compiler: improve typo detection
parent
52fc16bc8a
commit
075a8e5ccb
|
@ -1015,12 +1015,12 @@ fn (f &Fn) find_misspelled_local_var(name string, min_match f32) string {
|
|||
mut closest := f32(0)
|
||||
mut closest_var := ''
|
||||
for var in f.local_vars {
|
||||
n := '${f.mod}.$var.name'
|
||||
if var.name == '' || !name.starts_with(f.mod) || (n.len - name.len > 3 || name.len - n.len > 3) { continue }
|
||||
p := strings.dice_coefficient(name, n)
|
||||
n := name.all_after('.')
|
||||
if var.name == '' || (n.len - var.name.len > 2 || var.name.len - n.len > 2) { continue }
|
||||
p := strings.dice_coefficient(var.name, n)
|
||||
if p > closest {
|
||||
closest = p
|
||||
closest_var = n
|
||||
closest_var = var.name
|
||||
}
|
||||
}
|
||||
return if closest >= min_match { closest_var } else { '' }
|
||||
|
|
|
@ -856,11 +856,11 @@ fn (t &Type) contains_field_type(typ string) bool {
|
|||
fn (table &Table) identify_typo(name string, current_fn &Fn, fit &FileImportTable) string {
|
||||
// dont check if so short
|
||||
if name.len < 2 { return '' }
|
||||
min_match := 0.8 // for dice coefficient between 0.0 - 1.0
|
||||
min_match := 0.50 // for dice coefficient between 0.0 - 1.0
|
||||
name_orig := name.replace('__', '.').replace('_dot_', '.')
|
||||
mut output := ''
|
||||
// check functions
|
||||
mut n := table.find_misspelled_fn(name_orig, min_match)
|
||||
mut n := table.find_misspelled_fn(name, fit, min_match)
|
||||
if n != '' {
|
||||
output += '\n * function: `$n`'
|
||||
}
|
||||
|
@ -878,16 +878,27 @@ fn (table &Table) identify_typo(name string, current_fn &Fn, fit &FileImportTabl
|
|||
}
|
||||
|
||||
// find function with closest name to `name`
|
||||
fn (table &Table) find_misspelled_fn(name string, min_match f32) string {
|
||||
fn (table &Table) find_misspelled_fn(name string, fit &FileImportTable, min_match f32) string {
|
||||
mut closest := f32(0)
|
||||
mut closest_fn := ''
|
||||
is_main_fn := name.starts_with('main__')
|
||||
n1 := if is_main_fn { name.right(6) } else { name }
|
||||
for _, f in table.fns {
|
||||
n := '${f.mod}.$f.name'
|
||||
if !name.starts_with(f.mod) || (n.len - name.len > 3 || name.len - n.len > 3) { continue }
|
||||
p := strings.dice_coefficient(name, n)
|
||||
if n1.len - f.name.len > 2 || f.name.len - n1.len > 2 { continue }
|
||||
if !(f.mod in ['', 'main', 'builtin']) {
|
||||
mut mod_imported := false
|
||||
for _, m in fit.imports {
|
||||
if f.mod == m {
|
||||
mod_imported = true
|
||||
break
|
||||
}
|
||||
}
|
||||
if !mod_imported { continue }
|
||||
}
|
||||
p := strings.dice_coefficient(n1, f.name)
|
||||
if p > closest {
|
||||
closest = p
|
||||
closest_fn = n
|
||||
closest_fn = f.name
|
||||
}
|
||||
}
|
||||
return if closest >= min_match { closest_fn } else { '' }
|
||||
|
@ -899,7 +910,7 @@ fn (table &Table) find_misspelled_imported_mod(name string, fit &FileImportTable
|
|||
mut closest_mod := ''
|
||||
for alias, mod in fit.imports {
|
||||
n := '${fit.module_name}.$alias'
|
||||
if !name.starts_with(fit.module_name) || (n.len - name.len > 3 || name.len - n.len > 3) { continue }
|
||||
if !name.starts_with(fit.module_name) || (n.len - name.len > 2 || name.len - n.len > 2) { continue }
|
||||
p := strings.dice_coefficient(name, n)
|
||||
if p > closest {
|
||||
closest = p
|
||||
|
|
|
@ -40,22 +40,21 @@ pub fn dice_coefficient(s1, s2 string) f32 {
|
|||
if s1.len == 0 || s2.len == 0 { return 0.0 }
|
||||
if s1 == s2 { return 1.0 }
|
||||
if s1.len < 2 || s2.len < 2 { return 0.0 }
|
||||
a := if s1.len > s2.len { s1 } else { s2 }
|
||||
b := if a == s1 { s2 } else { s1 }
|
||||
mut first_bigrams := map[string]int
|
||||
for i := 0; i < s1.len-1; i++ {
|
||||
a := s1[i]
|
||||
b := s1[i+1]
|
||||
bigram := (a+b).str()
|
||||
for i := 0; i < a.len-1; i++ {
|
||||
bigram := a.substr(i, i+2)
|
||||
first_bigrams[bigram] = if bigram in first_bigrams { first_bigrams[bigram]+1 } else { 1 }
|
||||
}
|
||||
mut intersection_size := 0
|
||||
for i := 0; i < s2.len-1; i++ {
|
||||
a := s2[i]
|
||||
b := s2[i+1]
|
||||
bigram := (a+b).str()
|
||||
for i := 0; i < b.len-1; i++ {
|
||||
bigram := b.substr(i, i+2)
|
||||
count := if bigram in first_bigrams { first_bigrams[bigram] } else { 0 }
|
||||
if count > 0 {
|
||||
first_bigrams[bigram] = count - 1
|
||||
intersection_size++
|
||||
}
|
||||
}
|
||||
return (2.0 * intersection_size) / (f32(s1.len) + f32(s2.len) - 2)
|
||||
return (2.0 * intersection_size) / (f32(a.len) + f32(b.len) - 2)
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue