1
0
mirror of https://github.com/vlang/v.git synced 2023-08-10 21:13:21 +03:00
v/vlib/builtin/string.v

1503 lines
28 KiB
V
Raw Normal View History

2020-01-23 23:04:46 +03:00
// Copyright (c) 2019-2020 Alexander Medvednikov. All rights reserved.
2019-06-23 05:21:30 +03:00
// Use of this source code is governed by an MIT license
// that can be found in the LICENSE file.
2019-06-22 21:20:28 +03:00
module builtin
2019-11-01 00:28:52 +03:00
import strconv
/*
2019-10-04 15:48:09 +03:00
NB: A V string should be/is immutable from the point of view of
V user programs after it is first created. A V string is
also slightly larger than the equivalent C string because
the V string also has an integer length attached.
2019-10-04 15:48:09 +03:00
This tradeoff is made, since V strings are created just *once*,
but potentially used *many times* over their lifetime.
2019-10-04 15:48:09 +03:00
The V string implementation uses a struct, that has a .str field,
which points to a C style 0 terminated memory block. Although not
2019-10-04 15:48:09 +03:00
strictly necessary from the V point of view, that additional 0
is *very useful for C interoperability*.
2019-10-04 15:48:09 +03:00
The V string implementation also has an integer .len field,
containing the length of the .str field, excluding the
terminating 0 (just like the C's strlen(s) would do).
2019-10-04 15:48:09 +03:00
The 0 ending of .str, and the .len field, mean that in practice:
a) a V string s can be used very easily, wherever a
C string is needed, just by passing s.str,
without a need for further conversion/copying.
2019-10-04 15:48:09 +03:00
b) where strlen(s) is needed, you can just pass s.len,
without having to constantly recompute the length of s
*over and over again* like some C programs do. This is because
V strings are immutable and so their length does not change.
2019-10-04 15:48:09 +03:00
Ordinary V code *does not need* to be concerned with the
additional 0 in the .str field. The 0 *must* be put there by the
low level string creating functions inside this module.
2019-10-04 15:48:09 +03:00
Failing to do this will lead to programs that work most of the
time, when used with pure V functions, but fail in strange ways,
when used with modules using C functions (for example os and so on).
*/
2019-10-24 12:47:21 +03:00
pub struct string {
2019-06-22 21:20:28 +03:00
pub:
str byteptr // points to a C style 0 terminated string of bytes.
2019-12-19 23:52:45 +03:00
len int // the length of the .str field, excluding the ending 0 byte. It is always equal to strlen(.str).
mut:
is_lit int
}
// mut:
// hash_cache int
//
// NB string.is_lit is an enumeration of the following:
// .is_lit == 0 => a fresh string, should be freed by autofree
// .is_lit == 1 => a literal string from .rodata, should NOT be freed
// .is_lit == -98761234 => already freed string, protects against double frees.
// ^^^^^^^^^ calling free on these is a bug.
// Any other value means that the string has been corrupted.
2019-06-22 21:20:28 +03:00
2019-10-24 12:47:21 +03:00
pub struct ustring {
pub mut:
2019-06-22 21:20:28 +03:00
s string
runes []int
len int
}
[unsafe_fn]
2019-09-15 15:36:05 +03:00
pub fn vstrlen(s byteptr) int {
return unsafe {C.strlen(charptr(s))}
}
2019-09-15 15:36:05 +03:00
2019-08-17 22:19:37 +03:00
// Converts a C string to a V string.
// String data is reused, not copied.
[unsafe_fn]
pub fn tos(s byteptr, len int) string {
2019-06-22 21:20:28 +03:00
// This should never happen.
2019-10-04 15:48:09 +03:00
if s == 0 {
2019-06-22 21:20:28 +03:00
panic('tos(): nil string')
}
2019-12-19 23:52:45 +03:00
return string{
2019-06-22 21:20:28 +03:00
str: s
len: len
}
}
pub fn tos_clone(s byteptr) string {
2019-08-16 15:05:11 +03:00
return tos2(s).clone()
2019-06-22 21:20:28 +03:00
}
2019-08-17 22:19:37 +03:00
// Same as `tos`, but calculates the length. Called by `string(bytes)` casts.
// Used only internally.
2019-10-31 13:08:01 +03:00
pub fn tos2(s byteptr) string {
2019-10-04 15:48:09 +03:00
if s == 0 {
2019-06-22 21:20:28 +03:00
panic('tos2: nil string')
}
2019-12-19 23:52:45 +03:00
return string{
2019-10-04 15:48:09 +03:00
str: s
len: vstrlen(s)
}
}
2019-11-24 06:27:02 +03:00
// Same as `tos2`, but for char*, to avoid warnings
2019-12-02 11:37:35 +03:00
pub fn tos3(s charptr) string {
2019-10-04 15:48:09 +03:00
if s == 0 {
panic('tos3: nil string')
}
2019-12-19 23:52:45 +03:00
return string{
2019-10-04 15:48:09 +03:00
str: byteptr(s)
len: unsafe {C.strlen(s)}
2019-10-04 15:48:09 +03:00
}
2019-06-22 21:20:28 +03:00
}
2020-05-07 19:05:54 +03:00
pub fn tos_lit(s charptr) string {
return string{
str: byteptr(s)
len: unsafe {C.strlen(s)}
is_lit: 1
2020-05-07 19:05:54 +03:00
}
}
// string.clone_static returns an independent copy of a given array
// It should be used only in -autofree generated code.
fn (a string) clone_static() string {
return a.clone()
}
pub fn (a string) clone() string {
2019-12-19 23:52:45 +03:00
mut b := string{
str: unsafe {malloc(a.len + 1)}
2020-05-18 22:38:06 +03:00
len: a.len
2019-06-22 21:20:28 +03:00
}
unsafe {
C.memcpy(b.str, a.str, a.len)
b.str[a.len] = `\0`
}
2019-06-22 21:20:28 +03:00
return b
}
2019-08-17 22:19:37 +03:00
/*
pub fn (s string) cstr() byteptr {
2019-06-22 21:20:28 +03:00
clone := s.clone()
return clone.str
}
2019-08-17 22:19:37 +03:00
*/
2019-11-28 11:46:52 +03:00
// cstring_to_vstring creates a copy of cstr and turns it into a v string
pub fn cstring_to_vstring(cstr byteptr) string {
return tos_clone(cstr)
2019-11-28 11:46:52 +03:00
}
2019-11-11 18:43:22 +03:00
pub fn (s string) replace_once(rep, with string) string {
2019-12-19 23:52:45 +03:00
index := s.index(rep) or {
return s.clone()
2019-12-19 23:52:45 +03:00
}
return s.substr(0, index) + with + s.substr(index + rep.len, s.len)
2019-11-11 18:43:22 +03:00
}
2019-06-22 21:20:28 +03:00
pub fn (s string) replace(rep, with string) string {
if s.len == 0 || rep.len == 0 {
return s.clone()
2019-06-22 21:20:28 +03:00
}
// TODO PERF Allocating ints is expensive. Should be a stack array
// Get locations of all reps within this string
mut idxs := []int{}
2019-12-10 18:50:21 +03:00
mut idx := 0
2019-06-29 18:29:29 +03:00
for {
2019-12-10 18:50:21 +03:00
idx = s.index_after(rep, idx)
if idx == -1 {
break
}
idxs << idx
2019-12-27 07:20:06 +03:00
idx += rep.len
2019-06-22 21:20:28 +03:00
}
// Dont change the string if there's nothing to replace
if idxs.len == 0 {
return s.clone()
2019-06-22 21:20:28 +03:00
}
// Now we know the number of replacements we need to do and we can calc the len of the new string
new_len := s.len + idxs.len * (with.len - rep.len)
2019-12-19 23:52:45 +03:00
mut b := malloc(new_len + 1) // add a newline just in case
2019-06-22 21:20:28 +03:00
// Fill the new string
mut idx_pos := 0
mut cur_idx := idxs[idx_pos]
mut b_i := 0
2019-06-22 21:20:28 +03:00
for i := 0; i < s.len; i++ {
if i == cur_idx {
2020-04-07 19:51:39 +03:00
// Reached the location of rep, replace it with "with"
for j in 0..with.len {
unsafe {
b[b_i] = with[j]
}
2019-06-22 21:20:28 +03:00
b_i++
}
// Skip the length of rep, since we just replaced it with "with"
i += rep.len - 1
// Go to the next index
idx_pos++
if idx_pos < idxs.len {
2019-12-10 14:32:12 +03:00
cur_idx = idxs[idx_pos]
}
}
else {
2020-04-07 19:51:39 +03:00
// Rep doesnt start here, just copy
unsafe {
b[b_i] = s[i]
}
2019-12-10 14:32:12 +03:00
b_i++
}
}
unsafe {
b[new_len] = `\0`
return tos(b, new_len)
}
2019-12-10 14:32:12 +03:00
}
struct RepIndex {
2019-12-19 23:52:45 +03:00
idx int
2019-12-10 14:32:12 +03:00
val_idx int
}
2020-02-04 19:44:39 +03:00
fn compare_rep_index(a, b &RepIndex) int {
if a.idx < b.idx {
return -1
}
if a.idx > b.idx {
return 1
}
return 0
}
2020-05-17 14:51:18 +03:00
fn (mut a []RepIndex) sort() {
2019-12-10 14:32:12 +03:00
a.sort_with_compare(compare_rep_index)
}
// TODO
/*
fn (a RepIndex) < (b RepIndex) bool {
return a.idx < b.idx
}
*/
2019-12-19 23:52:45 +03:00
2019-12-10 14:32:12 +03:00
pub fn (s string) replace_each(vals []string) string {
if s.len == 0 || vals.len == 0 {
return s
}
if vals.len % 2 != 0 {
println('string.replace_each(): odd number of strings')
2019-12-10 14:32:12 +03:00
return s
}
// `rep` - string to replace
// `with` - string to replace with
// Remember positions of all rep strings, and calculate the length
// of the new string to do just one allocation.
mut new_len := s.len
mut idxs := []RepIndex{}
2019-12-10 14:32:12 +03:00
mut idx := 0
2019-12-19 23:52:45 +03:00
for rep_i := 0; rep_i < vals.len; rep_i += 2 {
2019-12-10 14:32:12 +03:00
// vals: ['rep1, 'with1', 'rep2', 'with2']
rep := vals[rep_i]
2019-12-19 23:52:45 +03:00
with := vals[rep_i + 1]
2019-12-10 14:32:12 +03:00
for {
idx = s.index_after(rep, idx)
if idx == -1 {
break
}
// We need to remember both the position in the string,
// and which rep/with pair it refers to.
