lib/string: add `Text::escape_to_mk` to escape to Makefiles
[nit.git] / lib / standard / string.nit
index 011cfa1..0ffb303 100644 (file)
@@ -181,8 +181,25 @@ abstract class Text
        #
        #     assert "abcd".has_substring("bc",1)            ==  true
        #     assert "abcd".has_substring("bc",2)            ==  false
+       #
+       # Returns true iff all characters of `str` are presents
+       # at the expected index in `self.`
+       # The first character of `str` being at `pos`, the second
+       # character being at `pos+1` and so on...
+       #
+       # This means that all characters of `str` need to be inside `self`.
+       #
+       #     assert "abcd".has_substring("xab", -1)         == false
+       #     assert "abcd".has_substring("cdx", 2)          == false
+       #
+       # And that the empty string is always a valid substring.
+       #
+       #     assert "abcd".has_substring("", 2)             == true
+       #     assert "abcd".has_substring("", 200)           == true
        fun has_substring(str: String, pos: Int): Bool
        do
+               if str.is_empty then return true
+               if pos < 0 or pos + str.length > length then return false
                var myiter = self.chars.iterator_from(pos)
                var itsiter = str.chars.iterator
                while myiter.is_ok and itsiter.is_ok do
@@ -369,6 +386,33 @@ abstract class Text
        #     assert "\na\nb\tc\t".trim          == "a\nb\tc"
        fun trim: SELFTYPE do return (self.l_trim).r_trim
 
+       # Justify a self in a space of `length`
+       #
+       # `left` is the space ratio on the left side.
+       # * 0.0 for left-justified (no space at the left)
+       # * 1.0 for right-justified (all spaces at the left)
+       # * 0.5 for centered (half the spaces at the left)
+       #
+       #     assert "hello".justify(10, 0.0)  == "hello     "
+       #     assert "hello".justify(10, 1.0)  == "     hello"
+       #     assert "hello".justify(10, 0.5)  == "  hello   "
+       #
+       # If `length` is not enough, `self` is returned as is.
+       #
+       #     assert "hello".justify(2, 0.0)   == "hello"
+       #
+       # REQUIRE: left >= 0.0 and left <= 1.0
+       # ENSURE: `self.length <= length implies result.length == length`
+       # ENSURE: `self.length >= length implies result == self
+       fun justify(length: Int, left: Float): SELFTYPE
+       do
+               var diff = length - self.length
+               if diff <= 0 then return self
+               assert left >= 0.0 and left <= 1.0
+               var before = (diff.to_f * left).to_i
+               return " " * before + self + " " * (diff-before)
+       end
+
        # Mangle a string to be a unique string only made of alphanumeric characters
        fun to_cmangle: String
        do
@@ -450,6 +494,50 @@ abstract class Text
        #     assert "\n\"'\\\{\}".escape_to_nit      == "\\n\\\"\\'\\\\\\\{\\\}"
        fun escape_to_nit: String do return escape_more_to_c("\{\}")
 
+       # Escape to POSIX Shell (sh).
+       #
+       # Abort if the text contains a null byte.
+       #
+       #     assert "\n\"'\\\{\}0".escape_to_sh == "'\n\"'\\''\\\{\}0'"
+       fun escape_to_sh: String do
+               var b = new FlatBuffer
+               b.chars.add '\''
+               for i in [0..length[ do
+                       var c = chars[i]
+                       if c == '\'' then
+                               b.append("'\\''")
+                       else
+                               assert without_null_byte: c != '\0'
+                               b.add(c)
+                       end
+               end
+               b.chars.add '\''
+               return b.to_s
+       end
+
+       # Escape to include in a Makefile
+       #
+       # Unfortunately, some characters are not escapable in Makefile.
+       # These characters are `;`, `|`, `\`, and the non-printable ones.
+       # They will be rendered as `"?{hex}"`.
+       fun escape_to_mk: String do
+               var b = new FlatBuffer
+               for i in [0..length[ do
+                       var c = chars[i]
+                       if c == '$' then
+                               b.append("$$")
+                       else if c == ':' or c == ' ' or c == '#' then
+                               b.add('\\')
+                               b.add(c)
+                       else if c.ascii < 32 or c == ';' or c == '|' or c == '\\' or c == '=' then
+                               b.append("?{c.ascii.to_base(16, false)}")
+                       else
+                               b.add(c)
+                       end
+               end
+               return b.to_s
+       end
+
        # Return a string where Nit escape sequences are transformed.
        #
        #     var s = "\\n"
@@ -659,7 +747,11 @@ end
 abstract class FlatText
        super Text
 
-       private var items: NativeString
+       # Underlying C-String (`char*`)
+       #
+       # Warning : Might be void in some subclasses, be sure to check
+       # if set before using it.
+       private var items: NativeString is noinit
 
