js: fix string.bytes codegen, readline, add tests for `strings` (#12060)

vlib/builtin/js/array.js.v

@@ -45,9 +45,9 @@ fn (mut a array_buffer) set(ix int, val voidptr) {
 #array_buffer.prototype.set = function(ix,val) { array_buffer_set(this,ix,val); }
 
 struct array {
-mut:
+pub mut:
 	arr array_buffer
-pub:
+
 	len int
 	cap int
 }
@@ -67,6 +67,14 @@ fn v_sort(mut arr array, comparator fn (voidptr, voidptr) int) {
 	}
 }
 
+// trim trims the array length to "index" without modifying the allocated data. If "index" is greater
+// than len nothing will be changed.
+pub fn (mut a array) trim(index int) {
+	if index < a.len {
+		a.len = index
+	}
+}
+
 #function flatIntoArray(target, source, sourceLength, targetIndex, depth) {
 #"use strict";
 #
@@ -368,7 +376,7 @@ pub fn (a &array) free() {
 // todo: once (a []byte) will work rewrite this
 pub fn (a array) bytestr() string {
 	res := ''
-	#a.arr.arr.forEach((item) => res.str += String.fromCharCode(+item))
+	#for (let i = 0;i < a.arr.len.valueOf();i++) res.str += String.fromCharCode(a.arr.get(new int(i)))
 
 	return res
 }

vlib/builtin/js/string.js.v

@@ -71,7 +71,7 @@ pub fn (s string) split(dot string) []string {
 pub fn (s string) bytes() []byte {
 	sep := ''
 	mut arr := s.str.split(sep.str).map(it.charCodeAt(0))
-	#arr = new array(arr)
+	#arr = new array(new array_buffer({arr,index_start: new int(0),len: new int(arr.length)}))
 
 	return arr
 }
@@ -735,11 +735,21 @@ pub fn (s string) index_after(p string, start int) int {
 
 pub fn (s string) split_into_lines() []string {
 	mut res := []string{}
-	#let i = 0
-	#s.str.split('\n').forEach((str) => {
-	#res.arr[i] = new string(str);
-	#})
-
+	if s.len == 0 {
+		return res
+	}
+	mut start := 0
+	mut end := 0
+	for i := 0; i < s.len; i++ {
+		if s[i] == 10 {
+			end = if i > 0 && s[i - 1] == 13 { i - 1 } else { i }
+			res << if start == end { '' } else { s[start..end] }
+			start = i + 1
+		}
+	}
+	if start < s.len {
+		res << s[start..]
+	}
 	return res
 }
 

vlib/readline/readline_js.js.v

@@ -0,0 +1,32 @@
+module readline
+
+#const $readline = require('readline')
+
+struct Termios {}
+
+// Only use standard os.get_line
+// Need implementation for readline capabilities
+//
+// read_line_utf8 blocks execution in a loop and awaits user input
+// characters from a terminal until `EOF` or `Enter` key is encountered
+// in the input stream.
+// read_line_utf8 returns the complete input line as an UTF-8 encoded `[]rune` or
+// an error if the line is empty.
+// The `prompt` `string` is output as a prefix text for the input capturing.
+// read_line_utf8 is the main method of the `readline` module and `Readline` struct.
+
+pub fn (mut r Readline) read_line(prompt string) ?string {
+	res := ''
+	print(prompt)
+	#const rl = $readline.createInterface({input: $process.stdin,output: $process.stdout,prompt: prompt.str})
+	#rl.prompt()
+	#rl.on('line', function (ans) { rl.prompt(); res.str = ans; rl.close();})
+
+	return res
+}
+
+pub fn read_line(prompt string) ?string {
+	mut r := Readline{}
+	s := r.read_line(prompt) ?
+	return s
+}

vlib/strings/builder.js.v

@@ -3,32 +3,37 @@
 // that can be found in the LICENSE file.
 module strings
 
+/*
 pub struct Builder {
 mut:
 	buf []byte
 pub mut:
 	len          int
 	initial_size int = 1
-}
+}*/
+
+pub type Builder = []byte
 
 pub fn new_builder(initial_size int) Builder {
-	return Builder{[]byte{cap: initial_size}, 0, initial_size}
+	return []byte{cap: initial_size}
 }
 
 pub fn (mut b Builder) write_b(data byte) {
-	b.buf << data
+	b << data
 }
 
 pub fn (mut b Builder) write(data []byte) ?int {
 	if data.len == 0 {
 		return 0
 	}
-	b.buf << data
+	b << data
 	return data.len
 }
 
 pub fn (b &Builder) byte_at(n int) byte {
-	return b.buf[n]
+	unsafe {
+		return b[n]
+	}
 }
 
 pub fn (mut b Builder) write_string(s string) {
@@ -37,7 +42,7 @@ pub fn (mut b Builder) write_string(s string) {
 	}
 
 	for c in s {
-		b.buf << c
+		b << c
 	}
 }
 
@@ -46,14 +51,41 @@ pub fn (mut b Builder) writeln(s string) {
 		b.write_string(s)
 	}
 
