Day 860
Qutebrowser crashing - again
Ubuntu 18.04, qutebrowser etc, as usual. What helped was creating the environment with these options:
python3 scripts/mkvenv.py --pyqt-version 5.14
jq | less zsh alias
Should’ve done this a long time ago:
lq() {
jq . "$1" -C | less
}
kitty terminal copy url
From config; I should use them more.
# Select a filename and copy it
map kitty_mod+p>c kitten hints --type path --program @
#: Select a path/filename and open it with the default open program.
map kitty_mod+p>o kitten hints --type line --program -
update-alternatives & installing another gcc
Nicely described: How to switch between multiple GCC and G++ compiler versions on Ubuntu 20.04 LTS Focal Fossa - LinuxConfig.org
# install stuff
$ sudo apt -y install gcc-7 g++-7 gcc-8 g++-8 gcc-9 g++-9
# Add it to update-alternatives
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-7 7
sudo update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-7 7
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-8 8
sudo update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-8 8
sudo update-alternatives --install /usr/bin/gcc gcc /usr/bin/gcc-9 9
sudo update-alternatives --install /usr/bin/g++ g++ /usr/bin/g++-9 9
# choose the default one
$ sudo update-alternatives --config gcc
There are 3 choices for the alternative gcc (providing /usr/bin/gcc).
Selection Path Priority Status
------------------------------------------------------------
0 /usr/bin/gcc-9 9 auto mode
1 /usr/bin/gcc-7 7 manual mode
* 2 /usr/bin/gcc-8 8 manual mode
3 /usr/bin/gcc-9 9 manual mode
Press to keep the current choice[*], or type selection number:
From the docs:
--install link name path priority
Python pip
Editable installations (pip install -e .) are a thing. TODO - learn more about them.
Qutebrowser config - adding bindings for tabs 20-30
Given that the standard ones are not enough for me, and even my additional ones for 10-20 are not enough, added a third level:
config.bind('1', 'tab-focus 1')
config.bind('2', 'tab-focus 2')
config.bind('3', 'tab-focus 3')
config.bind('4', 'tab-focus 4')
config.bind('5', 'tab-focus 5')
config.bind('6', 'tab-focus 6')
config.bind('7', 'tab-focus 7')
config.bind('8', 'tab-focus 8')
config.bind('9', 'tab-focus 9')
config.bind('0', 'tab-focus 10')
config.bind('<Alt-1>', 'tab-focus 11')
config.bind('<Alt-2>', 'tab-focus 12')
config.bind('<Alt-3>', 'tab-focus 13')
config.bind('<Alt-4>', 'tab-focus 14')
config.bind('<Alt-5>', 'tab-focus 15')
config.bind('<Alt-6>', 'tab-focus 16')
config.bind('<Alt-7>', 'tab-focus 17')
config.bind('<Alt-8>', 'tab-focus 18')
config.bind('<Alt-9>', 'tab-focus 19')
config.bind('<Alt-0>', 'tab-focus 20')
config.bind('<Alt-Ctrl-1>', 'tab-focus 21')
config.bind('<Alt-Ctrl-2>', 'tab-focus 22')
config.bind('<Alt-Ctrl-3>', 'tab-focus 23')
config.bind('<Alt-Ctrl-4>', 'tab-focus 24')
config.bind('<Alt-Ctrl-5>', 'tab-focus 25')
config.bind('<Alt-Ctrl-6>', 'tab-focus 26')
config.bind('<Alt-Ctrl-7>', 'tab-focus 27')
config.bind('<Alt-Ctrl-8>', 'tab-focus 28')
config.bind('<Alt-Ctrl-9>', 'tab-focus 29')
config.bind('<Alt-Ctrl-0>', 'tab-focus -1')
EDIT: Actually, to think of it, in for a penny, in for a pound!
for i in range(30, 60):
config.bind(','+str(i), 'tab-focus '+str(i))
Takes about 9 seconds to :config-source everything, but then works like a charm! And doesn’t seem to make anything else slower (strangely, even startup is as usual).
pycharm can parse markdown!
Opened a README.md, and see it being rendered nicely to the left. I can also edit it directly. Wow.
Website with references / cheat sheets for a lot of CLI programs
sed Cheat Sheet - very down-to-earth, “praxisnah”, I like it. Except for the idiotic scrolling override animations
jq basics - again
-
I should use
'for the filter,"for any string elements inside it -
select- Get full record if it matches something
jq '.results[] | select(.name == "John") | {age}' # Get age for 'John'
-
Value VS key-value
jq '.something'gets the content of fieldssomethingremoving the keyjq '. | {something}'gets key-value ofsomething- Sample:
$ jq '. | select(.tokens[0]=="Tel") | .tokens[]' mvs.json
"Tel"
":"
$ jq '. | select(.tokens[0]=="Tel") | .tokens' mvs.json
[
"Tel",
":"
]
$ jq '. | select(.tokens[0]=="Tel") | {tokens}' mvs.json
{
"tokens": [
"Tel",
":"
]
}
|keysto extract keys only
jq Cheet Sheet · GitHub also nice
TIl that you don’t need jq '. | keys', jq 'keys' etc is enough.
- `‘del(.tokens)’ to delete a key
- Indexing works like in Python, say
jq '.[-2:]' 'sort_by(.foo)'
I think now I’m ready for the holy of holies: jq 1.4 Manual
{user, title: .titles[]}will return an array of {user, title} for each value inside.titles[]!- Putting
()s around an expression means it’ll be evaluated.{(.user): .titles}will use the value of the keyuser!
$ jq '. | {(.id): .id}' mvs.json
{
"7574": "7574"
}
- Putting values inside strings with
\(foo)
$ echo "[1,2,3]" | jq '"A string \(.)"'
"A string [1,2,3]"
It’s basically synonymous to python3’s f"My f-{string}"
'.a=23'will produce an output with.abeing set to 23. Will be created if not there.- No “change” is being done, the actual value is the same;
.ain the same filter after a comma will still return the old value.
- No “change” is being done, the actual value is the same;
|=will “update” the value by running its previous value through the expression:
$ echo '{"one": 23,"two":2}' | jq '.one|=(. | tostring)'
{
"one": "23",
"two": 2
}
- slurp mode - instead of returning objects, return a list of objects! For more ‘correct’ json.
jq -sto use, and previosu input can be piped through to it!- `jq ‘.’ file | jq -s
- bash - jq: output array of json objects - Stack Overflow say
'[...]'can be used for the same thing. - though I can’t get this to work
Python JSON parser + jq compact mode
It didn’t read the jq-generated multi-line output without commas between items, but jq compact mode does one record (without comma and not as part of an array) per line, and this gets parsed correctly!
JQ compact mode is jq -c '.' sth.json
Before:
{
"id": "7575",
"ner_tags": [
"6",
"6"
],
"tokens": [
"Tel",
":"
]
}
After:
{"id":"7575","ner_tags":["6","6"],"tokens":["Tel",":"]}
Linux - creating a directory accessible to multiple users via a group
How to Create a Shared Directory for All Users in Linux
# Create the group
$sudo groupadd project
# Add user to this group
$sudo usermod -a -G project theuser
# Change the group of the directory
$ sudo chgrp -R project /var/www/reports/
# Turn on the `setGID` bit, so newly created subfiles inherit the same group as the directory
# And rwxrwx-rx
$ sudo chmod -R 2775 /var/www/reports/