Test data for line breaking
===========================

Files named ``*.in'' are input text. ``*.out'' are expected outputs.

Default configuration is as following. Overridden values are described
by each test data.

    charmax:        998
    colmin:         0
    colmax:         76
    context:        NONEASTASIAN
    format:         SIMPLE
    hangul as AL:   no
    legacy CM:      yes
    newline:        "\n"
    sizing:         UAX11
    tailoring EAW:  none
    tailoring LBC:  none
    urgent breaking:none
    preprocessing:  none

01 Generic
----------

ar.in
ar.out
    Arabic
el.in
el.out
    Greek
fr.in
fr.out
    French
ja.in
ja.out
    Japanese
ja-a.in
ja-a.out
    Japanese (annotated readings)
ja-k.in
    Japanese (kana transcribed)
ko.in
ko.out
    Korean
ru.in
ru.out
    Russian
th.in
th.out
    Thai
vi.in
vi.out
    Vietnamese
vi-decomp.in
vi-decomp.out
    Vietnamese (decomposed)
zh.in
zh.out
    Chinese Mandarin

02 Hangul text
--------------

amitagyong.in
amitagyong.out
    complex hangul syllables and conjoining jamo.
ko.al.out
    treat hangul syllables and conjoining jamo as alphabetic (AL).

03 Tailoring Line Breaking Classes
----------------------------------

ja-k.in
ja-k.out
    colmax:         72
ja-k.ns.out
    colmax:         72
    kana nonstarters (small kana) are assigned Line Breaking Class ``ID''.

04 Folding/unfolding
--------------------

fr.fixed.out
ja.fixed.out
    same as default but an empty line is inserted after each paragraph.
fr.flowed.out
ja.flowed.out
    RFC 3676 ``Format="FLOWED"; DelSp="YES"'' format.
fr.plain.out
ja.plain.out
    same as default.
quotes.in
    unfolded e-mail text with one problematic line.
quotes.norm.in
    unfolded e-mail text without problematic lines.
quotes.fixed.out
quotes.flowed.out
quotes.plain.out
    folded e-mail text by three methods as above.

05 Long lines
-------------

ecclesiazusae.in
ecclesiazusae.out
ecclesiazusae.CharactersMax.out
    charmax:        79
ecclesiazusae.ColumnsMax.out
    urgent breaking:FORCE
ecclesiazusae.ColumnsMin.out
    colmin:         7
    colmax:         66
    urgent breaking:FORCE

06 East Asian context
---------------------

fr.ea.out
    context:        EASTASIAN

07 n/a
------

08 n/a
------

09 URI
------

uri.in
uri.break.out
    colmax:         1
    preprocessing:  break URIs according to some CMoS rules.
uri.break.http.out
    colmax:         1
    preprocessing:  break HTTP URLs according to some CMoS rules; never break FTP URLs.
uri.nonbreak.out
    colmax:         1
    preprocessing:  never break URIs.

10 n/a
------

11 Formatting context
---------------------

fr.format.out
ja.format.out
    insert context names.
fr.newline.out
ko.newline.out
    trim spaces at end of each lines.

12 Indentation
--------------

fr.wrap.out
ja.wrap.out
    Paragraphs are indented by one horizontal tab ("\t") and other lines
    are indented by four spaces ("    ").

13 RFC 3676
-----------

fr.flowed.out
ja.flowed.out
    folding by RFC 3676 ``Format="FLOWED"; DelSp="YES"'' method.
flowedsp.in
flowedsp.out
    unfolding RFC 3676 ``Format="FLOWED"; DelSp="NO"'' (obsoleted
    RFC 2646) format.

14 Non-South East Asian context
-------------------------------

th.al.out
    treat South East Asian complex context (SA) as alphabetic (AL).

15 n/a
------

16 n/a
------

Other useful test data
----------------------

* LineBreakTest.txt file in Unicode Character Database (UCD) may be
  useful for regression test:
    http://www.unicode.org/Public/UCD/

$$