-	b.buf << 10
+	b << 10
 }
 
 pub fn (mut b Builder) str() string {
 	s := ''
 
-	#for (const c of b.val.buf.arr.arr)
+	#for (const c of b.val.arr.arr)
 	#s.str += String.fromCharCode(+c)
-
+	b.trim(0)
 	return s
 }
+
+pub fn (mut b Builder) cut_last(n int) string {
+	cut_pos := b.len - n
+	x := b[cut_pos..]
+	res := x.bytestr()
+	b.trim(cut_pos)
+	return res
+}
+
+pub fn (mut b Builder) go_back_to(pos int) {
+	b.trim(pos)
+}
+
+// go_back discards the last `n` bytes from the buffer
+pub fn (mut b Builder) go_back(n int) {
+	b.trim(b.len - n)
+}
+
+// cut_to cuts the string after `pos` and returns it.
+// if `pos` is superior to builder length, returns an empty string
+// and cancel further operations
+pub fn (mut b Builder) cut_to(pos int) string {
+	if pos > b.len {
+		return ''
+	}
+	return b.cut_last(b.len - pos)
+}

vlib/strings/builder_test.js.v

@@ -0,0 +1,94 @@
+import strings
+
+type MyInt = int
+
+const maxn = 100000
+
+fn test_sb() {
+	mut sb := strings.new_builder(100)
+	sb.write_string('hi')
+	sb.write_string('!')
+	sb.write_string('hello')
+	assert sb.len == 8
+	sb_end := sb.str()
+	assert sb_end == 'hi!hello'
+	assert sb.len == 0
+	///
+	sb = strings.new_builder(10)
+	sb.write_string('a')
+	sb.write_string('b')
+	assert sb.len == 2
+	assert sb.str() == 'ab'
+	// Test interpolation optimization
+	sb = strings.new_builder(10)
+	x := 10
+	y := MyInt(20)
+	sb.writeln('x = $x y = $y')
+	res := sb.str()
+	assert res[res.len - 1] == `\n`
+	println('"$res"')
+	assert res.trim_space() == 'x = 10 y = 20'
+	//
+	sb = strings.new_builder(10)
+	sb.write_string('x = $x y = $y')
+	assert sb.str() == 'x = 10 y = 20'
+	//$if !windows {
+	sb = strings.new_builder(10)
+	sb.write_string('123456')
+	last_2 := sb.cut_last(2)
+	assert last_2 == '56'
+	final_sb := sb.str()
+	assert final_sb == '1234'
+	//}
+}
+
+fn test_big_sb() {
+	mut sb := strings.new_builder(100)
+	mut sb2 := strings.new_builder(10000)
+	for i in 0 .. maxn {
+		sb.writeln(i.str())
+		sb2.write_string('+')
+	}
+	s := sb.str()
+	lines := s.split_into_lines()
+	assert lines.len == maxn
+	assert lines[0] == '0'
+	assert lines[1] == '1'
+	assert lines[777] == '777'
+	assert lines[98765] == '98765'
+	println(sb2.len)
+	assert sb2.len == maxn
+}
+
+fn test_byte_write() {
+	mut sb := strings.new_builder(100)
+	temp_str := 'byte testing'
+	mut count := 0
+	for word in temp_str {
+		sb.write_b(word)
+		count++
+		assert count == sb.len
+	}
+	sb_final := sb.str()
+	assert sb_final == temp_str
+}
+
+fn test_strings_builder_reuse() {
+	mut sb := strings.new_builder(256)
+	sb.write_string('world')
+	assert sb.str() == 'world'
+	sb.write_string('hello')
+	assert sb.str() == 'hello'
+}
+
+fn test_cut_to() {
+	mut sb := strings.new_builder(16)
+	sb.write_string('hello')
+	assert sb.cut_to(3) == 'lo'
+	assert sb.len == 3
+	assert sb.cut_to(3) == ''
+	assert sb.len == 3
+	assert sb.cut_to(0) == 'hel'
+	assert sb.cut_to(32) == ''
+	assert sb.len == 0
+}

