Update split() for CSV data.

10 files changed, 76 insertions, 18 deletions

diff --git a/ChangeLog b/ChangeLog
index 55d16d3c..99d9fbf9 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2023-03-17         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* field.c (do_split): Modify behavior for --csv.
+
 2023-03-16         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* awk.h (enum do_flag_values): Add DO_CSV.
diff --git a/field.c b/field.c
index 44c153dc..0f0f5192 100644
--- a/field.c
+++ b/field.c
@@ -1173,7 +1173,10 @@ do_split(int nargs)
 	if ((sep->flags & REGEX) != 0)
 		sep = sep->typed_re;
 
-	if (   (sep->re_flags & FS_DFLT) != 0
+	if (do_csv && (sep->re_flags & FS_DFLT) != 0 && nargs == 3) {
+		fs = NULL;
+		parseit = comma_parse_field;
+	} else if (   (sep->re_flags & FS_DFLT) != 0
 	    && current_field_sep() == Using_FS
 	    && ! RS_is_null) {
 		parseit = parse_field;
@@ -1194,15 +1197,6 @@ do_split(int nargs)
 		} else if (fs->stlen == 1 && (sep->re_flags & CONSTANT) == 0) {
 			if (fs->stptr[0] == ' ') {
 				parseit = def_parse_field;
-			} else if (fs->stptr[0] == ',' && ! do_posix) {
-				static bool warned = false;
-
-				parseit = comma_parse_field;
-
-				if (do_lint && ! warned) {
-					warned = true;
-					lintwarn(_("split: CSV parsing is a non-standard extension"));
-				}
 			} else
 				parseit = sc_parse_field;
 		} else {
diff --git a/pc/ChangeLog b/pc/ChangeLog
index 541be9c5..12c14c72 100644
--- a/pc/ChangeLog
+++ b/pc/ChangeLog
@@ -1,3 +1,7 @@
+2023-03-17         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* Makefile.tst: Regenerated.
+
 2023-03-16         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* Makefile.tst: Regenerated.
diff --git a/pc/Makefile.tst b/pc/Makefile.tst
index 72f8a9cb..7228004b 100644
--- a/pc/Makefile.tst
+++ b/pc/Makefile.tst
@@ -190,7 +190,7 @@ GAWK_EXT_TESTS = \
 	aadelete1 aadelete2 aarray1 aasort aasorti argtest arraysort \
 	arraysort2 arraytype asortbool backw badargs beginfile1 beginfile2 \
 	binmode1 charasbytes clos1way clos1way2 clos1way3 clos1way4 \
-	clos1way5 clos1way6 colonwarn commas crlf csv1 dbugeval dbugeval2 \
+	clos1way5 clos1way6 colonwarn commas crlf csv1 csv2 dbugeval dbugeval2 \
 	dbugeval3 dbugeval4 dbugtypedre1 dbugtypedre2 delsub \
 	devfd devfd1 devfd2 dfacheck1 dumpvars \
 	elemnew1 elemnew2 elemnew3 errno exit fieldwdth forcenum fpat1 fpat2 \
@@ -292,7 +292,7 @@ NEED_TRADITIONAL = litoct tradanch rscompat
 NEED_PMA = pma
 
 # List of tests that need --csv
-NEED_CSV = csv1
+NEED_CSV = csv1 csv2
 
 # Lists of tests that run a shell script
 RUN_SHELL = exit fflush localenl modifiers next randtest rtlen rtlen01
@@ -2727,6 +2727,11 @@ csv1:
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
 	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
 
+csv2:
+	@echo $@
+	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
+	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
+
 dbugeval2:
 	@echo $@
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --debug < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
diff --git a/test/ChangeLog b/test/ChangeLog
index f247c540..989d105b 100644
--- a/test/ChangeLog
+++ b/test/ChangeLog
@@ -1,3 +1,8 @@
+2023-03-17         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* Makefile.am (EXTRA_DIST, NEED_CSV, GAWK_EXT_TESTS): New test, csv2.
+	* csv2.awk, csv2.ok: New files.
+
 2023-03-16         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* Gentests: Handle NEED_CSV.
diff --git a/test/Makefile.am b/test/Makefile.am
index 7bb4c983..e0be391c 100644
--- a/test/Makefile.am
+++ b/test/Makefile.am
@@ -225,7 +225,8 @@ EXTRA_DIST = \
 	crlf.ok \
 	csv1.awk \
 	csv1.in \
-	csv1.ok \
+	csv2.awk \
+	csv2.ok \
 	datanonl.awk \
 	datanonl.in \
 	datanonl.ok \
@@ -1506,7 +1507,7 @@ GAWK_EXT_TESTS = \
 	aadelete1 aadelete2 aarray1 aasort aasorti argtest arraysort \
 	arraysort2 arraytype asortbool backw badargs beginfile1 beginfile2 \
 	binmode1 charasbytes clos1way clos1way2 clos1way3 clos1way4 \
-	clos1way5 clos1way6 colonwarn commas crlf csv1 dbugeval dbugeval2 \
+	clos1way5 clos1way6 colonwarn commas crlf csv1 csv2 dbugeval dbugeval2 \
 	dbugeval3 dbugeval4 dbugtypedre1 dbugtypedre2 delsub \
 	devfd devfd1 devfd2 dfacheck1 dumpvars \
 	elemnew1 elemnew2 elemnew3 errno exit fieldwdth forcenum fpat1 fpat2 \
@@ -1608,7 +1609,7 @@ NEED_TRADITIONAL = litoct tradanch rscompat
 NEED_PMA = pma
 
