Getting the current username, it sounds like a very easy task:

1
2

$ bash -c 'echo "$USER"'
chao

But depending on the environment variable is not reliable, it can be easily overridden:

1
2

$ USER=foo bash -c 'echo "$USER"'
foo

Or just simply fail to work:

1

$ docker run --rm bash:4.4.23 bash -c 'USER=foo echo "$USER"'

The USER environment was not set in /etc/profile in this case:

1
2
3
4
5
6

$ docker run --rm bash:4.4.23 bash -c 'cat /etc/profile'
export CHARSET=UTF-8
export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
export PAGER=less
export PS1='\h:\w\$ '
umask 022

This is especially important when writing shell script. You do not want to end up with a wrong username.

One solution is to use whoami, which prints the effective user:

1
2

$ USER=foo bash -c 'whoami'
chao

Another solution is to use id command:

1
2

$ USER=foo bash -c 'id -un'
chao

This will eliminate the overridden environment variable issue. Also, check out this StackOverflow post for others and POSIX compatible means:

https://stackoverflow.com/questions/19306771/get-current-users-username-in-bash

Print the JSON object in an array where the object key is equal to the specified value with jq:

1
2
3
4

$ echo '[{"key":"foo"},{"key":"bar"}]' | jq '.[] | select(.key == "foo")'
{
  "key": "foo"
}

Reference:

1
2

$ jq --version
jq-1.5

Some utilities print out JSON data (object) line by line. Line by line JSON is not really a valid JSON format, need to wrap it in bracket as an array.

For example, here are a few lines of JSON objects:

1
2
3
4

$ for i in $(seq 3); do echo '{"id":'$i'}'; done
{"id":1}
{"id":2}
{"id":3}

To combine them into a single array, we can use jq‘s -s, —slurp option:

1
2

$ for i in $(seq 3); do echo '{"id":'$i'}'; done | jq -cM -s
[{"id":1},{"id":2},{"id":3}]

The option reads the entire input stream into a large array. This is good until the input stream is too large to process, because jq is not producing output while slurping up the input. If the input is line by line JSON, and all you want to do is wrapping it into an array (with or even without the trailing newline), we can simply do the following:

1
2
3

$ for i in $(seq 3); do echo '{"id":'$i'}'; done |\
  awk 'BEGIN { printf "["; getline; printf "%s", $0 } { printf ",%s", $0 } END { print "]" }'
[{"id":1},{"id":2},{"id":3}]

This is with the newline printed. And it will solve the problem of a very large input. Data will simply stream out as it comes it. Try it with a forever while loop.

There is another option in jq named —stream, and it seems to be doing the same. But option —slurp overrides —stream, and the option itself is already so complicated.

In conclusion, you can use jq -s to wrap line by line JSON into an array. If you want to be safe, just use the awk example above.

I have the following directory structure:

1
2
3
4
5
6
7
8
9
10
11
12
13
14

$ tree -aF src/
src/
├── github.com/
│   ├── git/
│   │   └── git/
│   │       └── file
│   └── realguess/
│       ├── git/
│       │   └── file
│       └── hello/
│           └── hello.go
└── .gitignore
6 directories, 4 files

My intention is to keep the projects in src/github.com/realguess/ only. So, I have added the following gitignore file:

1
2
3

$ cat src/.gitignore
/github.com/
!/github.com/realguess/

But the directory is ignored:

1
2

$ git check-ignore -v github.com/realguess/
src/.gitignore:1:/github.com/      github.com/realguess/

It is not possible to re-include a file if a parent directory of that file is excluded. Git doesn’t list excluded directories for performance reasons, so any patterns on contained files have no effect, no matter where they are defined.[^1]

Uh-oh! The directory github.com/ is excluded. Therefore, it is not possible to re-include a file github.com/realguess (directory is a file). To fix it, just need to add the wildcard to the end of the parent directory:

1
2
3

$ cat src/.gitignore
/github.com/*
!/github.com/realguess/

By adding the wildcard character, it matches every file (including hidden ones) or directory under github.com/ directory, the directory itself as the parent of github.com/realguess/ is not excluded. Then, we use the negation pattern to re-include github.com/realguess/.

