First commit

FastChemTokenizer.py

@@ -0,0 +1,322 @@
+import torch
+import json
+import os
+from typing import List, Union, Optional, Tuple
+from transformers.tokenization_utils_base import BatchEncoding
+from functools import lru_cache
+
+# Copyright 2025 Genta Pramillean Bayu (@gbyuvd)
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+class TrieNode:
+    __slots__ = ['children', 'token_id']
+    def __init__(self):
+        self.children = {}
+        self.token_id = None  # If set, this node completes a valid token
+
+
+class FastChemTokenizer:
+    def __init__(self, token_to_id, model_max_length=512):
+        self.token_to_id = token_to_id
+        self.id_to_token = {v: k for k, v in token_to_id.items()}
+        # No more self.token_set — replaced by trie
+        self.model_max_length = model_max_length
+
+        # Precompute max token length for possible use & clarity
+        self.max_token_len = max(len(t) for t in token_to_id.keys())
+
+        # Build trie for fast longest-match lookup
+        self.trie_root = self._build_trie(token_to_id)
+
+        # Validate required special tokens
+        required_special_tokens = ["<s>", "</s>", "<pad>", "<unk>", "<mask>"]
+        for tok in required_special_tokens:
+            if tok not in token_to_id:
+                raise KeyError(f"Required special token '{tok}' not found in vocab.")
+
+        # Special token IDs
+        self.bos_token_id = token_to_id["<s>"]
+        self.eos_token_id = token_to_id["</s>"]
+        self.pad_token_id = token_to_id["<pad>"]
+        self.unk_token_id = token_to_id["<unk>"]
+        self.mask_token_id = token_to_id["<mask>"]
+
+        # Special tokens for convenience
+        self.bos_token = "<s>"
+        self.eos_token = "</s>"
+        self.pad_token = "<pad>"
+        self.unk_token = "<unk>"
+        self.mask_token = "<mask>"
+
+    def _build_trie(self, token_to_id):
+        root = TrieNode()
+        for token, tid in token_to_id.items():
+            node = root
+            for char in token:
+                if char not in node.children:
+                    node.children[char] = TrieNode()
+                node = node.children[char]
+            node.token_id = tid
+        return root
+
+    def __len__(self):
+        """Return vocab size — REQUIRED for HF compatibility."""
+        return len(self.token_to_id)
+
+    def __call__(self, text: Union[str, List[str]], text_pair: Optional[Union[str, List[str]]] = None, **kwargs) -> BatchEncoding:
+        if isinstance(text, list):
+            batch = [(t, p) if p is not None else t for t, p in zip(text, text_pair)] if text_pair else text
+            return self.batch_encode_plus(batch, **kwargs)
+        else:
+            return self.encode_plus(text=text, text_pair=text_pair, **kwargs)
+
+    @lru_cache(maxsize=10000)
+    def _cached_encode_str(self, s: str) -> Tuple[int, ...]:
+        return tuple(self._encode_core(s))
+
+    def _encode_core(self, text: str) -> List[int]:
+        """Core encoding logic using Trie — no caching."""
+        tokens = text
+        result_ids = []
+        i = 0
+        n = len(tokens)
+
+        while i < n:
+            node = self.trie_root
+            j = i
+            last_match_id = None
+            last_match_end = i
+
+            # Traverse trie while characters match
+            while j < n and tokens[j] in node.children:
+                node = node.children[tokens[j]]
+                j += 1
+                if node.token_id is not None:
+                    last_match_id = node.token_id
+                    last_match_end = j  # Remember end of valid token
+
+            if last_match_id is not None:
+                result_ids.append(last_match_id)
+                i = last_match_end
+            else:
+                # Fallback: encode single char
+                tok = tokens[i]
+                result_ids.append(self.token_to_id.get(tok, self.unk_token_id))
+                i += 1
+
+        return result_ids
+
+    def encode(self, text: str) -> List[int]:
+        """Public encode method — strips input and uses cache."""
+        return list(self._cached_encode_str(text.strip()))
+
