_and_hash("yCa" if case_sensitive else "yca")
     assert hashes[0][5] == _encode_and_hash("yCap" if case_sensitive else "ycap")
     assert hashes[0][6] == _encode_and_hash("yCaps" if case_sensitive else "ycaps")
-
     assert hashes[0][7] == _encode_and_hash("p")
     assert hashes[0][8] == _encode_and_hash("p")
     assert hashes[0][9] == _encode_and_hash("p")
@@ -1539,11 +1525,10 @@ def test_get_character_combination_hashes_good_case(en_tokenizer, case_sensitive
 
 def test_get_character_combination_hashes_good_case_partial(en_tokenizer):
     doc = en_tokenizer("spaCy✨ and Prodigy")
-    ps1, ps2, ps3, ps4 = get_search_char_byte_arrays("rp", False)
+    ps_search_chars, ps_width_offsets = get_search_char_byte_arrays("rp", False)
     hashes = doc.get_character_combination_hashes(
         cs=False,
         p_lengths=bytes(),
-        p_max_l=0,
         s_lengths=bytes(
             (
                 2,
@@ -1552,28 +1537,12 @@ def test_get_character_combination_hashes_good_case_partial(en_tokenizer):
                 5,
             )
         ),
-        s_max_l=5,
+        ps_search_chars=ps_search_chars,
-        ps_1byte_ch=ps1,
+        ps_width_offsets=ps_width_offsets,
-        ps_1byte_ch_l=len(ps1),
-        ps_2byte_ch=ps2,
-        ps_2byte_ch_l=len(ps2),
-        ps_3byte_ch=ps3,
-        ps_3byte_ch_l=len(ps3),
-        ps_4byte_ch=ps4,
-        ps_4byte_ch_l=len(ps4),
         ps_lengths=bytes((2,)),
-        ps_max_l=2,
+        ss_search_chars=bytes(),
-        ss_1byte_ch=bytes(),
+        ss_width_offsets=bytes(),
-        ss_1byte_ch_l=0,
-        ss_2byte_ch=bytes(),
-        ss_2byte_ch_l=0,
-        ss_3byte_ch=bytes(),
-        ss_3byte_ch_l=0,
-        ss_4byte_ch=bytes(),
-        ss_4byte_ch_l=0,
         ss_lengths=bytes(),
-        ss_max_l=0,
-        hashes_per_tok=5,
     )
 
     assert hashes[0][0] == _encode_and_hash("yc")
@@ -1607,30 +1576,13 @@ def test_get_character_combination_hashes_copying_in_middle(en_tokenizer):
             hashes = doc.get_character_combination_hashes(
                 cs=False,
                 p_lengths=bytes((p_length,)),
-                p_max_l=p_length,
                 s_lengths=bytes((s_length,)),
-                s_max_l=s_length,
+                ps_search_chars=bytes(),
-                ps_1byte_ch=bytes(),
+                ps_width_offsets=bytes(),
-                ps_1byte_ch_l=0,
-                ps_2byte_ch=bytes(),
-                ps_2byte_ch_l=0,
-                ps_3byte_ch=bytes(),
-                ps_3byte_ch_l=0,
-                ps_4byte_ch=bytes(),
-                ps_4byte_ch_l=0,
                 ps_lengths=bytes(),
-                ps_max_l=0,
+                ss_search_chars=bytes(),
-                ss_1byte_ch=bytes(),
+                ss_width_offsets=bytes(),
-                ss_1byte_ch_l=0,
-                ss_2byte_ch=bytes(),
-                ss_2byte_ch_l=0,
-                ss_3byte_ch=bytes(),
-                ss_3byte_ch_l=0,
-                ss_4byte_ch=bytes(),
-                ss_4byte_ch_l=0,
                 ss_lengths=bytes(),
-                ss_max_l=0,
-                hashes_per_tok=2,
             )
 
             assert hashes[0][0] == _encode_and_hash("sp𐌞cé"[:p_length])
@@ -1642,7 +1594,7 @@ def test_get_character_combination_hashes_turkish_i_with_dot(
     en_tokenizer, case_sensitive
 ):
     doc = en_tokenizer("İ".lower() + "İ")
-    s1, s2, s3, s4 = get_search_char_byte_arrays("İ", case_sensitive)
+    search_chars, width_offsets = get_search_char_byte_arrays("İ", case_sensitive)
     hashes = doc.get_character_combination_hashes(
         cs=case_sensitive,
         p_lengths=bytes(
@@ -1653,7 +1605,6 @@ def test_get_character_combination_hashes_turkish_i_with_dot(
                 4,
             )
         ),
-        p_max_l=4,
         s_lengths=bytes(
             (
                 1,
@@ -1662,15 +1613,8 @@ def test_get_character_combination_hashes_turkish_i_with_dot(
                 4,
             )
         ),
-        s_max_l=4,
+        ps_search_chars=search_chars,
-        ps_1byte_ch=s1,
+        ps_width_offsets=width_offsets,
-        ps_1byte_ch_l=len(s1),
-        ps_2byte_ch=s2,
-        ps_2byte_ch_l=len(s2),
-        ps_3byte_ch=s3,
-        ps_3byte_ch_l=len(s3),
-        ps_4byte_ch=s4,
-        ps_4byte_ch_l=len(s4),
         ps_lengths=bytes(
             (
                 1,
@@ -1679,15 +1623,8 @@ def test_get_character_combination_hashes_turkish_i_with_dot(
                 4,
             )
         ),
-        ps_max_l=4,
+        ss_search_chars=search_chars,
-        ss_1byte_ch=s1,
+        ss_width_offsets=width_offsets,
-        ss_1byte_ch_l=len(s1),
-        ss_2byte_ch=s2,
-        ss_2byte_ch_l=len(s2),
-        ss_3byte_ch=s3,
-        ss_3byte_ch_l=len(s3),
-        ss_4byte_ch=s4,
-        ss_4byte_ch_l=len(s4),
         ss_lengths=bytes(
             (
                 1,
@@ -1696,8 +1633,6 @@ def test_get_character_combination_hashes_turkish_i_with_dot(
                 4,
             )
         ),
-        ss_max_l=4,
-        hashes_per_tok=16,
     )
 
     COMBINING_DOT_ABOVE = b"\xcc\x87".decode("UTF-8")
@@ -1747,34 +1682,17 @@ def test_get_character_combination_hashes_string_store_spec_cases(
     assert len(long_word) > 255
     doc = en_tokenizer(" ".join((symbol, short_word, normal_word, long_word)))
     assert len(doc) == 4
-    ps1, ps2, ps3, ps4 = get_search_char_byte_arrays("E", case_sensitive)
+    ps_search_chars, ps_width_offsets = get_search_char_byte_arrays("E", case_sensitive)
     hashes = doc.get_character_combination_hashes(
         cs=case_sensitive,
         p_lengths=bytes((2,)),
-        p_max_l=2,
         s_lengths=bytes((2,)),
-        s_max_l=2,
+        ps_search_chars=ps_search_chars,
-        ps_1byte_ch=ps1,
+        ps_width_offsets=ps_width_offsets,
-        ps_1byte_ch_l=len(ps1),
-        ps_2byte_ch=ps2,
-        ps_2byte_ch_l=len(ps2),
-        ps_3byte_ch=ps3,
-        ps_3byte_ch_l=len(ps3),
-        ps_4byte_ch=ps4,
-        ps_4byte_ch_l=len(ps4),
         ps_lengths=bytes((2,)),
-        ps_max_l=2,
+        ss_search_chars=bytes(),
-        ss_1byte_ch=bytes(),
+        ss_width_offsets=bytes(),
-        ss_1byte_ch_l=0,
-        ss_2byte_ch=bytes(),
-        ss_2byte_ch_l=0,
-        ss_3byte_ch=bytes(),
-        ss_3byte_ch_l=0,
-        ss_4byte_ch=bytes(),
-        ss_4byte_ch_l=0,
         ss_lengths=bytes(),
-        ss_max_l=0,
-        hashes_per_tok=3,
     )
     assert hashes[0][0] == _encode_and_hash("FL" if case_sensitive else "fl")
     assert hashes[0][1] == _encode_and_hash("91")
@@ -1799,30 +1717,13 @@ def test_character_combination_hashes_empty_lengths(en_tokenizer):
         doc.get_character_combination_hashes(
             cs=True,
             p_lengths=bytes(),
-            p_max_l=0,
             s_lengths=bytes(),
-            s_max_l=0,
+            ps_search_chars=bytes(),
-            ps_1byte_ch=bytes(),
+            ps_width_offsets=bytes(),
-            ps_1byte_ch_l=0,
-            ps_2byte_ch=bytes(),
-            ps_2byte_ch_l=0,
-            ps_3byte_ch=bytes(),
-            ps_3byte_ch_l=0,
-            ps_4byte_ch=bytes(),
-            ps_4byte_ch_l=0,
             ps_lengths=bytes(),
-            ps_max_l=0,
+            ss_search_chars=bytes(),
-            ss_1byte_ch=bytes(),
+            ss_width_offsets=bytes(),
-            ss_1byte_ch_l=0,
-            ss_2byte_ch=bytes(),
-            ss_2byte_ch_l=0,
-            ss_3byte_ch=bytes(),
-            ss_3byte_ch_l=0,
-            ss_4byte_ch=bytes(),
-            ss_4byte_ch_l=0,
             ss_lengths=bytes(),
-            ss_max_l=0,
-            hashes_per_tok=0,
         ).shape
         == (1, 0)
     )