2020-03-16 17:46:09 +03:00
idxs << RepIndex {
2020-02-04 19:44:39 +03:00
idx:idx
2020-03-16 17:46:09 +03:00
val_idx:rep_i
}
2019-12-10 14:32:12 +03:00
idx++
new_len += with.len - rep.len
}
}
// Dont change the string if there's nothing to replace
if idxs.len == 0 {
return s
}
idxs.sort()
2019-12-19 23:52:45 +03:00
mut b := malloc(new_len + 1) // add a \0 just in case
2019-12-10 14:32:12 +03:00
// Fill the new string
mut idx_pos := 0
mut cur_idx := idxs[idx_pos]
mut b_i := 0
for i := 0; i < s.len; i++ {
if i == cur_idx.idx {
2020-04-07 19:51:39 +03:00
// Reached the location of rep, replace it with "with"
2019-12-10 14:32:12 +03:00
rep := vals[cur_idx.val_idx]
2019-12-19 23:52:45 +03:00
with := vals[cur_idx.val_idx + 1]
for j in 0..with.len {
unsafe {
b[b_i] = with[j]
}
2019-12-10 14:32:12 +03:00
b_i++
}
// Skip the length of rep, since we just replaced it with "with"
i += rep.len - 1
// Go to the next index
idx_pos++
if idx_pos < idxs.len {
2019-06-22 21:20:28 +03:00
cur_idx = idxs[idx_pos]
}
}
else {
2020-04-07 19:51:39 +03:00
// Rep doesnt start here, just copy
unsafe {
b[b_i] = s.str[i]
}
2019-06-22 21:20:28 +03:00
b_i++
}
}
unsafe {
b[new_len] = `\0`
return tos(b, new_len)
}
2019-06-22 21:20:28 +03:00
}
2019-12-09 17:10:44 +03:00
pub fn (s string) bool() bool {
return s == 'true' || s == 't' // TODO t for pg, remove
}
2019-10-25 23:41:18 +03:00
pub fn (s string) int() int {
2019-12-19 23:52:45 +03:00
return int(strconv.common_parse_int(s, 0, 32, false, false))
2019-10-25 23:41:18 +03:00
}
pub fn (s string) i64() i64 {
return strconv.common_parse_int(s, 0, 64, false, false)
}
2020-01-14 20:05:38 +03:00
pub fn (s string) i8() i8 {
return i8(strconv.common_parse_int(s, 0, 8, false, false))
}
pub fn (s string) i16() i16 {
return i16(strconv.common_parse_int(s, 0, 16, false, false))
}
pub fn (s string) f32() f32 {
2019-12-19 23:52:45 +03:00
// return C.atof(charptr(s.str))
2019-12-17 01:07:13 +03:00
return f32(strconv.atof64(s))
}
pub fn (s string) f64() f64 {
2019-12-19 23:52:45 +03:00
// return C.atof(charptr(s.str))
2019-12-17 01:07:13 +03:00
return strconv.atof64(s)
}
2020-01-14 20:05:38 +03:00
pub fn (s string) u16() u16 {
return u16(strconv.common_parse_uint(s, 0, 16, false, false))
}
pub fn (s string) u32() u32 {
return u32(strconv.common_parse_uint(s, 0, 32, false, false))
}
pub fn (s string) u64() u64 {
return strconv.common_parse_uint(s, 0, 64, false, false)
}
2019-06-22 21:20:28 +03:00
// ==
fn (s string) eq(a string) bool {
if s.str == 0 {
2019-12-19 23:52:45 +03:00
// should never happen
2019-06-22 21:20:28 +03:00
panic('string.eq(): nil string')
}
if s.len != a.len {
return false
}
unsafe {
return C.memcmp(s.str, a.str, a.len) == 0
}
2019-06-22 21:20:28 +03:00
}
// !=
fn (s string) ne(a string) bool {
return !s.eq(a)
}
2019-06-23 21:25:50 +03:00
// s < a
fn (s string) lt(a string) bool {
for i in 0..s.len {
2019-06-23 21:25:50 +03:00
if i >= a.len || s[i] > a[i] {
2019-06-22 21:20:28 +03:00
return false
}
2019-06-23 21:25:50 +03:00
else if s[i] < a[i] {
2019-06-22 21:20:28 +03:00
return true
}
}
2019-06-23 21:25:50 +03:00
if s.len < a.len {
return true
}
return false
2019-06-22 21:20:28 +03:00
}
// s <= a
fn (s string) le(a string) bool {
2019-06-23 21:25:50 +03:00
return s.lt(a) || s.eq(a)
2019-06-22 21:20:28 +03:00
}
// s > a
fn (s string) gt(a string) bool {
2019-06-23 21:25:50 +03:00
return !s.le(a)
}
// s >= a
fn (s string) ge(a string) bool {
return !s.lt(a)
2019-06-22 21:20:28 +03:00
}
// TODO `fn (s string) + (a string)` ? To be consistent with operator overloading syntax.
2019-07-20 17:43:33 +03:00
fn (s string) add(a string) string {
2019-06-22 21:20:28 +03:00
new_len := a.len + s.len
2019-12-19 23:52:45 +03:00
mut res := string{
str: malloc(new_len + 1)
2020-05-18 22:38:06 +03:00
len: new_len
2019-06-22 21:20:28 +03:00
}
for j in 0..s.len {
unsafe {
res.str[j] = s.str[j]
}
2019-06-22 21:20:28 +03:00
}
for j in 0..a.len {
unsafe {
res.str[s.len + j] = a.str[j]
}
}
unsafe {
res.str[new_len] = `\0` // V strings are not null terminated, but just in case
2019-06-22 21:20:28 +03:00
}
return res
}
pub fn (s string) split(delim string) []string {
2019-12-01 16:10:13 +03:00
return s.split_nth(delim, 0)
}
/*
2020-02-04 19:44:39 +03:00
split_nth - splits the string based on the passed `delim` substring.
It returns the first Nth parts. When N=0, return all the splits.
The last returned element has the remainder of the string, even if
the remainder contains more `delim` substrings.