+    def decode(self, token_ids: Union[List[int], torch.Tensor], skip_special_tokens: bool = False) -> str:
+        if isinstance(token_ids, torch.Tensor):
+            token_ids = token_ids.tolist()
+
+        if skip_special_tokens:
+            special_ids = {
+                self.bos_token_id,
+                self.eos_token_id,
+                self.pad_token_id,
+                self.mask_token_id,  
+            }
+        else:
+            special_ids = set()
+
+        tokens = []
+        for tid in token_ids:
+            if tid in special_ids:
+                continue
+            token = self.id_to_token.get(tid, self.unk_token)
+            tokens.append(token)
+
+        return "".join(tokens)
+
+    def decode_with_trace(self, token_ids: List[int]) -> None:
+        print(f"\n🔍 Decoding {len(token_ids)} tokens:")
+        for i, tid in enumerate(token_ids):
+            token = self.id_to_token.get(tid, self.unk_token)
+            print(f"  [{i:03d}] ID={tid:5d} → '{token}'")
+
+    def convert_ids_to_tokens(self, ids: List[int]) -> List[str]:
+        return [self.id_to_token.get(i, self.unk_token) for i in ids]
+
+    def convert_tokens_to_ids(self, tokens: List[str]) -> List[int]:
+        return [self.token_to_id.get(t, self.unk_token_id) for t in tokens]
+
+    def encode_plus(
+        self,
+        text: str,
+        text_pair: Optional[str] = None,
+        add_special_tokens: bool = True,
+        padding: Union[bool, str] = False,
+        truncation: bool = False,
+        max_length: Optional[int] = None,
+        return_tensors: Optional[str] = None,
+        return_attention_mask: bool = True,
+        return_token_type_ids: bool = True,
+    ) -> BatchEncoding:
+        if max_length is None:
+            max_length = self.model_max_length
+
+        ids_a = self.encode(text)
+
+        if text_pair is not None:
+            ids_b = self.encode(text_pair)
+        else:
+            ids_b = None
+
+        input_ids = []
+        token_type_ids = []
+
+        if add_special_tokens:
+            input_ids.append(self.bos_token_id)
+            token_type_ids.append(0)
+            if ids_b is not None:
+                input_ids.extend(ids_a)
+                token_type_ids.extend([0] * len(ids_a))
+                input_ids.append(self.eos_token_id)
+                token_type_ids.append(0)
+
+                input_ids.extend(ids_b)
+                token_type_ids.extend([1] * len(ids_b))
+                input_ids.append(self.eos_token_id)
+                token_type_ids.append(1)
+            else:
+                input_ids.extend(ids_a)
+                token_type_ids.extend([0] * len(ids_a))
+                input_ids.append(self.eos_token_id)
+                token_type_ids.append(0)
+        else:
+            input_ids = ids_a
+            token_type_ids = [0] * len(input_ids)
+            if ids_b is not None:
+                input_ids.extend(ids_b)
+                token_type_ids.extend([1] * len(ids_b))
+
+        if truncation and len(input_ids) > max_length:
+            input_ids = input_ids[:max_length]
+            token_type_ids = token_type_ids[:max_length]
+
+        if padding:
+            pad_len = max_length - len(input_ids)
+            if pad_len > 0:
+                input_ids.extend([self.pad_token_id] * pad_len)
+                token_type_ids.extend([0] * pad_len)
+
+        attention_mask = [1 if tid != self.pad_token_id else 0 for tid in input_ids]
+
+        encoded_dict = {
+            "input_ids": input_ids,
+            "attention_mask": attention_mask,
+        }
+        if return_token_type_ids:
+            encoded_dict["token_type_ids"] = token_type_ids
+
+        if return_tensors == "pt":
+            output = {}
+            for k, v in encoded_dict.items():
+                tensor = torch.tensor(v, dtype=torch.long)  #  Fixed: use torch.tensor, not as_tensor
+                if tensor.ndim == 1:
+                    tensor = tensor.unsqueeze(0)
+                output[k] = tensor
+        else:
+            output = encoded_dict
+
+        return BatchEncoding(output, tensor_type=return_tensors)
+
+    def batch_encode_plus(
+        self,
+        batch_text_or_text_pairs: List[Union[str, Tuple[str, str]]],
+        **kwargs
