function XMLReader(array $t_args, array $output)
{
$my_output = [];
foreach ($output as $key => $out) {
$name = $key;
$my_output[$name] = $out;
}
$xpath_row = $t_args['paths']['row'];
$xpath_columns = $t_args['paths']['columns'];
$className = generate_name('XMLReader');
?>
class <?php
echo $className;
?>
{
std::istream& my_stream;
pugi::xml_document doc;
const char* xpath_row;
std::vector<const char*> xpath_columns;
pugi::xpath_node_set rows;
pugi::xpath_node_set::const_iterator start;
pugi::xpath_node_set::const_iterator end;
<?php
\grokit\declareDictionaries($my_output);
?>
public:
<?php
echo $className;
?>
( GIStreamProxy& _stream ) :
my_stream(_stream.get_stream( ))
{
doc.load( my_stream );
xpath_row = "<?php
echo $xpath_row;
?>
";
rows = doc.select_nodes(xpath_row);
<?php
$push = 'xpath_columns.push_back("';
$close_brace = '");';
foreach ($xpath_columns as $col) {
?>
<?php
echo $push;
echo $col;
echo $close_brace;
?>
<?php
}
?>
// Capture the first and last row
start = rows.begin();
end = rows.end();
}
bool ProduceTuple( <?php
echo typed_ref_args($my_output);
?>
)
{
if( start == end ) {
return false;
}
else
{
pugi::xml_node xmlnode = start->node();
std::vector<const char*>::const_iterator it = xpath_columns.begin();
<?php
$col_num = 0;
$node_type = 'auto ';
$select_single_node = ' = xmlnode.select_single_node(';
$close = ');';
$iterator_increment = '++it;';
foreach ($my_output as $name => $type) {
?>
<?php
echo $node_type;
echo ' col_';
echo $col_num;
echo $select_single_node;
?>
*it<?php
echo $close;
?>
<?php
echo \grokit\fromStringDict($name, $type, 'col_' . $col_num . '.node().child_value()');
?>
;
<?php
$col_num = $col_num + 1;
?>
<?php
echo $iterator_increment;
?>
<?php
}
?>
++start;
return true;
}
}
};
<?php
$sys_headers = ['vector', 'string', 'iostream', 'fstream'];
return ['name' => $className, 'kind' => 'GI', 'output' => $my_output, 'system_headers' => $sys_headers, 'lib_headers' => ['pugixml.hpp']];
}
function CSVReader(array $t_args, array $output)
{
$my_output = [];
// Handle case where outputs are given as template arguments
// and not implied.
if (\count($output) == 0) {
grokit_assert(array_key_exists('output', $t_args), 'Did not receive any description of my output!');
$output_list = $t_args['output'];
grokit_assert(is_array($output_list), 'Expected list of types for template argument "output"');
$i = 1;
foreach ($outputs_list as $name => $out_type) {
grokit_assert(is_datatype($out_type) || is_identifier($out_type), 'Expected only types in the "output" list');
if (is_identifier($out_type)) {
$out_type = lookupType($out_type->value());
}
$name = 'val_' . $i;
$my_output[$name] = $out_type;
$i += 1;
}
} else {
foreach ($output as $key => $out) {
$name = $key;
$my_output[$name] = $out;
}
}
$debug = get_default($t_args, 'debug', 0);
$simple = get_default($t_args, 'simple', false);
$trimCR = get_default($t_args, 'trim.cr', false);
// Handle separator
$separator = ',';
if (array_key_exists('sep', $t_args) || array_key_exists('separator', $t_args)) {
$sep = get_first_key($t_args, ['sep', 'separator']);
grokit_assert(is_string($sep), "Got " . gettype($sep) . " instead of string for separator.");
if (strtolower($sep) === 'tab') {
$sep = '\\t';
}
grokit_assert($sep != "\n", 'CSV column delimiter cannot be new line');
// Scream if separator is longer than one character
grokit_assert(\strlen($sep) == 1 || $sep == '\\t', 'Expected string of length 1 for separator, got string <' . $sep . '> instead');
$separator = $sep;
}
// Handle quote character
$quotechar = '"';
if (array_key_exists('quote', $t_args) && !is_null($t_args['quote'])) {
grokit_assert(!$simple, 'Quote option not available for simple CSVReader');
$quote = $t_args['quote'];
grokit_assert(is_string($quote), "Got " . gettype($quote) . " instead of string for quote.");