*/
2019-12-01 16:10:13 +03:00
pub fn (s string) split_nth(delim string, nth int) []string {
mut res := []string{}
2019-12-01 16:10:13 +03:00
mut i := 0
2019-06-22 21:20:28 +03:00
if delim.len == 0 {
2019-12-01 16:10:13 +03:00
i = 1
2019-11-10 19:37:36 +03:00
for ch in s {
2019-12-01 16:10:13 +03:00
if nth > 0 && i >= nth {
res << s.substr(i, s.len)
break
}
2019-11-10 19:37:36 +03:00
res << ch.str()
2019-12-01 16:10:13 +03:00
i++
2019-11-10 19:37:36 +03:00
}
2019-06-22 21:20:28 +03:00
return res
}
2019-12-01 16:10:13 +03:00
mut start := 0
nth_1 := nth - 1
2019-12-01 16:10:13 +03:00
for i <= s.len {
mut is_delim := unsafe {s.str[i] == delim.str[0]}
2019-12-01 16:10:13 +03:00
mut j := 0
for is_delim && j < delim.len {
is_delim = is_delim && unsafe {s.str[i + j] == delim.str[j]}
2019-06-22 21:20:28 +03:00
j++
}
last := i == s.len - 1
2019-12-01 16:10:13 +03:00
if is_delim || last {
if !is_delim && last {
2019-06-22 21:20:28 +03:00
i++
}
mut val := s.substr(start, i)
2019-12-01 16:10:13 +03:00
if val.starts_with(delim) {
val = val.right(delim.len)
2019-06-22 21:20:28 +03:00
}
was_last := nth > 0 && res.len == nth_1
if was_last {
res << s.right(start)
break
}
2020-02-04 19:44:39 +03:00
2019-12-01 16:10:13 +03:00
res << val
start = i + delim.len
2019-06-22 21:20:28 +03:00
}
i++
}
2019-12-19 23:52:45 +03:00
if s.ends_with(delim) && (nth < 1 || res.len < nth) {
2019-12-01 16:10:13 +03:00
res << ''
2019-06-22 21:20:28 +03:00
}
return res
}
pub fn (s string) split_into_lines() []string {
mut res := []string{}
2019-06-22 21:20:28 +03:00
if s.len == 0 {
return res
}
mut start := 0
for i := 0; i < s.len; i++ {
is_lf := unsafe {s.str[i]} == `\n`
is_crlf := i != s.len - 1 && unsafe {s.str[i] == `\r` && s.str[i + 1] == `\n`}
2020-02-09 11:23:57 +03:00
is_eol := is_lf || is_crlf
is_last := if is_crlf {
i == s.len - 2
} else {
i == s.len - 1
}
if is_eol || is_last {
if is_last && !is_eol {
2019-06-22 21:20:28 +03:00
i++
}
2020-02-09 11:23:57 +03:00
2019-06-22 21:20:28 +03:00
line := s.substr(start, i)
res << line
2020-02-09 11:23:57 +03:00
if is_crlf {
i++
}
2019-06-22 21:20:28 +03:00
start = i + 1
}
}
return res
}
// 'hello'.left(2) => 'he'
fn (s string) left(n int) string {
2019-06-22 21:20:28 +03:00
if n >= s.len {
return s
}
return s.substr(0, n)
}
2019-12-19 23:52:45 +03:00
2019-08-26 13:32:53 +03:00
// 'hello'.right(2) => 'llo'
fn (s string) right(n int) string {
2019-06-22 21:20:28 +03:00
if n >= s.len {
return ''
}
return s.substr(n, s.len)
}
// used internally for [2..4]
fn (s string) substr2(start, _end int, end_max bool) string {
end := if end_max { s.len } else { _end }
return s.substr(start, end)
}
pub fn (s string) substr(start, end int) string {
$if !no_bounds_checking? {
if start > end || start > s.len || end > s.len || start < 0 || end < 0 {
panic('substr($start, $end) out of bounds (len=$s.len)')
}
2019-06-22 21:20:28 +03:00
}
len := end - start
2019-12-19 23:52:45 +03:00
mut res := string{
str: malloc(len + 1)
2020-05-18 22:38:06 +03:00
len: len
}
for i in 0..len {
unsafe {
res.str[i] = s.str[start + i]
}
}
unsafe {
res.str[len] = `\0`
}
2019-12-19 23:52:45 +03:00
/*
2019-06-22 21:20:28 +03:00
res := string {
str: s.str + start
len: len
}
2019-08-17 22:19:37 +03:00
*/
return res
2019-06-22 21:20:28 +03:00
}
pub fn (s string) index_old(p string) int {
2020-02-20 22:14:21 +03:00
if p.len > s.len || p.len == 0 {
2019-06-22 21:20:28 +03:00
return -1
}
2019-09-14 23:48:30 +03:00
mut i := 0
for i < s.len {
mut j := 0
for j < p.len && unsafe {s.str[i + j] == p.str[j]} {
2019-06-22 21:20:28 +03:00
j++
}
if j == p.len {
return i
2019-06-22 21:20:28 +03:00
}
2019-09-14 23:48:30 +03:00
i++
2019-06-22 21:20:28 +03:00
}
return -1
}
pub fn (s string) index(p string) ?int {
2020-02-20 22:14:21 +03:00
if p.len > s.len || p.len == 0 {
return none
}
mut i := 0
for i < s.len {
mut j := 0
for j < p.len && unsafe {s.str[i + j] == p.str[j]} {
j++
}
if j == p.len {
return i
}
i++
}
return none
}
2019-09-14 23:48:30 +03:00
// KMP search
2019-12-12 21:44:52 +03:00
fn (s string) index_kmp(p string) int {
2019-12-19 23:52:45 +03:00
if p.len > s.len {
return -1
}
mut prefix := []int{len:p.len}
2019-12-19 23:52:45 +03:00
mut j := 0
for i := 1; i < p.len; i++ {
for unsafe {p.str[j] != p.str[i]} && j > 0 {
2019-12-19 23:52:45 +03:00
j = prefix[j - 1]
}
if unsafe {p.str[j] == p.str[i]} {
2019-12-19 23:52:45 +03:00
j++
}
prefix[i] = j
}
j = 0
for i in 0..s.len {