+    ) -> BatchEncoding:
+        all_input_ids = []
+        all_attention_masks = []
+        all_token_type_ids = []
+
+        for item in batch_text_or_text_pairs:
+            if isinstance(item, tuple):
+                text, text_pair = item
+            else:
+                text, text_pair = item, None
+
+            encoded = self.encode_plus(
+                text=text,
+                text_pair=text_pair,
+                **kwargs
+            )
+            all_input_ids.append(encoded["input_ids"])
+            all_attention_masks.append(encoded["attention_mask"])
+            if "token_type_ids" in encoded:
+                all_token_type_ids.append(encoded["token_type_ids"])
+
+        batched = {
+            "input_ids": all_input_ids,
+            "attention_mask": all_attention_masks,
+        }
+        if all_token_type_ids:
+            batched["token_type_ids"] = all_token_type_ids
+
+        if kwargs.get("return_tensors") == "pt":
+            def to_tensor_list(lst):
+                # Use torch.tensor for safety — avoids "copy construct from tensor" warning
+                return [torch.tensor(item, dtype=torch.long) for item in lst]
+
+            batched = {
+                k: torch.nn.utils.rnn.pad_sequence(
+                    to_tensor_list(v),
+                    batch_first=True,
+                    padding_value=self.pad_token_id if k == "input_ids" else 0
+                )
+                for k, v in batched.items()
+            }
+
+        return BatchEncoding(batched, tensor_type=kwargs.get("return_tensors"))
+
+    # Save vocab to directory
+    def save_pretrained(self, save_directory: str):
+        """
+        Save tokenizer vocab as `vocab.json` in target directory.
+        Mimics Hugging Face convention.
+        """
+        if not os.path.exists(save_directory):
+            os.makedirs(save_directory)
+
+        vocab_file = os.path.join(save_directory, "vocab.json")
+
+        # Keys are strings, values are ints — JSON-safe
+        with open(vocab_file, "w", encoding="utf-8") as f:
+            json.dump(self.token_to_id, f, ensure_ascii=False, indent=2)
+
+        print(f"✅ Tokenizer vocab saved to: {vocab_file}")
+
+    # Load from pretrained directory
+    @classmethod
+    def from_pretrained(cls, pretrained_directory: str, model_max_length=512):
+        """
+        Load tokenizer from directory containing `vocab.json`.
+        """
+        vocab_file = os.path.join(pretrained_directory, "vocab.json")
+
+        if not os.path.exists(vocab_file):
+            raise FileNotFoundError(f"Vocab file not found: {vocab_file}")
+
+        with open(vocab_file, "r", encoding="utf-8") as f:
+            token_to_id = json.load(f)
+
+        # Convert keys to str (JSON loads as str anyway), values to int
+        token_to_id = {str(k): int(v) for k, v in token_to_id.items()}
+
+        return cls(token_to_id=token_to_id, model_max_length=model_max_length)

smitok/vocab.json

@@ -0,0 +1,1240 @@
+{
+  "<s>": 0,
+  "<pad>": 1,
+  "</s>": 2,
+  "<unk>": 3,
+  "<mask>": 4,
+  "cccc([N+](=O)[O-])": 5,
+  "ccc([N+](=O)[O-])": 6,
+  "([N+](=O)[O-])cc": 7,
+  "cc([N+](=O)[O-])": 8,
+  ")[C@H](CCCNC(=N": 9,
+  ")[C@H](CCCCN)NC": 10,
+  "[CH][CH][CH][CH": 11,
+  "C@@H](CCCNC(=N)": 12,
+  "c([N+](=O)[O-])": 13,
+  "](CCCNC(=N)N)NC": 14,
+  "N[C@@H](CCCNC(=": 15,
+  "[C@@H](CCCNC(=N": 16,
+  "O)[C@H](CCCNC(=": 17,
+  "@@H](CCCNC(=N)N": 18,
+  "([N+](=O)[O-])c": 19,
+  "COc1ccc(-c2coc3": 20,
+  "(=O)[C@H](CCCNC": 21,
+  "[C@@H](Cc1ccccc": 22,
+  "1ccc([N+](=O)[O": 23,
+  "(=O)[C@H](CCCCN": 24,
+  "N+](=O)[O-])cc1": 25,
+  "H](CCCNC(=N)N)": 26,
+  "c[nH]c2ccccc12": 27,
+  "N+](=O)[O-])cc": 28,
+  ")N[C@@H](CCCNC": 29,
+  "c[nH]c3ccccc23": 30,
+  ")N[C@@H](CCCCN": 31,
+  "@H](CCCNC(=N)N": 32,
+  "([N+](=O)[O-])": 33,
+  "COc1ccc(NC(=O)": 34,
+  "][CH][CH][CH][": 35,
+  "N[C@@H](CCCCN)": 36,
+  "O=[N+]([O-])c1": 37,
+  "C@H](CCCNC(=N)": 38,
+  "CH][CH][CH][C]": 39,
+  "COc1ccc(-c2ccc": 40,
+  "[nH]c2ccccc12)": 41,
+  "](CO)[C@@H](O": 42,
+  "COc1ccc(-c2cc": 43,
+  "ccc(S(=O)(=O)": 44,
+  "C@@H](NC(=O)[": 45,
+  "=[N+]([O-])c1": 46,
+  "H](CO)[C@@H](": 47,
+  "NC(=O)[C@@H](": 48,
+  "COc1ccc(C(=O)": 49,
+  "H](CCCNC(=N)N": 50,
