共计 8000 个字符,预计需要花费 20 分钟才能阅读完成。
《Terraform 101 从入门到实际》这本小册在南瓜慢说官方网站和 GitHub 两个中央同步更新,书中的示例代码也是放在 GitHub 上,不便大家参考查看。
Terraform 的函数
Terraform 为了让大家在表达式上能够更加灵便不便地进行计算,提供了大量的内置函数(Function)。目前并不反对自定义函数,只能应用 Terraform 自带的。应用函数的格局也很简略,间接写函数名 + 参数即可。如上面的函数为取最大值:
> max(34, 45, 232, 25)
232
这里把函数独自列成一章不是因为它很难了解,而因为它很罕用,值得把这些函数梳理一下,以便查问应用吧。
数值计算函数
绝对值 abs:
> abs(5)
5
> abs(-3.1415926)
3.1415926
> abs(0)
0
返回大于等于该数值的最小整数:
> ceil(3)
3
> ceil(3.1)
4
> ceil(2.9)
3
小于等于该数值的最大整数:
> floor(6)
6
> floor(6.9)
6
> floor(5.34)
5
对数函数:
> log(16, 2)
4
> log(9, 3)
2.0000000000000004
指数函数:
> pow(6, 2)
36
> pow(6, 1)
6
> pow(6, 0)
1
最大值、最小值:
> max(2, 98, 75, 4)
98
> min(2, 98, 75, 4)
2
字符串转换成整数,第二个参数为进制:
> parseint("16", 10)
16
> parseint("16", 16)
22
> parseint("FF", 16)
255
> parseint("1010", 2)
10
信号量函数:
> signum(6)
1
> signum(-6)
-1
> signum(0)
0
字符串函数
删去换行,在从文件中读取文本时十分有用:
> chomp("www.pkslow.com")
"www.pkslow.com"
> chomp("www.pkslow.com\n")
"www.pkslow.com"
> chomp("www.pkslow.com\n\n")
"www.pkslow.com"
> chomp("www.pkslow.com\n\n\r")
"www.pkslow.com"
> chomp("www.pkslow.com\n\n\ra")
<<EOT
www.pkslow.com
a
EOT
格式化输入:
> format("Hi, %s!", "Larry")
"Hi, Larry!"
> format("My name is %s, I'm %d","Larry", 18)"My name is Larry, I'm 18"
> format("The reuslt is %.2f", 3)
"The reuslt is 3.00"
> format("The reuslt is %.2f", 3.1415)
"The reuslt is 3.14"
> format("The reuslt is %8.2f", 3.1415)
"The reuslt is 3.14"
遍历格式化列表:
> formatlist("My name is %s, I'm %d %s.", ["Larry","Jeremy","Tailor"], [18, 28, 33],"in 2022")
tolist([
"My name is Larry, I'm 18 in 2022.","My name is Jeremy, I'm 28 in 2022.",
"My name is Tailor, I'm 33 in 2022.",
])
参数能够是 List,还能够是单个变量。
字符串连贯:
> join(".", ["www", "pkslow", "com"])
"www.pkslow.com"
> join(",", ["Larry", "Pkslow", "JJ"])
"Larry, Pkslow, JJ"
大小写字母转换:
> lower("Larry Nanhua DENG")
"larry nanhua deng"
> upper("Larry Nanhua DENG")
"LARRY NANHUA DENG"
首字母大写:
> title("larry")
"Larry"
替换:
> replace("www.larrydpk.com", "larrydpk", "pkslow")
"www.pkslow.com"
> replace("hello larry", "/la.*y/", "pkslow")
"hello pkslow"
宰割:
> split(".", "www.pklow.com")
tolist([
"www",
"pklow",
"com",
])
反转:
> strrev("pkslow")
"wolskp"
截取:
> substr("Larry Deng", 0, 5)
"Larry"
> substr("Larry Deng", -4, -1)
"Deng"
去除头尾某些特定字符,留神这里只有有对应字符就会删除:
> trim("?!what?!!!!!", "?!")
"what"
> trim("abaaaaabbLarry Dengaab", "ab")
"Larry Deng"
去除头尾特定字符串,留神与下面的区别:
> trimsuffix("?!what?!!!!!", "!!!")
"?!what?!!"
> trimprefix("?!what?!!!!!", "?!")
"what?!!!!!"