 # List of tests that need --csv
-NEED_CSV = csv1
+NEED_CSV = csv1 csv2
 
 # Lists of tests that run a shell script
 RUN_SHELL = exit fflush localenl modifiers next randtest rtlen rtlen01
diff --git a/test/Makefile.in b/test/Makefile.in
index 28c5ebc5..ce2c7744 100644
--- a/test/Makefile.in
+++ b/test/Makefile.in
@@ -489,7 +489,8 @@ EXTRA_DIST = \
 	crlf.ok \
 	csv1.awk \
 	csv1.in \
-	csv1.ok \
+	csv2.awk \
+	csv2.ok \
 	datanonl.awk \
 	datanonl.in \
 	datanonl.ok \
@@ -1770,7 +1771,7 @@ GAWK_EXT_TESTS = \
 	aadelete1 aadelete2 aarray1 aasort aasorti argtest arraysort \
 	arraysort2 arraytype asortbool backw badargs beginfile1 beginfile2 \
 	binmode1 charasbytes clos1way clos1way2 clos1way3 clos1way4 \
-	clos1way5 clos1way6 colonwarn commas crlf csv1 dbugeval dbugeval2 \
+	clos1way5 clos1way6 colonwarn commas crlf csv1 csv2 dbugeval dbugeval2 \
 	dbugeval3 dbugeval4 dbugtypedre1 dbugtypedre2 delsub \
 	devfd devfd1 devfd2 dfacheck1 dumpvars \
 	elemnew1 elemnew2 elemnew3 errno exit fieldwdth forcenum fpat1 fpat2 \
@@ -1872,7 +1873,7 @@ NEED_TRADITIONAL = litoct tradanch rscompat
 NEED_PMA = pma
 
 # List of tests that need --csv
-NEED_CSV = csv1
+NEED_CSV = csv1 csv2
 
 # Lists of tests that run a shell script
 RUN_SHELL = exit fflush localenl modifiers next randtest rtlen rtlen01
@@ -4490,6 +4491,11 @@ csv1:
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
 	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
 
+csv2:
+	@echo $@
+	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
+	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
+
 dbugeval2:
 	@echo $@
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --debug < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
diff --git a/test/Maketests b/test/Maketests
index d284aab6..ed1a050e 100644
--- a/test/Maketests
+++ b/test/Maketests
@@ -1417,6 +1417,11 @@ csv1:
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
 	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
 
+csv2:
+	@echo $@
+	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --csv >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
+	@-$(CMP) "$(srcdir)"/$@.ok _$@ && rm -f _$@
+
 dbugeval2:
 	@echo $@
 	@-AWKPATH="$(srcdir)" $(AWK) -f $@.awk  --debug < "$(srcdir)"/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >>_$@
diff --git a/test/csv2.awk b/test/csv2.awk
new file mode 100644
index 00000000..93cbca65
--- /dev/null
+++ b/test/csv2.awk
@@ -0,0 +1,34 @@
+BEGIN {
+	text[1] = "a";			expected[1] = 1
+	text[2] = "  a";		expected[2] = 1
+	text[3] = ",a";			expected[3] = 2
+	text[4] = " , a";		expected[4] = 2
+	text[5] = "a,b";		expected[5] = 2
+	text[6] = "a,b,c";		expected[6] = 3
+	text[7] = "\"\"";		expected[7] = 1
+	text[8] = "\"abc\"";		expected[8] = 1
+	text[9] = "\"a\"\"b\"";		expected[9] = 1
+	text[10] = "\"a\",\"b\"";	expected[10] = 2
+	text[11] = "a\"\"b";		expected[11] = 1
+	text[12] = "\"a,b\"";		expected[12] = 1
+	text[13] = "\"\"\"\"";		expected[13] = 1
+	text[14] = "\"\"\"\"\"\"";	expected[14] = 1
+	text[15] = "\"\"\"x\"\"\"";	expected[15] = 1
+	text[16] = ",,\"\"";		expected[16] = 3
+	text[17] = "a\"\"b";		expected[17] = 1
+	text[18] = "a\"b";		expected[18] = 1
+	text[19] = "a''b";		expected[19] = 1
+	text[20] = "\"abc";		expected[20] = 1
+	text[21] = ",,";		expected[21] = 3
+	text[22] = "a,";		expected[22] = 2
+	text[23] = "\"\",";		expected[23] = 2
+	text[24] = ",";			expected[24] = 2
+	text[25] = "\"abc\",def";	expected[25] = 2
+
+	for (i = 1; i <= length(text); i++) {
+		n = split(text[i], array)
+		if (n != expected[i])
+			printf("text[%d] = <%s>, expected %d, got %d\n",
+			       i, text[i], expected[i], n)
+	}
+}
diff --git a/test/csv2.ok b/test/csv2.ok
new file mode 100644
index 00000000..e69de29b
--- /dev/null
+++ b/test/csv2.ok