for unsafe {p.str[j] != s.str[i]} && j > 0 {
2019-12-19 23:52:45 +03:00
j = prefix[j - 1]
}
if unsafe {p.str[j] == s.str[i]} {
2019-12-19 23:52:45 +03:00
j++
}
if j == p.len {
return i - p.len + 1
}
}
return -1
2019-09-14 23:48:30 +03:00
}
2019-08-01 16:01:03 +03:00
pub fn (s string) index_any(chars string) int {
for c in chars {
2019-12-19 23:52:45 +03:00
index := s.index(c.str()) or {
continue
}
return index
2019-08-01 16:01:03 +03:00
}
return -1
}
2019-12-12 21:44:52 +03:00
pub fn (s string) last_index(p string) ?int {
2020-02-20 22:14:21 +03:00
if p.len > s.len || p.len == 0 {
2019-12-12 21:44:52 +03:00
return none
2019-06-22 21:20:28 +03:00
}
mut i := s.len - p.len
for i >= 0 {
mut j := 0
for j < p.len && unsafe {s.str[i + j] == p.str[j]} {
2019-06-22 21:20:28 +03:00
j++
}
if j == p.len {
return i
}
i--
}
2019-12-12 21:44:52 +03:00
return none
2019-06-22 21:20:28 +03:00
}
pub fn (s string) index_after(p string, start int) int {
if p.len > s.len {
return -1
}
mut strt := start
if start < 0 {
strt = 0
}
if start >= s.len {
return -1
}
mut i := strt
for i < s.len {
mut j := 0
mut ii := i
for j < p.len && unsafe {s.str[ii] == p.str[j]} {
2019-06-22 21:20:28 +03:00
j++
ii++
}
if j == p.len {
return i
}
i++
}
return -1
}
pub fn (s string) index_byte(c byte) int {
for i in 0..s.len {
if unsafe {s.str[i]} == c {
return i
}
}
return -1
}
pub fn (s string) last_index_byte(c byte) int {
2019-12-19 23:52:45 +03:00
for i := s.len - 1; i >= 0; i-- {
if unsafe {s.str[i] == c} {
return i
}
}
return -1
}
2019-08-03 00:30:22 +03:00
// counts occurrences of substr in s
pub fn (s string) count(substr string) int {
if s.len == 0 || substr.len == 0 {
return 0
}
2019-08-26 13:32:53 +03:00
if substr.len > s.len {
return 0
}
2019-08-03 00:30:22 +03:00
mut n := 0
2019-08-03 01:18:19 +03:00
mut i := 0
2019-08-03 00:30:22 +03:00
for {
i = s.index_after(substr, i)
2019-08-03 00:30:22 +03:00
if i == -1 {
return n
}
2019-08-03 01:18:19 +03:00
i += substr.len
2019-08-03 00:30:22 +03:00
n++
}
return 0 // TODO can never get here - v doesn't know that
2019-08-03 00:30:22 +03:00
}
2019-06-22 21:20:28 +03:00
pub fn (s string) contains(p string) bool {
if p.len == 0 {
return true
}
s.index(p) or {
return false
}
return true
2019-06-22 21:20:28 +03:00
}
pub fn (s string) starts_with(p string) bool {
if p.len > s.len {
2019-12-19 23:52:45 +03:00
return false
}
for i in 0..p.len {
if unsafe {s.str[i] != p.str[i]} {
return false
}
}
return true
2019-06-22 21:20:28 +03:00
}
pub fn (s string) ends_with(p string) bool {
if p.len > s.len {
return false
}
for i in 0..p.len {
if p[i] != s[s.len - p.len + i] {
return false
}
2019-12-12 21:44:52 +03:00
}
return true
2019-06-22 21:20:28 +03:00
}
// TODO only works with ASCII
pub fn (s string) to_lower() string {
unsafe {
mut b := malloc(s.len + 1)
for i in 0..s.len {
b[i] = byte(C.tolower(s.str[i]))
}
return tos(b, s.len)
2019-06-22 21:20:28 +03:00
}
}
pub fn (s string) is_lower() bool {
for i in 0..s.len {
if s[i] >= `A` && s[i] <= `Z` {
return false
}
}
return true
}
2019-06-22 21:20:28 +03:00
pub fn (s string) to_upper() string {
unsafe {
mut b := malloc(s.len + 1)
for i in 0..s.len {
b[i] = byte(C.toupper(s.str[i]))
}
return tos(b, s.len)
2019-06-22 21:20:28 +03:00
}
}
pub fn (s string) is_upper() bool {
for i in 0..s.len {
if s[i] >= `a` && s[i] <= `z` {
return false
}
}
return true
}
2019-08-26 13:32:53 +03:00
pub fn (s string) capitalize() string {
2020-02-20 13:33:38 +03:00
if s.len == 0 {
return ''
}
2019-08-26 13:32:53 +03:00
sl := s.to_lower()
2019-12-19 23:52:45 +03:00
cap := sl[0].str().to_upper() + sl.right(1)
2019-09-01 22:37:22 +03:00
return cap
2019-08-26 13:32:53 +03:00
}
pub fn (s string) is_capital() bool {
if s.len == 0 || !(s[0] >= `A` && s[0] <= `Z`) {
return false
}
for i in 1..s.len {
if s[i] >= `A` && s[i] <= `Z` {
return false
}
}
return true
}
2019-08-26 13:32:53 +03:00
pub fn (s string) title() string {
2019-12-19 23:52:45 +03:00
words := s.split(' ')
mut tit := []string{}
2019-08-26 13:32:53 +03:00
for word in words {
tit << word.capitalize()
}
title := tit.join(' ')
return title
2019-08-26 13:32:53 +03:00
}
pub fn (s string) is_title() bool {
words := s.split(' ')
for word in words {
if !word.is_capital() {
return false
}
}
return true
}
2019-06-22 21:20:28 +03:00
// 'hey [man] how you doin'
// find_between('[', ']') == 'man'
pub fn (s string) find_between(start, end string) string {