// Scream if separator is longer than one character
grokit_assert(\strlen($quote) == 1, 'Expected string of length 1 for quote character, got string <' . $quote . '> instead');
$quotechar = $quote;
}
$quotechar = addcslashes($quotechar, '\\\'');
// Handle escape character
$escapeChar = '\\';
if (array_key_exists('escape', $t_args) && !is_null($t_args['escape'])) {
grokit_assert(!$simple, 'Escape option not available for simple CSVReader');
$escape = $t_args['escape'];
grokit_assert(is_string($escape), 'Got ' . gettype($escape) . ' instead of string for escape character.');
grokit_assert(\strlen($escape) == 1, 'Expected string of length 1 for escape character, got string <' . $escape . '> instead');
$escapeChar = $escape;
}
$escapeChar = addcslashes($escapeChar, '\\\'');
// Handle header lines
$headerLines = 0;
if (array_key_exists('skip', $t_args)) {
$headerLines = $t_args['skip'];
grokit_assert(is_int($headerLines), 'Got ' . gettype($headerLines) . ' instead of int for number of lines to skip.');
grokit_assert($headerLines >= 0, 'Cannot skip a negative number of lines.');
}
// Maximum number of lines to read
$maxLines = get_default($t_args, 'n', -1);
grokit_assert(is_int($maxLines), 'Got ' . gettype($maxLines) . ' instead of int for template argument "n"');
$nullArg = get_first_key_default($t_args, ['nullable'], false);
$nullable = [];
$nullStr = [];
foreach ($my_output as $name => $type) {
$nullable[$name] = false;
}
if ($nullArg === true) {
foreach ($my_output as $name => $type) {
$nullable[$name] = true;
$nullStr[$name] = 'NULL';
}
} else {
if (is_array($nullArg)) {
foreach ($nullArg as $n => $v) {
// If nullable value is an associative mapping, the value is either true/false
// or the value of the null string
if (is_string($n)) {
grokit_assert(is_string($v) || is_bool($v), 'CSVReader: nullable associative mapping must have string or boolean values');
grokit_assert(array_key_exists($n, $nullable), 'CSVReader: cannot make unknown attribute ' . $n . ' nullable');
if (is_bool($v)) {
$nullable[$n] = $v;
$nullStr[$n] = 'NULL';
} else {
$nullable[$n] = true;
$nullStr[$n] = $v;
}
} else {
if (is_array($v)) {
grokit_assert(array_key_exists('attr', $v), 'CSVReader: Name of nullable attribute not specified');
$attrName = $v['attr']->name();
$nullable[$attrName] = true;
$nullStr[$attrName] = array_key_exists('null', $v) ? $v['null'] : 'NULL';
} else {
// Otherwise, it's just nullable
$attrName = $v->name();
grokit_assert(array_key_exists($attrName, $nullable), 'CSVReader: cannot make unknown attribute ' . $v . ' nullable');
$nullable[$attrName] = true;
$nullStr[$attrName] = 'NULL';
}
}
}
} else {
if ($nullArg === false) {
// Nothing
} else {
if (is_string($nullArg)) {
foreach ($my_output as $name => $type) {
$nullable[$name] = true;
$nullStr[$name] = $nullArg;
}
} else {
grokit_error('Template argument "nullable" must be boolean or array, ' . typeof($nullArg) . ' given');
}
}
}
}
// Come up with a name for ourselves
$className = generate_name('CSVReader');
if ($debug >= 2) {
foreach ($my_output as $name => $type) {
fwrite(STDERR, "CSVReader: {$name} is nullable: " . ($nullable[$name] ? 'true' : 'false') . PHP_EOL);
}
}
?>
class <?php
echo $className;
?>
{
std::istream& my_stream;
std::string fileName;
// Template parameters
static constexpr size_t MAX_LINES = <?php
echo $maxLines;
?>
;
static constexpr size_t HEADER_LINES = <?php
echo $headerLines;
?>
;
static constexpr char DELIMITER = '<?php
echo $separator;
?>
';
<?php
if (!$simple) {
?>
static constexpr char QUOTE_CHAR = '<?php
echo $quotechar;
?>
';
static constexpr char ESCAPE_CHAR = '<?php
echo $escapeChar;
?>
';
typedef boost::escaped_list_separator<char> separator;
typedef boost::tokenizer< separator > Tokenizer;
separator my_separator;
Tokenizer my_tokenizer;
<?php
}
?>
// Prevent having to allocate this every time.