+  "N[C@@H](CCCNC": 51,
+  "(=O)[C@@H](NC": 52,
+  "1ccc(S(=O)(=O": 53,
+  "O)[C@@H](NC(=": 54,
+  "[CH][CH][CH][": 55,
+  "COc1ccc2[nH]c": 56,
+  "COc1ccc(/C=C/": 57,
+  "[nH]c2ccccc12": 58,
+  ")[C@@H](NC(=O": 59,
+  "CO)[C@@H](O)[": 60,
+  "1c[nH]c2ccccc": 61,
+  "N+](=O)[O-])c": 62,
+  "@@H](NC(=O)[C": 63,
+  "](CCCNC(=N)N)": 64,
+  "COc1ccc(-c2nc": 65,
+  "COc1cc(C(=O)": 66,
+  "(=O)[C@H](Cc": 67,
+  "COc1ccc2[nH]": 68,
+  "N+](=O)[O-])": 69,
+  "COc1ccc(/C=C": 70,
+  "CH][CH][CH][": 71,
+  "COc1ccc(OC)c": 72,
+  "NC(=O)[C@H](": 73,
+  "COC(=O)[C@@H": 74,
+  "COC(=O)c1ccc": 75,
+  "CC(=O)Nc1ccc": 76,
+  "COc1cc(/C=C/": 77,
+  "H](O)[C@@H](": 78,
+  ")[C@@H](O)[C": 79,
+  "(=O)[C@H](CC": 80,
+  "COc1cccc(-c2": 81,
+  "@H](CO)[C@@H": 82,
+  "@H](NC(=O)[C": 83,
+  "CCCCCCCCCCCC": 84,
+  "COc1cc2c(cc1": 85,
+  ")NC(=O)[C@@H": 86,
+  "CCOC(=O)c1cc": 87,
+  "CC(=O)OC1CC2": 88,
+  "C@H](CO)[C@@": 89,
+  "Cc1ccc(C(=O)": 90,
+  "COc1ccc2c(c1": 91,
+  "@@H](O)[C@@H": 92,
+  "NC(=O)[C@@H]": 93,
+  ".[Br-].[Br-]": 94,
+  "C@H](NC(=O)[": 95,
+  ")[C@@H](NC(=": 96,
+  "](NC(=O)[C@H": 97,
+  "H](O)[C@H](": 98,
+  "](NC(=O)[C@": 99,
+  "[C@H](NC(=O": 100,
+  "[CH][CH][CH": 101,
+  ")NC(=O)[C@@": 102,
+  "C@H](O)[C@@": 103,
+  "@H](O)[C@@H": 104,
+  "COc1ccc(-c2": 105,
+  "ccc(C(F)(F)": 106,
+  "O=C(Nc1ccc(": 107,
+  ")N[C@@H](CC": 108,
+  ")[C@H](O)[C": 109,
+  ")N[C@@H](Cc": 110,
+  "COC(=O)c1cc": 111,
+  "C@@H](O)[C@": 112,
+  "O)[C@H](Cc1": 113,
+  "(=O)[C@@H](": 114,
+  "COc1ccc(-n2": 115,
+  "O)[C@H](O)[": 116,
+  ")NC(=O)[C@H": 117,
+  "COc1ccc(/C=": 118,
+  "+](=O)[O-])": 119,
+  "Cc1cc(C(=O)": 120,
+  "COC(=O)[C@H": 121,
+  "[C@H](CO)[C": 122,
+  "](O)[C@H](O": 123,
+  "COc1cc(/C=C": 124,
+  "H](Cc1c[nH]": 125,
+  "CCS(=O)(=O)": 126,
+  "(S(=O)(=O)N": 127,
+  "C[C@]12CC[C": 128,
+  "COc1cc(OC)c": 129,
+  "O=C(Nc1cccc": 130,
+  "CCOC(=O)c1c": 131,
+  "H](O)[C@@H]": 132,
+  "N[C@@H](CC(": 133,
+  "O[C@H](CO)[": 134,
+  "C@@H](NC(=O": 135,
+  "O)[C@H](CC(": 136,
+  ")[C@@H](O)[": 137,
+  "CC(=O)OC1CC": 138,
+  "COc1ccccc1C": 139,
+  "@@H](O)[C@H": 140,
+  "C=CC(=O)Nc1": 141,
+  "N[C@@H](Cc1": 142,
+  "O=C(COc1ccc": 143,
+  "(=O)N[C@@H": 144,
+  "COc1cc(/C=": 145,
+  "CC(C)(C)c1": 146,
+  "Na+].[Na+]": 147,
+  "(=O)[C@H](": 148,
+  "C@H](C(=O)": 149,
+  "Cc1c[nH]c2": 150,
+  "a+].[Na+].": 151,
+  "](O)[C@H](": 152,
+  "O=S(=O)(c1": 153,
+  "C[C@@H]1CN": 154,
+  "COc1cc2c(c": 155,
+  "COc1cc(OC)": 156,
+  "[C@H](C(=O": 157,
+  "CS(=O)(=O)": 158,
+  "@@H](O)[C@": 159,
+  ")[C@H](Cc1": 160,
+  "O)N[C@@H](": 161,
+  "NC(=O)[C@H": 162,
+  "C(=O)N[C@@": 163,
+  "C@H](NC(=O": 164,
+  "O=C(Nc1ccc": 165,
+  "S(=O)(=O)c": 166,
+  "COc1ccc2nc": 167,
+  "CC(C)[C@@H": 168,
+  "COc1ccc(C=": 169,
+  "COc1ccccc1": 170,
+  "+].[Na+].[": 171,
+  ")NC(=O)[C@": 172,
+  "CCOC(=O)[C": 173,
+  "COc1ccc(CC": 174,
+  "(=O)[C@@H]": 175,
+  "COc1ccc(C2": 176,
+  "(F)(F)F)cc": 177,
+  "COc1cc(O)c": 178,
+  "C@H](O)[C@": 179,
+  "CC(=O)c1c(": 180,
+  "COc1cccc(-": 181,
+  "N[C@H](C(=": 182,
+  "(S(=O)(=O)": 183,
+  "CN(C)c1ccc": 184,
+  "H](O)[C@H]": 185,
+  "COc1ccc([C": 186,
+  "CC(C)=CCc1": 187,
+  "CCOC(=O)C1": 188,
+  "Cc1ccc(-c2": 189,
+  "O=c1cc(-c2": 190,
+  "Nc1ncnc2c1": 191,
+  "COc1cccc(C": 192,
+  "COc1cc(-c2": 193,
+  "NC(=O)[C@@": 194,
+  "CC(=O)OCC1": 195,
+  "COc1ccc(CN": 196,
+  "COc1ccc(N2": 197,
+  "CCOC(=O)c1": 198,
+  "S(=O)(=O)N": 199,
+  "O=C(c1ccc(": 200,
+  "OC(CO)C(O)": 201,
+  "[C@H](NC(=": 202,
+  "CC(=O)OC1C": 203,
+  "CC(C)c1ccc": 204,
+  "@H](O)[C@H": 205,
+  "O)[C@H](Cc": 206,
+  "FC(F)(F)c1": 207,
+  "O)[C@H](CC": 208,
+  "N[C@@H](CC": 209,
+  "[C@@H](Cc1": 210,
+  "COc1ccc(-": 211,
+  "CC(=O)Oc1": 212,
+  "O)[C@@H]1": 213,
+  "Cc1cccc(C": 214,
+  "COC(=O)CC": 215,
+  "Cc1cc(=O)": 216,
+  "O)N[C@H](": 217,
+  "CCOc1ccc(": 218,
+  "N+](=O)[O": 219,
+  "(F)(F)F)c": 220,
+  "N#Cc1cccc": 221,
+  "[C@@H](Cc": 222,
+  "COc1cccc2": 223,
+  "COC(=O)C(": 224,
+  "(=O)N[C@H": 225,
+  ")NC(=O)[C": 226,
+  "COC(=O)[C": 227,
+  "O=C(c1ccc": 228,
+  "COc1ccc(N": 229,
+  ")[C@H](Cc": 230,
+  ")C(=O)N[C": 231,
+  "CC(C)(C)C": 232,
+  "CNC(=O)c1": 233,
+  ")C(F)(F)F": 234,
+  "CC(C)[C@H": 235,
+  "Cc1ccccc1": 236,
+  "CC(=O)N[C": 237,
+  ")N[C@@H](": 238,
+  "@H](NC(=O": 239,
