Add ECMA definition of symmetry operation regexp and associated tests

GNUmakefile

@@ -55,6 +55,9 @@
 #   - tests/generated/identifiers.ere
 #   - tests/generated/numbers.ere
 #   - tests/generated/strings.ere
+#   - tests/generated/symops.pcre
+#   - tests/generated/symop_definitions.pcre
+#   - tests/generated/symops.ecma
 #
 #
 # Targets for testing / auditing the specification

optimade.rst

@@ -4095,35 +4095,48 @@ The strings below contain Extended Regular Expressions (EREs) to recognize ident
 The Symmetry Operation String Regular Expressions
 -------------------------------------------------
 
-Symmetry operation strings that comprise the :property:`space\_group\_symmetry\_operations\_xyz` property MUST conform to the following regular expressions.
-The regular expressions are recorded in the Perl Compatible Regular Expression (PCRE) syntax, with `Perl extensions <https://perldoc.perl.org/perlre>`__ used for readability.
-The :val:`symop_definitions` section defines several variables in Perl syntax that capture common parts of the regular expressions (REs) and need to be interpolated into the final REs used for matching.
-The :val:`symops` section contains the REs themselves.
-The whitespace characters in these definitions are not significant; if used in Perl programs, these expressions MUST be processed with the :code:`/x` RE option.
-A working example of these REs in action can be found in the :code:`tests/cases/pcre_symops_001.sh` and other test cases.
+Symmetry operation strings that comprise the :property:`space\_group\_symmetry\_operation\_xyz` property MUST conform to the following regular expressions.
+The regular expressions are recorded below in two forms, one in a more readable form using variables and the other as an explicit pattern compatible with the `OPTIMADE Regular Expression Format`_.
 
-.. code:: PCRE
+- Perl Compatible Regular Expression (PCRE) syntax, with `Perl extensions <https://perldoc.perl.org/perlre>`__ used for readability and expressivity.
+  The :val:`symop_definitions` section defines several variables in Perl syntax that capture common parts of the regular expressions (REs) and need to be interpolated into the final REs used for matching.
+  The :val:`symops` section contains the REs themselves.
+  The whitespace characters in these definitions are not significant; if used in Perl programs, these expressions MUST be processed with the :code:`/x` RE modifier.
+  A working example of these REs in action can be found in the :code:`tests/cases/pcre_symops_001.sh` and other test cases.
 
-    #BEGIN PCRE symop_definitions
 
-    $translations = '1\/2|[12]\/3|[1-3]\/4|[1-5]\/6';
+  .. code:: PCRE
 
-    $symop_translation_appended = "[-+]? [xyz] ([-+][xyz])? ([-+] ($translations) )?";
-    $symop_translation_prepended = "[-+]? ($translations) ([-+] [xyz] ([-+][xyz])? )?";
+     #BEGIN PCRE symop_definitions
 
-    $symop_re = "($symop_translation_appended|$symop_translation_prepended)";
+     $translations = '1\/2|[12]\/3|[1-3]\/4|[1-5]\/6';
 
-    #END PCRE symop_definitions
+     $symop_translation_appended = "[-+]? [xyz] ([-+][xyz])? ([-+] ($translations) )?";
+     $symop_translation_prepended = "[-+]? ($translations) ([-+] [xyz] ([-+][xyz])? )?";
 
-.. code:: PCRE
+     $symop_re = "($symop_translation_appended|$symop_translation_prepended)";
 
-    #BEGIN PCRE symops
+     #END PCRE symop_definitions
 
-    ^ # From the beginning of the string...
-    ($symop_re)(,$symop_re){2}
-    $ # ... match to the very end of the string
+  .. code:: PCRE
 
-    #END PCRE symops
+     #BEGIN PCRE symops
+
+     ^ # From the beginning of the string...
+     ($symop_re)(,$symop_re){2}
+     $ # ... match to the very end of the string
+
+     #END PCRE symops
+
+- The regular expression is also provided in an expanded form:
+
+  .. code:: ECMA
+
+     #BEGIN ECMA symops
+
+     ^([-+]?[xyz]([-+][xyz])?([-+](1/2|[12]/3|[1-3]/4|[1-5]/6))?|[-+]?(1/2|[12]/3|[1-3]/4|[1-5]/6)([-+][xyz]([-+][xyz])?)?),([-+]?[xyz]([-+][xyz])?([-+](1/2|[12]/3|[1-3]/4|[1-5]/6))?|[-+]?(1/2|[12]/3|[1-3]/4|[1-5]/6)([-+][xyz]([-+][xyz])?)?),([-+]?[xyz]([-+][xyz])?([-+](1/2|[12]/3|[1-3]/4|[1-5]/6))?|[-+]?(1/2|[12]/3|[1-3]/4|[1-5]/6)([-+][xyz]([-+][xyz])?)?)$
+
+     #END ECMA symops
 
 OPTIMADE JSON lines partial data format
 ---------------------------------------

tests/cases/ecma_symops_001.sh

@@ -0,0 +1,23 @@
+#! /bin/sh
+
+# Test case: test if the provided ECMA-compatible regular expression correctly
+# recognises symmetry operation strings.
+
+#BEGIN DEPEND
+
+INPUT_GRAMMAR=tests/generated/symops.ecma
+
+#END DEPEND
+
+
+/usr/bin/env python << EOF
+import re
+import sys
+with open("${INPUT_GRAMMAR}") as f:
+    expression = [line.strip() for line in f.readlines() if line.strip() and not line.strip().startswith("#")][0]
+
+with open("tests/inputs/symops.lst") as cases:
+    for case in cases:
+        if re.match(expression, case):
+            print(case, end="")
+EOF

tests/cases/symops_pcre_to_ecma_001.sh

@@ -0,0 +1,44 @@
+#! /bin/sh
+
+# Test case: tests the equivalence of the provided PCRE and ECMA regular
+# expressions used in the validation of space group symmetry operations.
+# The equivalence is tested by translating the expression from PCRE to the
+# subset of the ECMA 262 dialect supported by OPTIMADE.
+
+set -ue
+
+#BEGIN DEPEND
+
+INPUT_PCRE_DEFS=tests/generated/symop_definitions.pcre
+INPUT_PCRE_GRAMMAR=tests/generated/symops.pcre
+INPUT_ECMA_GRAMMAR=tests/generated/symops.ecma
+
+#END DEPEND
+
+PCRE_REGEX=$( \
+    grep -v -e '^ *#' -e '^\s+$' ${INPUT_PCRE_GRAMMAR} | \
+    perl -ne 's/\s+//g; s/#.*//; s/[\$]$/\\\$/; print;' | \
+    perl -ne 's/^[\^][(](.+)[)][(](.+)[)]\{2\}\\[\$]$/^$1$2$2$3\\\$/; print;' \
+)
+
+EXPANDED_PCRE_REGEX=$( \
+    perl -I. -w \
+        -e "require '${INPUT_PCRE_DEFS}';" \
+        -e "my \$extended_regex = \"${PCRE_REGEX}\";" \
+        -e '$extended_regex =~ s/[\s\\]+//g;' \
+        -e 'print $extended_regex;' \
+)
+
+ECMA_REGEX=$( \
+    grep -v -e '^ *#' -e '^\s+$' ${INPUT_ECMA_GRAMMAR} | \
+    perl -n -e 's/\s+//g; print;' \
+)
+
+if [ "${EXPANDED_PCRE_REGEX}" = "${ECMA_REGEX}" ]
+then
+    printf '%s\n' 'PASS: expanded regular expressions match.'
+else
+    printf '%s\n' 'FAIL: expanded regular expressions do not match.'
+    echo "PCRE: ${EXPANDED_PCRE_REGEX}"
+    echo "ECMA: ${ECMA_REGEX}"
+fi

tests/makefiles/Makelocal-grammars

@@ -20,7 +20,7 @@ EBNF_FILES    ?= ${GRAMMARS:%=${GRAMMAR_DIR}/%.ebnf}
 GRAMMAR_FILES ?= ${EBNF_FILES:%.ebnf=%.g}
 
 REGEXPS = $(sort $(shell awk '/^ *${RE_START_STRING}/{print $$3}' ${RST_FILES} | tr -d "\r"))
-REGEXP_FILES = ${REGEXPS:%=${GRAMMAR_DIR}/%.ere} ${REGEXPS:%=${GRAMMAR_DIR}/%.pcre}
+REGEXP_FILES = ${REGEXPS:%=${GRAMMAR_DIR}/%.ere} ${REGEXPS:%=${GRAMMAR_DIR}/%.pcre} ${REGEXPS:%=${GRAMMAR_DIR}/%.ecma}
 
 GRAMMAR_DEPENDENCIES = .grammars.d
 
@@ -47,6 +47,8 @@ ${GRAMMAR_DEPENDENCIES}: ${RST_FILES}
 		$^ | tr -d "\r" >> $@
 	awk '/^ *${RE_START_STRING} PCRE/{print "${GRAMMAR_DIR}/"$$3".pcre:", FILENAME}' \
 		$^ | tr -d "\r" >> $@
+	awk '/^ *${RE_START_STRING} ECMA/{print "${GRAMMAR_DIR}/"$$3".ecma:", FILENAME}' \
+		$^ | tr -d "\r" >> $@
 
 ${GRAMMAR_DIR}/%.ebnf:
 	awk '/^ *${GRAMMAR_START_STRING} $*/,/^ *${GRAMMAR_END_STRING} $*/' $< \
@@ -60,6 +62,10 @@ ${GRAMMAR_DIR}/%.pcre:
 	awk '/^ *${RE_START_STRING} PCRE $*/,/^ *${RE_END_STRING} PCRE $*/' $< \
 		| sed 's/^    //' | tr -d "\r" > $@
 
+${GRAMMAR_DIR}/%.ecma:
+	awk '/^ *${RE_START_STRING} ECMA $*/,/^ *${RE_END_STRING} ECMA $*/' $< \
+		| sed 's/^    //' | tr -d "\r" > $@
+
 .PHONY: tools
 
 ${GRAMMAR_DIR}/%.g: ${GRAMMAR_DIR}/%.ebnf | tools