I am trying to write an awk command to take certain fields out of one file and write them to a new file. I am able to to separate the fields (comma delineate) and print it to stdout but I can't figure out how to get it into the new file in the correct format. Here is my current command:
我试图写一个awk命令从一个文件中取出某些字段并将它们写入一个新文件。我能够分隔字段（逗号描述）并将其打印到stdout但我无法弄清楚如何以正确的格式将其放入新文件中。这是我当前的命令：
$ awk '{FS =","};{print $3}' test.log > test1.log
This puts the third field of each line in the new file but inserts 2 blank lines at the beginning of the new file and then a blank line in between each field. I would also like the keep the comma separators but can't figure that out ether. I'm fairly new to bash and awk so any help is appreciated!
这会将每行的第三个字段放在新文件中，但在新文件的开头插入2个空行，然后在每个字段之间插入一个空行。我也想保留逗号分隔符，但无法想象以太。我是bash和awk的新手，所以任何帮助都表示赞赏！
Here a chunk of my sample data:
这是我的样本数据的一大块：
10.60.3.109, anonymous, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3007]USER, anonymous, -,
10.60.3.109, -, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 530, 1326, [3007]PASS, IEUser@, -,
10.60.3.109, anonymous, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3008]USER, anonymous, -,
10.60.3.109, -, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 530, 1326, [3008]PASS, IEUser@, -,
10.60.3.109, anonymous, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3009]USER, anonymous, -,
10.60.3.109, -, 12/1/2003, 6:56:20, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 530, 1326, [3009]PASS, IEUser@, -,
10.60.3.109, sally, 12/1/2003, 6:56:26, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3010]USER, sally, -,
10.60.3.109, -, 12/1/2003, 6:56:26, MSFTPSVC1, CSS, ipaddr, 16, 0, 0, 530, 1326, [3010]PASS, -, -,
10.60.3.109, sally, 12/1/2003, 6:56:31, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3011]USER, sally, -,
10.60.3.109, -, 12/1/2003, 6:56:31, MSFTPSVC1, CSS, ipaddr, 31, 0, 0, 530, 1326, [3011]PASS, -, -,
10.60.3.109, sally, 12/1/2003, 6:56:36, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3012]USER, sally, -,
10.60.3.109, -, 12/1/2003, 6:56:36, MSFTPSVC1, CSS, ipaddr, 15, 0, 0, 530, 1326, [3012]PASS, -, -,
10.60.3.109, sally, 12/1/2003, 6:56:40, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3013]USER, sally, -,
10.60.3.109, sally, 12/1/2003, 6:56:40, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 230, 0, [3013]PASS, -, -,
10.60.3.109, sally, 12/1/2003, 6:56:40, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3014]USER, sally, -,
10.60.3.109, sally, 12/1/2003, 6:56:40, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 230, 0, [3014]PASS, -, -,
10.60.3.109, sally, 12/1/2003, 6:56:47, MSFTPSVC1, CSS, ipaddr, 0, 0, 0, 331, 0, [3015]USER, sally, -,
1 个解决方案
#1
2
awk '{FS =","};{print $3}' test.log > test1.log
The first action is {FS = ","} and it is being executed for each line. It should probably be preceded by BEGIN. Because the field separator is currently not set until after the first line is already read and split, the first line is most likely read as a single field, and hence $3 is empty, leading to the blank line.
第一个动作是{FS =“，”}，它正在为每一行执行。它可能应该在BEGIN之前。由于字段分隔符当前未在第一行已被读取和拆分之后设置，因此第一行很可能被读取为单个字段，因此$ 3为空，从而导致空白行。
The second action is ;, which prints the entire line doesn't seem to do anything. You don't normally use a semicolon like that; it normally only appears inside the braces of an action.
第二个动作是;，打印整行不似乎做任何事情。你通常不会使用这样的分号;它通常只出现在动作的大括号内。
The third action is {print $3}, which prints the 3rd field (when there is a third field to print).
第三个动作是{print $ 3}，它打印第3个字段（当有第三个字段要打印时）。
I'm not yet sure I understand the double blank line at the beginning of your output, unless there's a blank line at the beginning of your input. Nor am I reproducing the alternating blank lines, this using the BSD awk from Mac OS X 10.10.1 Yosemite, but also using GNU awk 3.1.7.
我还不确定我理解输出开头的双空行，除非输入开头有空行。我也没有使用Mac OS X 10.10.1优胜美地的BSD awk再现交替的空白行，但也使用GNU awk 3.1.7。
You probably want:
你可能想要：
awk 'BEGIN {FS = ","} {print $3}' test.log > test1.log
or (preferable in my view):
或（我认为最好）：
awk -F"," '{print $3}' test.log > test1.log
where the quotes around the comma are optional (since comma is not a shell metacharacter). But if you're later going to be selecting multiple fields, you probably want to do as jaypal singh suggested in a comment, and use:
逗号周围的引号是可选的（因为逗号不是shell元字符）。但是如果你以后要选择多个字段，你可能想像jaypal singh在评论中建议的那样做，并使用：
awk 'BEGIN {OFS = FS = ","} {print $3, $5, $9}' test.log > test1.log
for whatever permutation of the fields you want printed.
无论您想要打印哪些字段的排列。
#1
2
awk '{FS =","};{print $3}' test.log > test1.log
The first action is {FS = ","} and it is being executed for each line. It should probably be preceded by BEGIN. Because the field separator is currently not set until after the first line is already read and split, the first line is most likely read as a single field, and hence $3 is empty, leading to the blank line.
第一个动作是{FS =“，”}，它正在为每一行执行。它可能应该在BEGIN之前。由于字段分隔符当前未在第一行已被读取和拆分之后设置，因此第一行很可能被读取为单个字段，因此$ 3为空，从而导致空白行。
The second action is ;, which prints the entire line doesn't seem to do anything. You don't normally use a semicolon like that; it normally only appears inside the braces of an action.
第二个动作是;，打印整行不似乎做任何事情。你通常不会使用这样的分号;它通常只出现在动作的大括号内。
The third action is {print $3}, which prints the 3rd field (when there is a third field to print).
第三个动作是{print $ 3}，它打印第3个字段（当有第三个字段要打印时）。
I'm not yet sure I understand the double blank line at the beginning of your output, unless there's a blank line at the beginning of your input. Nor am I reproducing the alternating blank lines, this using the BSD awk from Mac OS X 10.10.1 Yosemite, but also using GNU awk 3.1.7.
我还不确定我理解输出开头的双空行，除非输入开头有空行。我也没有使用Mac OS X 10.10.1优胜美地的BSD awk再现交替的空白行，但也使用GNU awk 3.1.7。
You probably want:
你可能想要：
awk 'BEGIN {FS = ","} {print $3}' test.log > test1.log
or (preferable in my view):
或（我认为最好）：
awk -F"," '{print $3}' test.log > test1.log
where the quotes around the comma are optional (since comma is not a shell metacharacter). But if you're later going to be selecting multiple fields, you probably want to do as jaypal singh suggested in a comment, and use:
逗号周围的引号是可选的（因为逗号不是shell元字符）。但是如果你以后要选择多个字段，你可能想像jaypal singh在评论中建议的那样做，并使用：
awk 'BEGIN {OFS = FS = ","} {print $3, $5, $9}' test.log > test1.log
for whatever permutation of the fields you want printed.
无论您想要打印哪些字段的排列。