I needed to write a script to enter multi-line input to a program (psql).

After a bit of googling, I found the following syntax works:

cat <开发者_如何学JAVA< EOF | psql ---params
BEGIN;

`pg_dump ----something`

update table .... statement ...;

END;
EOF

This correctly constructs the multi-line string (from BEGIN; to END;, inclusive) and pipes it as an input to psql.

But I have no idea how/why it works, can some one please explain?

I'm referring mainly to cat << EOF, I know > outputs to a file, >> appends to a file, < reads input from file.

What does << exactly do?

And is there a man page for it?

---

The cat <<EOF syntax is very useful when working with multi-line text in Bash, eg. when assigning multi-line string to a shell variable, file or a pipe.

Examples of cat <<EOF syntax usage in Bash:

1. Assign multi-line string to a shell variable

$ sql=$(cat <<EOF
SELECT foo, bar FROM db
WHERE foo='baz'
EOF
)

The $sql variable now holds the new-line characters too. You can verify with echo -e "$sql".

2. Pass multi-line string to a file in Bash

$ cat <<EOF > print.sh
#!/bin/bash
echo \$PWD
echo $PWD
EOF

The print.sh file now contains:

#!/bin/bash
echo $PWD
echo /home/user

3. Pass multi-line string to a pipe in Bash

$ cat <<EOF | grep 'b' | tee b.txt
foo
bar
baz
EOF

The b.txt file contains bar and baz lines. The same output is printed to stdout.

---

This is called heredoc format to provide a string into stdin. See https://en.wikipedia.org/wiki/Here_document#Unix_shells for more details.

---

From man bash:

Here Documents

This type of redirection instructs the shell to read input from the current source until a line containing only word (with no trailing blanks) is seen.

All of the lines read up to that point are then used as the standard input for a command.

The format of here-documents is:

          <<[-]word
                  here-document
          delimiter

No parameter expansion, command substitution, arithmetic expansion, or pathname expansion is performed on word. If any characters in word are quoted, the delimiter is the result of quote removal on word, and the lines in the here-document are not expanded. If word is unquoted, all lines of the here-document are subjected to parameter expansion, command substitution, and arithmetic expansion. In the latter case, the character sequence \<newline> is ignored, and \ must be used to quote the characters \, $, and `.

If the redirection operator is <<-, then all leading tab characters are stripped from input lines and the line containing delimiter. This allows here-documents within shell scripts to be indented in a natural fashion.

---

In your case, "EOF" is known as a "Here Tag". Basically <<Here tells the shell that you are going to enter a multiline string until the "tag" Here. You can name this tag as you want, it's often EOF or STOP.

Some rules about the Here tags:

1. The tag can be any string, uppercase or lowercase, though most people use uppercase by convention.
2. The tag will not be considered as a Here tag if there are other words in that line. In this case, it will merely be considered part of the string. The tag should be by itself on a separate line, to be considered a tag.
3. The tag should have no leading or trailing spaces in that line to be considered a tag. Otherwise it will be considered as part of the string.

example:

$ cat >> test <<HERE
> Hello world HERE <-- Not by itself on a separate line -> not considered end of string
> This is a test
>  HERE <-- Leading space, so not considered end of string
> and a new line
> HERE <-- Now we have the end of the string

---

POSIX 7

kennytm quoted man bash, but most of that is also POSIX 7: http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_07_04 :

The redirection operators "<<" and "<<-" both allow redirection of lines contained in a shell input file, known as a "here-document", to the input of a command.

The here-document shall be treated as a single word that begins after the next <newline> and continues until there is a line containing only the delimiter and a <newline>, with no <blank> characters in between. Then the next here-document starts, if there is one. The format is as follows:

[n]<<word
    here-document
delimiter

where the optional n represents the file descriptor number. If the number is omitted, the here-document refers to standard input (file descriptor 0).

If any character in word is quoted, the delimiter shall be formed by performing quote removal on word, and the here-document lines shall not be expanded. Otherwise, the delimiter shall be the word itself.

If no characters in word are quoted, all lines of the here-document shall be expanded for parameter expansion, command substitution, and arithmetic expansion. In this case, the <backslash> in the input behaves as the <backslash> inside double-quotes (see Double-Quotes). However, the double-quote character ( '"' ) shall not be treated specially within a here-document, except when the double-quote appears within "$()", "``", or "${}".

If the redirection symbol is "<<-", all leading <tab> characters shall be stripped from input lines and the line containing the trailing delimiter. If more than one "<<" or "<<-" operator is specified on a line, the here-document associated with the first operator shall be supplied first by the application and shall be read first by the shell.

When a here-document is read from a terminal device and the shell is interactive, it shall write the contents of the variable PS2, processed as described in Shell Variables, to standard error before reading each line of input until the delimiter has been recognized.

Examples

Some examples not yet given.