spacy/tests/test_util.py

@@ -36,6 +36,30 @@ def test_get_search_char_byte_arrays_all_widths(case_sensitive):
         assert width_offsets == b"\x00\x02\x04\x07\x0b"
 
 
+@pytest.mark.parametrize("case_sensitive", [True, False])
+def test_get_search_char_byte_arrays_widths_1_and_3(case_sensitive):
+    search_chars, width_offsets = spacy.util.get_search_char_byte_arrays(
+        "B—", case_sensitive
+    )
+    if case_sensitive:
+        assert search_chars == "B—".encode("utf-8")
+    else:
+        assert search_chars == "b—".encode("utf-8")
+    assert width_offsets == b"\x00\x01\x01\x04\x04"
+
+
+@pytest.mark.parametrize("case_sensitive", [True, False])
+def test_get_search_char_byte_arrays_widths_1_and_4(case_sensitive):
+    search_chars, width_offsets = spacy.util.get_search_char_byte_arrays(
+        "B𐌞", case_sensitive
+    )
+    if case_sensitive:
+        assert search_chars == "B𐌞".encode("utf-8")
+    else:
+        assert search_chars == "b𐌞".encode("utf-8")
+    assert width_offsets == b"\x00\x01\x01\x01\x05"
+
+
 @pytest.mark.parametrize("case_sensitive", [True, False])
 def test_turkish_i_with_dot(case_sensitive):
     search_chars, width_offsets = spacy.util.get_search_char_byte_arrays(

spacy/tokens/doc.pxd

@@ -43,34 +43,28 @@ cdef int [:,:] _get_lca_matrix(Doc, int start, int end)
 cdef void _set_prefix_lengths(
     const unsigned char* tok_str,
     const int tok_str_l,
-    unsigned char* pref_l_buf,
     const int p_max_l, 
+    unsigned char* pref_l_buf,
 ) nogil
 
 
 cdef void _set_suffix_lengths(
     const unsigned char* tok_str,
     const int tok_str_l,
-    unsigned char* suff_l_buf,
     const int s_max_l, 
+    unsigned char* suff_l_buf,
 ) nogil
 
 
 cdef void _search_for_chars(
     const unsigned char* tok_str,
     const int tok_str_l,
-    const unsigned char* s_1byte_ch,
+    const unsigned char* search_chars,
-    const int s_1byte_ch_l,
+    const unsigned char* width_offsets,
-    const unsigned char* s_2byte_ch,
+    const int max_res_l,
-    const int s_2byte_ch_l,
+    const bint suffs_not_prefs,
-    const unsigned char* s_3byte_ch,
-    const int s_3byte_ch_l,
-    const unsigned char* s_4byte_ch,
-    const int s_4byte_ch_l,
     unsigned char* res_buf,
-    int max_res_l,
     unsigned char* l_buf,
-    bint suffs_not_prefs
 ) nogil
 
