After including the use of glib we might as well replace inih
with the glib key file parser.
All configuraiton file parsing has been reworked and also the options
parsing has been cleaned up, resulting in better and stricter
configuration file and option value checks.
Compared to old, configuration files now requires any default
configurations to be put in a group/section named [default].
Configuration file keywords such as "enable", "disable", "on",
"off", "yes", "no", "0", "1" have been retired. Now only "true" and
"false" apply to boolean configuration options. This is done to simplify
things and avoid any confusion.
The pattern option feature has been reworked so now the user can now
access the full match string and any matching subexpression using the
%mN syntax.
For example:
[usb devices]
pattern = usb([0-9]*)
device = /dev/ttyUSB%m1
Then when using tio:
$ tio usb12
%m0 = 'usb12' // Full match string
%m1 = 12 // First match subexpression
Which results in device = /dev/ttyUSB12
* Rename --list-devices to --list
* Rename --no-autoconnect to --no-reconnect
* Switch -l and -L options
* -l now lists available serial devices
* -L enables log to file
* Add option --auto-connect <strategy>
* Supported strategies:
* "new" - Waits to connect first new appearing serial device
* "latest" - Connects to latest registered serial device
* "direct" - Connect directly to specified serial device (default)
* Add options to exclude serial devices from auto connect strategy by
pattern
* Supported exclude options:
* --exclude-devices <pattern>
Example: '--exclude-devices "/dev/ttyUSB2,/dev/ttyS?"'
* --exclude-drivers <pattern>
Example: '--exclude-drivers "cdc_acm"'
* --exclude-tids <pattern>
Example: '--exclude-tids "yW07,bCC2"'
* Patterns support '*' and '?'
* Connect to same port/device combination via unique topology ID (TID)
* Topology ID is a 4 digit base62 encoded hash of a device topology
string coming from the Linux kernel. This means that whenever you
plug in the same e.g. USB serial port device to the same USB hub
port connected via the exact same hub topology all the way to your
computer, you will get the same unique TID.
* Useful for stable reconnections when serial device has no serial
device by ID
* For now, only tested on Linux.
* Reworked and improved listing of serial devices to show serial devices:
* By device
* Including TID, uptime, driver, and description.
* Sorted by uptime (newest device listed last)
* By unique topology ID
* By ID
* By path
* Add script interface 'list = tty_search()' for searching for serial
devices.
Add --log-append option which makes tio append to any existing log file.
This also changes the default behaviour of tio from appending to
overwriting any existing log file. Now you have to use this new option
to make tio append.
Add key command 'ctrl-t f' which will toggle log on/off.
If no log filename has been specified via the 'log-filename' option then
tio will automatically generate a new log filename every time the log
feature is toggled on. Meaning, when toggled multiple times, multiple
log files will be generated.
However, if a log filename has been specified, tio will only write and
append to that same file.
Replace flushing/writing of log at every log write operation with line
buffering, meaning log will be written line by line to make it more I/O
friendly but still update frequently.
Clean up so that only the following error related printing functions are
used: tio_error_printf(), tio_error_printf_silent(),
tio_warning_printf().
A session mode switch is introduced for error printing so that it will
print error messages with better formatting depending on in or out of
session.
Logfiles are important to see what happend, in particular if something
unexpected happened; so we want to make sure that the logfile is flushed
to disk.
Before this change, the logfile was typically written at the end in
a large chunk as the default (large) buffering applied. Now, characters are
written out ASAP, so it is possible to get a live-view with a
tail -f <logfile>
Renamed to "tio" because it is shorter and this new name also more
precisely reflects what the program is - a simple TTY terminal I/O
application.
"tio" can be considered short for terminal I/O or TTY I/O or a
combination of the two, whichever you prefer.
Also, wanted to avoid naming conflicts with other projects.
