From: NeilBrown <neil@brown.name>
Date: Tue, 28 May 2019 05:57:09 +0000 (+1000)
Subject: Add test code for the scanner
X-Git-Url: https://ocean-lang.org/code/?p=ocean;a=commitdiff_plain;h=4175113ddadf8ce701903fd784b5b5420d3f93f5

Add test code for the scanner

Currently only 85% coverage.  I haven't tested error yet,
or various other things.  I have found bugs though!

Signed-off-by: NeilBrown <neil@brown.name>
---

diff --git a/csrc/scanner-tests.mdc b/csrc/scanner-tests.mdc
new file mode 100644
index 0000000..3646208
--- /dev/null
+++ b/csrc/scanner-tests.mdc
@@ -0,0 +1,515 @@
+# Scanner test code
+
+The focus of these tests is coverage of the code in scanner.mdc
+with the aim of being able to detect regressions.
+
+We have a few different input files, and scan each with a variety
+of different setting and ensure the output is as expected.
+
+The inputs are in sectoins called "test: name" and the expected
+outputs are in "test: name,arg,arg,arg".  As we cannot (yet)
+extract a list of section names, we need to tell the make script
+about each test.
+
+###### File: scanner-tests.mk
+
+	scanner_tests :=
+	## test list
+
+	tests:: scanner_test_suite
+	scanner_test_suite : md2c coverage_scanner
+		@rm -rf coverage; mkdir -p coverage
+		@cp *.gcno coverage
+		@for T in $(scanner_tests); do \
+		    echo -n "Tests $$T ... ";  \
+		    i="$IFS"; IFS=,; set $$T; IFS="$$i"; t=$$1; shift; \
+		    ./md2c scanner-tests.mdc "output: $$T" | grep -v '^#' > .tmp.want; \
+		    echo '~~~~~~~' > .tmp.code ;\
+		    ./md2c scanner-tests.mdc "test: $$t" | grep -v '^#' >> .tmp.code; \
+		    ./coverage_scanner --file .tmp.code $${1+"$$@"} > .tmp.have; \
+		    if ! cmp -s .tmp.want .tmp.have; then \
+		        echo "FAILED"; diff -u .tmp.want .tmp.have; exit 1; fi ; \
+		    echo "PASSED"; \
+		done
+
+		@gcov -o coverage scanner.c libscanner.c > /dev/null 2> /dev/null
+		@mv *.gcov coverage; [ -f .gcov ] && mv .gcov coverage || true
+		@awk '/NOTEST/ { next } /^ *[1-9]/ {ran+=1} /^ *###/ {skip+=1} \
+		    END {printf "coverage: %6.2f%%\n", ran * 100 / (ran + skip); \
+		         if (ran < (ran + skip) *0.85) exit(1) }' \
+		        coverage/scanner.mdc.gcov
+		@rm -f .tmp*
+
+	coverage_scanner: scanner.c libscanner.c
+		$(CC) $(CFLAGS) --coverage -fprofile-dir=coverage -o coverage_scanner \
+			scanner.c libscanner.c \
+			libmdcode.o libnumber.o libstring.o -licuuc -lgmp
+
+
+## Basic tests
+
+Some simple tests... maybe all tests are simple.
+
+###### test list
+	scanner_tests += "test1,if,then,+,-"
+	scanner_tests += "test1,if,then,+,-,/"
+	scanner_tests += "test1,--ignore-indent,if,then,+,-,/"
+	scanner_tests += "test1,--ignore-indent,--ignore-newline,if,then,+,-,/"
+	scanner_tests += "test1,-Cc,--ignore-indent,--ignore-newline,if,then,+,-,/"
+	scanner_tests += "test1,-CcSz,--ignore-indent,--ignore-newline,if,then,+,-,/"
+
+###### test: test1
+
+	A B C
+	  D E
+	   F G
+	 H
+	I
+
+	A
+	  B
+	    C
+	    D
+	  E
+	F
+	  G
+	H
+
+	if else then fi while
+	1234,  1.234 -123.456e45
+	0x1234 +  0x543p3
+	"This is a string" &"so is this"a
+	a = """
+	   This is a multi-
+	   line string that
+	   has three lines
+	   """bb
+	I think # some comments are like this
+	and // some are like this
+	and some /* can go
+	over multiplt
+	lines */
+	divident /+ divisor
+
+###### output: test1,if,then,+,-
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	3:2 in()
+	3:2 ident(D)
+	3:4 ident(E)
+	4:3 in()
+	4:3 ident(F)
+	4:5 ident(G)
+	5:1 newline()
+	5:1 out()
+	5:1 newline()
+	5:1 out()
+	5:1 in()
+	5:1 ident(H)
+	6:0 newline()
+	6:0 out()
+	6:0 newline()
+	6:0 ident(I)
+	8:0 newline()
+	8:0 newline()
+	8:0 ident(A)
+	9:2 in()
+	9:2 ident(B)
+	10:4 in()
+	10:4 ident(C)
+	11:4 newline()
+	11:4 ident(D)
+	12:2 newline()
+	12:2 out()
+	12:2 newline()
+	12:2 ident(E)
+	13:0 newline()
+	13:0 out()
+	13:0 newline()
+	13:0 ident(F)
+	14:2 in()
+	14:2 ident(G)
+	15:0 newline()
+	15:0 out()
+	15:0 newline()
+	15:0 ident(H)
+	17:0 newline()
+	17:0 newline()
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	18:0 newline()
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	19:0 newline()
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	20:0 newline()
+	20:0 string("This is a string")  This is a string
+	20:19 mark(&)
+	20:20 string("so is this"a) a so is this
+	21:0 newline()
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mstring("""\x0a   This is a m..) bb This is a multi-\x0al..