+  "CC1(C)CCC": 240,
+  "C=C1C(=O)": 241,
+  ")[C@@H](O": 242,
+  "ccc(Cl)cc": 243,
+  "COC(=O)C1": 244,
+  "COC(=O)c1": 245,
+  "CC(=O)OC1": 246,
+  "COc1ccc(C": 247,
+  "COc1cc(C2": 248,
+  "S(=O)(=O)": 249,
+  "Cc1cc(C)c": 250,
+  "CCOC(=O)C": 251,
+  "COc1cccc(": 252,
+  "Cc1cc(O)c": 253,
+  "O)[C@@H](": 254,
+  "F)(F)F)cc": 255,
+  "COc1cc(O)": 256,
+  "C(=O)N[C@": 257,
+  "CC(=O)OCC": 258,
+  "(=O)N[C@@": 259,
+  "(C(F)(F)F": 260,
+  "[Na+].[Na": 261,
+  "CC(=O)Nc1": 262,
+  "C(F)(F)F)": 263,
+  "a+].[Na+]": 264,
+  "@H](O)[C@": 265,
+  "](O)[C@H]": 266,
+  "(=O)(=O)c": 267,
+  "O=c1[nH]c": 268,
+  "COc1cc2c(": 269,
+  "COc1ccc2c": 270,
+  ".[Na+].[N": 271,
+  ")N[C@H](C": 272,
+  "CC(=O)O[C": 273,
+  "O=C(COc1": 274,
+  "(=O)(=O)": 275,
+  "O=C1NC(=": 276,
+  "C=C(C)C1": 277,
+  "C[C@@H](": 278,
+  ")N[C@H](": 279,
+  "O=C(O)c1": 280,
+  "Cc1cc(O)": 281,
+  "CC(=O)NC": 282,
+  "CC(=O)N1": 283,
+  "Cc1cc2c(": 284,
+  "[C@@H](C": 285,
+  "(=O)N[C@": 286,
+  "COC(=O)N": 287,
+  ")[C@@H](": 288,
+  "COc1cc(C": 289,
+  "CCOC(=O)": 290,
+  "O=C(NCc1": 291,
+  "CCCC(=O)": 292,
+  "CCOc1ccc": 293,
+  "O=C(CCc1": 294,
+  "O=C(O)CC": 295,
+  "O=C(O)[C": 296,
+  "N#Cc1ccc": 297,
+  "CN1C(=O)": 298,
+  "COC(=O)C": 299,
+  "CCc1ccc(": 300,
+  "COc1cccc": 301,
+  ".[Na+].[": 302,
+  "(C(F)(F)": 303,
+  "O=C(CSc1": 304,
+  "Cc1ccc(C": 305,
+  "Cc1cccc(": 306,
+  "Cc1cc(C)": 307,
+  "(O)C(O)C": 308,
+  "Cc1ccc(N": 309,
+  "CC(=O)c1": 310,
+  "COc1ccc(": 311,
+  "C=CC(=O)": 312,
+  "CCNC(=O)": 313,
+  "CCC(=O)N": 314,
+  "COc1c2c(": 315,
+  "CC1CCC2(": 316,
+  "C(=O)N[C": 317,
+  "(=O)[C@H": 318,
+  "O)N[C@@H": 319,
+  ")[C@@H]1": 320,
+  "O=C(c1cc": 321,
+  "CCCCCCCC": 322,
+  "COc1ccc2": 323,
+  "CC(C)(C)": 324,
+  "C(F)(F)F": 325,
+  "NC(=O)c1": 326,
+  "F)(F)F)c": 327,
+  "CC(=O)OC": 328,
+  ")[C@H](O": 329,
+  "O)[C@H](": 330,
+  "(F)(F)F)": 331,
+  "CC(C)C[C": 332,
+  "Cc1ccc(-": 333,
+  "O=S(=O)(": 334,
+  "C[C@@H]1": 335,
+  "O)[C@@H]": 336,
+  "COc1c(O)": 337,
+  "H](O)[C@": 338,
+  "CC1(C)CC": 339,
+  "CC1(C)[C": 340,
+  "(NC(=O)": 341,
+  "CC[C@@H": 342,
+  "S(=O)(=": 343,
+  "C[C@H](": 344,
+  "CCCc1cc": 345,
+  "CCN(CC)": 346,
+  "C@@H](C": 347,
+  "a+].[Na": 348,
+  "CCC(=O)": 349,
+  "CC(=O)N": 350,
+  "[C@@H](": 351,
+  "(=O)N[C": 352,
+  "Clc1ccc": 353,
+  "CC(C)CN": 354,
+  "CCCCCCC": 355,
+  "CC1CCC2": 356,
+  "Cc1cc(N": 357,
+  "Cc1nc(N": 358,
+  "CC(C)c1": 359,
+  "O)N[C@@": 360,
+  "C(F)(F)": 361,
+  "C=C1CCC": 362,
+  "2ccccc2": 363,
+  "CC(C)CC": 364,
+  "O=C1Nc2": 365,
+  "O=C(NC1": 366,
+  "C[C@H]1": 367,
+  "[C@H](C": 368,
+  "(F)(F)F": 369,
+  ")[C@H](": 370,
+  ")N[C@@H": 371,
+  "F)(F)F)": 372,
+  "CCOc1cc": 373,
+  "Cc1ccc2": 374,
+  "CCC(C)C": 375,
+  "C(=O)N[": 376,
+  "CC1=C(C": 377,
+  "Cc1cccc": 378,
+  "COc1cc2": 379,
+  "Cc1cc(C": 380,
+  "COC(=O)": 381,
+  "O=C(NCC": 382,
+  "O=C(Cc1": 383,
+  ")C(=O)N": 384,
+  "C=C(C)C": 385,
+  "O)C(O)C": 386,
+  "O=C(Nc1": 387,
+  "CN1CCN(": 388,
+  "O=C(N[C": 389,
+  "Cc1ccc(": 390,
+  "C[C@]12": 391,
+  "CN(C)c1": 392,
+  ")[C@@H]": 393,
+  "CCc1ccc": 394,
+  "(=O)[C@": 395,
+  "CC1(C)C": 396,
+  "O=C(CN1": 397,
+  "COc1cc(": 398,
+  "N[C@H](": 399,
+  "COc1ccc": 400,
+  "O=C1C(=": 401,
+  "CC12CCC": 402,
+  "O=C(O)C": 403,
+  "[C@@H]2": 404,
+  "CNC(=O)": 405,
+  "O)N[C@H": 406,
+  "O)[C@H]": 407,
+  "1ccccc1": 408,
+  "[C@@H]1": 409,
+  "CC(C)[C": 410,
+  "CC(C)C(": 411,
+  "COc1c(": 412,
+  ")C(=O)": 413,
+  "CC(C)N": 414,
+  "O)(=O)": 415,
+  "CC1CCC": 416,
+  "F)(F)F": 417,
+  "[C@H]2": 418,
+  "C=C(C)": 419,
+  "[C@@H]": 420,
+  "CC(C)C": 421,
+  "CCc1nc": 422,
+  "CC(C)=": 423,
+  "CCCCCC": 424,
+  "C@H](C": 425,
+  "CCCCc1": 426,
+  "Cc1nc2": 427,
+  "Cc1nnc": 428,
+  "CCc1cc": 429,
+  "c1ccc(": 430,
+  "O=C1C=": 431,
+  "CC1OC(": 432,
+  "Fc1ccc": 433,
+  "(=O)c1": 434,
+  "C@@H]2": 435,
+  "C@H](O": 436,
+  "C[C@@H": 437,
+  "NC(=O)": 438,
+  "CCCCN(": 439,
+  "C=CCc1": 440,
+  "Cc1oc2": 441,
+  "O)[C@@": 442,
+  "C=C1CC": 443,
+  "CC[C@H": 444,
+  ")N[C@@": 445,
+  ")[C@H]": 446,
+  "CC(C)(": 447,
+  "(C)(C)": 448,
+  "O=C(O)": 449,
+  "(=O)[C": 450,
+  "Cc1cc(": 451,
+  "C[C@@]": 452,
+  "N=C(N)": 453,
+  "Cc1cc2": 454,
+  "N#Cc1c": 455,
+  "C@@H]1": 456,
+  "(F)(F)": 457,
+  "O=C1c2": 458,
