From b66112cb90d7223a7de7c2f61c49c6d71a071334 Mon Sep 17 00:00:00 2001
From: richardpaulhudson <richard@explosion.ai>
Date: Fri, 9 Sep 2022 21:15:38 +0200
Subject: [PATCH] Intermediate state

---
 spacy/tokens/doc.pxd |  2 ++
 spacy/tokens/doc.pyx | 42 ++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 44 insertions(+)

diff --git a/spacy/tokens/doc.pxd b/spacy/tokens/doc.pxd
index 57d087958..ae5e945f5 100644
--- a/spacy/tokens/doc.pxd
+++ b/spacy/tokens/doc.pxd
@@ -9,6 +9,7 @@ from ..attrs cimport attr_id_t
 
 cdef attr_t get_token_attr(const TokenC* token, attr_id_t feat_name) nogil
 cdef attr_t get_token_attr_for_matcher(const TokenC* token, attr_id_t feat_name) nogil
+cdef np.ndarray init_array(int num_tokens, int length)
 
 
 ctypedef const LexemeC* const_Lexeme_ptr
@@ -69,3 +70,4 @@ cdef class Doc:
     cdef int push_back(self, LexemeOrToken lex_or_tok, bint has_space) except -1
 
     cpdef np.ndarray to_array(self, object features)
+
diff --git a/spacy/tokens/doc.pyx b/spacy/tokens/doc.pyx
index 7ba9a3341..987690dc3 100644
--- a/spacy/tokens/doc.pyx
+++ b/spacy/tokens/doc.pyx
@@ -94,6 +94,12 @@ cdef attr_t get_token_attr_for_matcher(const TokenC* token, attr_id_t feat_name)
         return get_token_attr(token, feat_name)
 
 
+cdef np.ndarray init_array(int num_tokens, int length):
+    cdef np.ndarray output = numpy.zeros((num_tokens, length), dtype='uint8')
+    output.fill(255)
+    return output
+
+
 class SetEntsDefault(str, Enum):
     blocked = "blocked"
     missing = "missing"
@@ -1734,6 +1740,42 @@ cdef class Doc:
                     j += 1
         return output
 
+    def get_suffixes(self, int min_length, int max_length, special_chars:str, int sc_min_length, int sc_max_length):
+        """
+        TODO
+        """
+        byte_strings = [token.orth_.encode('utf8') for token in self]
+        special_chars_enc = special_chars.encode('utf8')
+        cdef num_tokens = len(byte_strings)
+        outputs = []        
+        for length in range(min_length, max_length+1):
+            outputs.append(init_array(num_tokens, length))
+        for length in range(sc_min_length, sc_max_length+1):
+            outputs.append(init_array(num_tokens, length))
+        
+        cdef int token_i, sc_char_i, idx
+        cdef bytes byte_string
+        cdef unsigned char utf8_char
+        cdef num_normal_arr = 1 + max_length - min_length
+        cdef num_sc_arr = 1 + sc_max_length - sc_min_length
+        for token_i, byte_string in enumerate(byte_strings):
+            sc_char_i = 0
+            idx = 0
+            while (idx < max_length or sc_char_i < sc_max_length) and idx < len(byte_string):
+                this_char = <unsigned char>byte_string[len(byte_string) - (1 + idx)]
+                for normal_arr_i in range(num_normal_arr - 1, -1, -1):
+                    if idx >= normal_arr_i + min_length:
+                        break
+                    outputs[normal_arr_i][token_i, idx] = this_char
+                if this_char in special_chars_enc:
+                    for sc_arr_i in range(num_sc_arr - 1, -1, -1):
+                        if sc_char_i >= sc_arr_i + sc_min_length:
+                            break
+                        outputs[sc_arr_i + num_normal_arr][token_i, sc_char_i] = this_char
+                    sc_char_i += 1
+                idx += 1
+        return outputs
+
     @staticmethod
     def _get_array_attrs():
         attrs = [LENGTH, SPACY]