Fix array indexing to disallow numeric string.

9 files changed, 147 insertions, 2 deletions

diff --git a/ChangeLog b/ChangeLog
index 5b99d92c..44828d42 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2016-05-12         Arnold Robbins       <arnold@skeeve.com>
+
+	* str_array.c (str_lookup): Remove MAYBE_NUM from subscript flags.
+	Bug reported by Andres Legarra <Andres.Legarra@toulouse.inra.fr>.
+
 2016-05-12  Eli Zaretskii  <eliz@gnu.org>
 
 	* nonposix.h: Add prototypes for Posix functions emulated in pc/*
diff --git a/str_array.c b/str_array.c
index db24980a..9514d515 100644
--- a/str_array.c
+++ b/str_array.c
@@ -168,6 +168,15 @@ str_lookup(NODE *symbol, NODE *subs)
 		subs = dupnode(subs);
 	}
 
+	/*
+	 * Repeat after me: "Array indices are always strings."
+	 * "Array indices are always strings."
+	 * "Array indices are always strings."
+	 * "Array indices are always strings."
+	 * ....
+	 */
+	subs->flags &= ~MAYBE_NUM;
+
 	getbucket(b);
 	b->ahnext = symbol->buckets[hash1];
 	symbol->buckets[hash1] = b;
diff --git a/test/ChangeLog b/test/ChangeLog
index 11908b40..e9ea5770 100644
--- a/test/ChangeLog
+++ b/test/ChangeLog
@@ -1,3 +1,8 @@
+2016-05-12         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* Makefile.am (arrayind1): New test.
+	* arrayind1.awk, arrayind1.in, arrayind1.ok: New files.
+
 2016-04-27         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* Makefile.am (rscompat): New test.
diff --git a/test/Makefile.am b/test/Makefile.am
index 844c87fb..07637e12 100644
--- a/test/Makefile.am
+++ b/test/Makefile.am
@@ -52,6 +52,9 @@ EXTRA_DIST = \
 	argarray.ok \
 	argtest.awk \
 	argtest.ok \
+	arrayind1.awk \
+	arrayind1.in \
+	arrayind1.ok \
 	arrayparm.awk \
 	arrayparm.ok \
 	arrayprm2.awk \
@@ -1041,7 +1044,7 @@ CLEANFILES = core core.* fmtspcl.ok
 
 # try to keep these sorted. each letter starts a new line
 BASIC_TESTS = \
-	addcomma anchgsub argarray arrayparm arrayprm2 arrayprm3 \
+	addcomma anchgsub argarray arrayind1 arrayparm arrayprm2 arrayprm3 \
 	arrayref arrymem1 arryref2 arryref3 arryref4 arryref5 arynasty \
 	arynocls aryprm1 aryprm2 aryprm3 aryprm4 aryprm5 aryprm6 aryprm7 \
 	aryprm8 aryprm9 arysubnm asgext awkpath \
diff --git a/test/Makefile.in b/test/Makefile.in
index 5e55b1b6..cb42bfed 100644
--- a/test/Makefile.in
+++ b/test/Makefile.in
@@ -309,6 +309,9 @@ EXTRA_DIST = \
 	argarray.ok \
 	argtest.awk \
 	argtest.ok \
+	arrayind1.awk \
+	arrayind1.in \
+	arrayind1.ok \
 	arrayparm.awk \
 	arrayparm.ok \
 	arrayprm2.awk \
@@ -1297,7 +1300,7 @@ CLEANFILES = core core.* fmtspcl.ok
 
 # try to keep these sorted. each letter starts a new line
 BASIC_TESTS = \
-	addcomma anchgsub argarray arrayparm arrayprm2 arrayprm3 \
+	addcomma anchgsub argarray arrayind1 arrayparm arrayprm2 arrayprm3 \
 	arrayref arrymem1 arryref2 arryref3 arryref4 arryref5 arynasty \
 	arynocls aryprm1 aryprm2 aryprm3 aryprm4 aryprm5 aryprm6 aryprm7 \
 	aryprm8 aryprm9 arysubnm asgext awkpath \
@@ -2598,6 +2601,11 @@ anchgsub:
 	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
 	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
 
+arrayind1:
+	@echo $@
+	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
+	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
+
 arrayparm:
 	@echo $@
 	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
diff --git a/test/Maketests b/test/Maketests
index 16b1e826..7f9fe1c5 100644
--- a/test/Maketests
+++ b/test/Maketests
@@ -10,6 +10,11 @@ anchgsub:
 	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
 	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
 