Double check it:

1
2

$ git check-ignore -v github.com/realguess/
1

The debugging command returns non-zero status code, which indicates that the directory is not ignored.

Using the Node.js repository as an example:

1
2
3

$ git remote -v
origin  https://github.com/nodejs/node.git (fetch)
origin  https://github.com/nodejs/node.git (push)

If we would like to list all tags with v0.12 versions, we could do:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

$ git tag -l 'v0.12.*'
v0.12.0
v0.12.1
v0.12.10
v0.12.11
v0.12.12
v0.12.13
v0.12.14
v0.12.15
v0.12.2
v0.12.3
v0.12.4
v0.12.5
v0.12.6
v0.12.7
v0.12.8
v0.12.8-rc.1
v0.12.9

However, v0.12.2 should come after v0.12.1.

To fix it, we use the sort command with option:

1
2

-V, --version-sort
    natural sort of (version) numbers within text

Thus:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

$ git tag -l 'v0.12.*' | sort --version-sort
v0.12.0
v0.12.1
v0.12.2
v0.12.3
v0.12.4
v0.12.5
v0.12.6
v0.12.7
v0.12.8
v0.12.8-rc.1
v0.12.9
v0.12.10
v0.12.11
v0.12.12
v0.12.13
v0.12.14
v0.12.15

This is a Node.js starting script to stream HTTP request directly into response:

1
2
3

require('http').createServer((req, res) => {
  req.pipe(res); // Pipe request directly to response
}).listen(3000);

It behaves almost like an echo, you get back whatever you sent. For example, use HTTPie to make a request to the above server:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

$ echo foo | http --verbose --stream :3000 Content-Type:text/plain
POST / HTTP/1.1
Accept: application/json, */*
Accept-Encoding: gzip, deflate
Connection: keep-alive
Content-Length: 4
Host: localhost:3000
User-Agent: HTTPie/0.9.6
Content-Type: text/plain
foo
HTTP/1.1 200 OK
Connection: keep-alive
Transfer-Encoding: chunked
foo

We can also add the Content-Type response header to echo back what the entire media type is after assembling all chunks.

1
2
3
4
5
6
7

require('http').createServer((req, res) => {
  req.pipe(res); // Pipe request directly to response
  if (req.headers['content-type']) {
    res.setHeader('Content-Type', req.headers['content-type']);
  }
}).listen(3000);

The response should have the Content-Type field as below:

1
2
3
4
5
6

HTTP/1.1 200 OK
Connection: keep-alive
Content-Type: text/plain
Transfer-Encoding: chunked
foo

Notice that instead of usual Content-Length in the response header, we’ve got Transfer-Encoding: chunked. The default transfer encoding for Node.js HTTP is chunked:

Sending a ‘Content-length’ header will disable the default chunked encoding.[^1]

About transfer encoding:

Chunked transfer encoding is a data transfer mechanism in version 1.1 of the Hypertext Transfer Protocol (HTTP) in which data is sent in a series of “chunks”. It uses the Transfer-Encoding HTTP header in place of the Content-Length header, which the earlier version of the protocol would otherwise require. Because the Content-Length header is not used, the sender does not need to know the length of the content before it starts transmitting a response to the receiver. Senders can begin transmitting dynamically-generated content before knowing the total size of that content. … The size of each chunk is sent right before the chunk itself so that the receiver can tell when it has finished receiving data for that chunk. The data transfer is terminated by a final chunk of length zero.[^2]

With the above starting script, now you can attach some transform streams to manipulate the request and stream back in chunked response.

Settings:

1
2
3
4

$ node --version
v6.3.1
$ http --version
0.9.6

[^1]: HTTP, Node.js API Docs

[^2]: Chunked transfer encoding, Wikipedia

Escape characters are part of the syntax for many programming languages, data formats, and communication protocols. For a given alphabet an escape character’s purpose is to start character sequences (so named escape sequences), which have to be interpreted differently from the same characters occurring without the prefixed escape character.[^2]

JSON or JavaScript Object Notation is a data interchange format. It has an escape character as well.

