Linux bash script recipes
19. 9. 2022
Publikované 5. 12. 2019
API JUSTICE.CZ - parser obchodního rejstříku České republiky
Parser pre vyhľadávanie základných údajov o českých obchodných spoločnostiach a podnikateľoch - názov subjektu, IČO a adresa. Na rozdiel od ARES sú odpovede rýchlejšie a umožňujú aj čiastkové vyhľadanie podľa mena (názvu) subjektu (fulltext). (doplnené: vyhľadávanie v ARES databáze bolo od 12/2018 vypnuté kvôli nejasnostiam v súvislosti s GDPR).
Nevýhodou implementovaného parsovania je nestabilné usporiadanie adries - niekedy sú adresy bez PSČ, inokedy je PSČ na konci, alebo pred názvom obce - čo výrazne sťažuje odladenie kódu. Taktiež striktná viazanosť na štruktúru vráteného HTML kódu znamená, že ak sa zmení usporiadanie DOM elementov, je nutné parser adekvátne upraviť.
Na rozdiel od parsera slovenského obchodného registra ORSR.sk nie je implementované parsovanie detailov subjektu - štatutári, sídlo, predmet činnosti apod. XML štruktúra DOM elementov je v prípade justice.cz o niečo zložitejšia (má väčšiu hĺbku), než v prípade slovenského obchodného registra, ktorý používa tabuľkový layout. Dá sa to dorobiť v prípade potreby, zatiaľ som to však nepotreboval :-)
Inštalácia cez composer - Github.com
> composer require "lubosdz/parser-justice-cz" : "~1.0.0"
Demo - vyhledávání v obchodním rejstříku CZ
Nalezeno / Results:
HTML:
<div class="form-group">
<label for="byName">podle názvu společnosti (vložte nejmíň 3 znaky):</label>
<input type="text" id="byName" placeholder="např. Moje zuby" maxlength="30" class="form-control" />
<div id="response-message"></div>
</div>JS / jQuery:
$("#byName").autocomplete({
// solution #1 - simpler e.g. without spinner
// 'source' => '/justice/find-by-company-name', // implement your own JusticeController
// solution #2 - advanced e.g. add spinner or custom "no match" message
'source' => function(request, response) {
var msg = $("#response-message");
$("#byName").parent().find("label")
.append("<span class='spinner-border spinner-border-sm'></span>");
$.ajax({
url: "/justice/find-by-company-name", // implement your own JusticeController
data: { term: request.term },
success: function (data) {
$("span.spinner-border").remove();
if (!data.length) {
msg.hide().html("Žádné záznamy pro: <b>"+request.term+"</b>").fadeIn();
}
response ($.map(data, function(item) {
return {
label: item.label,
value: item.value
}
}));
}
})
},
'minLength' => 3,
'delay' => 750,
'select' => function(event, ui) {
event.preventDefault(); // prevent from setting JSON string into search field
var msg = $("#response-message"),
companyName = ui.item.label.split(" (IČO")[0];
$(this).val(companyName);
try {
var data = $.parseJSON( ui.item.value ),
html = JSON.stringify(data, null, 4); // JSON pretty print
msg.html("<h3>Společnost: "+companyName+" </h3>"+html);
} catch (error) {
msg.hide().html("<span class='text-danger'>Chyba! "+error+"</span>").fadeIn();
}
},
'focus' => function(event, ui) {
event.preventDefault();
$(this).val(ui.item.label.split(" (IČO")[0]); // set company name into search field
},
});PHP:
<?php
/**
* PHP parser pre vyhladanie subjektov a vypis adresy subjektu a IČO z Veřejnýho rejstříka ČR
* Lookup service for Czech Business Directory / Public Business Register (www.justice.cz)
* ------------------------------------------------------------------
* Demo:
* https://synet.sk/blog/php/365-API-justice-cz-parser-obchodniho-rejstriku
* Author (c) lubosdz@gmail.com
* ------------------------------------------------------------------
* Disclaimer / Prehlásenie:
* Kód poskytnutý je bez záruky a môže kedykoľvek prestať fungovať.
* Jeho funkčnosť je striktne naviazaná na generovanú štruktúru HTML elementov.
