X-Git-Url: http://nitlanguage.org diff --git a/lib/standard/string.nit b/lib/standard/string.nit index 07a7e1f..d41ab1a 100644 --- a/lib/standard/string.nit +++ b/lib/standard/string.nit @@ -178,7 +178,7 @@ abstract class Text # As with substring, a `from` index < 0 will be replaced by 0 fun substring_from(from: Int): SELFTYPE do - if from > self.length then return empty + if from >= self.length then return empty if from < 0 then from = 0 return substring(from, length - from) end @@ -247,8 +247,8 @@ abstract class Text var i = 0 var neg = false - for c in self.chars - do + for j in [0..length[ do + var c = chars[j] var v = c.to_i if v > base then if neg then @@ -278,12 +278,10 @@ abstract class Text fun is_numeric: Bool do var has_point_or_comma = false - for i in self.chars - do - if not i.is_numeric - then - if (i == '.' or i == ',') and not has_point_or_comma - then + for i in [0..length[ do + var c = chars[i] + if not c.is_numeric then + if (c == '.' or c == ',') and not has_point_or_comma then has_point_or_comma = true else return false @@ -300,7 +298,8 @@ abstract class Text # assert "0G".is_hex == false fun is_hex: Bool do - for c in self.chars do + for i in [0..length[ do + var c = chars[i] if not (c >= 'a' and c <= 'f') and not (c >= 'A' and c <= 'F') and not (c >= '0' and c <= '9') then return false @@ -316,7 +315,8 @@ abstract class Text # assert "Hello World".is_upper == false fun is_upper: Bool do - for char in self.chars do + for i in [0..length[ do + var char = chars[i] if char.is_lower then return false end return true @@ -329,7 +329,8 @@ abstract class Text # assert "Hello World".is_lower == false fun is_lower: Bool do - for char in self.chars do + for i in [0..length[ do + var char = chars[i] if char.is_upper then return false end return true @@ -363,7 +364,7 @@ abstract class Text if iter.item.ascii > 32 then break iter.next end - if iter.index == length then return self.empty + if iter.index < 0 then return self.empty return self.substring(0, iter.index + 1) end @@ -379,7 +380,8 @@ abstract class Text do var res = new FlatBuffer var underscore = false - for c in self.chars do + for i in [0..length[ do + var c = chars[i] if (c >= 'a' and c <= 'z') or (c >='A' and c <= 'Z') then res.add(c) underscore = false @@ -412,7 +414,8 @@ abstract class Text fun escape_to_c: String do var b = new FlatBuffer - for c in self.chars do + for i in [0..length[ do + var c = chars[i] if c == '\n' then b.append("\\n") else if c == '\0' then @@ -464,7 +467,8 @@ abstract class Text do var res = new FlatBuffer.with_capacity(self.length) var was_slash = false - for c in chars do + for i in [0..length[ do + var c = chars[i] if not was_slash then if c == '\\' then was_slash = true @@ -498,7 +502,8 @@ abstract class Text do var buf = new FlatBuffer - for c in self.chars do + for i in [0..length[ do + var c = chars[i] if (c >= '0' and c <= '9') or (c >= 'a' and c <= 'z') or (c >= 'A' and c <= 'Z') or @@ -554,6 +559,29 @@ abstract class Text return buf.to_s end + # Escape the four characters `<`, `>`, `&`, and `"` with their html counterpart + # + # assert "a&b->\"x\"".html_escape == "a&b->"x"" + fun html_escape: SELFTYPE + do + var buf = new FlatBuffer + + for i in [0..length[ do + var c = chars[i] + if c == '&' then + buf.append "&" + else if c == '<' then + buf.append "<" + else if c == '>' then + buf.append ">" + else if c == '"' then + buf.append """ + else buf.add c + end + + return buf.to_s + end + # Equality of text # Two pieces of text are equals if thez have the same characters in the same order. # @@ -613,7 +641,8 @@ abstract class Text # djb2 hash algorithm var h = 5381 - for char in self.chars do + for i in [0..length[ do + var char = chars[i] h = h.lshift(5) + h + char.ascii end @@ -739,7 +768,8 @@ abstract class String var new_str = new FlatBuffer.with_capacity(self.length) var is_first_char = true - for char in self.chars do + for i in [0..length[ do + var char = chars[i] if is_first_char then new_str.add(char.to_lower) is_first_char = false @@ -777,7 +807,8 @@ abstract class String var is_first_char = true var follows_us = false - for char in self.chars do + for i in [0..length[ do + var char = chars[i] if is_first_char then new_str.add(char) is_first_char = false @@ -825,6 +856,15 @@ class FlatString redef var chars: SequenceRead[Char] = new FlatStringCharView(self) + redef fun [](index) + do + # Check that the index (+ index_from) is not larger than indexTo + # In other terms, if the index is valid + assert index >= 0 + assert (index + index_from) <= index_to + return items[index + index_from] + end + ################################################ # AbstractString specific methods # ################################################ @@ -931,15 +971,15 @@ class FlatString redef fun to_cstring: NativeString do - if real_items != null then return real_items.as(not null) - if index_from > 0 or index_to != items.cstring_length - 1 then + if