+  "(C(=O)": 459,
+  "CN(Cc1": 460,
+  "C(=O)N": 461,
+  "CCCC[C": 462,
+  "O=c1cc": 463,
+  "Cc1nc(": 464,
+  "O=C(NC": 465,
+  "CC(=O)": 466,
+  "CN(C)C": 467,
+  ")C(O)C": 468,
+  ")(F)F)": 469,
+  "O=C(c1": 470,
+  "[C@H](": 471,
+  "C@@H](": 472,
+  "Nc1ccc": 473,
+  "COc1cc": 474,
+  "O=C(CC": 475,
+  "CC1(C)": 476,
+  ")NC(=O": 477,
+  "Cc1ccc": 478,
+  "O=c1c2": 479,
+  "CC1=CC": 480,
+  "O=C1NC": 481,
+  "Oc1ccc": 482,
+  "[C@H]1": 483,
+  "(=O)N[": 484,
+  "Cc1c(C": 485,
+  "O=C1CC": 486,
+  "O=C(O": 487,
+  "N[C@H": 488,
+  "c1ccc": 489,
+  "CCN(C": 490,
+  "CC(O)": 491,
+  ")(F)F": 492,
+  "@@H]1": 493,
+  "COCCN": 494,
+  "O=C(C": 495,
+  "CCCc1": 496,
+  "CC1CC": 497,
+  "Cn1nc": 498,
+  "CN1CC": 499,
+  "Cc1oc": 500,
+  "(=O)N": 501,
+  "O=C(N": 502,
+  "@@H](": 503,
+  "Cc1cn": 504,
+  "C@@H]": 505,
+  "C=C1C": 506,
+  "O=c1c": 507,
+  "CCn1c": 508,
+  "COC1=": 509,
+  "[C@H]": 510,
+  "CC1(C": 511,
+  "Cc1cc": 512,
+  "CCC[C": 513,
+  "Cc1c(": 514,
+  "CCC(C": 515,
+  "C(=O)": 516,
+  "C[C@H": 517,
+  "C@H](": 518,
+  "F)(F)": 519,
+  ")[C@@": 520,
+  "[nH]c": 521,
+  "C/C(=": 522,
+  "C[C@]": 523,
+  "C(O)C": 524,
+  "C=CC1": 525,
+  "CN(C)": 526,
+  "Cc1nn": 527,
+  "CCCN(": 528,
+  "N#Cc1": 529,
+  "CC1=C": 530,
+  "ccccc": 531,
+  "(=O)c": 532,
+  "COc1c": 533,
+  "Cn1cc": 534,
+  "C=C[C": 535,
+  "Cc1nc": 536,
+  "CC(C)": 537,
+  "CCOc1": 538,
+  "CCCCC": 539,
+  "O)[C@": 540,
+  "C@H]1": 541,
+  "1ccc(": 542,
+  "(=O)O": 543,
+  "(=O)C": 544,
+  "Nc1nc": 545,
+  "C=C(C": 546,
+  "CCCCN": 547,
+  "C/C=C": 548,
+  "(=O)[": 549,
+  "O)N[C": 550,
+  "C@H]2": 551,
+  "COC[C": 552,
+  "O=C([": 553,
+  "C=CCN": 554,
+  "CCC1(": 555,
+  ")C(O)": 556,
+  "N#CC1": 557,
+  "O=C1C": 558,
+  "COC1C": 559,
+  "COCC": 560,
+  "COC1": 561,
+  "O=C1": 562,
+  "COc1": 563,
+  "CC(C": 564,
+  "(Cl)": 565,
+  "CCC(": 566,
+  "CN(C": 567,
+  "[nH]": 568,
+  ")cc1": 569,
+  "OC[C": 570,
+  "O)[C": 571,
+  "@H]1": 572,
+  "O)C(": 573,
+  "CN1C": 574,
+  "(C)C": 575,
+  ")[C@": 576,
+  "1ccc": 577,
+  "C=CC": 578,
+  "C=C1": 579,
+  "CSc1": 580,
+  "@@H]": 581,
+  "CO[C": 582,
+  "CC1=": 583,
+  "CCCC": 584,
+  "(O)C": 585,
+  "O)c1": 586,
+  "CCn1": 587,
+  "[C@@": 588,
+  "ccc(": 589,
+  "CCN1": 590,
+  "CC/C": 591,
+  "O=C(": 592,
+  "C@H]": 593,
+  "O)N[": 594,
+  "Cn1c": 595,
+  "CCc1": 596,
+  "CCC1": 597,
+  "CC12": 598,
+  "CN[C": 599,
+  "CC1C": 600,
+  "CC=C": 601,
+  "C=C(": 602,
+  "CCCN": 603,
+  "Cc1c": 604,
+  "CC1(": 605,
+  "@H](": 606,
+  "CNc1": 607,
+  "CNCC": 608,
+  "O=c1": 609,
+  "Clc1": 610,
+  "N#CC": 611,
+  "(=O)": 612,
+  ")N[C": 613,
+  "C(O)": 614,
+  "CC(O": 615,
+  ")C(=": 616,
+  "CC[C": 617,
+  "CCN(": 618,
+  "c1cc": 619,
+  "(C(=": 620,
+  "nH]c": 621,
+  "(-c2": 622,
+  "N[C": 623,
+  "O=C": 624,
+  "cc(": 625,
+  "CCN": 626,
+  ")[C": 627,
+  "O)N": 628,
+  ")c2": 629,
+  "=C(": 630,
+  "O)[": 631,
+  "(-c": 632,
+  ")CC": 633,
+  "(O)": 634,
+  "[C@": 635,
+  "C@@": 636,
+  "@H]": 637,
+  "H]1": 638,
+  ")C(": 639,
+  "cc1": 640,
+  "C)C": 641,
+  "C/C": 642,
+  "O=S": 643,
+  "Cc1": 644,
+  "CCC": 645,
+  "CC=": 646,
+  "NC(": 647,
+  "CNC": 648,
+  "C[C": 649,
+  "nH]": 650,
+  "C(C": 651,
+  "N#C": 652,
+  "CO/": 653,
+  ")c(": 654,
+  "ccc": 655,
+  "H](": 656,
+  ")cc": 657,
+  "O)c": 658,
+  "(C)": 659,
+  ")c1": 660,
+  "O)C": 661,
+  "COC": 662,
+  "CC(": 663,
+  "O[C": 664,
+  "C=C": 665,
+  "CN(": 666,
+  "Cl.": 667,
+  "Fc1": 668,
+  "CC1": 669,
+  "Nc1": 670,
+  "CCO": 671,
+  "Cn1": 672,
+  "Oc1": 673,
+  "CN1": 674,
+  "C[N": 675,
+  "C@": 676,
+  "H]": 677,
+  ")[": 678,
+  "OC": 679,
+  "CC": 680,
+  "C[": 681,
+  "O=": 682,
+  "CO": 683,
+  "NC": 684,
+  "3)": 685,
+  "=C": 686,
+  "@@": 687,
+  "F)": 688,
+  "c3": 689,
+  ")N": 690,
+  "](": 691,
+  "(O": 692,
+  "c(": 693,
+  "c1": 694,
+  "N#": 695,
+  "CN": 696,
+  "C=": 697,
+  "2)": 698,
+  "[C": 699,
+  ")C": 700,
+  "(C": 701,
+  "(=": 702,
+  "@H": 703,
+  "c2": 704,
+  "C(": 705,
+  "cc": 706,
+  "C)": 707,
+  ")c": 708,
+  "O)": 709,
+  "C#": 710,
+  "2c": 711,
+  "@": 712,
+  "]": 713,
+  "4": 714,
+  "F": 715,
+  "H": 716,
+  "=": 717,
+  "2": 718,
+  "1": 719,
+  "(": 720,
+  "3": 721,
+  ")": 722,
+  "N": 723,
+  "[": 724,
+  "C": 725,
+  "CNN": 726,
