Log message:
*: recursive bump for icu 74.1

Log message:
*: bump for openssl 3

Log message:
revbump after textproc/icu update

Log message:
massive revision bump after textproc/icu update

Log message:
*: bump PKGREVISION for libunistring shlib major bump

Log message:
Update libzrtpcpp to version 4.7.0

Pkgsrc changes:
Github and pkglint fixes.

Changes from NEWS:

## GNU ZRTP 4.6.6 ##

Small fix in zrtp/crypto/zrtpDh.cpp to fix a small memory leak.

## GNU ZRTP 4.6.5 ##

Cleanup compiler flags, reduce visibility for Android build,
check some buffer length. No functional enhancements, no changes
in API.

## GNU ZRTP 4.6.4 ##

Some fixes to slience Windows C/C++ compiler, fix a few include
statements when using openSSL, small fixes to check disclosure
flag. Reset valid flags when adding a new cache record to avoid
wrong security message.

## GNU ZRTP 4.6.3 ##

A small fix inside the ZRTP main module to ignore malformed
DH1 packets and avoid an NULL pointer access.

## GNU ZRTP 4.6.2 ##

A small fix in the ZrtpCWrapper to fix an issue within 4.6.1
;-)

## GNU ZRTP 4.6.1 ##

A small fix in the ZrtpCWrapper to initialize and use the ZRTP
master instance in case of multi-stream usage. Does not affect
the main ZRTP usage, only projects that use the wrapper such
as PJSIP or Gstreamer projects.

These project should re-compile if they use the multi-stream
feature.

## GNU ZRTP 4.6.0 ##

Only a small add-on to the code to implement handling of the
disclosure flag. See RFC6189, chapter 11 for more details
about the disclosure flag.

Because the API changed, thus it's necessary to recompile
applications that use the new library version.

## GNU ZRTP 4.5.0 ##

Added a new SAS algorithm 'B32E' that uses 32 Unicode Emoji
code points instead of 32 ASCII characters. Application that
are able to display Emojis may use this new SAS algorithm to
display nice Emojis instead of 'boring' ASCII letters and
digits.

Some technical details:

* the 32 selected emojis are easily distinguishable, known to
  everyone, not offending etc, and use standard Unicode code
  points
* select colored emojis that look good on white and on black
  backgrounds (most emojis look good on white only)
* select emojis that are available on iOS, Android, Mac OS X
  (Windows not checked)
* the resulting SAS string is UTF-8 encoded, suitable for most
  platforms except Java.

To use the codes for Java the application needs to translate the
UTF-8 encoding into UTF-16 encoding. Because most of the emojis
are Unicode supplementary characters the UTF-8 to UTF-16 conversion
must generate the necessary UTF-16 surrogate pairs.

To support the UTF-8 / UTF-16 conversion the common directory
contains conversion functions that I extracted from ICU C/C++
library source.

Because the API changed, thus it's necessary to recompile
applications that use the new library version.

Log message:
revbump for textproc/icu update

Log message:
revbump for icu and libffi