2019-12-19 23:52:45 +03:00
start_pos := s.index(start) or {
return ''
}
2019-06-22 21:20:28 +03:00
// First get everything to the right of 'start'
val := s.right(start_pos + start.len)
2019-12-19 23:52:45 +03:00
end_pos := val.index(end) or {
return val
}
2019-06-22 21:20:28 +03:00
return val.left(end_pos)
}
// TODO generic
fn (ar []string) contains(val string) bool {
2019-06-22 21:20:28 +03:00
for s in ar {
if s == val {
return true
}
}
return false
}
2019-08-17 22:19:37 +03:00
/*
2019-06-30 14:06:46 +03:00
pub fn (a []string) to_c() voidptr {
2019-08-17 22:19:37 +03:00
mut res := malloc(sizeof(byteptr) * a.len)
for i in 0..a.len {
2019-06-22 21:20:28 +03:00
val := a[i]
2019-08-17 22:19:37 +03:00
res[i] = val.str
2019-06-22 21:20:28 +03:00
}
2019-08-17 22:19:37 +03:00
return res
2019-06-22 21:20:28 +03:00
}
2019-08-17 22:19:37 +03:00
*/
2019-06-22 21:20:28 +03:00
pub fn (c byte) is_space() bool {
// 0x0085 is NEXT LINE (NEL)
// 0x00a0 is NO-BREAK SPACE
return c in [` `, `\n`, `\t`, `\v`, `\f`, `\r`, 0x85, 0xa0]
2019-06-22 21:20:28 +03:00
}
pub fn (s string) trim_space() string {
return s.trim(' \n\t\v\f\r')
2019-06-22 21:20:28 +03:00
}
pub fn (s string) trim(cutset string) string {
if s.len < 1 || cutset.len < 1 {
return s
2019-06-22 21:20:28 +03:00
}
cs_arr := cutset.bytes()
mut pos_left := 0
mut pos_right := s.len - 1
mut cs_match := true
for pos_left <= s.len && pos_right >= -1 && cs_match {
cs_match = false
if s[pos_left] in cs_arr {
pos_left++
cs_match = true
}
if s[pos_right] in cs_arr {
pos_right--
cs_match = true
}
if pos_left > pos_right {
return ''
}
2019-06-22 21:20:28 +03:00
}
2019-12-19 23:52:45 +03:00
return s.substr(pos_left, pos_right + 1)
2019-06-22 21:20:28 +03:00
}
pub fn (s string) trim_left(cutset string) string {
if s.len < 1 || cutset.len < 1 {
2019-06-22 21:20:28 +03:00
return s
}
cs_arr := cutset.bytes()
mut pos := 0
2020-04-06 15:30:25 +03:00
for pos < s.len && s[pos] in cs_arr {
pos++
2019-06-22 21:20:28 +03:00
}
return s.right(pos)
2019-06-22 21:20:28 +03:00
}
pub fn (s string) trim_right(cutset string) string {
if s.len < 1 || cutset.len < 1 {
2019-08-17 22:19:37 +03:00
return s
}
cs_arr := cutset.bytes()
mut pos := s.len - 1
2020-04-06 15:30:25 +03:00
for pos >= 0 && s[pos] in cs_arr {
2019-08-17 22:19:37 +03:00
pos--
}
2020-04-06 15:30:25 +03:00
return if pos < 0 { '' } else { s.left(pos + 1) }
2019-06-22 21:20:28 +03:00
}
pub fn (s string) trim_prefix(str string) string {
if s.starts_with(str) {
return s[str.len..]
}
return s
}
pub fn (s string) trim_suffix(str string) string {
if s.ends_with(str) {
return s[..s.len-str.len]
}
return s
}
pub fn compare_strings(a, b &string) int {
if a.lt(b) {
2019-06-22 21:20:28 +03:00
return -1
}
if a.gt(b) {
2019-06-22 21:20:28 +03:00
return 1
}
return 0
}
2019-09-01 22:51:16 +03:00
fn compare_strings_by_len(a, b &string) int {
2019-06-22 21:20:28 +03:00
if a.len < b.len {
return -1
}
if a.len > b.len {
return 1
}
return 0
}
2019-09-01 22:51:16 +03:00
fn compare_lower_strings(a, b &string) int {
2019-06-22 21:20:28 +03:00
aa := a.to_lower()
2019-06-25 07:29:02 +03:00
bb := b.to_lower()
2019-06-22 21:20:28 +03:00
return compare_strings(aa, bb)
}
2020-05-17 14:51:18 +03:00
pub fn (mut s []string) sort() {
2019-06-22 21:20:28 +03:00
s.sort_with_compare(compare_strings)
}
2020-05-17 14:51:18 +03:00
pub fn (mut s []string) sort_ignore_case() {
2019-06-22 21:20:28 +03:00
s.sort_with_compare(compare_lower_strings)
}
2020-05-17 14:51:18 +03:00
pub fn (mut s []string) sort_by_len() {
2019-06-22 21:20:28 +03:00
s.sort_with_compare(compare_strings_by_len)
}
2020-04-29 14:51:42 +03:00
pub fn (s string) str() string {
return s
}
pub fn (s ustring) str() string {
return s.s
}
pub fn (s string) ustring() ustring {
2019-12-19 23:52:45 +03:00
mut res := ustring{
2019-06-22 21:20:28 +03:00
s: s
// runes will have at least s.len elements, save reallocations
// TODO use VLA for small strings?
2020-01-12 21:59:57 +03:00
runes: __new_array(0, s.len, int(sizeof(int)))
2019-06-22 21:20:28 +03:00
}
for i := 0; i < s.len; i++ {
char_len := utf8_char_len(unsafe {s.str[i]})
2019-06-22 21:20:28 +03:00
res.runes << i
i += char_len - 1
res.len++
}
return res
}
// A hack that allows to create ustring without allocations.
// It's called from functions like draw_text() where we know that the string is going to be freed
// right away. Uses global buffer for storing runes []int array.
