PoHeader.php
Same filename in other branches
Namespace
Drupal\Component\GettextFile
-
core/
lib/ Drupal/ Component/ Gettext/ PoHeader.php
View source
<?php
namespace Drupal\Component\Gettext;
/**
* Gettext PO header handler.
*
* Possible Gettext PO header elements are explained in
* http://www.gnu.org/software/gettext/manual/gettext.html#Header-Entry,
* but we only support a subset of these directly.
*
* Example header:
*
* "Project-Id-Version: Drupal core (7.11)\n"
* "POT-Creation-Date: 2012-02-12 22:59+0000\n"
* "PO-Revision-Date: YYYY-mm-DD HH:MM+ZZZZ\n"
* "Language-Team: Catalan\n"
* "MIME-Version: 1.0\n"
* "Content-Type: text/plain; charset=utf-8\n"
* "Content-Transfer-Encoding: 8bit\n"
* "Plural-Forms: nplurals=2; plural=(n>1);\n"
*/
class PoHeader {
/**
* Language code.
*
* @var string
*/
protected $langcode;
/**
* Formula for the plural form.
*
* @var string
*/
protected $pluralForms;
/**
* Author(s) of the file.
*
* @var string
*/
protected $authors;
/**
* Date the po file got created.
*
* @var string
*/
protected $poDate;
/**
* Human readable language name.
*
* @var string
*/
protected $languageName;
/**
* Name of the project the translation belongs to.
*
* @var string
*/
protected $projectName;
/**
* Constructor, creates a PoHeader with default values.
*
* @param string $langcode
* Language code.
*/
public function __construct($langcode = NULL) {
$this->langcode = $langcode;
// Ignore errors when run during site installation before
// date_default_timezone_set() is called.
$this->poDate = @date("Y-m-d H:iO");
$this->pluralForms = 'nplurals=2; plural=(n > 1);';
}
/**
* Gets the plural form.
*
* @return string
* Plural form component from the header, for example:
* 'nplurals=2; plural=(n > 1);'.
*/
public function getPluralForms() {
return $this->pluralForms;
}
/**
* Set the human readable language name.
*
* @param string $languageName
* Human readable language name.
*/
public function setLanguageName($languageName) {
$this->languageName = $languageName;
}
/**
* Gets the human readable language name.
*
* @return string
* The human readable language name.
*/
public function getLanguageName() {
return $this->languageName;
}
/**
* Set the project name.
*
* @param string $projectName
* Human readable project name.
*/
public function setProjectName($projectName) {
$this->projectName = $projectName;
}
/**
* Gets the project name.
*
* @return string
* The human readable project name.
*/
public function getProjectName() {
return $this->projectName;
}
/**
* Populate internal values from a string.
*
* @param string $header
* Full header string with key-value pairs.
*/
public function setFromString($header) {
// Get an array of all header values for processing.
$values = $this->parseHeader($header);
// There is only one value relevant for our header implementation when
// reading, and that is the plural formula.
if (!empty($values['Plural-Forms'])) {
$this->pluralForms = $values['Plural-Forms'];
}
}
/**
* Generate a Gettext PO formatted header string based on data set earlier.
*/
public function __toString() {
$output = '';
$isTemplate = empty($this->languageName);
$output .= '# ' . ($isTemplate ? 'LANGUAGE' : $this->languageName) . ' translation of ' . ($isTemplate ? 'PROJECT' : $this->projectName) . "\n";
if (!empty($this->authors)) {
$output .= '# Generated by ' . implode("\n# ", $this->authors) . "\n";
}
$output .= "#\n";
// Add the actual header information.