In many programming languages such as C, Perl, and PHP and in Unix scripting languages, the backslash is an escape character, used to indicate that the character following it should be treated specially (if it would otherwise be treated normally), or normally (if it would otherwise be treated specially).[^3]

JavaScript also uses backslash as an escape character. JSON is based on a subset of the JavaScript Programming Language, therefore, JSON also uses backslash as the escape character:

A string is a sequence of zero or more Unicode characters, wrapped in double quotes, using backslash escapes.[^1]

A character can be:

• Any Unicode character except " or \ or control character
• \"
• \\
• \/
• \b
• \f
• \n
• \r
• \t
• \u + four-hex-digits

Only a few characters can be escaped in JSON. If the character is not one of the listed:

1
2

$ cat data.json
"\a"

it returns a SyntaxError[^4]:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

$ node -e 'console.log(require("./data.json"))'
module.js:561
    throw err;
    ^
SyntaxError: /home/chao/tmp/js/data.json: Unexpected token a in JSON at position 2
    at Object.parse (native)
    at Object.Module._extensions..json (module.js:558:27)
    at Module.load (module.js:458:32)
    at tryModuleLoad (module.js:417:12)
    at Function.Module._load (module.js:409:3)
    at Module.require (module.js:468:17)
    at require (internal/module.js:20:19)
    at [eval]:1:13
    at ContextifyScript.Script.runInThisContext (vm.js:25:33)
    at Object.exports.runInThisContext (vm.js:77:17)

One liner to create a Node.js HTTP server:

1

require('http').createServer((req, res) => res.end('OK')).listen(3000);

Also see the previous one liners for older Node.js versions.

When you have a release for the first time, how do you name your release version based on SemVer? 1.0.0, 0.1.0 or 0.0.1?

I like to start off with 0.1.0. Because version 0.0.0 means there is nothing. Then from nothing to something, that’s breaking change, hence version 0.1.0.

A changelog starter to be consistent with my coding style:

• Two header levels only with setext style.
• Use underlines for first-level headers and dashes for second-level headers.
• Each release is a section, hence use second-level header.
• Each category in a release is not a section, hence use emphasis with double asterisks.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82

Changelog
=========
Guideline:
1. Record notable changes.
2. List releases in reversed chronological order, newest on top.
3. Write all dates in `YYYY-MM-DD` format.
4. Follow [SemVer] format for each release.
5. Creating a new release by copying the **Example** section (on the bottom)
   to the top and rename it **Unreleased**.
6. Keep an **Unreleased** section on top for tracking any changes in order to
   minimize required effort.
7. Change _Unreleased_ to a release name with the date when ready.
_The guideline attempts to follow [Keep a CHANGELOG][1]._
[SemVer]: http://semver.org/
[1]: http://keepachangelog.com/
0.1.0 / 2015-01-01
------------------
**Security**
 - Inform users to upgrade for vulnerabilities.
**Removed**
 - Remove a deprecated feature in this release.
**Deprecated**
 - Deprecate a feature in the upcoming release.
**Changed**
 - Change an existing functionality.
**Added**
 - Add a new feature.
**Fixed**
 - Fix a bug.
Example / YYYY-MM-DD
--------------------
**Security**
 - Inform users to upgrade for vulnerabilities.
 - Each sentence must use the present tense and ends with a period.
**Removed**
 - Remove a deprecated feature in this release.
 - Basic Authentication is dropped in favor of Digest Authentication.
**Deprecated**
 - Deprecate a feature in the upcoming release.
 - Basic Authentication will be replaced by Digest Authentication in the next
   release.
**Changed**
 - Change an existing functionality.
 - User session data is added to the user object upon initialization.
**Added**
 - Add a new feature.
 - A new endpoint `/users` is added to handle user CRUD.
**Fixed**
 - Fix a bug.
 - Add the default case to handle unknown cases in user group assigning.