 

spacy/tokens/doc.pyx

@@ -1739,30 +1739,13 @@ cdef class Doc:
         *,
         const bint cs, 
         const unsigned char* p_lengths,
-        const int p_max_l,
         const unsigned char* s_lengths,
-        const int s_max_l,
+        const unsigned char* ps_search_chars,
-        const unsigned char* ps_1byte_ch,
+        const unsigned char* ps_width_offsets,
-        const int ps_1byte_ch_l,
-        const unsigned char* ps_2byte_ch,
-        const int ps_2byte_ch_l,
-        const unsigned char* ps_3byte_ch,
-        const int ps_3byte_ch_l,
-        const unsigned char* ps_4byte_ch,
-        const int ps_4byte_ch_l,
         const unsigned char* ps_lengths,
-        const int ps_max_l,
+        const unsigned char* ss_search_chars,
-        const unsigned char* ss_1byte_ch,
+        const unsigned char* ss_width_offsets,
-        const int ss_1byte_ch_l,
-        const unsigned char* ss_2byte_ch,
-        const int ss_2byte_ch_l,
-        const unsigned char* ss_3byte_ch,
-        const int ss_3byte_ch_l,
-        const unsigned char* ss_4byte_ch,
-        const int ss_4byte_ch_l,
         const unsigned char* ss_lengths,
-        const int ss_max_l,
-        const int hashes_per_tok
     ):
         """
         Returns a 2D NumPy array where the rows represent tokens and the columns represent hashes of various character combinations 
@@ -1778,31 +1761,43 @@ cdef class Doc:
         cs: if *False*, hashes are generated based on the lower-case version of each token.
         p_lengths: an array of single-byte values specifying the lengths of prefixes to be hashed in ascending order. 
             For example, if *p_lengths==[2, 3]*, the prefixes hashed for "spaCy" would be "sp" and "spa".
-        p_max_l: the value of *p_lengths[-1]*, or *0* if *p_lengths==None*. Passed in for speed.
         s_lengths: an array of single-byte values specifying the lengths of suffixes to be hashed in ascending order. 
-            For example, if *s_lengths==[2, 3]* and *cs == True*, the suffixes hashed for "spaCy" would be "Cy" and "aCy".
+            For example, if *s_lengths==[2, 3]* and *cs == True*, the suffixes hashed for "spaCy" would be "yC" and "yCa".
-        s_max_l: the value of *s_lengths[-1]*, or *0* if *s_lengths==None*. Passed in for speed.
+        ps_search_chars: a byte array containing, in numerical order, UTF-8 characters to search for within each token, 
-        ps_<n>byte_ch: a byte array containing in order n-byte-wide characters to search for within each token, 
             starting at the beginning.
-        ps_<n>byte_ch_l: the length of *ps_<n>byte_ch*. Passed in for speed.
+        ps_width_offsets: an array of single-byte values [1-char-start, 2-char-start, 3-char-start, 4-char-start, 4-char-end]
+            specifying the offsets within *ps_search_chars* that contain UTF-8 characters with the specified widths.
         ps_lengths: an array of single-byte values specifying the lengths of search results (from the beginning) to be hashed 
             in ascending order. For example, if *ps_lengths==[1, 2]*, *ps_search=="aC" and *cs==False*, the searched strings 
             hashed for "spaCy" would be "a" and "ac".
-        ps_max_l: the value of *ps_lengths[-1]*, or *0* if *ps_lengths==None*. Passed in for speed.
+        ss_search_chars: a byte array containing, in numerical order, UTF-8 characters to search for within each token, 
-        ss_<n>byte_ch: a byte array containing in order n-byte-wide characters to search for within each token, 
             starting at the end.
-        ss_<n>byte_ch_l: the length of *ss_<n>byte_ch*. Passed in for speed.
+        ss_width_offsets: an array of single-byte values [1-char-start, 2-char-start, 3-char-start, 4-char-start, 4-char-end]
+            specifying the offsets within *ss_search_chars* that contain UTF-8 characters with the specified widths.
         ss_lengths: an array of single-byte values specifying the lengths of search results (from the end) to be hashed
              in ascending order. For example, if *ss_lengths==[1, 2]*, *ss_search=="aC" and *cs==False*, the searched strings 
              hashed for "spaCy" would be "c" and "ca".
-        ss_max_l: the value of *ss_lengths[-1]*, or *0* if *ss_lengths==None*. Passed in for speed.
+    
-        hashes_per_tok: the total number of hashes produced for each token. Passed in for speed.
+        Many of the buffers passed into and used by this method contain single-byte numerical values. This takes advantage of 
+        the fact that we are hashing short affixes and searching for small groups of characters; the calling code is responsible
+        lengths being passed in cannot exceed 63 and that *_search_chars buffers are never longer than 255.
         """
 