+	26:0 newline()
+	26:0 ident(I)
+	26:2 ident(think)
+	26:8 lcomment(# some comments ar..)
+	27:0 newline()
+	27:0 ident(and)
+	27:4 lcomment(// some are like t..)
+	28:0 newline()
+	28:0 ident(and)
+	28:4 ident(some)
+	28:9 bcomment(/* can go\x0aover mul..)
+	31:0 newline()
+	31:0 ident(divident)
+	31:9 mark(/+)
+	31:12 ident(divisor)
+	32:0 newline()
+	32:0 eof()
+
+###### output: test1,if,then,+,-,/
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	3:2 in()
+	3:2 ident(D)
+	3:4 ident(E)
+	4:3 in()
+	4:3 ident(F)
+	4:5 ident(G)
+	5:1 newline()
+	5:1 out()
+	5:1 newline()
+	5:1 out()
+	5:1 in()
+	5:1 ident(H)
+	6:0 newline()
+	6:0 out()
+	6:0 newline()
+	6:0 ident(I)
+	8:0 newline()
+	8:0 newline()
+	8:0 ident(A)
+	9:2 in()
+	9:2 ident(B)
+	10:4 in()
+	10:4 ident(C)
+	11:4 newline()
+	11:4 ident(D)
+	12:2 newline()
+	12:2 out()
+	12:2 newline()
+	12:2 ident(E)
+	13:0 newline()
+	13:0 out()
+	13:0 newline()
+	13:0 ident(F)
+	14:2 in()
+	14:2 ident(G)
+	15:0 newline()
+	15:0 out()
+	15:0 newline()
+	15:0 ident(H)
+	17:0 newline()
+	17:0 newline()
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	18:0 newline()
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	19:0 newline()
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	20:0 newline()
+	20:0 string("This is a string")  This is a string
+	20:19 mark(&)
+	20:20 string("so is this"a) a so is this
+	21:0 newline()
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mstring("""\x0a   This is a m..) bb This is a multi-\x0al..
+	26:0 newline()
+	26:0 ident(I)
+	26:2 ident(think)
+	26:8 lcomment(# some comments ar..)
+	27:0 newline()
+	27:0 ident(and)
+	27:4 lcomment(// some are like t..)
+	28:0 newline()
+	28:0 ident(and)
+	28:4 ident(some)
+	28:9 bcomment(/* can go\x0aover mul..)
+	31:0 newline()
+	31:0 ident(divident)
+	31:9 /
+	31:10 +
+	31:12 ident(divisor)
+	32:0 newline()
+	32:0 eof()
+
+###### output: test1,--ignore-indent,if,then,+,-,/
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	2:5 newline()
+	3:2 ident(D)
+	3:4 ident(E)
+	3:5 newline()
+	4:3 ident(F)
+	4:5 ident(G)
+	4:6 newline()
+	5:1 ident(H)
+	5:2 newline()
+	6:0 ident(I)
+	6:1 newline()
+	7:0 newline()
+	8:0 ident(A)
+	8:1 newline()
+	9:2 ident(B)
+	9:3 newline()
+	10:4 ident(C)
+	10:5 newline()
+	11:4 ident(D)
+	11:5 newline()
+	12:2 ident(E)
+	12:3 newline()
+	13:0 ident(F)
+	13:1 newline()
+	14:2 ident(G)
+	14:3 newline()
+	15:0 ident(H)
+	15:1 newline()
+	16:0 newline()
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	17:21 newline()
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	18:24 newline()
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	19:17 newline()
+	20:0 string("This is a string")  This is a string
+	20:19 mark(&)
+	20:20 string("so is this"a) a so is this
+	20:33 newline()
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mstring("""\x0a   This is a m..) bb This is a multi-\x0al..
+	25:8 newline()
+	26:0 ident(I)
+	26:2 ident(think)
+	26:8 lcomment(# some comments ar..)
