Moves regular expressions to a dataclass

Author: gnikit

fortls/constants.py

@@ -4,6 +4,8 @@
 import sys
 from dataclasses import dataclass, field
 
+from fortls.regex_patterns import FortranRegularExpressions
+
 PY3K = sys.version_info >= (3, 0)
 
 log = logging.getLogger(__name__)
@@ -81,3 +83,6 @@ class FUN_sig:
     def __post_init__(self):
         if not self.result.name:
             self.result.name = self.name
+
+
+FRegex = FortranRegularExpressions()

fortls/helper_functions.py

@@ -3,19 +3,7 @@
 import os
 from pathlib import Path
 
-from fortls.constants import KEYWORD_ID_DICT, KEYWORD_LIST, log, sort_keywords
-from fortls.regex_patterns import (
-    DQ_STRING_REGEX,
-    FIXED_COMMENT_LINE_MATCH,
-    FREE_FORMAT_TEST,
-    LINE_LABEL_REGEX,
-    LOGICAL_REGEX,
-    NAT_VAR_REGEX,
-    NUMBER_REGEX,
-    OBJBREAK_REGEX,
-    SQ_STRING_REGEX,
-    WORD_REGEX,
-)
+from fortls.constants import KEYWORD_ID_DICT, KEYWORD_LIST, FRegex, log, sort_keywords
 
 
 def expand_name(line: str, char_poss: int) -> str:
@@ -35,7 +23,13 @@ def expand_name(line: str, char_poss: int) -> str:
     """
     # The order here is important.
     # WORD will capture substrings in logical and strings
-    regexs = [LOGICAL_REGEX, SQ_STRING_REGEX, DQ_STRING_REGEX, WORD_REGEX, NUMBER_REGEX]
+    regexs = [
+        FRegex.LOGICAL,
+        FRegex.SQ_STRING,
+        FRegex.DQ_STRING,
+        FRegex.WORD,
+        FRegex.NUMBER,
+    ]
     for r in regexs:
         for num_match in r.finditer(line):
             if num_match.start(0) <= char_poss <= num_match.end(0):
@@ -59,13 +53,13 @@ def detect_fixed_format(file_lines: list[str]) -> bool:
         True if file_lines are of Fixed Fortran style
     """
     for line in file_lines:
-        if FREE_FORMAT_TEST.match(line):
+        if FRegex.FREE_FORMAT_TEST.match(line):
             return False
-        tmp_match = NAT_VAR_REGEX.match(line)
+        tmp_match = FRegex.VAR.match(line)
         if tmp_match and tmp_match.start(1) < 6:
             return False
         # Trailing ampersand indicates free or intersection format
-        if not FIXED_COMMENT_LINE_MATCH.match(line):
+        if not FRegex.FIXED_COMMENT.match(line):
             line_end = line.split("!")[0].strip()
             if len(line_end) > 0 and line_end[-1] == "&":
                 return False
@@ -86,7 +80,7 @@ def strip_line_label(line: str) -> tuple[str, str | None]:
         Output string, Line label returns None if no line label present
     """
 
-    match = LINE_LABEL_REGEX.match(line)
+    match = FRegex.LINE_LABEL.match(line)
     if match is None:
         return line, None
     else:
@@ -118,11 +112,11 @@ def repl_dq(m):
         return '"{0}"'.format(" " * (len(m.group()) - 2))
 
     if maintain_len:
-        out_line = SQ_STRING_REGEX.sub(repl_sq, in_line)
-        out_line = DQ_STRING_REGEX.sub(repl_dq, out_line)
+        out_line = FRegex.SQ_STRING.sub(repl_sq, in_line)
+        out_line = FRegex.DQ_STRING.sub(repl_dq, out_line)
     else:
-        out_line = SQ_STRING_REGEX.sub("", in_line)
-        out_line = DQ_STRING_REGEX.sub("", out_line)
+        out_line = FRegex.SQ_STRING.sub("", in_line)
+        out_line = FRegex.DQ_STRING.sub("", out_line)
     return out_line
 