vlib/strings/similarity_test.js.v

@@ -0,0 +1,13 @@
+import strings
+
+fn test_levenshtein_distance() {
+	assert strings.levenshtein_distance('', '') == 0
+	assert strings.levenshtein_distance('one', 'one') == 0
+	assert strings.levenshtein_distance('', 'two') == 3
+	assert strings.levenshtein_distance('three', '') == 5
+	assert strings.levenshtein_distance('bananna', '') == 7
+	assert strings.levenshtein_distance('cats', 'hats') == 1
+	assert strings.levenshtein_distance('hugs', 'shrugs') == 2
+	assert strings.levenshtein_distance('broom', 'shroom') == 2
+	assert strings.levenshtein_distance('flomax', 'volmax') == 3
+}

vlib/strings/strings.js.v

@@ -9,10 +9,8 @@ pub fn repeat(c byte, n int) string {
 }
 
 pub fn repeat_string(s string, n int) string {
-	return ''
-	/*
-	// TODO: uncomment this. It is commented for now, so that `v doc strings` works
-	res := # s.repeat(n)
+	res := ''
+	#res.str = s.str.repeat(n.valueOf())
+
 	return res
-	*/
 }

vlib/strings/strings_test.js.v

@@ -0,0 +1,14 @@
+import strings
+
+fn test_repeat() {
+	assert strings.repeat(`x`, 10) == 'xxxxxxxxxx'
+	assert strings.repeat(`a`, 1) == 'a'
+	assert strings.repeat(`a`, 0) == ''
+}
+
+fn test_repeat_string() {
+	assert strings.repeat_string('abc', 3) == 'abcabcabc'
+	assert strings.repeat_string('abc', 1) == 'abc'
+	assert strings.repeat_string('abc', 0) == ''
+	assert strings.repeat_string('', 200) == ''
+}

vlib/strings/textscanner/textscanner_test.js.v

@@ -0,0 +1,159 @@
+import strings.textscanner
+
+fn test_remaining() {
+	mut s := textscanner.new('abc')
+	assert s.remaining() == 3
+	s.next()
+	s.next()
+	assert s.remaining() == 1
+	s.next()
+	assert s.remaining() == 0
+	s.next()
+	s.next()
+	assert s.remaining() == 0
+	s.reset()
+	assert s.remaining() == 3
+}
+
+fn test_next() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	assert s.next() == `b`
+	assert s.next() == `c`
+	assert s.next() == -1
+	assert s.next() == -1
+	assert s.next() == -1
+}
+
+fn test_skip() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	s.skip()
+	assert s.next() == `c`
+	assert s.next() == -1
+}
+
+fn test_skip_n() {
+	mut s := textscanner.new('abc')
+	s.skip_n(2)
+	assert s.next() == `c`
+	assert s.next() == -1
+}
+
+fn test_peek() {
+	mut s := textscanner.new('abc')
+	assert s.peek() == `a`
+	assert s.peek() == `a`
+	assert s.peek() == `a`
+	//
+	assert s.next() == `a`
+	assert s.next() == `b`
+	assert s.next() == `c`
+	assert s.next() == -1
+}
+
+fn test_peek_n() {
+	mut s := textscanner.new('abc')
+	assert s.peek_n(0) == `a`
+	assert s.peek_n(1) == `b`
+	assert s.peek_n(2) == `c`
+	assert s.peek_n(3) == -1
+	assert s.peek_n(4) == -1
+	//
+	assert s.next() == `a`
+	assert s.next() == `b`
+	assert s.next() == `c`
+	assert s.next() == -1
+}
+
+fn test_back() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	s.back()
+	assert s.next() == `a`
+	assert s.next() == `b`
+	s.back()
+	assert s.next() == `b`
+	assert s.next() == `c`
+	assert s.next() == -1
+}
+
+fn test_back_n() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	s.back_n(10)
+	assert s.next() == `a`
+	assert s.next() == `b`
+	assert s.next() == `c`
+	s.back_n(2)
+	assert s.next() == `b`
+}
+
+fn test_peek_back() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	assert s.next() == `b`
+	// check that calling .peek_back() multiple times
+	// does not change the state:
+	assert s.peek_back() == `a`
+	assert s.peek_back() == `a`
+	assert s.peek_back() == `a`
+	// advance, then peek_back again:
+	assert s.next() == `c`
+	assert s.peek_back() == `b`
+	// peeking before the start:
+	s.reset()
+	assert s.peek_back() == -1
+	// peeking right at the end:
+	s.goto_end()
+	assert s.peek_back() == `b`
+}
+
+fn test_peek_back_n() {
+	mut s := textscanner.new('abc')
+	s.goto_end()
+	assert s.peek_back_n(0) == `c`
+	assert s.peek_back_n(1) == `b`
+	assert s.peek_back_n(2) == `a`
+	assert s.peek_back_n(3) == -1
+	assert s.peek_back_n(4) == -1
+}
+
+fn test_reset() {
+	mut s := textscanner.new('abc')
+	assert s.next() == `a`
+	s.next()
+	s.next()
+	assert s.next() == -1
+	s.reset()
+	assert s.next() == `a`
+}
+
+fn test_current() {
+	mut s := textscanner.new('abc')
+	assert s.current() == -1
+	assert s.next() == `a`
+	assert s.current() == `a`
+	assert s.current() == `a`
+	assert s.peek_back() == -1
+	assert s.next() == `b`
+	assert s.current() == `b`
+	assert s.current() == `b`
+	assert s.peek_back() == `a`
+	assert s.next() == `c`
+	assert s.current() == `c`
+	assert s.next() == -1
+	assert s.current() == `c`
+	assert s.next() == -1
+	assert s.current() == `c`
+	s.reset()
+	assert s.current() == -1
+	assert s.next() == `a`
+	assert s.current() == `a`
+}
+
+fn test_goto_end() {
+	mut s := textscanner.new('abc')
+	s.goto_end()
+	assert s.current() == `c`
+}