+	26:37 newline()
+	27:0 ident(and)
+	27:4 lcomment(// some are like t..)
+	27:25 newline()
+	28:0 ident(and)
+	28:4 ident(some)
+	28:9 bcomment(/* can go\x0aover mul..)
+	30:8 newline()
+	31:0 ident(divident)
+	31:9 /
+	31:10 +
+	31:12 ident(divisor)
+	31:19 newline()
+	32:0 eof()
+
+###### output: test1,--ignore-indent,--ignore-newline,if,then,+,-,/
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	3:2 ident(D)
+	3:4 ident(E)
+	4:3 ident(F)
+	4:5 ident(G)
+	5:1 ident(H)
+	6:0 ident(I)
+	8:0 ident(A)
+	9:2 ident(B)
+	10:4 ident(C)
+	11:4 ident(D)
+	12:2 ident(E)
+	13:0 ident(F)
+	14:2 ident(G)
+	15:0 ident(H)
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	20:0 string("This is a string")  This is a string
+	20:19 mark(&)
+	20:20 string("so is this"a) a so is this
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mstring("""\x0a   This is a m..) bb This is a multi-\x0al..
+	26:0 ident(I)
+	26:2 ident(think)
+	26:8 lcomment(# some comments ar..)
+	27:0 ident(and)
+	27:4 lcomment(// some are like t..)
+	28:0 ident(and)
+	28:4 ident(some)
+	28:9 bcomment(/* can go\x0aover mul..)
+	31:0 ident(divident)
+	31:9 /
+	31:10 +
+	31:12 ident(divisor)
+	32:0 eof()
+
+###### output: test1,-Cc,--ignore-indent,--ignore-newline,if,then,+,-,/
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	3:2 ident(D)
+	3:4 ident(E)
+	4:3 ident(F)
+	4:5 ident(G)
+	5:1 ident(H)
+	6:0 ident(I)
+	8:0 ident(A)
+	9:2 ident(B)
+	10:4 ident(C)
+	11:4 ident(D)
+	12:2 ident(E)
+	13:0 ident(F)
+	14:2 ident(G)
+	15:0 ident(H)
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	20:0 string("This is a string")  This is a string
+	20:19 mark(&)
+	20:20 string("so is this"a) a so is this
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mstring("""\x0a   This is a m..) bb This is a multi-\x0al..
+	26:0 ident(I)
+	26:2 ident(think)
+	27:0 ident(and)
+	28:0 ident(and)
+	28:4 ident(some)
+	31:0 ident(divident)
+	31:9 /
+	31:10 +
+	31:12 ident(divisor)
+	32:0 eof()
+
+###### output: test1,-CcSz,--ignore-indent,--ignore-newline,if,then,+,-,/
+	Tokenizing: 
+	2:0 ident(A)
+	2:2 ident(B)
+	2:4 ident(C)
+	3:2 ident(D)
+	3:4 ident(E)
+	4:3 ident(F)
+	4:5 ident(G)
+	5:1 ident(H)
+	6:0 ident(I)
+	8:0 ident(A)
+	9:2 ident(B)
+	10:4 ident(C)
+	11:4 ident(D)
+	12:2 ident(E)
+	13:0 ident(F)
+	14:2 ident(G)
+	15:0 ident(H)
+	17:0 if
+	17:3 ident(else)
+	17:8 then
+	17:13 ident(fi)
+	17:16 ident(while)
+	18:0 number(1234)  1234
+	18:4 mark(,)
+	18:7 number(1.234 )  617/500
+	18:13 -
+	18:14 number(123.456e45)  123456000000000000000000000000000000000000000000
+	19:0 number(0x1234 )  4660
+	19:7 +
+	19:10 number(0x543p3)  10776
+	20:0 mark(")
+	20:1 ident(This)
+	20:6 ident(is)
+	20:9 ident(a)
+	20:11 ident(string)
+	20:17 mark(")
+	20:19 mark(&")
+	20:21 ident(so)
+	20:24 ident(is)
+	20:27 ident(this)
+	20:31 mark(")
+	20:32 ident(a)
+	21:0 ident(a)
+	21:2 mark(=)
+	21:4 mark(""")
+	22:3 ident(This)
+	22:8 ident(is)
+	22:11 ident(a)
+	22:13 ident(multi)
+	22:18 -
+	23:3 ident(line)
+	23:8 ident(string)
+	23:15 ident(that)
+	24:3 ident(has)
+	24:7 ident(three)
+	24:13 ident(lines)
+	25:3 mark(""")
+	25:6 ident(bb)
+	26:0 ident(I)
+	26:2 ident(think)
+	27:0 ident(and)
+	28:0 ident(and)
+	28:4 ident(some)
+	31:0 ident(divident)
+	31:9 /
+	31:10 +
+	31:12 ident(divisor)
+	32:0 eof()