去除头尾的空格、换行等空串:
> trimspace("Larry Deng \n\r")
"Larry Deng"
正则匹配,上面的例子是匹配第一个和匹配所有:
> regex("[a-z\\.]+", "2021www.pkslow.com2022larry deng 31415926")
"www.pkslow.com"
> regexall("[a-z\\.]+", "2021www.pkslow.com2022larry deng 31415926")
tolist([
"www.pkslow.com",
"larry",
"deng",
])
更多正则匹配语法可参考:https://www.terraform.io/lang…
汇合类函数
alltrue
:判断列表是否全为真,空列表间接返回 true。只能是 bool 类型或者对应的字符串。
> alltrue([true, "true"])
true
> alltrue([true, "true", false])
false
> alltrue([])
true
> alltrue([1])
╷
│ Error: Invalid function argument
│
│ on <console-input> line 1:
│ (source code not available)
│
│ Invalid value for "list" parameter: element 0: bool required.
anytrue
:判断列表是否有真,只有有一个为真就返回 true。空列表为 false。
> anytrue([true])
true
> anytrue([true, false])
true
> anytrue([false, false])
false
> anytrue([])
false
chunklist
分片:依据分片数来对列表进行切分。
> chunklist(["www", "pkslow", "com", "Larry", "Deng"], 3)
tolist([
tolist([
"www",
"pkslow",
"com",
]),
tolist([
"Larry",
"Deng",
]),
])
coalesce
返回第一个非空元素:
> coalesce("","a","b")"a"> coalesce("", "","b")"b"
coalescelist
返回第一个非空列表:
> coalescelist([], ["pkslow"])
["pkslow",]
从字符串列表里把空的去掉:
> compact(["","www","", "pkslow", "com"])
tolist([
"www",
"pkslow",
"com",
])
concat
连贯多个列表:
> concat([1, 2, 3], [4, 5, 6])
[
1,
2,
3,
4,
5,
6,
]
contains
判断是否存在某个元素:
> contains(["www", "pkslow", "com"], "pkslow")
true
> contains(["www", "pkslow", "com"], "Larry")
false
distinct
去除反复元素:
> distinct([1, 2, 2, 1, 3, 8, 1, 10])
tolist([
1,
2,
3,
8,
10,
])
element
获取列表的某个元素:
> element(["a", "b", "c"], 1)
"b"
> element(["a", "b", "c"], 2)
"c"
> element(["a", "b", "c"], 3)
"a"
> element(["a", "b", "c"], 4)
"b"
flatten
把内嵌的列表都开展成一个列表:
> flatten([1, 2, 3, [1], [[6]]])
[
1,
2,
3,
1,
6,
]
index
获取列表中的元素的索引值:
> index(["www", "pkslow", "com"], "pkslow")
1
keys
获取 map 的所有 key 值:
> keys({name="Larry", age=18, webSite="www.pkslow.com"})
[
"age",
"name",
"webSite",
]
values
获取 map 的 value 值:
> values({name="Larry", age=18, webSite="www.pkslow.com"})
[
18,
"Larry",
"www.pkslow.com",
]
length
获取字符串、列表、Map 等的长度:
> length([])
0
> length(["pkslow"])
1
> length(["pkslow", "com"])
2
> length({pkslow = "com"})
1
> length("pkslow")
6
lookup(map, key, default)
依据 key 值在 map 中找到对应的 value 值,如果没有则返回默认值:
> lookup({name = "Larry", age = 18}, "age", 1)
18
> lookup({name = "Larry", age = 18}, "myAge", 1)
1
matchkeys(valueslist, keyslist, searchset)
对 key 值进行匹配。匹配到 key 值后,返回对应的 Value 值。
> matchkeys(["a", "b", "c", "d"], [1, 2, 3, 4], [2, 4])
tolist([
"b",
"d",
])
merge
合并 Map,key 雷同的会被最初的笼罩:
> merge({name = "Larry", webSite = "pkslow.com"}, {age = 18})
{
"age" = 18
"name" = "Larry"
"webSite" = "pkslow.com"
}
> merge({name = "Larry", webSite = "pkslow.com"}, {age = 18}, {age = 13})
{
"age" = 13
"name" = "Larry"
"webSite" = "pkslow.com"
}
one
取汇合的一个元素,如果为空则返回 null;如果只有一个元素,则返回该元素;如果多个元素,则报错:
> one([])
null
> one(["pkslow"])
"pkslow"
> one(["pkslow", "com"])
╷
│ Error: Invalid function argument
│
│ on <console-input> line 1:
│ (source code not available)
│
│ Invalid value for "list" parameter: must be a list, set, or tuple value with either zero or one elements.
