The best and simplest way to get input from a user in the CLI with only PHP is to use fgetc() function with the STDIN constant:
<?php
echo 'Are you sure you want to quit? (y/n) ';
$input = fgetc(STDIN);
if ($input == 'y')
{
exit(0);
}
?>
PHP - Manual: fgetc
2024-12-22
(PHP 4, PHP 5, PHP 7, PHP 8)
fgetc — 从文件指针中读取字符
$handle
): string从文件句柄中获取一个字符。
返回一个包含有一个字符的字符串,该字符从 handle
指向的文件中得到。
碰到 EOF 则返回 false
。
示例 #1 一个 fgetc() 例子
<?php
$fp = fopen('somefile.txt', 'r');
if (!$fp) {
echo 'Could not open file somefile.txt';
}
while (false !== ($char = fgetc($fp))) {
echo "$char\n";
}
?>
注意: 此函数可安全用于二进制对象。
The best and simplest way to get input from a user in the CLI with only PHP is to use fgetc() function with the STDIN constant:
<?php
echo 'Are you sure you want to quit? (y/n) ';
$input = fgetc(STDIN);
if ($input == 'y')
{
exit(0);
}
?>
I was using command-line PHP to create an interactive script and wanted the user to enter just one character of input - in response a Yes/No question. Had some trouble finding a way to do so using fgets(), fgetc(), various suggestions using readline(), popen(), etc. Came up with the following that works quite nicely:
$ans = strtolower( trim( `bash -c "read -n 1 -t 10 ANS ; echo \\\$ANS"` ) );
To read a single key-press in CLI mode, you can either use ncurses (which will probably require additional modules for PHP) or get nasty with the *nix "/bin/stty" command)
<?php
function stty($options) {
exec($cmd = "/bin/stty $options", $output, $el);
$el AND die("exec($cmd) failed");
return implode(" ", $output);
}
function getchar($echo = false) {
$echo = $echo ? "" : "-echo";
# Get original settings
$stty_settings = preg_replace("#.*; ?#s", "", stty("--all"));
# Set new ones
stty("cbreak $echo");
# Get characters until a PERIOD is typed,
# showing their hexidecimal ordinal values.
printf("> ");
do {
printf("%02x ", ord($c = fgetc(STDIN)));
} while ($c != '.');
# Return settings
stty($stty_settings);
}
getchar();
?>
If you call fgetc(HANDLE) multiple times you must clear buffer before, thus get all unwanted chars to the new new line character.
while (true) {
//clear buffer - read all unwanted characters
while(fgetc(STDIN) != "\n");
//get first character from STDIN
$first = fgetc(STDIN);
}
You can't just simple print separated characters of a text which is encoded in multibyte character set like this;
Because fgetc() will break each multibyte character on its every byte. Consider this example:
<?php
$path = 'foo/cyrillic.txt';
$handle = fopen($path, 'rb');
while (FALSE !== ($ch = fgetc($handle))) {
$curs = ftell($hanlde);
print "[$curs:] $ch\n";
}
/* The result will be something like this:
<
[1]: <
[2]: h
[3]: 2
[4]: >
[5]: �
[6]: �
[7]: �
[8]: �
[9]: �
[10]: �
[11]:
[12]: �
[13]: �
[14]: �
[15]: �
[16]: �
*/ ?>
I don't think this is the best, but it can be a workaround:
<?php
$path = 'path/to/your/file.ext';
if (!$handle = fopen($path, 'rb')) {
echo "Can't open ($path) file';
exit;
}
$mbch = ''; // keeps the first byte of 2-byte cyrillic letters
while (FALSE !== ($ch = fgetc($handle))) {
//check for the sign of 2-byte cyrillic letters
if (empty($mbch) && (FALSE !== array_search(ord($ch), Array(208,209,129)))) {
$mbch = $ch; // keep the first byte
continue;
}
$curs = ftell($handle);
print "[$curs]: " . $mbch . $ch . PHP_EOL;
// or print "[$curs]: $mbch$ch\n";
if (!empty($mbch)) $mbch = ''; // erase the byte after using
}
?>