+arrayind1:
+	@echo $@
+	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
+	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
+
 arrayparm:
 	@echo $@
 	@AWKPATH="$(srcdir)" $(AWK) -f $@.awk  >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
diff --git a/test/arrayind1.awk b/test/arrayind1.awk
new file mode 100755
index 00000000..29fe80bc
--- /dev/null
+++ b/test/arrayind1.awk
@@ -0,0 +1,99 @@
+#!/usr/local/bin/awk -f
+# this script renums pedigrees with metafounders
+# so that they are added *before*regular animals
+# mf are ascertained because they are not in the 1st column
+# also ordered by generation assuming that all metafounders
+# start at generation 1
+#
+#
+	function maxval (a,b) { 
+		max=b 
+		if (a>b){max=a} 
+		return max 
+	}
+BEGIN{
+	pos[0]=0 # actual position in the original data file
+	changed=1
+	posout[0]=0 # position in the new file
+	last=0
+	cum[0]=0 #number of times it acts as parent
+	gen[0]=0
+}
+# read and store pedigree file
+{
+	pos[$1]=NR
+	dam[$1]=$2
+	sire[$1]=$3
+	included[$1]=0
+	gen[$1]=9999
+}
+END{
+    #printf("%10s%10s\n",NR,"animals") > "/dev/stderr" 
+    # find out who is a genetic group
+    for (x in sire){
+        if(! (sire[x] in pos)){ 
+            is_group[sire[x]]=1 
+		#printf("%10s%10s\n","sire",sire[x]) > "/dev/stderr"
+		cum[sire[x],"sire"]++
+	}
+    }
+    for (x in dam){
+        if(! (dam[x] in pos)){ 
+            is_group[dam[x]]=1 
+		#printf("%10s%10s\n","dam",dam[x]) > "/dev/stderr"
+		cum[dam[x],"dam"]++
+        }
+    }
+    # compute numbers, but don't write them out
+    #printf("%16s\n","info on mf") > "/dev/stderr"
+    for (x in is_group){
+        nmf++
+        posout[x]=posout[last]+1
+	printf("%16s%16s%16s%16s%16s%16s%16s\n", posout[x],0,0,x,0,0,0)
+        included[x]=1
+        gen[x]=0
+        last=x
+        cumgen[0]++
+	#printf("%10s%16s%10s%16s%10s%10s%10s\n","group",last," included as",posout[x]," as sire, as dam",cum[x,"sire"],cum[x,"dam"]) > "/dev/stderr"
+    }
+    # regular individuals
+	printf("%16s\n","recoding animals") > "/dev/stderr"
+	iter=1
+	#--> comment next line and this element of the associative array becomes null
+	#printf("%s%16s\n","at the beginning it should be empty: ",posout["00000779770060"]) > "/dev/stderr"
+	#-->
+	while (changed){
+		changed=0
+		for (x in pos){
+			if(x!=0 && !included[x]){
+			# change to number of generation
+				#	if(included[dam[x]] && included[sire[x]]){
+						if((gen[dam[x]]<iter) && (gen[sire[x]]<iter)){
+							# the new code is actually the order animals are printed
+							posout[x]=posout[last]+1
+							gen[x]=maxval(gen[dam[x]],gen[sire[x]])+1
+							printf("%16s%16s%16s%16s%16s%16s%16s\n", posout[x],posout[dam[x]],posout[sire[x]],x,dam[x],sire[x],gen[x])
+							included[x]=1
+							changed=1
+							last=x
+							nanim++
+							cumgen[gen[x]]++
+						}
+				#	}
+			}
+        	}
+		iter++
+		#-->
+		printf("%s%16s\n","in the loop: ",posout["00000779770060"]) > "/dev/stderr"
+		#-->
+		#printf("%10s%10s%10s%10s%16s%16s\n","round",iter,"included",posout[last],"last ",last) > "/dev/stderr"
+	}
+	#printf("%10s%10s%10s%10s%10s%10s\n","metafounders:",nmf," animals",nanim," total",posout[last]) > "/dev/stderr"
+	#printf("%16s%16s\n","pseudogenerations:",gen[last]) > "/dev/stderr"
+	    for (x in cumgen){
+		#printf("%10s%16s%10s%16s\n","pseudogeneration:",x," including: ",cumgen[x]) > "/dev/stderr"
+	    }
+	#--> this is the guilty line
+	printf("%s%16s\n","at the end:  ",posout["00000779770060"]) > "/dev/stderr"
+	#-->
+}
diff --git a/test/arrayind1.in b/test/arrayind1.in
new file mode 100644
index 00000000..915d8657
--- /dev/null
+++ b/test/arrayind1.in
@@ -0,0 +1,2 @@
+64000757790003 00000779770060 00000757740011 1979 2 1975 0
+00000779770060 1975su1du1 1975su1du1 1975 1 1975 1975su1du1
diff --git a/test/arrayind1.ok b/test/arrayind1.ok
new file mode 100644
index 00000000..c68346ea
--- /dev/null
+++ b/test/arrayind1.ok
@@ -0,0 +1,9 @@
+recoding animals
+in the loop:                3
+in the loop:                3
+in the loop:                3
+at the end:                 3
+               1               0               0  00000757740011               0               0               0
+               2               0               0      1975su1du1               0               0               0
+               3               2               2  00000779770060      1975su1du1      1975su1du1               1
+               4               3               1  64000757790003  00000779770060  00000757740011               2