__global g_ustring_runes []int
2019-12-19 23:52:45 +03:00
pub fn (s string) ustring_tmp() ustring {
if g_ustring_runes.len == 0 {
g_ustring_runes = __new_array(0, 128, int(sizeof(int)))
}
2019-12-19 23:52:45 +03:00
mut res := ustring{
2019-06-22 21:20:28 +03:00
s: s
}
res.runes = g_ustring_runes
res.runes.len = s.len
2019-06-22 21:20:28 +03:00
mut j := 0
for i := 0; i < s.len; i++ {
char_len := utf8_char_len(unsafe {s.str[i]})
2019-06-22 21:20:28 +03:00
res.runes[j] = i
j++
i += char_len - 1
res.len++
}
return res
}
fn (u ustring) eq(a ustring) bool {
if u.len != a.len || u.s != a.s {
return false
}
return true
}
fn (u ustring) ne(a ustring) bool {
return !u.eq(a)
}
fn (u ustring) lt(a ustring) bool {
return u.s < a.s
}
fn (u ustring) le(a ustring) bool {
return u.lt(a) || u.eq(a)
}
fn (u ustring) gt(a ustring) bool {
return !u.le(a)
}
fn (u ustring) ge(a ustring) bool {
return !u.lt(a)
}
pub fn (u ustring) add(a ustring) ustring {
2019-12-19 23:52:45 +03:00
mut res := ustring{
s: u.s + a.s
runes: __new_array(0, u.s.len + a.s.len, int(sizeof(int)))
}
mut j := 0
for i := 0; i < u.s.len; i++ {
char_len := utf8_char_len(unsafe {u.s.str[i]})
res.runes << j
i += char_len - 1
j += char_len
res.len++
}
for i := 0; i < a.s.len; i++ {
char_len := utf8_char_len(unsafe {a.s.str[i]})
res.runes << j
i += char_len - 1
j += char_len
res.len++
}
return res
}
pub fn (u ustring) index_after(p ustring, start int) int {
if p.len > u.len {
return -1
}
mut strt := start
if start < 0 {
strt = 0
}
if start > u.len {
return -1
}
mut i := strt
for i < u.len {
mut j := 0
mut ii := i
for j < p.len && u.at(ii) == p.at(j) {
j++
ii++
}
if j == p.len {
return i
}
i++
}
return -1
}
// counts occurrences of substr in s
pub fn (u ustring) count(substr ustring) int {
if u.len == 0 || substr.len == 0 {
return 0
}
if substr.len > u.len {
return 0
}
mut n := 0
mut i := 0
for {
i = u.index_after(substr, i)
if i == -1 {
return n
}
i += substr.len
n++
}
return 0 // TODO can never get here - v doesn't know that
}
2019-08-06 19:13:04 +03:00
pub fn (u ustring) substr(_start, _end int) string {
$if !no_bounds_checking? {
if _start > _end || _start > u.len || _end > u.len || _start < 0 || _end < 0 {
panic('substr($_start, $_end) out of bounds (len=$u.len)')
}
}
2019-12-19 23:52:45 +03:00
end := if _end >= u.len { u.s.len } else { u.runes[_end] }
return u.s.substr(u.runes[_start], end)
2019-06-22 21:20:28 +03:00
}
2019-07-07 22:46:21 +03:00
pub fn (u ustring) left(pos int) string {
if pos >= u.len {
return u.s
}
2019-06-22 21:20:28 +03:00
return u.substr(0, pos)
}
2019-07-07 22:46:21 +03:00
pub fn (u ustring) right(pos int) string {
if pos >= u.len {
return ''
}
2019-06-22 21:20:28 +03:00
return u.substr(pos, u.len)
}
fn (s string) at(idx int) byte {
$if !no_bounds_checking? {
if idx < 0 || idx >= s.len {
panic('string index out of range: $idx / $s.len')
}
2019-06-22 21:20:28 +03:00
}
unsafe {
return s.str[idx]
}
2019-06-22 21:20:28 +03:00
}
pub fn (u ustring) at(idx int) string {
$if !no_bounds_checking? {
if idx < 0 || idx >= u.len {
panic('string index out of range: $idx / $u.runes.len')
}
}
2019-06-22 21:20:28 +03:00
return u.substr(idx, idx + 1)
}
[unsafe_fn]
fn (u &ustring) free() {
2020-07-11 14:22:16 +03:00
$if prealloc {
return
}
unsafe {
u.runes.free()
}
2019-06-22 21:20:28 +03:00
}
pub fn (c byte) is_digit() bool {
2019-06-22 21:20:28 +03:00
return c >= `0` && c <= `9`
}
2019-07-24 01:06:48 +03:00
pub fn (c byte) is_hex_digit() bool {
return c.is_digit() || (c >= `a` && c <= `f`) || (c >= `A` && c <= `F`)
}
pub fn (c byte) is_oct_digit() bool {
return c >= `0` && c <= `7`
}
pub fn (c byte) is_bin_digit() bool {
return c == `0` || c == `1`
}
pub fn (c byte) is_letter() bool {
2019-06-22 21:20:28 +03:00
return (c >= `a` && c <= `z`) || (c >= `A` && c <= `Z`)
}
pub fn (s &string) free() {
2020-07-11 14:22:16 +03:00
$if prealloc {
return
}
if s.is_lit == -98761234 {
C.printf('double string.free() detected\n')
return
}
if s.is_lit == 1 || s.len == 0 {
2020-06-12 20:20:51 +03:00
return
}
2019-07-25 01:25:21 +03:00
free(s.str)
s.is_lit = -98761234
2019-06-22 21:20:28 +03:00
}
// all_before('23:34:45.234', '.') == '23:34:45'
pub fn (s string) all_before(dot string) string {
2019-12-19 23:52:45 +03:00
pos := s.index(dot) or {
return s
}
2019-06-22 21:20:28 +03:00
return s.left(pos)
}
pub fn (s string) all_before_last(dot string) string {
2019-12-12 21:44:52 +03:00
pos := s.last_index(dot) or {
2019-06-22 21:20:28 +03:00
return s
}
return s.left(pos)
}
pub fn (s string) all_after(dot string) string {
pos := s.index(dot) or {
return s
}
return s.right(pos + dot.len)
}
pub fn (s string) all_after_last(dot string) string {
2019-12-12 21:44:52 +03:00
pos := s.last_index(dot) or {
2019-06-22 21:20:28 +03:00
return s
}
return s.right(pos + dot.len)
}
pub fn (s string) after(dot string) string {
return s.all_after_last(dot)
}
2020-02-18 22:20:15 +03:00
pub fn (s string) after_char(dot byte) string {
mut pos := 0
for i, c in s {
if c == dot {
pos = i
break
}
}
if pos == 0 {
return s
}
return s.right(pos+1)
}
2019-06-22 21:20:28 +03:00
// fn (s []string) substr(a, b int) string {
// return join_strings(s.slice_fast(a, b))
// }
2019-06-30 14:06:46 +03:00
pub fn (a []string) join(del string) string {
2019-06-22 21:20:28 +03:00
if a.len == 0 {
return ''
}
mut len := 0
for val in a {
2019-06-22 21:20:28 +03:00
len += val.len + del.len
}
len -= del.len
// Allocate enough memory
mut res := ''
res.len = len
res.str = malloc(res.len + 1)
mut idx := 0
// Go thru every string and copy its every char one by one
for i, val in a {
for j in 0..val.len {
unsafe {
res.str[idx] = val.str[j]
}
2019-06-22 21:20:28 +03:00
idx++
}
// Add del if it's not last
if i != a.len - 1 {
for k in 0..del.len {
unsafe {
res.str[idx] = del.str[k]
}
2019-06-22 21:20:28 +03:00
idx++
}
}
}
unsafe {
res.str[res.len] = `\0`
}
2019-06-22 21:20:28 +03:00
return res
}
2019-06-30 14:06:46 +03:00
pub fn (s []string) join_lines() string {
2019-06-22 21:20:28 +03:00
return s.join('\n')
}
2019-10-27 22:29:55 +03:00
// reverse will return a new reversed string.