╵
range
生成程序列表:
range(max)
range(start, limit)
range(start, limit, step)
> range(3)
tolist([
0,
1,
2,
])
> range(1, 6)
tolist([
1,
2,
3,
4,
5,
])
> range(1, 6, 2)
tolist([
1,
3,
5,
])
reverse
反转列表:
> reverse([1, 2, 3, 4])
[
4,
3,
2,
1,
]
setintersection
对 set 求交加:
> setintersection([1, 2, 3], [2, 3, 4], [2, 3, 6])
toset([
2,
3,
])
setproduct
列出所有组合可能:
> setproduct(["Larry", "Harry"], ["Deng", "Potter"])
tolist([
[
"Larry",
"Deng",
],
[
"Larry",
"Potter",
],
[
"Harry",
"Deng",
],
[
"Harry",
"Potter",
],
])
setsubtract
:set 的减法
> setsubtract([1, 2, 3], [3, 4])
toset([
1,
2,
])
# 求不同
> setunion(setsubtract(["a", "b", "c"], ["a", "c", "d"]), setsubtract(["a", "c", "d"], ["a", "b", "c"]))
[
"b",
"d",
]
setunion
:set 的加法
> setunion([1, 2, 3], [3, 4])
toset([
1,
2,
3,
4,
])
slice(list, startindex, endindex)
截取列表局部,包含 startindex,但不包含 endindex:
> slice(["a", "b", "c", "d", "e"], 1, 4)
[
"b",
"c",
"d",
]
sort
对列表中的字符串进行排序,要留神如果输出的是数字,会先转化为字符串再排序:
> sort(["larry", "pkslow", "com", "deng"])
tolist([
"com",
"deng",
"larry",
"pkslow",
])
> sort([3, 6, 1, 9, 12, 79, 22])
tolist([
"1",
"12",
"22",
"3",
"6",
"79",
"9",
])
sum
求和:
> sum([3, 1.2, 9, 17.3, 2.2])
32.7
transpose
对 Map 的 key 和 value 进行换位:
> transpose({"a" = ["1", "2"], "b" = ["2", "3"]})
tomap({
"1" = tolist(["a",])
"2" = tolist([
"a",
"b",
])
"3" = tolist(["b",])
})
zipmap
依据 key 和 value 的列表按一对一关系生成 Map:
> zipmap(["age", "name"], [18, "Larry Deng"])
{
"age" = 18
"name" = "Larry Deng"
}
加密解密
Base64:
> base64encode("pkslow")
"cGtzbG93"
> base64decode("cGtzbG93")
"pkslow"
> textencodebase64("pkslow", "UTF-8")
"cGtzbG93"
> textdecodebase64("cGtzbG93", "UTF-8")
"pkslow"
csv 文本解析:
> csvdecode("seq,name,age\n1,larry,18\n2,pkslow,3\n3,Jeremy,29")
tolist([
{
"age" = "18"
"name" = "larry"
"seq" = "1"
},
{
"age" = "3"
"name" = "pkslow"
"seq" = "2"
},
{
"age" = "29"
"name" = "Jeremy"
"seq" = "3"
},
])
Json 解析:
> jsonencode({"name"="Larry", "age"=18})
"{\"age\":18,\"name\":\"Larry\"}"
> jsondecode("{\"age\":18,\"name\":\"Larry\"}")
{
"age" = 18
"name" = "Larry"
}
URL:
> urlencode("Larry Deng/a/:/./@")
"Larry+Deng%2Fa%2F%3A%2F.%2F%40"
YAML:
> yamlencode({"a":"b", "c":"d"})
"a": "b"
"c": "d"
> yamlencode({"foo":[1, 2, 3], "bar": "baz"})
"bar": "baz"
"foo":
- 1
- 2
- 3
> yamlencode({"foo":[1, {"a":"b","c":"d"}, 3], "bar": "baz"})
"bar": "baz"
"foo":
- 1
- "a": "b"
"c": "d"
- 3
> yamldecode("hello: world")
{"hello" = "world"}
> yamldecode("true")
true
> yamldecode("{a: &foo [1, 2, 3], b: *foo}")
{
"a" = [
1,
2,
3,
]
"b" = [
1,
2,
3,
]
}
文件解决:
获取绝对路径:
> abspath(path.root)
"/Users/larry"
获取门路中的目录,或者是文件名:
> dirname("/home/larry/soft/terraform")
"/home/larry/soft"
> dirname("/home/larry/soft/terraform/")
"/home/larry/soft/terraform"
> basename("/home/larry/soft/terraform")
"terraform"
> basename("/home/larry/soft/terraform/")
"terraform"
判断文件是否存在,并获取文件内容:
> fileexists("/Users/larry/.bash_profile")
true
> file("/Users/larry/.bash_profile")
> filebase64("/Users/larry/.bash_profile")
依据模式匹配所有文件:
> fileset("/Users/larry", "*.bash*")
toset([
".bash_history",
".bash_profile",
".bash_profile.backup",
])
templatefile(path, vars)
模板化文件:指定文件和变量,把变量值替换掉模板中的变量。
工夫函数
获取以后工夫,并格式化显示,格局请参考:https://www.terraform.io/lang…
> formatdate("YYYY-MM-DD hh:mm:ss / D MMMM YYYY", timestamp())
"2022-03-05 08:25:48 / 5 March 2022"
> formatdate("EEEE, DD-MMM-YY hh:mm:ss ZZZ", "2018-01-02T23:12:01Z")
"Tuesday, 02-Jan-18 23:12:01 UTC"
工夫加减:
> timeadd(timestamp(), "24h")
"2022-03-06T08:28:52Z"
> timeadd(timestamp(), "-24h10m")
"2022-03-04T08:19:08Z"
反对的单位有:"ns"
, "us"
(or "µs"
), "ms"
, "s"
, "m"
, and "h"
.
其它
加密:
> md5("www.pkslow.com")
"97e164b60faf4d7875c2a8a5bc3f2245"
UUID:
> uuid()
"049bf418-15d1-e034-28db-92945067dcf6"
IP:
> cidrsubnet("172.16.0.0/12", 4, 2)
"172.18.0.0/16"
更多请参考官网。