[Template merge] Simplify test script, use helper scripts in the core…

… instead

m4/giella-macros.m4

@@ -88,7 +88,7 @@ AC_MSG_RESULT([$GIELLA_CORE])
 ###############################################################
 ### This is the version of the Giella Core that we require. ###
 ### UPDATE AS NEEDED.
-_giella_core_min_version=0.17.1
+_giella_core_min_version=0.18.0
 
 # GIELLA_CORE/GTCORE env. variable, required by the infrastructure to find scripts:
 AC_ARG_VAR([GIELLA_CORE], [directory for the Giella infra core scripts and other required resources])

test/src/morphology/Makefile.am

@@ -2,6 +2,8 @@
 ## Copyright: Sámediggi/Divvun/UiT
 ## Licence: GPL v3+
 
+TESTS_ENVIRONMENT = export GIELLA_CORE=$(GIELLA_CORE);
+
 ######## Test targets: ###########
 
 # List here (space separated) all test scripts that should be run

test/src/morphology/tag_test.sh

@@ -11,41 +11,29 @@ lexctags=$(mktemp -t giella-tag_test.XXXXXXXXXXX)
 roottags=$(mktemp -t giella-tag_test.XXXXXXXXXXX)
 trap 'rm -f "${lexctags}" "${roottags}"' EXIT
 
-# Ensure we're in langs/sme:
-# cd "$(dirname "$0")"/../../.. || exit 1
+# Get giella-core from the test environment:
+giella_core=$GIELLA_CORE
 
+# If verbose:
 if [[ $1 == "-v" ]]; then
     echo "$0: Are there tags not declared in root.lexc or misspelled?"
 fi
 
+# Extract USED tags:
 sed -e '1,/LEXICON Root/d' < \
     ../../../src/fst/lexicon.tmp.lexc | # Extract all lines after LEXICON Root
-    cut -d'!' -f1                     | # get rid of comments
-    grep ';'                          | # Get only lines with 
-    cut -d';' -f1                     | # get everything in front of ;
-    tr ' \t' '\n'                     | # turn all spaces and tabs into newlines
-    grep -E '[\+|@]'                  | # grep relevant stuff only
-    cut -d':' -f1                     | # get rid of surface side
-    sed 's/+/¢+/g'                    | # prepare for tag isolation
-    sed 's/@@/@¢@/g'                  | # prepare for flag diacritic splitting
-    tr '¢"#' '\n'                     | # replace some symbols with newlines = split tags
-    grep -E '^(\+|@)[A-Za-z]'         | # grep only relevant lines / symbols
-    perl -pe "s#^(\+[^@]+)@#\1\n@#g"  | # do a final split of +Tag@C.FLAG@
-    perl -pe "s#^(@[^@]+@)#\1\n#g"    | # do a final cleanup of @C.FLAG@abc
-    grep -E '^(\+|@)[A-Za-z]'         | # grep only relevant lines / symbols
-    sed 's/\-$//'                     | # Get rid of final hyphens, they are bogus
+    ${giella_core}/scripts/extract-used-tags.sh | # Extract tags, local mods after this line:
     LC_ALL=no_NO.UTF8 sort -u         \
     > "${lexctags}"
 
+# Extract DEFINED tags:
 sed -n '/LEXICON Root/q;p' \
     ../../../src/fst/lexicon.tmp.lexc | # Extract all lines before LEXICON Root
-    cut -d'!' -f1                     | # Remove comments
-    sed 's/Multichar_Symbols//'       | # Remove the string Multichar_Symbols
-    tr ' \t' '\n'                     | # Change all spaces and tabs to newlines
-    grep -E '^(\+|@)'                 | # Extract tags and flag diacritics
+    ${giella_core}/scripts/extract-defined-tags.sh | # Extract tags, local mods after this line:
     LC_ALL=no_NO.UTF8 sort -u         \
     > "${roottags}"
 
+# Compare the two sets of tags, report and fail if there is a diff:
 check=$(LC_ALL=no_NO.UTF8 comm -23 "${lexctags}" "${roottags}")
 if [[ -n "${check}" ]]; then
     echo "$0: Have a look at these:"
@@ -54,6 +42,3 @@ if [[ -n "${check}" ]]; then
 elif [[ $1 == "-v" ]]; then
     echo "$0: No errors found."
 fi
-
-#cat src/fst/clitics.lexc src/fst/compounding.lexc src/fst/affixes/*lexc |cut -d '!' -f1 | grep ';' |tr -s ' ' | sed 's/^ //' |grep ':' |cut -d ':' -f1 | sed 's/\+/¢+/g' | tr '¢' '\n' |sort | uniq -c |sort -n |less
-# visuell test: cat src/fst/clitics.lexc src/fst/compounding.lexc src/fst/affixes/*lexc |cut -d '!' -f1 | grep ';' |tr -s ' ' | sed 's/^ //' |grep ':' |cut -d ':' -f1 | tr -d '0' | sed 's/\+/¢+/g' | tr '¢' '\n' |egrep -v '^(\+|\@|<)' |grep -v '^$' |less