Fix one_of to generate regex internally, even if caseless or as_keyword given as True

3 files changed, 43 insertions, 22 deletions

diff --git a/CHANGES b/CHANGES
index a1f1d37..e80bead 100644
--- a/CHANGES
+++ b/CHANGES
@@ -2,33 +2,40 @@
 Change Log
 ==========
 
+Version 3.0.2 -
+---------------
+- Performance enhancement to `one_of` to always generate `regex`, even
+  if `caseless` or `as_keyword` args are given as `True` (unless explicitly
+  disabled by passing `use_regex=True`.
+
+
 Version 3.0.1 -
 ---------------
-- Fixed bug where Word(max=n) did not match word groups less than length 'n'.
+- Fixed bug where `Word(max=n)` did not match word groups less than length 'n'.
   Thanks to Joachim Metz for catching this!
 
-- Fixed bug where ParseResults accidentally created recursive contents.
+- Fixed bug where `ParseResults` accidentally created recursive contents.
   Joachim Metz on this one also!
 
-- Fixed bug where warn_on_multiple_string_args_to_oneof warning is raised
+- Fixed bug where `warn_on_multiple_string_args_to_oneof` warning is raised
   even when not enabled.
 
 
 Version 3.0.0 -
 ---------------
 - A consolidated list of all the changes in the 3.0.0 release can be found in
-  docs/whats_new_in_3_0_0.rst.
+  `docs/whats_new_in_3_0_0.rst`.
   (https://github.com/pyparsing/pyparsing/blob/master/docs/whats_new_in_3_0_0.rst)
 
 
 Version 3.0.0.final -
 ---------------------
-- Added support for python -W warning option to call enable_all_warnings() at startup.
-  Also detects setting of PYPARSINGENABLEALLWARNINGS environment variable to any non-blank
+- Added support for python `-W` warning option to call `enable_all_warnings`() at startup.
+  Also detects setting of `PYPARSINGENABLEALLWARNINGS` environment variable to any non-blank
   value.
 
 - Fixed named results returned by `url` to match fields as they would be parsed
-  using urllib.parse.urlparse.
+  using `urllib.parse.urlparse`.
 
 - Early response to `with_line_numbers` was positive, with some requested enhancements:
   . added a trailing "|" at the end of each line (to show presence of trailing spaces);
@@ -40,8 +47,8 @@ Version 3.0.0.final -
   . added mark_control argument to support highlighting of control characters using
     '.' or Unicode symbols, such as "␍" and "␊".
 
-- Modified helpers common_html_entity and replace_html_entity() to use the HTML
-  entity definitions from html.entities.html5.
+- Modified helpers `common_html_entity` and `replace_html_entity()` to use the HTML
+  entity definitions from `html.entities.html5`.
 
 - Updated the class diagram in the pyparsing docs directory, along with the supporting
   .puml file (PlantUML markup) used to create the diagram.
diff --git a/pyparsing/__init__.py b/pyparsing/__init__.py
index b1486ab..bf0fe81 100644
--- a/pyparsing/__init__.py
+++ b/pyparsing/__init__.py
@@ -105,7 +105,7 @@ __version__ = "{}.{}.{}".format(*__version_info__[:3]) + (
     ),
     "",
 )[__version_info__.release_level == "final"]
-__version_time__ = "25 October 2021 15:10 UTC"
+__version_time__ = "26 October 2021 20:39 UTC"
 __versionTime__ = __version_time__
 __author__ = "Paul McGuire <ptmcg.gm+pyparsing@gmail.com>"
 
diff --git a/pyparsing/helpers.py b/pyparsing/helpers.py
index 2d112de..0dde451 100644
--- a/pyparsing/helpers.py
+++ b/pyparsing/helpers.py
@@ -1,5 +1,6 @@
 # helpers.py
 import html.entities
+import re
 
 from . import __diag__
 from .core import *
@@ -253,9 +254,9 @@ def one_of(
     if not symbols:
         return NoMatch()
 
-    if not asKeyword:
-        # if not producing keywords, need to reorder to take care to avoid masking
-        # longer choices with shorter ones
+    # reorder given symbols to take care to avoid masking longer choices with shorter ones
+    # (but only if the given symbols are not just single characters)
+    if any(len(sym) > 1 for sym in symbols):
         i = 0
         while i < len(symbols) - 1:
             cur = symbols[i]
@@ -270,17 +271,30 @@ def one_of(
             else:
                 i += 1
 
-    if not (caseless or asKeyword) and useRegex:
-        # ~ print(strs, "->", "|".join([_escapeRegexChars(sym) for sym in symbols]))
+    if useRegex:
+        re_flags: int = re.IGNORECASE if caseless else 0
+
         try:
-            if len(symbols) == len("".join(symbols)):
-                return Regex(
-                    "[%s]" % "".join(_escapeRegexRangeChars(sym) for sym in symbols)
-                ).set_name(" | ".join(symbols))
+            if all(len(sym) == 1 for sym in symbols):
+                # symbols are just single characters, create range regex pattern
+                patt = "[{}]".format("".join(_escapeRegexRangeChars(sym) for sym in symbols))
             else:
-                return Regex("|".join(re.escape(sym) for sym in symbols)).set_name(
-                    " | ".join(symbols)
-                )
+                patt = "|".join(re.escape(sym) for sym in symbols)
+
+            # wrap with \b word break markers if defining as keywords
+            if asKeyword:
+                patt = r"\b(:?{})\b".format(patt)
+
+            ret = Regex(patt, flags=re_flags).set_name(" | ".join(symbols))
+
+            if caseless:
+                # add parse action to return symbols as specified, not in random
+                # casing as found in input string
+                symbol_map = {sym.lower(): sym for sym in symbols}
+                ret.add_parse_action(lambda s, l, t: symbol_map[t[0].lower()])
+
+            return ret
+
         except sre_constants.error:
             warnings.warn(
                 "Exception creating Regex for one_of, building MatchFirst", stacklevel=2