vlib/v/gen/js/builtin_types.v

@@ -111,10 +111,18 @@ fn (mut g JsGen) sym_to_js_typ(sym ast.TypeSymbol) string {
 	return styp
 }
 
+/*
 pub fn (mut g JsGen) base_type(t ast.Type) string {
 	mut styp := g.cc_type(t, true)
 	return styp
 }
+*/
+fn (mut g JsGen) base_type(_t ast.Type) string {
+	t := g.unwrap_generic(_t)
+	share := t.share()
+	mut styp := if share == .atomic_t { t.atomic_typename() } else { g.cc_type(t, true) }
+	return styp
+}
 
 pub fn (mut g JsGen) typ(t ast.Type) string {
 	sym := g.table.get_final_type_symbol(t)

vlib/v/gen/js/fn.v

@@ -106,7 +106,7 @@ fn (mut g JsGen) method_call(node ast.CallExpr) {
 		g.inc_indent()
 		g.writeln('try {')
 		g.inc_indent()
-		g.write('return builtin.unwrap(')
+		g.write('return unwrap(')
 	}
 	if node.name == 'str' {
 		mut rec_type := node.receiver_type

vlib/v/gen/js/infix.v

@@ -8,7 +8,8 @@ fn (mut g JsGen) gen_plain_infix_expr(node ast.InfixExpr) {
 	l_sym := g.table.get_final_type_symbol(it.left_type)
 	r_sym := g.table.get_final_type_symbol(it.right_type)
 	greater_typ := g.greater_typ(it.left_type, it.right_type)
-	g.write('new ${g.typ(greater_typ)}( ')
+	cast_ty := if greater_typ == it.left_type { l_sym.cname } else { r_sym.cname }
+	g.write('new ${g.js_name(cast_ty)}( ')
 	g.cast_stack << greater_typ
 	if (l_sym.kind == .i64 || l_sym.kind == .u64) || (r_sym.kind == .i64 || r_sym.kind == .u64) {
 		g.write('BigInt(')

vlib/v/gen/js/js.v

@@ -1398,7 +1398,7 @@ fn (mut g JsGen) gen_expr_stmt_no_semi(it ast.ExprStmt) {
 
 // cc_type whether to prefix 'struct' or not (C__Foo -> struct Foo)
 fn (mut g JsGen) cc_type(typ ast.Type, is_prefix_struct bool) string {
-	sym := g.table.get_final_type_symbol(g.unwrap_generic(typ))
+	sym := g.table.get_type_symbol(g.unwrap_generic(typ))
 	mut styp := sym.cname
 	match mut sym.info {
 		ast.Struct, ast.Interface, ast.SumType {
@@ -2306,10 +2306,10 @@ fn (mut g JsGen) gen_index_expr(expr ast.IndexExpr) {
 	left_typ := g.table.get_type_symbol(expr.left_type)
 	// TODO: Handle splice setting if it's implemented
 	if expr.index is ast.RangeExpr {
-		if left_typ.kind == .array {
-			g.write('array_slice(')
-		} else {
+		if left_typ.kind == .string {
 			g.write('string_slice(')
+		} else {
+			g.write('array_slice(')
 		}
 		g.expr(expr.left)
 		if expr.left_type.is_ptr() {