Quotes prevent parameter expansion

Without quotes:

a=0
cat <<EOF
$a
EOF

Output:

0

With quotes:

a=0
cat <<'EOF'
$a
EOF

or (ugly but valid):

a=0
cat <<E"O"F
$a
EOF

Outputs:

$a

Hyphen removes leading tabs

Without hyphen:

cat <<EOF
<tab>a
EOF

where <tab> is a literal tab, and can be inserted with Ctrl + V <tab>

Output:

<tab>a

With hyphen:

cat <<-EOF
<tab>a
<tab>EOF

Output:

a

This exists of course so that you can indent your cat like the surrounding code, which is easier to read and maintain. E.g.:

if true; then
    cat <<-EOF
    a
    EOF
fi

Unfortunately, this does not work for space characters: POSIX favored tab indentation here. Yikes.

---

Using tee instead of cat

Not exactly as an answer to the original question, but I wanted to share this anyway: I had the need to create a config file in a directory that required root rights.

The following does not work for that case:

$ sudo cat <<EOF >/etc/somedir/foo.conf
# my config file
foo=bar
EOF

because the redirection is handled outside of the sudo context.

I ended up using this instead:

$ sudo tee <<EOF /etc/somedir/foo.conf >/dev/null
# my config file
foo=bar
EOF

---

<< EoF basically means:

<< - "read the multi-line input that begins from the next line onward, and treat it as if it's code in a separate file"

EoF - "stop reading immediately after the word EoF is found in the multi-line input"

As other answers have explained, the multi-line input is called a Here Document

A Here Document is often used to generate output to be passed to a subsequent process. For example cat << EoF can be used to generate a desired output, using a Here Document.

Here's an example of using a Here Document to create a text document on the fly:

cat << EoF > ./my-document.txt
Hello world
Have a nice day
EoF

---

A little extension to the above answers. The trailing > directs the input into the file, overwriting existing content. However, one particularly convenient use is the double arrow >> that appends, adding your new content to the end of the file, as in:

cat <<EOF >> /etc/fstab
data_server:/var/sharedServer/authority/cert /var/sharedFolder/sometin/authority/cert nfs
data_server:/var/sharedServer/cert   /var/sharedFolder/sometin/vsdc/cert nfs
EOF

This extends your fstab without you having to worry about accidentally modifying any of its contents.

---

note to mention that cat << \EOT (see the backslash) will not expand any variables inside, while cat << EOT will do.

examples:

FOO="bar"

cat << \EOT > foobar.txt
echo "$FOO"
EOT

will output: echo $FOO

while:

FOO="bar"

cat << EOT > foobar.txt
echo "$FOO"
EOT

will output: echo "bar"

---

Long story short, EOF marker(but a different literal can be used as well) is a heredoc format that allows you to provide your input as multiline. A lot of confusion comes from how cat actually works it seems. You can use cat with >> or > as follows:

$ cat >> temp.txt
line 1
line 2

While cat can be used this way when writing manually into console, it's not convenient if I want to provide the input in a more declarative way so that it can be reused by tools and also to keep indentations, whitespaces, etc.
Heredoc allows to define your entire input as if you are not working with stdin but typing in a separate text editor. This is what Wikipedia article means by:

it is a section of a source code file that is treated as if it were a separate file.

---

Example to create a json file:

cat << EoF > ./allaccess.json
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:*"
      ],
      "Resource": [
        "arn:aws:s3:::*"
      ]
    }
  ]
}
EoF

As a result:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": [
        "s3:*"
      ],
      "Resource": [
        "arn:aws:s3:::*"
      ]
    }
  ]
}

---

This isn't necessarily an answer to the original question, but a sharing of some results from my own testing. This:

<<test > print.sh
#!/bin/bash
echo \$PWD
echo $PWD
test

will produce the same file as:

cat <<test > print.sh
#!/bin/bash
echo \$PWD
echo $PWD
test

So, I don't see the point of using the cat command.

---

Worth noting that here docs work in bash loops too. This example shows how-to get the column list of table:

export postgres_db_name='my_db'
export table_name='my_table_name'

# start copy 
while read -r c; do test -z "$c" || echo $table_name.$c , ; done < <(cat << EOF | psql -t -q -d $postgres_db_name -v table_name="${table_name:-}"
SELECT column_name
FROM information_schema.columns
WHERE 1=1
AND table_schema = 'public'
AND table_name   =:'table_name'  ;
EOF
)
# stop copy , now paste straight into the bash shell ...

output: 
my_table_name.guid ,
my_table_name.id ,
my_table_name.level ,
my_table_name.seq ,

or even without the new line

while read -r c; do test -z "$c" || echo $table_name.$c , | perl -ne 
's/\n//gm;print' ; done < <(cat << EOF | psql -t -q -d $postgres_db_name -v table_name="${table_name:-}"
 SELECT column_name
 FROM information_schema.columns
 WHERE 1=1
 AND table_schema = 'public'
 AND table_name   =:'table_name'  ;
 EOF
 )

 # output: daily_issues.guid ,daily_issues.id ,daily_issues.level ,daily_issues.seq ,daily_issues.prio ,daily_issues.weight ,daily_issues.status ,daily_issues.category ,daily_issues.name ,daily_issues.description ,daily_issues.type ,daily_issues.owner