real_items != null then + return real_items.as(not null) + else var newItems = calloc_string(length + 1) self.items.copy_to(newItems, length, index_from, 0) newItems[length] = '\0' self.real_items = newItems return newItems end - return items end redef fun ==(other) @@ -1021,8 +1061,9 @@ class FlatString s.items.copy_to(target_string, its_length, 0, my_length) else var curr_pos = my_length - for i in s.chars do - target_string[curr_pos] = i + for i in [0..s.length[ do + var c = s.chars[i] + target_string[curr_pos] = c curr_pos += 1 end end @@ -1095,7 +1136,7 @@ private class FlatStringReverseIterator curr_pos = pos + tgt.index_from end - redef fun is_ok do return curr_pos >= 0 + redef fun is_ok do return curr_pos >= target.index_from redef fun item do return target_items[curr_pos] @@ -1160,6 +1201,14 @@ abstract class Buffer # Specific implementations MUST set this to `true` in order to invalidate caches protected var is_dirty = true + # Copy-On-Write flag + # + # If the `Buffer` was to_s'd, the next in-place altering + # operation will cause the current `Buffer` to be re-allocated. + # + # The flag will then be set at `false`. + protected var written = false + # Modifies the char contained at pos `index` # # DEPRECATED : Use self.chars.[]= instead @@ -1246,6 +1295,24 @@ class FlatBuffer redef fun substrings do return new FlatSubstringsIter(self) + # Re-copies the `NativeString` into a new one and sets it as the new `Buffer` + # + # This happens when an operation modifies the current `Buffer` and + # the Copy-On-Write flag `written` is set at true. + private fun reset do + var nns = new NativeString(capacity) + items.copy_to(nns, length, 0, 0) + items = nns + written = false + end + + redef fun [](index) + do + assert index >= 0 + assert index < length + return items[index] + end + redef fun []=(index, item) do is_dirty = true @@ -1253,6 +1320,7 @@ class FlatBuffer add(item) return end + if written then reset assert index >= 0 and index < length items[index] = item end @@ -1267,6 +1335,7 @@ class FlatBuffer redef fun clear do is_dirty = true + if written then reset length = 0 end @@ -1277,6 +1346,9 @@ class FlatBuffer var c = capacity if cap <= c then return while c <= cap do c = c * 2 + 2 + # The COW flag can be set at false here, since + # it does a copy of the current `Buffer` + written = false var a = calloc_string(c+1) if length > 0 then items.copy_to(a, length, 0, 0) items = a @@ -1285,7 +1357,8 @@ class FlatBuffer redef fun to_s: String do - return to_cstring.to_s_with_length(length) + written = true + return new FlatString.with_infos(items, length, 0, length - 1) end redef fun to_cstring @@ -1314,8 +1387,9 @@ class FlatBuffer s.items.copy_to(items, length, 0, 0) else var curr_pos = 0 - for i in s.chars do - items[curr_pos] = i + for i in [0..s.length[ do + var c = s.chars[i] + items[curr_pos] = c curr_pos += 1 end end @@ -1343,8 +1417,9 @@ class FlatBuffer s.items.copy_to(items, sl, 0, length) else var curr_pos = self.length - for i in s.chars do - items[curr_pos] = i + for i in [0..s.length[ do + var c = s.chars[i] + items[curr_pos] = c curr_pos += 1 end end @@ -1381,6 +1456,7 @@ class FlatBuffer redef fun reverse do + written = false var ns = calloc_string(capacity) var si = length - 1 var ni = 0 @@ -1403,6 +1479,7 @@ class FlatBuffer redef fun upper do + if written then reset var it = items var id = length - 1 while id >= 0 do @@ -1413,6 +1490,7 @@ class FlatBuffer redef fun lower do + if written then reset var it = items var id = length - 1 while id >= 0 do @@ -1787,18 +1865,51 @@ redef class Collection[E] end redef class Array[E] + # Fast implementation redef fun to_s do - var s = new FlatBuffer - var i = 0 var l = length + if l == 0 then return "" + if l == 1 then if self[0] == null then return "" else return self[0].to_s + var its = _items + var na = new NativeArray[String](l) + var i = 0 + var sl = 0 + var mypos = 0 while i < l do - var e = self[i] - if e != null then s.append(e.to_s) + var itsi = its[i] + if itsi == null then + i += 1 + continue + end + var tmp = itsi.to_s + sl += tmp.length + na[mypos] = tmp i += 1 + mypos += 1 end - return s.to_s + var ns = new NativeString(sl + 1) + ns[sl] = '\0' + i = 0 + var off = 0 + while i < mypos do + var tmp = na[i] + var tpl = tmp.length + if tmp isa FlatString then + tmp.items.copy_to(ns, tpl, tmp.index_from, off) + off += tpl + else + for j in tmp.substrings do + var s = j.as(FlatString) + var slen = s.length + s.items.copy_to(ns, slen, s.index_from, off) + off += slen + end + end + i += 1 + end + return ns.to_s_with_length(sl) end end @@ -1843,7 +1954,8 @@ end # Native strings are simple C char * extern class NativeString `{ char* `} super StringCapable - + # Creates a new NativeString with a capacity of `length` + new(length: Int) is intern fun [](index: Int): Char is intern fun []=(index: Int, item: Char) is intern fun copy_to(dest: NativeString, length: Int, from: Int, to: Int) is intern