+        # Work out lengths
+        cdef int p_lengths_l = strlen(<char*> p_lengths)
+        cdef int s_lengths_l = strlen(<char*> s_lengths)
+        cdef int ps_lengths_l = strlen(<char*> ps_lengths)
+        cdef int ss_lengths_l = strlen(<char*> ss_lengths)
+        cdef int hashes_per_tok = p_lengths_l + s_lengths_l + ps_lengths_l + ss_lengths_l
+        cdef int p_max_l = p_lengths[p_lengths_l - 1] if p_lengths_l > 0 else 0
+        cdef int s_max_l = s_lengths[s_lengths_l - 1] if s_lengths_l > 0 else 0
+        cdef int ps_max_l = ps_lengths[ps_lengths_l - 1] if ps_lengths_l > 0 else 0
+        cdef int ss_max_l = ss_lengths[ss_lengths_l - 1] if ss_lengths_l > 0 else 0
+        
         # Define / allocate buffers
         cdef Pool mem = Pool()
         cdef unsigned char* pref_l_buf = <unsigned char*> mem.alloc(p_max_l, 1)
-        cdef unsigned char* suff_l_buf = <unsigned char*> mem.alloc(p_max_l, 1)
+        cdef unsigned char* suff_l_buf = <unsigned char*> mem.alloc(s_max_l, 1)
         cdef unsigned char* ps_res_buf = <unsigned char*> mem.alloc(ps_max_l, 4)
         cdef unsigned char* ps_l_buf = <unsigned char*> mem.alloc(ps_max_l, 1)
         cdef unsigned char* ss_res_buf = <unsigned char*> mem.alloc(ss_max_l, 4)
@@ -1813,7 +1808,7 @@ cdef class Doc:
          
         # Define working variables
         cdef TokenC tok_c
-        cdef int hash_idx, tok_i, tok_str_l
+        cdef int tok_i, tok_str_l
         cdef attr_t num_tok_attr
         cdef const unsigned char* tok_str
         cdef np.uint32_t* w_hashes_ptr = hashes_ptr
@@ -1825,21 +1820,21 @@ cdef class Doc:
             tok_str_l = strlen(<char*> tok_str)
             
             if p_max_l > 0:
-                _set_prefix_lengths(tok_str, tok_str_l, pref_l_buf, p_max_l)
+                _set_prefix_lengths(tok_str, tok_str_l, p_max_l, pref_l_buf)
                 w_hashes_ptr += _write_hashes(tok_str, p_lengths, pref_l_buf, 0, w_hashes_ptr)
 
             if s_max_l > 0:
-                _set_suffix_lengths(tok_str, tok_str_l, suff_l_buf, s_max_l)
+                _set_suffix_lengths(tok_str, tok_str_l, s_max_l, suff_l_buf)
                 w_hashes_ptr += _write_hashes(tok_str, s_lengths, suff_l_buf, tok_str_l - 1, w_hashes_ptr)
             
             if ps_max_l > 0:
-                _search_for_chars(tok_str, tok_str_l, ps_1byte_ch, ps_1byte_ch_l, ps_2byte_ch, ps_2byte_ch_l, 
+                _search_for_chars(tok_str, tok_str_l, ps_search_chars, ps_width_offsets, 
-                    ps_3byte_ch, ps_3byte_ch_l, ps_4byte_ch, ps_4byte_ch_l, ps_res_buf, ps_max_l, ps_l_buf, False)
+                    ps_max_l, False, ps_res_buf, ps_l_buf)
                 w_hashes_ptr += _write_hashes(ps_res_buf, ps_lengths, ps_l_buf, 0, w_hashes_ptr)
 
             if ss_max_l > 0:
-                _search_for_chars(tok_str, tok_str_l, ss_1byte_ch, ss_1byte_ch_l, ss_2byte_ch, ss_2byte_ch_l, 
+                _search_for_chars(tok_str, tok_str_l, ss_search_chars, ss_width_offsets, 
-                    ss_3byte_ch, ss_3byte_ch_l, ss_4byte_ch, ss_4byte_ch_l, ss_res_buf, ss_max_l, ss_l_buf, True)
+                    ss_max_l, True, ss_res_buf, ss_l_buf)
                 w_hashes_ptr += _write_hashes(ss_res_buf, ss_lengths, ss_l_buf, 0, w_hashes_ptr)
         
         cdef np.ndarray[np.uint32_t, ndim=2] hashes = numpy.empty(
@@ -2033,17 +2028,17 @@ cdef int [:,:] _get_lca_matrix(Doc doc, int start, int end):
 cdef void _set_prefix_lengths(
     const unsigned char* tok_str,
     const int tok_str_l,
-    unsigned char* pref_l_buf,
     const int p_max_l, 
+    unsigned char* pref_l_buf,
 ) nogil:
     """ Populate *pref_l_buf*, which has length *pref_l*, with the byte lengths of the first *pref_l* characters within *tok_str*. 
         Lengths that are greater than the character length of the whole word are populated with the byte length of the whole word.
 
         tok_str: a UTF-8 representation of a string.
         tok_str_l: the length of *tok_str*.
+        p_max_l: the number of characters to process at the beginning of the word.
         pref_l_buf: a buffer of length *p_max_l* in which to store the lengths. The calling code ensures that lengths
             greater than 255 cannot occur.
-        p_max_l: the number of characters to process at the beginning of the word.
     """
     cdef int tok_str_idx = 1, pref_l_buf_idx = 0
 
@@ -2066,17 +2061,17 @@ cdef void _set_prefix_lengths(
 cdef void _set_suffix_lengths(
     const unsigned char* tok_str,
     const int tok_str_l,
+    const int s_max_l,
     unsigned char* suff_l_buf,
-    const int s_max_l, 
 ) nogil:
     """ Populate *suff_l_buf*, which has length *suff_l*, with the byte lengths of the last *suff_l* characters within *tok_str*. 
         Lengths that are greater than the character length of the whole word are populated with the byte length of the whole word.
 
         tok_str: a UTF-8 representation of a string.
         tok_str_l: the length of *tok_str*.
+        s_max_l: the number of characters to process at the end of the word.
         suff_l_buf: a buffer of length *s_max_l* in which to store the lengths. The calling code ensures that lengths
             greater than 255 cannot occur.
-        s_max_l: the number of characters to process at the end of the word.
     """
     cdef int tok_str_idx = tok_str_l - 1, suff_l_buf_idx = 0
 
@@ -2096,67 +2091,48 @@ cdef void _set_suffix_lengths(
 cdef void _search_for_chars(
     const unsigned char* tok_str,
     const int tok_str_l,
-    const unsigned char* s_1byte_ch,
+    const unsigned char* search_chars,
-    const int s_1byte_ch_l,
+    const unsigned char* width_offsets,
-    const unsigned char* s_2byte_ch,
+    const int max_res_l,
-    const int s_2byte_ch_l,
+    const bint suffs_not_prefs,
-    const unsigned char* s_3byte_ch,
-    const int s_3byte_ch_l,
-    const unsigned char* s_4byte_ch,
-    const int s_4byte_ch_l,
     unsigned char* res_buf,
-    int max_res_l,
     unsigned char* l_buf,
-    bint suffs_not_prefs
 ) nogil:
-    """ Search *tok_str* within a string for characters within the *s_<n>byte_ch> buffers, starting at the 
+    """ Search *tok_str* within a string for characters within *search_chars*, starting at the 
         beginning or end depending on the value of *suffs_not_prefs*. Wherever a character matches,
-        it is added to *res_buf* and the byte length up to that point is added to *len_buf*. When nothing
+        it is added to *res_buf* and the byte length up to that point is added to *l_buf*. When nothing
-        more is found, the remainder of *len_buf* is populated wth the byte length from the last result,
+        more is found, the remainder of *l_buf* is populated wth the byte length from the last result,
         which may be *0* if the search was not successful.
 
         tok_str: a UTF-8 representation of a string.
         tok_str_l: the length of *tok_str*.
-        s_<n>byte_ch: a byte array containing in order n-byte-wide characters to search for.
+        search_chars: a byte array containing, in numerical order, UTF-8 characters to search for within *tok_str*. 
-        res_buf: the buffer in which to place the search results.
+        width_offsets: an array of single-byte values [1-char-start, 2-char-start, 3-char-start, 4-char-start, 4-char-end]
+            specifying the offsets within *search_chars* that contain UTF-8 characters with the specified widths.
         max_res_l: the maximum number of found characters to place in *res_buf*.
-        l_buf: a buffer of length *max_res_l* in which to store the byte lengths.
-            The calling code ensures that lengths greater than 255 cannot occur. 
         suffs_not_prefs: if *True*, searching starts from the end of the word; 
             if *False*, from the beginning.
+        res_buf: the buffer in which to place the search results.
+        l_buf: a buffer of length *max_res_l* in which to store the byte lengths.
+            The calling code ensures that lengths greater than 255 cannot occur. 
     """
     cdef int res_buf_idx = 0, l_buf_idx = 0, ch_wdth, tok_start_idx, search_char_idx
-    cdef int search_chars_l
-    cdef const unsigned char* search_chars
-
     cdef int last_tok_str_idx = tok_str_l if suffs_not_prefs else 0
     cdef int this_tok_str_idx = tok_str_l - 1 if suffs_not_prefs else 1
     