+  "CCCCO": 727,
+  "NCCOc": 728,
+  "ccncn": 729,
+  "CCNCC": 730,
+  "COCCCCN": 731,
+  "IC": 732,
+  "COCCOCCNC": 733,
+  "CCCCCn": 734,
+  "CCCNCC": 735,
+  "COCCSc": 736,
+  "(-[": 737,
+  "CCCCCc": 738,
+  "OCCOCC": 739,
+  "CSCC": 740,
+  "OCCS": 741,
+  ")(=": 742,
+  "Sc": 743,
+  "OCCN": 744,
+  "NCCCNC": 745,
+  "NCCSc": 746,
+  "nsc": 747,
+  "OCCSCC": 748,
+  "([": 749,
+  ")=": 750,
+  "SCC": 751,
+  "35": 752,
+  "NCCCc": 753,
+  "-])=": 754,
+  "OCCOCCS": 755,
+  "CCOCCNc": 756,
+  "OCCNc": 757,
+  "]([": 758,
+  "ccoc": 759,
+  "SCCCOc": 760,
+  "CCCCCCCCCCC": 761,
+  "CSc": 762,
+  "42": 763,
+  "Cc": 764,
+  "scnc": 765,
+  "+][": 766,
+  "ssc": 767,
+  "CSCCNC": 768,
+  "Br": 769,
+  "NCCCS": 770,
+  "ncoc": 771,
+  "CSCc": 772,
+  "31": 773,
+  "A": 774,
+  "COCCOCCN": 775,
+  "OCOC": 776,
+  "OP": 777,
+  "I": 778,
+  "cnccc": 779,
+  "OCCCCN": 780,
+  "21": 781,
+  "ccsc": 782,
+  "NCCSC": 783,
+  "nncs": 784,
+  "SCCNC": 785,
+  "COc": 786,
+  "CCCCCSc": 787,
+  "CCCl": 788,
+  "CCOCCSc": 789,
+  "nnn": 790,
+  "ncnn": 791,
+  "OCc": 792,
+  "OCCCc": 793,
+  "+]=[": 794,
+  "ClCC": 795,
+  "NCCS": 796,
+  "cccn": 797,
+  "nn": 798,
+  "CCCCCOc": 799,
+  "COCCCN": 800,
+  "NCCCCn": 801,
+  "CCCCSC": 802,
+  "NCCO": 803,
+  "]/": 804,
+  "CSC": 805,
+  "B": 806,
+  "CCCS": 807,
+  "@@]([": 808,
+  "CCCSCC": 809,
+  "6": 810,
+  "nscc": 811,
+  "M": 812,
+  "p": 813,
+  "COCCOCC": 814,
+  "NCN": 815,
+  "cs": 816,
+  "OCCOCCOCCOCCO": 817,
+  "NCC": 818,
+  "CCOCCCC": 819,
+  "COCCCC": 820,
+  "Nc": 821,
+  "NCCC": 822,
+  "nsnc": 823,
+  "CCCOc": 824,
+  "OCCSCc": 825,
+  "K": 826,
+  "ncncc": 827,
+  "@]": 828,
+  "COCCCNc": 829,
+  "CSCCc": 830,
+  "CS": 831,
+  "CCCCCO": 832,
+  "COCCOc": 833,
+  "CCNCc": 834,
+  "onc": 835,
+  "e": 836,
+  "CCOCCCNc": 837,
+  "OCn": 838,
+  "+]=": 839,
+  "CCOCCCc": 840,
+  "NCCNC": 841,
+  "CNn": 842,
+  "Nn": 843,
+  "CCOCCS": 844,
+  "SCCN": 845,
+  "CCCCl": 846,
+  "cncs": 847,
+  "CONC": 848,
+  "CCOCCNC": 849,
+  "nnnn": 850,
+  "R": 851,
+  "ncc": 852,
+  "CCCNc": 853,
+  "-])(": 854,
+  "CCCCNS": 855,
+  "o": 856,
+  "SCCSc": 857,
+  "CSCCOc": 858,
+  "9": 859,
+  "NCCCNS": 860,
+  "+]": 861,
+  "COCCc": 862,
+  "CNCCc": 863,
+  "COn": 864,
+  "Z": 865,
+  "CCl": 866,
+  "coc": 867,
+  "CCCSc": 868,
+  "NOC": 869,
+  "NS": 870,
+  ")([": 871,
+  "CCBr": 872,
+  "5": 873,
+  "nncc": 874,
+  "CI": 875,
+  "OCN": 876,
+  "CCCCCS": 877,
+  "NCCCCOc": 878,
+  "CCCSCc": 879,
+  "CSCCCNC": 880,
+  "COCCS": 881,
+  "CSCCCCNC": 882,
+  "BrCC": 883,
+  "(/": 884,
+  "CCNC": 885,
+  "On": 886,
+  "NCCc": 887,
+  "+": 888,
+  "#": 889,
+  "\\[": 890,
+  "CSCCCC": 891,
+  "CCOc": 892,
+  "NSC": 893,
+  "SH": 894,
+  "@](=": 895,
+  "-][": 896,
+  "COCCCOc": 897,
+  "ccnc": 898,
+  "CCCCSc": 899,
+  "FCCC": 900,
+  "CSCCN": 901,
+  "cncc": 902,
+  "SCCS": 903,
+  "sccc": 904,
+  "CSCCCN": 905,
+  "Cn": 906,
+  "nccs": 907,
+  "CSCCC": 908,
+  "OCCCO": 909,
+  "SCCCO": 910,
+  "COCOc": 911,
+  "OCCNC": 912,
+  "CCOCCOCC": 913,
+  "NCCSCC": 914,
+  "ccncc": 915,
+  "NO": 916,
+  "OCCBr": 917,
+  "+])(": 918,
+  "snnc": 919,
+  "SCCO": 920,
+  "g": 921,
+  "8": 922,
+  ")(": 923,
+  "ON": 924,
+  "ClCCCSc": 925,
+  "(-": 926,
+  "OCCOC": 927,
+  "CCCCCCN": 928,
+  "NNc": 929,
+  "CSCCOC": 930,
+  "NOCC": 931,
+  "occc": 932,
+  "nH": 933,
+  "COCCCn": 934,
+  "/[": 935,
+  "CCSc": 936,
+  "cccnc": 937,
+  "NCCOCCO": 938,
+  "ccs": 939,
+  "ncco": 940,
+  "CCOCCOC": 941,
+  "ONC": 942,
+  "l": 943,
+  "ccno": 944,
+  "SCCc": 945,
+  "nncn": 946,
+  "OCCSC": 947,
+  "csc": 948,
+  "T": 949,
+  "-])/": 950,
+  "CCCSC": 951,
+  "COCCOC": 952,
+  "nccn": 953,
+  "OCC": 954,
+  "r": 955,
+  "ccnnc": 956,
+  "][": 957,
+  "SCCOC": 958,
+  "CCOC": 959,
+  "nccc": 960,
+  "CCCCOCc": 961,
+  "SCCCS": 962,
+  "CCSCCOC": 963,
+  "ccnn": 964,
+  "NCc": 965,
+  "CCCCOC": 966,
+  "t": 967,
+  "CCOCCCn": 968,
+  "312": 969,
+  "OH": 970,
+  "cnnc": 971,
+  "CCCCn": 972,
+  "NCCNc": 973,
+  "CCSCCOc": 974,
+  "CCOCCOc": 975,
+  "+]\\": 976,
+  "OS": 977,
+  "43": 978,
+  "CCCF": 979,
+  "45": 980,
+  "34": 981,
+  "13": 982,
+  "COCCO": 983,
+  "NCCCCN": 984,
+  "nnccc": 985,
+  "SCc": 986,
+  "CH": 987,
+  "CCCNCc": 988,