std::string line;
std::vector<std::string> tokens;
size_t count;
<?php
\grokit\declareDictionaries($my_output);
?>
public:
<?php
echo $className;
?>
( GIStreamProxy& _stream ) :
my_stream(_stream.get_stream())
, fileName(_stream.get_file_name())
<?php
if (!$simple) {
?>
, my_separator(ESCAPE_CHAR, DELIMITER, QUOTE_CHAR)
, my_tokenizer(std::string(""))
<?php
}
?>
, count(0)
{
<?php
if ($headerLines > 0) {
?>
for( size_t i = 0; i < HEADER_LINES; ++i ) {
FATALIF( !getline( my_stream, line ), "CSV Reader reached end of file before finishing header.\n" );
}
<?php
}
// If headerLines > 0
?>
}
// >
bool ProduceTuple( <?php
echo typed_ref_args($my_output);
?>
) {
if (count < MAX_LINES) { //>
count++;
} else {
return false;
}
if( getline( my_stream, line ) ) {
<?php
if ($trimCR) {
?>
if( line.back() == '\r' ) {
line.pop_back();
}
<?php
}
// if trimCR
if (!$simple) {
if ($debug >= 1) {
?>
try {
<?php
}
// if debug >= 1
?>
my_tokenizer.assign( line, my_separator );
<?php
if ($debug >= 1) {
?>
} catch(...) {
FATAL("CSVReader for file %s failed on line: %s", fileName.c_str(), line.c_str());
}
<?php
}
// if debug >= 1
?>
Tokenizer::iterator it = my_tokenizer.begin();
<?php
foreach ($my_output as $name => $type) {
if ($nullable[$name]) {
// nullable
?>
<?php
\grokit\fromStringNullable($name, $type, 'it->c_str()', true, $nullStr[$name]);
?>
<?php
} else {
// not nullable
?>
<?php
echo \grokit\fromStringDict($name, $type, 'it->c_str()');
?>
;
<?php
}
// end nullable check
?>
++it;
<?php
}
// foreach output
} else {
?>
for( char & c : line ) {
if( c == DELIMITER )
c = '\0';
}
const char * ptr = line.c_str();
<?php
$first = true;
foreach ($my_output as $name => $type) {
if ($first) {
$first = false;
} else {
?>
while( *(ptr++) != '\0' )
; // Advance past next delimiter
<?php
}
// not first output
if ($nullable[$name]) {
?>
<?php
echo \grokit\fromStringNullable($name, $type, 'ptr', true, $nullStr[$name]);
} else {
// not nullable
?>
<?php
echo \grokit\fromStringDict($name, $type, 'ptr');
?>
;
<?php
}
// if nullable
}
// foreach output
}
// if simple reader
?>
return true;
}
else {
return false;
}
}
<?php
\grokit\declareDictionaryGetters($my_output);
?>
};
<?php
$sys_headers = ['vector', 'string', 'iostream', 'cstdint'];
if (!$simple) {
$sys_headers[] = 'boost/tokenizer.hpp';
}
return ['name' => $className, 'kind' => 'GI', 'output' => $my_output, 'system_headers' => $sys_headers, 'user_headers' => ['GIStreamInfo.h', 'Dictionary.h', 'DictionaryManager.h']];
}