$output .= "msgid \"\"\n";
$output .= "msgstr \"\"\n";
$output .= "\"Project-Id-Version: PROJECT VERSION\\n\"\n";
$output .= "\"POT-Creation-Date: " . $this->poDate . "\\n\"\n";
$output .= "\"PO-Revision-Date: " . $this->poDate . "\\n\"\n";
$output .= "\"Last-Translator: NAME <EMAIL@ADDRESS>\\n\"\n";
$output .= "\"Language-Team: LANGUAGE <EMAIL@ADDRESS>\\n\"\n";
$output .= "\"MIME-Version: 1.0\\n\"\n";
$output .= "\"Content-Type: text/plain; charset=utf-8\\n\"\n";
$output .= "\"Content-Transfer-Encoding: 8bit\\n\"\n";
$output .= "\"Plural-Forms: " . $this->pluralForms . "\\n\"\n";
$output .= "\n";
return $output;
}
/**
* Parses a Plural-Forms entry from a Gettext Portable Object file header.
*
* @param string $pluralforms
* The Plural-Forms entry value.
*
* @return array|bool
* An indexed array of parsed plural formula data. Containing:
* - 'nplurals': The number of plural forms defined by the plural formula.
* - 'plurals': Array of plural positions keyed by plural value.
* False when there is no plural string.
*
* @throws \Exception
*/
public function parsePluralForms($pluralforms) {
$plurals = [];
// First, delete all whitespace.
$pluralforms = strtr($pluralforms, [
" " => "",
"\t" => "",
]);
// Select the parts that define nplurals and plural.
$nplurals = strstr($pluralforms, "nplurals=");
if (strpos($nplurals, ";")) {
// We want the string from the 10th char, because "nplurals=" length is 9.
$nplurals = substr($nplurals, 9, strpos($nplurals, ";") - 9);
}
else {
return FALSE;
}
$plural = strstr($pluralforms, "plural=");
if (strpos($plural, ";")) {
// We want the string from the 8th char, because "plural=" length is 7.
$plural = substr($plural, 7, strpos($plural, ";") - 7);
}
else {
return FALSE;
}
// If the number of plurals is zero, we return a default result.
if ($nplurals == 0) {
return [
$nplurals,
[
'default' => 0,
],
];
}
// Calculate possible plural positions of different plural values. All known
// plural formula's are repetitive above 100.
// For data compression we store the last position the array value
// changes and store it as default.
$element_stack = $this->parseArithmetic($plural);
if ($element_stack !== FALSE) {
for ($i = 0; $i <= 199; $i++) {
$plurals[$i] = $this->evaluatePlural($element_stack, $i);
}
$default = $plurals[$i - 1];
$plurals = array_filter($plurals, function ($value) use ($default) {
return $value != $default;
});
$plurals['default'] = $default;
return [
$nplurals,
$plurals,
];
}
else {
throw new \Exception('The plural formula could not be parsed.');
}
}
/**
* Parses a Gettext Portable Object file header.
*
* @param string $header
* A string containing the complete header.
*
* @return array
* An associative array of key-value pairs.
*/
private function parseHeader($header) {
$header_parsed = [];
$lines = array_map('trim', explode("\n", $header));
foreach ($lines as $line) {
if ($line) {
[
$tag,
$contents,
] = explode(":", $line, 2);
$header_parsed[trim($tag)] = trim($contents);
}
}
return $header_parsed;
}
/**
* Parses and sanitizes an arithmetic formula into a plural element stack.
*
* While parsing, we ensure, that the operators have the right
* precedence and associativity.
*
* @param string $string
* A string containing the arithmetic formula.
*
* @return array|bool
* A stack of values and operations to be evaluated. False if the formula
* could not be parsed.
*/
private function parseArithmetic($string) {
// Operator precedence table.
$precedence = [
"(" => -1,
")" => -1,
"?" => 1,
":" => 1,
"||" => 3,
"&&" => 4,
"==" => 5,
"!=" => 5,
"<" => 6,
">" => 6,
"<=" => 6,
">=" => 6,
"+" => 7,
"-" => 7,
"*" => 8,
"/" => 8,
"%" => 8,
];
// Right associativity.
$right_associativity = [
"?" => 1,
":" => 1,
];
$tokens = $this->tokenizeFormula($string);
// Parse by converting into infix notation then back into postfix
// Operator stack - holds math operators and symbols.
$operator_stack = [];
// Element Stack - holds data to be operated on.
$element_stack = [];
foreach ($tokens as $token) {
$current_token = $token;
// Numbers and the $n variable are simply pushed into $element_stack.
if (is_numeric($token)) {
$element_stack[] = $current_token;
}
elseif ($current_token == "n") {
$element_stack[] = '$n';
}
elseif ($current_token == "(") {
$operator_stack[] = $current_token;
}
elseif ($current_token == ")") {
$topop = array_pop($operator_stack);
while (isset($topop) && $topop != "(") {
$element_stack[] = $topop;
$topop = array_pop($operator_stack);
}
}
elseif (!empty($precedence[$current_token])) {
// If it's an operator, then pop from $operator_stack into
// $element_stack until the precedence in $operator_stack is less
// than current, then push into $operator_stack.
$topop = array_pop($operator_stack);
while (isset($topop) && $precedence[$topop] >= $precedence[$current_token] && !($precedence[$topop] == $precedence[$current_token] && !empty($right_associativity[$topop]) && !empty($right_associativity[$current_token]))) {
$element_stack[] = $topop;
$topop = array_pop($operator_stack);
}
if ($topop) {
// Return element to top.
$operator_stack[] = $topop;
}
// Parentheses are not needed.
$operator_stack[] = $current_token;
}
else {
return FALSE;
}
}
// Flush operator stack.
$topop = array_pop($operator_stack);
while ($topop != NULL) {
$element_stack[] = $topop;
$topop = array_pop($operator_stack);
}
$return = $element_stack;
// Now validate stack.
$previous_size = count($element_stack) + 1;
while (count($element_stack) < $previous_size) {
$previous_size = count($element_stack);
for ($i = 2; $i < count($element_stack); $i++) {
$op = $element_stack[$i];
if (!empty($precedence[$op])) {
if ($op == ":") {
$f = $element_stack[$i - 2] . "):" . $element_stack[$i - 1] . ")";
}
elseif ($op == "?") {
$f = "(" . $element_stack[$i - 2] . "?(" . $element_stack[$i - 1];
}
else {
$f = "(" . $element_stack[$i - 2] . $op . $element_stack[$i - 1] . ")";
}
array_splice($element_stack, $i - 2, 3, $f);
break;
}
}
}
// If only one element is left, the number of operators is appropriate.
return count($element_stack) == 1 ? $return : FALSE;
}
/**
* Tokenize the formula.
*
* @param string $formula
* A string containing the arithmetic formula.
*
* @return array
* List of arithmetic tokens identified in the formula.
*/
private function tokenizeFormula($formula) {
$formula = str_replace(" ", "", $formula);
$tokens = [];
for ($i = 0; $i < strlen($formula); $i++) {
if (is_numeric($formula[$i])) {
$num = $formula[$i];
$j = $i + 1;
while ($j < strlen($formula) && is_numeric($formula[$j])) {
$num .= $formula[$j];
$j++;
}
$i = $j - 1;
$tokens[] = $num;
}
elseif ($pos = strpos(" =<>!&|", $formula[$i])) {
$next = $formula[$i + 1];
switch ($pos) {
case 1:
case 2:
case 3:
case 4:
if ($next == '=') {
$tokens[] = $formula[$i] . '=';
$i++;
}
else {
$tokens[] = $formula[$i];
}
break;
case 5:
if ($next == '&') {
$tokens[] = '&&';
$i++;
}
else {
$tokens[] = $formula[$i];
}
break;
case 6:
if ($next == '|') {
$tokens[] = '||';
$i++;
}
else {
$tokens[] = $formula[$i];
}
break;
}
}
else {
$tokens[] = $formula[$i];
}
}
return $tokens;
}
/**
* Evaluate the plural element stack using a plural value.
*
* Using an element stack, which represents a plural formula, we calculate
* which plural string should be used for a given plural value.
*
* An example of plural formula parting and evaluation:
* Plural formula: 'n!=1'
* This formula is parsed by parseArithmetic() to a stack (array) of elements:
* array(
* 0 => '$n',
* 1 => '1',
* 2 => '!=',
* );
* The evaluatePlural() method evaluates the $element_stack using the plural
* value $n. Before the actual evaluation, the '$n' in the array is replaced
* by the value of $n.
* For example: $n = 2 results in:
* array(
* 0 => '2',
* 1 => '1',
* 2 => '!=',
* );
* The stack is processed until only one element is (the result) is left. In
* every iteration the top elements of the stack, up until the first operator,
* are evaluated. After evaluation the arguments and the operator itself are
* removed and replaced by the evaluation result. This is typically 2
* arguments and 1 element for the operator.
* Because the operator is '!=' the example stack is evaluated as:
* $f = (int) 2 != 1;
* The resulting stack is:
* array(
* 0 => 1,
* );
* With only one element left in the stack (the final result) the loop is
* terminated and the result is returned.
*
* @param array $element_stack
* Array of plural formula values and operators create by parseArithmetic().
* @param int $n
* The @count number for which we are determining the right plural position.
*
* @return int
* Number of the plural string to be used for the given plural value.
*
* @see parseArithmetic()
*
* @throws \Exception
*/
protected function evaluatePlural($element_stack, $n) {
$count = count($element_stack);
$limit = $count;
// Replace the '$n' value in the formula by the plural value.
for ($i = 0; $i < $count; $i++) {
if ($element_stack[$i] === '$n') {
$element_stack[$i] = $n;
}
}
// We process the stack until only one element is (the result) is left.
// We limit the number of evaluation cycles to prevent an endless loop in
// case the stack contains an error.
while (isset($element_stack[1])) {
for ($i = 2; $i < $count; $i++) {
// There's no point in checking non-symbols. Also, switch(TRUE) would
// match any case and so it would break.
if (is_bool($element_stack[$i]) || is_numeric($element_stack[$i])) {
continue;
}
$f = NULL;
$length = 3;
$delta = 2;
switch ($element_stack[$i]) {
case '==':
$f = $element_stack[$i - 2] == $element_stack[$i - 1];
break;
case '!=':
$f = $element_stack[$i - 2] != $element_stack[$i - 1];
break;
case '<=':
$f = $element_stack[$i - 2] <= $element_stack[$i - 1];
break;
case '>=':
$f = $element_stack[$i - 2] >= $element_stack[$i - 1];
break;
case '<':
$f = $element_stack[$i - 2] < $element_stack[$i - 1];
break;
case '>':
$f = $element_stack[$i - 2] > $element_stack[$i - 1];
break;
case '+':
$f = $element_stack[$i - 2] + $element_stack[$i - 1];
break;
case '-':
$f = $element_stack[$i - 2] - $element_stack[$i - 1];
break;
case '*':
$f = $element_stack[$i - 2] * $element_stack[$i - 1];
break;
case '/':
$f = $element_stack[$i - 2] / $element_stack[$i - 1];
break;
case '%':
$f = $element_stack[$i - 2] % $element_stack[$i - 1];
break;
case '&&':
$f = $element_stack[$i - 2] && $element_stack[$i - 1];
break;
case '||':
$f = $element_stack[$i - 2] || $element_stack[$i - 1];
break;
case ':':
$f = $element_stack[$i - 3] ? $element_stack[$i - 2] : $element_stack[$i - 1];
// This operator has 3 preceding elements, instead of the default 2.
$length = 5;
$delta = 3;
break;
}
// If the element is an operator we remove the processed elements and
// store the result.
if (isset($f)) {
array_splice($element_stack, $i - $delta, $length, $f);
break;
}
}
}
if (!$limit) {
throw new \Exception('The plural formula could not be evaluated.');
}
return (int) $element_stack[0];
}
}
Classes
Title | Deprecated | Summary |
---|---|---|
PoHeader | Gettext PO header handler. |
Buggy or inaccurate documentation? Please file an issue. Need support? Need help programming? Connect with the Drupal community.