        # Real items, used as cache for to_cstring is called
        private var real_items: nullable NativeString = null
@@ -689,11 +781,6 @@ private abstract class StringCharView
 
        private var target: SELFTYPE
 
-       private init(tgt: SELFTYPE)
-       do
-               target = tgt
-       end
-
        redef fun is_empty do return target.is_empty
 
        redef fun length do return target.length
@@ -735,6 +822,8 @@ abstract class String
 
        fun insert_at(s: String, pos: Int): SELFTYPE is abstract
 
+       redef fun substrings: Iterator[String] is abstract
+
        # Returns a reversed version of self
        #
        #     assert "hello".reversed  == "olleh"
@@ -826,6 +915,38 @@ abstract class String
 
                return new_str.to_s
        end
+
+       # Returns a capitalized `self`
+       #
+       # Letters that follow a letter are lowercased
+       # Letters that follow a non-letter are upcased.
+       #
+       # SEE : `Char::is_letter` for the definition of letter.
+       #
+       #    assert "jAVASCRIPT".capitalized == "Javascript"
+       #    assert "i am root".capitalized == "I Am Root"
+       #    assert "ab_c -ab0c ab\nc".capitalized == "Ab_C -Ab0C Ab\nC"
+       fun capitalized: SELFTYPE do
+               if length == 0 then return self
+
+               var buf = new FlatBuffer.with_capacity(length)
+
+               var curr = chars[0].to_upper
+               var prev = curr
+               buf[0] = curr
+
+               for i in [1 .. length[ do
+                       prev = curr
+                       curr = self[i]
+                       if prev.is_letter then
+                               buf[i] = curr.to_lower
+                       else
+                               buf[i] = curr.to_upper
+                       end
+               end
+
+               return buf.to_s
+       end
 end
 
 private class FlatSubstringsIter
@@ -833,8 +954,6 @@ private class FlatSubstringsIter
 
        var tgt: nullable FlatText
 
-       init(tgt: FlatText) do self.tgt = tgt
-
        redef fun item do
                assert is_ok
                return tgt.as(not null)
@@ -851,10 +970,10 @@ class FlatString
        super String
 
        # Index in _items of the start of the string
-       private var index_from: Int
+       private var index_from: Int is noinit
 
        # Indes in _items of the last item of the string
-       private var index_to: Int
+       private var index_to: Int is noinit
 
        redef var chars: SequenceRead[Char] = new FlatStringCharView(self)
 
@@ -973,15 +1092,15 @@ 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)
@@ -1203,6 +1322,14 @@ 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
@@ -1265,6 +1392,38 @@ abstract class Buffer
        #     assert b == "hello world!"
        fun lower is abstract
 
+       # Capitalizes each word in `self`
+       #
+       # Letters that follow a letter are lowercased
+       # Letters that follow a non-letter are upcased.
+       #
+       # SEE: `Char::is_letter` for the definition of a letter.
+       #
+       #    var b = new FlatBuffer.from("jAVAsCriPt")"
+       #    b.capitalize
+       #    assert b == "Javascript"
+       #    b = new FlatBuffer.from("i am root")
+       #    b.capitalize
+       #    assert b == "I Am Root"
+       #    b = new FlatBuffer.from("ab_c -ab0c ab\nc")
+       #    b.capitalize
+       #    assert b == "Ab_C -Ab0C Ab\nC"
+       fun capitalize do
+               if length == 0 then return
+               var c = self[0].to_upper
+               self[0] = c
+               var prev: Char = c
+               for i in [1 .. length[ do
+                       prev = c
+                       c = self[i]
+                       if prev.is_letter then
+                               self[i] = c.to_lower
+                       else
+                               self[i] = c.to_upper
+                       end
+               end
+       end
+
        redef fun hash
        do
                if is_dirty then hash_cache = null
@@ -1289,6 +1448,17 @@ 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
@@ -1303,6 +1473,7 @@ class FlatBuffer
                        add(item)
                        return
                end
+               if written then reset
                assert index >= 0 and index < length
                items[index] = item
        end
@@ -1317,6 +1488,7 @@ class FlatBuffer
 
        redef fun clear do
                is_dirty = true
+               if written then reset
                length = 0
        end
 
@@ -1327,6 +1499,9 @@ 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
@@ -1335,7 +1510,9 @@ class FlatBuffer
 
        redef fun to_s: String
        do
-               return to_cstring.to_s_with_length(length)
+               written = true
+               if length == 0 then items = new NativeString(1)
+               return new FlatString.with_infos(items, length, 0, length - 1)
        end
 
        redef fun to_cstring
@@ -1433,6 +1610,7 @@ class FlatBuffer
 
        redef fun reverse
        do
+               written = false
                var ns = calloc_string(capacity)
                var si = length - 1
                var ni = 0
@@ -1455,6 +1633,7 @@ class FlatBuffer
 
        redef fun upper
        do
+               if written then reset
                var it = items
                var id = length - 1
                while id >= 0 do
@@ -1465,6 +1644,7 @@ class FlatBuffer
 