 
@@ -183,7 +177,7 @@ def find_word_in_line(line: str, word: str) -> tuple[int, int]:
         start and end positions (indices) of the word if not found it returns
         -1, len(word) -1"""
     i0 = -1
-    for poss_name in WORD_REGEX.finditer(line):
+    for poss_name in FRegex.WORD.finditer(line):
         if poss_name.group() == word:
             i0 = poss_name.start()
             break
@@ -417,7 +411,7 @@ def get_var_stack(line):
         final_var += line[section[0] : section[1]]
     #
     if final_var is not None:
-        final_op_split = OBJBREAK_REGEX.split(final_var)
+        final_op_split = FRegex.OBJBREAK.split(final_var)
         return final_op_split[-1].split("%")
     else:
         return None

fortls/langserver.py

@@ -21,6 +21,7 @@
     SELECT_TYPE_ID,
     SUBROUTINE_TYPE_ID,
     VAR_TYPE_ID,
+    FRegex,
     log,
 )
 from fortls.helper_functions import (
@@ -47,14 +48,7 @@
     get_use_tree,
 )
 from fortls.parse_fortran import fortran_file, get_line_context, process_file
-from fortls.regex_patterns import (
-    DQ_STRING_REGEX,
-    INT_STMNT_REGEX,
-    LOGICAL_REGEX,
-    NUMBER_REGEX,
-    SQ_STRING_REGEX,
-    src_file_exts,
-)
+from fortls.regex_patterns import src_file_exts
 from fortls.version import __version__
 
 # Global regexes
@@ -540,6 +534,7 @@ def build_comp(
         if self.autocomplete_no_prefix:
             var_prefix = ""
         # Suggestions for user-defined type members
+        scope_list = []
         if is_member:
             curr_scope = file_obj.ast.get_inner_scope(ac_line + 1)
             type_scope = climb_type_tree(var_stack, curr_scope, self.obj_tree)
@@ -758,14 +753,16 @@ def get_definition(
             # Return a dummy object with the correct type & position in the doc
             if hover_req and curr_scope:
                 var_type = None
-                if NUMBER_REGEX.match(def_name):
+                if FRegex.NUMBER.match(def_name):
                     if any(s in def_name for s in [".", "e", "d"]):
                         var_type = f"{FORTRAN_LITERAL}REAL"
                     else:
                         var_type = f"{FORTRAN_LITERAL}INTEGER"
-                elif LOGICAL_REGEX.match(def_name):
+                elif FRegex.LOGICAL.match(def_name):
                     var_type = f"{FORTRAN_LITERAL}LOGICAL"
-                elif SQ_STRING_REGEX.match(def_name) or DQ_STRING_REGEX.match(def_name):
+                elif FRegex.SQ_STRING.match(def_name) or FRegex.DQ_STRING.match(
+                    def_name
+                ):
                     var_type = f"{FORTRAN_LITERAL}STRING"
                 if var_type:
                     return fortran_var(
@@ -853,7 +850,7 @@ def check_optional(arg, params):
                         break
         # Check keywords
         if (var_obj is None) and (
-            INT_STMNT_REGEX.match(line_prefix[:sub_end]) is not None
+            FRegex.INT_STMNT.match(line_prefix[:sub_end]) is not None
         ):
             key = sub_name.lower()
             for candidate in get_intrinsic_keywords(self.statements, self.keywords, 0):

fortls/objects.py

@@ -24,10 +24,10 @@
     SUBROUTINE_TYPE_ID,
     VAR_TYPE_ID,
     WHERE_TYPE_ID,
+    FRegex,
 )
 from fortls.helper_functions import get_keywords, get_paren_substring, get_var_stack
 from fortls.jsonrpc import path_to_uri
-from fortls.regex_patterns import CLASS_VAR_REGEX, DEF_KIND_REGEX
 
 # Helper types
 VAR_info = NamedTuple(
@@ -1595,7 +1595,7 @@ def __init__(
         self.desc: str = var_desc
         self.keywords: list = keywords
         self.keyword_info: dict = keyword_info
-        self.callable: bool = CLASS_VAR_REGEX.match(var_desc) is not None
+        self.callable: bool = FRegex.CLASS_VAR.match(var_desc) is not None
         self.children: list = []
         self.use: list[USE_line] = []
         self.link_obj = None
@@ -1718,7 +1718,7 @@ def set_external_attr(self):
 
     def check_definition(self, obj_tree, known_types={}, interface=False):
         # Check for type definition in scope
-        type_match = DEF_KIND_REGEX.match(self.desc)
+        type_match = FRegex.DEF_KIND.match(self.desc)
         if type_match is not None:
             var_type = type_match.group(1).strip().lower()
             if var_type == "procedure":