-    while True:
+    while this_tok_str_idx >= 0 and this_tok_str_idx <= tok_str_l:
-        if (
+        if ( 
-            this_tok_str_idx == tok_str_l or 
+            (this_tok_str_idx == tok_str_l) or 
-            (tok_str[this_tok_str_idx] & 0xc0) != 0x80 # not continuation character, always applies to [0].
+            ((tok_str[this_tok_str_idx] & 0xc0) != 0x80) # not continuation character, always applies to [0].
         ):
             if this_tok_str_idx > last_tok_str_idx:
                 ch_wdth = this_tok_str_idx - last_tok_str_idx
             else:
                 ch_wdth = last_tok_str_idx - this_tok_str_idx
-            if ch_wdth == 1:
+            
-                search_chars = s_1byte_ch
-                search_chars_l = s_1byte_ch_l
-            elif ch_wdth == 2:
-                search_chars = s_2byte_ch
-                search_chars_l = s_2byte_ch_l
-            elif ch_wdth == 3:
-                search_chars = s_3byte_ch
-                search_chars_l = s_3byte_ch_l
-            else:
-                search_chars = s_4byte_ch
-                search_chars_l = s_4byte_ch_l
             tok_start_idx = this_tok_str_idx if suffs_not_prefs else last_tok_str_idx
-
+            search_char_idx = width_offsets[ch_wdth - 1]
-            search_char_idx = 0
+            while search_char_idx < width_offsets[ch_wdth]:
-            while search_char_idx < search_chars_l:
                 cmp_result = memcmp(&tok_str[tok_start_idx], &search_chars[search_char_idx], ch_wdth)
                 if cmp_result == 0:
                     memcpy(res_buf + res_buf_idx, &search_chars[search_char_idx], ch_wdth)
@@ -2171,12 +2147,8 @@ cdef void _search_for_chars(
             last_tok_str_idx = this_tok_str_idx
         if suffs_not_prefs:
             this_tok_str_idx -= 1
-            if this_tok_str_idx < 0:
-                break
         else:
             this_tok_str_idx += 1
-            if this_tok_str_idx > tok_str_l:
-                break    
 
     # fill in unused characters in the length buffer
     memset(l_buf + l_buf_idx, res_buf_idx, max_res_l - l_buf_idx)

spacy/util.py

@@ -1757,10 +1757,10 @@ def get_search_char_byte_arrays(
         search_char_string = search_char_string.lower()
     ordered_search_char_string = "".join(sorted(set(search_char_string)))
     search_chars = ordered_search_char_string.encode("UTF-8")
-    width_offsets = [0, -1, -1, -1, -1]
+    width_offsets = [-1] * 5
-    working_start = -1
+    working_start = 0
-    working_width = 1
+    working_width = 0
-    for idx in range(len(search_chars) + 1):
+    for idx in range(1, len(search_chars) + 1):
         if (
             idx == len(search_chars)
             or search_chars[idx] & 0xC0 != 0x80  # not continuation byte
@@ -1769,11 +1769,10 @@ def get_search_char_byte_arrays(
             if this_width > 4 or this_width < working_width:
                 raise RuntimeError(Errors.E1050)
             if this_width > working_width:
-                width_offsets[this_width - 1] = working_start
+                for i in range(working_width, 5):
+                    width_offsets[i] = working_start
                 working_width = this_width
             working_start = idx
-    width_offsets[this_width] = idx
+    for i in range(this_width, 5):
-    for i in range(5):
+        width_offsets[i] = idx
-        if width_offsets[i] == -1:
-            width_offsets[i] = width_offsets[i - 1]
     return search_chars, bytes((width_offsets))