pub fn (s string) reverse() string {
2020-02-20 22:14:21 +03:00
if s.len == 0 || s.len == 1 {
return s
2019-12-16 21:29:32 +03:00
}
2019-12-19 23:52:45 +03:00
mut res := string{
str: malloc(s.len)
2020-05-18 22:38:06 +03:00
len: s.len
2019-06-27 03:03:19 +03:00
}
for i := s.len - 1; i >= 0; i-- {
unsafe {
res.str[s.len - i - 1] = s[i]
}
2019-06-27 03:03:19 +03:00
}
return res
}
2019-10-27 22:29:55 +03:00
// limit returns a portion of the string, starting at `0` and extending for a given number of characters afterward.
2019-06-22 21:20:28 +03:00
// 'hello'.limit(2) => 'he'
// 'hi'.limit(10) => 'hi'
pub fn (s string) limit(max int) string {
2019-06-22 21:20:28 +03:00
u := s.ustring()
if u.len <= max {
return s
}
return u.substr(0, max)
}
[deprecated]
pub fn (c byte) is_white() bool {
2020-06-23 17:25:24 +03:00
eprintln('warning: `string.is_white` has been deprecated, use `string.is_space` instead')
return c.is_space()
2019-06-22 21:20:28 +03:00
}
pub fn (s string) hash() int {
2019-12-19 23:52:45 +03:00
// mut h := s.hash_cache
mut h := u32(0)
2019-08-17 22:19:37 +03:00
if h == 0 && s.len > 0 {
for c in s {
h = h * 31 + u32(c)
2019-07-10 10:48:10 +03:00
}
2019-08-17 22:19:37 +03:00
}
return int(h)
2019-06-22 21:20:28 +03:00
}
2019-07-15 18:49:01 +03:00
pub fn (s string) bytes() []byte {
if s.len == 0 {
2019-11-14 09:18:07 +03:00
return []
2019-07-15 18:49:01 +03:00
}
mut buf := []byte{ len:s.len }
unsafe {
C.memcpy(buf.data, s.str, s.len)
}
2019-07-15 18:49:01 +03:00
return buf
}
2019-09-26 22:54:53 +03:00
2019-10-27 22:29:55 +03:00
// repeat returns a new string with a specified number of copies of the string it was called on.
2019-09-26 22:54:53 +03:00
pub fn (s string) repeat(count int) string {
if count < 0 {
panic('string.repeat: count is negative: $count')
} else if count == 0 {
return ''
} else if count == 1 {
2019-09-26 22:54:53 +03:00
return s
}
mut ret := malloc(s.len * count + 1)
2019-12-19 23:52:45 +03:00
for i in 0 .. count {
for j in 0 .. s.len {
unsafe {
ret[i * s.len + j] = s[j]
}
}
2019-09-26 22:54:53 +03:00
}
unsafe {
ret[s.len * count] = 0
}
2019-09-26 22:54:53 +03:00
return string(ret)
}
2020-03-15 07:46:12 +03:00
pub fn (s string) fields() []string {
2020-06-09 11:15:17 +03:00
// TODO do this in a better way
return s.replace('\t', ' ').split(' ')
}
2020-06-17 13:26:55 +03:00
pub fn (s string) map(func fn(byte) byte) string {
unsafe {
mut res := malloc(s.len + 1)
for i in 0..s.len {
res[i] = func(s[i])
}
return tos(res, s.len)
}
2020-06-17 13:26:55 +03:00
}
2020-07-12 21:07:28 +03:00
pub fn (s string) filter(func fn(b byte) bool) string {
mut new_len := 0
mut buf := malloc(s.len + 1)
for i in 0 .. s.len {
mut b := s[i]
if func(b) {
buf[new_len] = b
new_len++
}
}
buf[new_len] = 0
return string(buf, new_len)
}
2020-03-15 07:46:12 +03:00
// Allows multi-line strings to be formatted in a way that removes white-space
// before a delimeter. by default `|` is used.
// Note: the delimiter has to be a byte at this time. That means surrounding
// the value in ``.
2020-03-16 17:46:09 +03:00
//
2020-03-15 07:46:12 +03:00
// Example:
// st := 'Hello there,
// |this is a string,
// | Everything before the first | is removed'.strip_margin()
// Returns:
// Hello there,
// this is a string,
// Everything before the first | is removed
pub fn (s string) strip_margin() string {
return s.strip_margin_custom(`|`)
}
pub fn (s string) strip_margin_custom(del byte) string {
mut sep := del
if sep.is_space() {
eprintln("Warning: `strip_margin` cannot use white-space as a delimiter")
eprintln(" Defaulting to `|`")
sep = `|`
2020-03-15 07:46:12 +03:00
}
// don't know how much space the resulting string will be, but the max it
// can be is this big
mut ret := malloc(s.len + 1)
mut count := 0
for i := 0; i < s.len; i++ {
if s[i] in [`\n`, `\r`] {
unsafe {
ret[count] = s[i]
}
2020-03-16 17:46:09 +03:00
count++
// CRLF
if s[i] == `\r` && i < s.len - 1 && s[i+1] == `\n` {
unsafe {
ret[count] = s[i+1]
}
2020-03-15 07:46:12 +03:00
count++
2020-03-16 17:46:09 +03:00
i++
2020-03-15 07:46:12 +03:00
}
2020-03-16 17:46:09 +03:00
2020-03-15 07:46:12 +03:00
for s[i] != sep {
i++
if i >= s.len {
break
}
}
} else {
unsafe {
ret[count] = s[i]
}
2020-03-15 07:46:12 +03:00
count++
}
}
unsafe {
ret[count] = 0
}
2020-03-15 07:46:12 +03:00
return string(ret)
}