        redef fun lower
        do
+               if written then reset
                var it = items
                var id = length - 1
                while id >= 0 do
@@ -1678,12 +1858,12 @@ redef class Int
 end
 
 redef class Float
-       # Pretty print self, print needoed decimals up to a max of 3.
+       # Pretty representation of `self`, with decimals as needed from 1 to a maximum of 3
        #
-       #     assert 12.34.to_s        == "12.34"
-       #     assert (-0120.03450).to_s  == "-120.035"
+       #     assert 12.34.to_s       == "12.34"
+       #     assert (-0120.030).to_s == "-120.03"
        #
-       # see `to_precision` for a different precision.
+       # see `to_precision` for a custom precision.
        redef fun to_s do
                var str = to_precision( 3 )
                if is_inf != 0 or is_nan then return str
@@ -1702,13 +1882,15 @@ redef class Float
                return str
        end
 
-       # `self` representation with `nb` digits after the '.'.
+       # `String` representation of `self` with the given number of `decimals`
        #
-       #     assert 12.345.to_precision(1) == "12.3"
-       #     assert 12.345.to_precision(2) == "12.35"
-       #     assert 12.345.to_precision(3) == "12.345"
-       #     assert 12.345.to_precision(4) == "12.3450"
-       fun to_precision(nb: Int): String
+       #     assert 12.345.to_precision(0)    == "12"
+       #     assert 12.345.to_precision(3)    == "12.345"
+       #     assert (-12.345).to_precision(3) == "-12.345"
+       #     assert (-0.123).to_precision(3)  == "-0.123"
+       #     assert 0.999.to_precision(2)     == "1.00"
+       #     assert 0.999.to_precision(4)     == "0.9990"
+       fun to_precision(decimals: Int): String
        do
                if is_nan then return "nan"
 
@@ -1719,25 +1901,34 @@ redef class Float
                        return  "-inf"
                end
 
-               if nb == 0 then return self.to_i.to_s
+               if decimals == 0 then return self.to_i.to_s
                var f = self
-               for i in [0..nb[ do f = f * 10.0
+               for i in [0..decimals[ do f = f * 10.0
                if self > 0.0 then
                        f = f + 0.5
                else
                        f = f - 0.5
                end
                var i = f.to_i
-               if i == 0 then return "0.0"
-               var s = i.to_s
+               if i == 0 then return "0." + "0"*decimals
+
+               # Prepare both parts of the float, before and after the "."
+               var s = i.abs.to_s
                var sl = s.length
-               if sl > nb then
-                       var p1 = s.substring(0, s.length-nb)
-                       var p2 = s.substring(s.length-nb, nb)
-                       return p1 + "." + p2
+               var p1
+               var p2
+               if sl > decimals then
+                       # Has something before the "."
+                       p1 = s.substring(0, sl-decimals)
+                       p2 = s.substring(sl-decimals, decimals)
                else
-                       return "0." + ("0"*(nb-sl)) + s
+                       p1 = "0"
+                       p2 = "0"*(decimals-sl) + s
                end
+
+               if i < 0 then p1 = "-" + p1
+
+               return p1 + "." + p2
        end
 
        # `self` representation with `nb` digits after the '.'.
@@ -1952,7 +2143,8 @@ extern class NativeString `{ char* `}
        fun to_s_with_length(length: Int): FlatString
        do
                assert length >= 0
-               return new FlatString.with_infos(self, length, 0, length - 1)
+               var str = new FlatString.with_infos(self, length, 0, length - 1)
+               return str
        end
 
        fun to_s_with_copy: FlatString
@@ -1960,7 +2152,10 @@ extern class NativeString `{ char* `}
                var length = cstring_length
                var new_self = calloc_string(length + 1)
                copy_to(new_self, length, 0, 0)
-               return new FlatString.with_infos(new_self, length, 0, length - 1)
+               var str = new FlatString.with_infos(new_self, length, 0, length - 1)
+               new_self[length] = '\0'
+               str.real_items = new_self
+               return str
        end
 end
 
@@ -2018,7 +2213,8 @@ end
 #
 # Note: it caching is not usefull, see `alpha_comparator`
 class CachedAlphaComparator
-       super Comparator[Object]
+       super Comparator
+       redef type COMPARED: Object
 
        private var cache = new HashMap[Object, String]
 
@@ -2036,7 +2232,7 @@ end
 
 # see `alpha_comparator`
 private class AlphaComparator
-       super Comparator[Object]
+       super Comparator
        redef fun compare(a, b) do return a.to_s <=> b.to_s
 end
 
@@ -2048,7 +2244,7 @@ end
 #     var a = [1, 2, 3, 10, 20]
 #     alpha_comparator.sort(a)
 #     assert a == [1, 10, 2, 20, 3]
-fun alpha_comparator: Comparator[Object] do return once new AlphaComparator
+fun alpha_comparator: Comparator do return once new AlphaComparator
 
 # The arguments of the program as given by the OS
 fun args: Sequence[String]