+  "+](/[": 989,
+  "SCn": 990,
+  "](/": 991,
+  "OCCF": 992,
+  "cncnc": 993,
+  "snc": 994,
+  "OCCCC": 995,
+  "OCCCn": 996,
+  "Brc": 997,
+  "ncn": 998,
+  "csnn": 999,
+  "SCCOc": 1000,
+  "+](=": 1001,
+  "CCOCCN": 1002,
+  "+](-": 1003,
+  "sccn": 1004,
+  "CCCCS": 1005,
+  "scc": 1006,
+  "nc": 1007,
+  "cncn": 1008,
+  "+](/": 1009,
+  "CNc": 1010,
+  "CCCCCCO": 1011,
+  "%": 1012,
+  "BrCCC": 1013,
+  "-]/": 1014,
+  "NNN": 1015,
+  "@@](=": 1016,
+  "-])[": 1017,
+  "NOCc": 1018,
+  "CCCOCc": 1019,
+  "ClC": 1020,
+  "cnccn": 1021,
+  "nnco": 1022,
+  "CCCOC": 1023,
+  "CSCCO": 1024,
+  "NCCCn": 1025,
+  "L": 1026,
+  ")(/": 1027,
+  "OCCCNC": 1028,
+  "occ": 1029,
+  "CCCNS": 1030,
+  "OCCc": 1031,
+  "41": 1032,
+  "-[": 1033,
+  "O": 1034,
+  "OCCCCC": 1035,
+  "CCc": 1036,
+  "COCCCS": 1037,
+  "CCSCC": 1038,
+  "NCCCOC": 1039,
+  "CCCBr": 1040,
+  "CCCCNC": 1041,
+  "NCCCSC": 1042,
+  "COCCNS": 1043,
+  "nccnc": 1044,
+  "SCCn": 1045,
+  "NOc": 1046,
+  "@](": 1047,
+  "OCCCl": 1048,
+  "=[": 1049,
+  "COCCCCC": 1050,
+  "CON": 1051,
+  "0": 1052,
+  ")/[": 1053,
+  "ccn": 1054,
+  "CCCc": 1055,
+  "a": 1056,
+  "NN": 1057,
+  "on": 1058,
+  ")/": 1059,
+  "+]/": 1060,
+  "ocnc": 1061,
+  "cnco": 1062,
+  "CNS": 1063,
+  "nnsc": 1064,
+  ")=[": 1065,
+  "CCSCc": 1066,
+  "-]=[": 1067,
+  "-": 1068,
+  "CCNc": 1069,
+  "CCCn": 1070,
+  "conc": 1071,
+  "CCOCc": 1072,
+  "ssnc": 1073,
+  "NCCCOCC": 1074,
+  "OCCCS": 1075,
+  "oc": 1076,
+  "S": 1077,
+  "CCONC": 1078,
+  "CCCCOc": 1079,
+  "CCCOCC": 1080,
+  "+])": 1081,
+  "CBr": 1082,
+  "BrCc": 1083,
+  "SCCC": 1084,
+  "CCOCCn": 1085,
+  "CCn": 1086,
+  "NNC": 1087,
+  "CCOP": 1088,
+  "CCOCCCNC": 1089,
+  "co": 1090,
+  "NCCCSc": 1091,
+  "cnsn": 1092,
+  "COCCNc": 1093,
+  "OCCSc": 1094,
+  "COCO": 1095,
+  "COS": 1096,
+  "NCCCCC": 1097,
+  ".": 1098,
+  "CCCCCCS": 1099,
+  "CSCCCNc": 1100,
+  "-])([": 1101,
+  ")-": 1102,
+  "NCCCO": 1103,
+  "b": 1104,
+  "NNS": 1105,
+  "CSCCn": 1106,
+  "sn": 1107,
+  "cnns": 1108,
+  "G": 1109,
+  "COCCn": 1110,
+  "CCS": 1111,
+  "OCCCSC": 1112,
+  "CCCCCOC": 1113,
+  "NCCOCC": 1114,
+  "OCCNS": 1115,
+  "CCOCCO": 1116,
+  "cscc": 1117,
+  "23": 1118,
+  "cnsc": 1119,
+  "(\\": 1120,
+  "@@]": 1121,
+  "COCCCNC": 1122,
+  "OCO": 1123,
+  "CCOCCC": 1124,
+  "FC": 1125,
+  "ccon": 1126,
+  "-])=[": 1127,
+  "OCCCOc": 1128,
+  "OCCCSc": 1129,
+  "c": 1130,
+  "CCCCCNC": 1131,
+  "cnnn": 1132,
+  "OCCn": 1133,
+  "CNCc": 1134,
+  "cccs": 1135,
+  "SC": 1136,
+  "24": 1137,
+  "OCCCN": 1138,
+  "NCCCN": 1139,
+  "-])": 1140,
+  "n": 1141,
+  "cnn": 1142,
+  "OCCOc": 1143,
+  "NCCCOc": 1144,
+  "CCSC": 1145,
+  "OCCCNc": 1146,
+  "+]([": 1147,
+  "NH": 1148,
+  "sc": 1149,
+  "ccco": 1150,
+  "CSCN": 1151,
+  "CCSCCC": 1152,
+  "oncc": 1153,
+  "OCCC": 1154,
+  "CCOCCOCCOCC": 1155,
+  "SCCCC": 1156,
+  "CCOCC": 1157,
+  "cccnn": 1158,
+  "CCOCCOCc": 1159,
+  "cn": 1160,
+  "NOCCc": 1161,
+  "CSCCSC": 1162,
+  "cscn": 1163,
+  "COCCSCc": 1164,
+  "CCSS": 1165,
+  "CCCNC": 1166,
+  "P": 1167,
+  "CCCCCCNC": 1168,
+  "+](": 1169,
+  "ncsc": 1170,
+  "CCNS": 1171,
+  "ncccc": 1172,
+  "noc": 1173,
+  "E": 1174,
+  "COP": 1175,
+  "12": 1176,
+  "cccc": 1177,
+  "123": 1178,
+  "CCOCCCN": 1179,
+  "7": 1180,
+  ")\\": 1181,
+  "@@](": 1182,
+  "i": 1183,
+  "ClCc": 1184,
+  "54": 1185,
+  "NCCNS": 1186,
+  "COCc": 1187,
+  "COCCC": 1188,
+  "(/[": 1189,
+  "X": 1190,
+  "NCCN": 1191,
+  "OCCCCn": 1192,
+  "CSCCS": 1193,
+  "no": 1194,
+  "NCCCC": 1195,
+  "CCCCc": 1196,
+  "\\": 1197,
+  "NCCSCc": 1198,
+  "-]": 1199,
+  "Ic": 1200,
+  "nonc": 1201,
+  "s": 1202,
+  "COCCNCc": 1203,
+  "CCCCCN": 1204,
+  "SCN": 1205,
+  "ncccn": 1206,
+  "NCCOCc": 1207,
+  "cnoc": 1208,
+  "cnc": 1209,
+  "SN": 1210,
+  "/": 1211,
+  "ccccn": 1212,
+  "COCCNC": 1213,
+  "COCCCOC": 1214,
+  "CNCCN": 1215,
+  "nnc": 1216,
+  "NCCOC": 1217,
+  "ncon": 1218,
+  "ns": 1219,
+  "Cl": 1220,
+  "CNCCC": 1221,
+  "OCCO": 1222,
+  "NCCn": 1223,
+  "nnnc": 1224,
+  "OCCOCCOCCO": 1225,
+  "SCCCc": 1226,
+  "CCCCOCC": 1227,
+  "(=[": 1228,
+  "ncnc": 1229,
+  "Oc": 1230,
+  "@]([": 1231,
+  "NCCCCl": 1232,
+  "NCCCNc": 1233,
+  "NCCCCCC": 1234,
+  "CCCO": 1235,
+  "CCCCNc": 1236,
+  "32": 1237
+}