CHANGES (p.zed.2.0.6.doc.CHANGES)

add wanted_column support for wide width character
Zed_lines: get_idx_by_width set row column_width return the offset of the character at [row, column_width]

Zed_string
- exception Invalid of string * string raised when an invalid Zed_char sequence is encounted
- next_ofs : t -> int -> int returns the offset of the next zchar in t
- prev_ofs : t -> int -> int returns the offset of the prev zchar in t

Zed_char: add an indv_combining option to the transforming functions(of_uChars, zChars_of_uChars, of_utf8) to determine whether to extract individual combining marks from the parameter (#18)
Zed_char: clarify some documentation comments (#18)

module Zed_char
module Zed_string
Zed_cursor
- column_display: Zed_cursor.t -> int React.signal
- get_column: Zed_cursor.t -> int
- coordinates_display: Zed_cursor.t -> (int * int) React.signal
- get_coordinates: Zed_cursor.t -> int * int
Zed_edit
- regexp_word_core: Zed_re.Core.t
- regexp_word_raw: Zed_re.raw.t
- match_by_regexp_core
- match_by_regexp_raw

Zed_rope
- Zed_rope.empty is a function now
- Other functions in this module take Zed_char.t or Zed_string.t as arguemnts instead of UChar.t or Zed_utf8.t
- module Zipper is divided into two modules, Zip and Zip_raw, to navigate over a rope by Zed_char.t or UChar.t, respectively
- module Text is divided into three modules, Text, Text_core, Text_raw, to manager Zed_rope by Zed_char.t, the core UChar.t of a Zed_char.t and raw UChar.t, respectively
Zed_re is therefore divided into two modules: Core and Raw
Zed_cursor: the type changes is defined as a structure and has two more fields: added_width and removed_width

Switch to jbuilder (Rudi Gringberg, #4)
Make {delete_,kill_,}{next,prev}_word consistent near the start/end of the buffer (Fabian (github use copy), #5)

Zed_rope fixes:
- rev_map: fix recursion
- enforce evaluation order in map & rev_map