skf - simple kanji filter

This is patch level 15 of skf-2.10 series.

changelog for skf 2.10.15 unstable; urgency=low

• [skf:41024] fix double free on ruby extension.

• fix: man page formatting.

• fix: unnecessary directory creation during install

• Change Unicode Data to 15.0.0

• [skf:46032] fix -q option bug.

• remove CHANGES_ja.txt.

changelog for skf 2.10.14 unstable; urgency=low

• [skf:39882] process surrogate pairs if python LWL input is 2BYTE_KIND.

• Detect surrogate pair in 2BYTE_KIND mode, and switch to 4BYTE_KIND.

• assumes MIME output/UTF-7 is ascii, not target codeset.

• native UTF-16/32 endianness follows machine endianness.

• fix codeset trailing spaces handling bug.

changelog for skf 2.10.13 unstable; urgency=low

• [skf:39882] python unicode pep393 support

• add _DEFAULT_SOURCE to make compiler warning suppressed.

• fix resetting debug mode on LWL call.

• fix py-out-binary does not work.

• delete lineend from charset name string in LWL extension.

• fix perl/ruby/python sitearchdir does not return a correct path.

• fix utf32beb/leb isn't treated as Unic*de.

• change Unicode Data to 13.0.0

• added --fix-ms-radical-bug option

changelog for skf 2.10.12 unstable; urgency=medium

• [skf:39695] make localedir option effective

• [skf:39836] support behaviour change in ruby2.7

• fix: udhdata allocate error handling

• fix: make oldincludedir/docdir option effective

• fix: add support ruby2.8, python 3.9 to configure

• fix: use python-config for python 2.7

• change unicode data to 12.1.0

• some cosmetic bug fixed.

changelog for skf 2.10.11

• security fix punycode or uri as ic crashes

• fix: inconsistent nkf-help message after 2.10.8 fix

• fix: compile fails if ARIB_SUPPORT is disabled

• fix: re-enabled --enable-python2 option in configure

• added cesu-8, mutf8, utf7-imap support to man page

• fix some codes does not decode correctly in rot mode

changelog for skf 2.10.10

• [skf:39111] fix: man directory setting through configure.

• add configuration display for INPLACESTAR

• security fix hungup in GSM udh detection routine.

• fix: CRLF in utf-16le detection failure

• fix: CRLF in utf-32le/32be detection failure

• added a guard for vni parsing.

• fix: -r with SJIS-in does not work.

• fix: --set option with separate arguments error handling bug.

• fix: --preferred-subst-char does not work

• fix: out_codeset display in debug mode does not work

• fix: rot char in be-in deleted from stream due to wrong bound check.

• fix: GSM UDH detect return range inconsistent.

• add cesu-8 and java modified utf8 for codeset.

• make python3 as default.

changelog for skf 2.10.9

• [skf:38830] _POSIX_SOURCE definition hides mkstemp

• new japanese era support

• [skf:38933] nkf in-place compatibility. Suppress '* suffix support.

• fix various issues in code detection logic.

• fix no-utf7 & detect-x0201-kana was overturned

changelog for skf 2.10.8.2

• Fix: lwl compile error [skf:38817]

• Fix: ruby 2.6.0 extension compile error

changelog for skf 2.10.8.1

• remove debug output

• fb-html changes to decimal from hex.

changelog for skf 2.10.8

• [skf:38570] do not treat ruby "8-bit binary" as codeset.

• don't bite arguments after --

• add -t option for nkf compatibility

• undocumented nkf behavior support: suffix=*, -t2, -L0, -g0-2

• re-enable -n option for nkf-compatibility

• kicks pre-undefined characters in -i, -o option for nkf-compatibility

• security hungup fixed if file open failed in in-place option.

• security fix crash in in-place option.

• fix codeset inquiry in in-place option.

• tries to unlink temporal file when in-place rename failed.

• added -fb-java, fb-perl, fb-skip support for nkf compatibility.

• added -fb-python.

• retain input files when suffix is added in --in-place, overwrite

changelog for skf 2.10.7

• fix: keis strout doesn't convert ascii strings to EBCDICs
• fix: ebcdic force-lineend changed to NEL
• fix: inquiry output code control. fixes to "LOCALE" with default to utf-8.

• UTF-7-IMAP conversion input/output added.

• python 3.6, 3.7 support.

• Update Unicodedata to Unicode 11.0.

• fix: sup_space_conv setting condition.
• fix: B-right/V strout doesn't convert ascii strings to EBCDICs
• fix: punycode with idn prohibit char causes character drop.
• fix: similar bug in hz & zW EOF handling.

• fix: disable MIME cp932 related cname.

• Security fix clashes related space-convert
• fix: convert-html-uri does not work.

changelog for skf 2.10.6

• fix: locale output doesn't use nl_langinfo result

• fix: set-lang doesn't work.

• fix: cname containing '.' isn't recognized.

• fix: rfc1554 iso-2022-jp-2-kr doesn't work.

• fix: ka + handakuten outputs garbage.

• fix: -v doesn't reflect language setting.

• fix: euc-jp-ms etc. get wrong codeset

• fix: crash on bright/v output with html-sanitize

• fix: couldn't return from unic*de variants by standard return.

• fix: JIS X-0221(iso 10646) calling sequence error.

• fix: JIS X-0211 pass through state change.

• fix: cp932/cp51932 gaiji area output

• fix: gsm 03.38 UDH nullified and table added. register tags and variation selector supplement as a likely code

• fix: Added TSCII 1.7, Burgarian MIK as separate codeset

• fix: koi-8, bright, koi-8-u output error

• fix: make ruby2.5 aware.

changelog for skf 2.10.5

• fix: added repeating to disabled macro definition.

• security fix sjis-oldntt crashes in private area.

• fix: -i, -o option revived with new spec.

• fix: some option doesn't work for racing.

• fix: convert-html-uri does not work.

• fix: latin2html output goes without encode redirection.

• fix: some man page glitches.

• fix: halfwidth mime character count correction.

• fix: convert-html does not handle non-BMP range.

• fix: cns11463 single shift2 fails.

• fix: iso-2022 announcer doesn't properly processed

• fix: cannot return to iso-2022 from unicode.

• fix: trim cp932 to cp51932 gaiji convert condition

• fix: B-Right/V latin output conversion fixed.

• fix: [skf:37767] configure portability fix.

changelog for skf 2.10.4

• security fix crash in unicode decomposition

• security fix crash in U+1f1xx conversion

• security fix crash due to wrong iso-2022-jp-ms table definition

• fix: crash if locale output specified and codeset is unknown.

• fix: --set-g* crashes if following string is not charset

• LWL(like ruby) extension compile

• various bug fix. see debian/changelog

changelog for skf 2.10.3

• fix error message for "-o" option

• fix obsolete option action properly

• fix octal encoding output wrong code on tail.

• fix to suppress MIME encoding on transparent mode

• fix X-0213 reverse conversion table

• nkf-compat hankaku-kana output condition fix

• added cp65001 as UTF-8 alias, and also cp65000+cp1200+1201

• fix B-right output

• fix zW output space escape condition.

changelog for skf 2.10.2

• fix iso-2022-ms table

• fix skf-packedstatus duplicate bit assign

• Changed UnicodeData to Unicode 10.0

• suppress some codeset for MIME use

• fix: iso-2022-jp-2-2004 doesn't show in MIME

• input side cp932 gaiji-area fix. delete cp50221/2 gaiji definition

changelog for skf.2.10.1

• fix: Q-encode overdecoding on nkf-compatible mode

• fix: doesn't detect binary name change on systems with binary-suffix.

• fix: man pages and documents