* Autor nie je povinný udržiavať kód aktuálny a funkčný, ani neposkytuje ku nemu žiadnu podporu.
* Autor nezodpovedá za nesprávne použitie kódu.
* ------------------------------------------------------------------
* Usage example:
* $connector = new ConnectorJusticeCz;
* $out = $connector->findByNazev('auto');
* $out = $connector->findForAutocomplete('auto');
* $out = $connector->findByIco('44315945');
* $out = $connector->getDetailByICO('44315945');
* echo '<pre>'.print_r($out, 1).'</pre>';
*/
namespace lubosdz\parserJusticeCz;
class ConnectorJusticeCz
{
/**
* @var string URL endpoint base
*/
const URL_SERVER = 'https://or.justice.cz/ias/ui/rejstrik-$firma';
/**
* @var int Max. company label length for autocomplete options
*/
public $labelMaxChars = 30;
/**
* Find company/subject by ICO (company ID), partial ICO not allowed, used by autocomplete
* @param string $ico Full company identifier 8-digits, e.g. 29243831 or 64612023
*/
public function findByIco($ico)
{
$response = [];
$ico = preg_replace('/[^\d]/', '', $ico);
if (preg_match('/^\d{8}$/', $ico)) {
$url = self::URL_SERVER.'?ico='.$ico;
$response = file_get_contents($url);
$response = self::extractSubjects($response);
}
return $response;
}
/**
* Find company by partial name (fulltext)
* @param string $nazev, partial company name e.g. "pojisteni"
*/
public function findByNazev($nazev)
{
$response = [];
if ($nazev) {
$nazev = trim($nazev);
$url = self::URL_SERVER.'?nazev='.urlencode($nazev);
$response = file_get_contents($url);
$response = self::extractSubjects($response);
}
return $response;
}
/**
* Find company/subject by ICO (company ID), partial ICO not allowed, used for direct query
* @param string $ico Full company identifier 8-digits, e.g. 29243831 or 64612023
*/
public function getDetailByICO($ico)
{
$response = [];
$ico = preg_replace('/[^\d]/', '', $ico);
if (preg_match('/^\d{8}$/', $ico)) {
$url = self::URL_SERVER.'?ico='.$ico;
$response = file_get_contents($url);
if($response){
$response = self::extractSubjects($response);
if(!empty($response[0])){
$response = $response[0];
}
}
}
return $response;
}
/**
* Return options for autocomplete list
* @param string $term Searched matching string
* @param int $size How many items to return, 1 - 50 (server returns max. 50 items)
* @param string $valueAs Any key returned from the response (ie. ico or addr_full), if invalid or empty, return JSON encoded all attributes (default)
*/
public function findForAutocomplete($term, $size = 10, $valueAs = 'json')
{
$out = [];
$size = intval($size);
if ($term && $size > 0 && mb_strlen($term, 'utf-8') >= 3) {
// Justice requires at least 3 chars
if (preg_match('/^\d{8}$/', $term)) {
$subjects = $this->findByIco($term);
} else {
$subjects = $this->findByNazev($term);
}
}
if (!empty($subjects) && is_array($subjects)) {
$subjects = array_slice($subjects, 0, $size); // return first $size matches
foreach ($subjects as $id => $subject) {
// add shortened company name for labels
$shortName = $subject['name'];
if (mb_strlen($shortName, 'utf-8') > $this->labelMaxChars) {
$shortName = mb_substr($shortName, 0, $this->labelMaxChars - 3, 'utf-8').'..';
}
$subjects[$id]['shortname'] = $shortName;
}
foreach ($subjects as $subject) {
if(!empty($subject['ico'])){
$out[] = [
'label' => "{$subject['shortname']} (IČO: {$subject['ico']})",
// single attribute only - deprecated
// 'value' => $subject['ico'],
// single attribute or all attributes to save further request for reading company details
'value' => $subject[$valueAs] ?? json_encode($subject, JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES),
];
}
}
}
return $out;
}
/**
* Find list of matching subjects with ICO and the address
* @param string $html HTML response from server justice.cz
* @return array(name, ico, city, addr_city, ..)
*/
protected static function extractSubjects($html)
{
if (!$html) {
throw new \Exception("No HTML content.");
}
// ensure valid XHTML markup
if (!extension_loaded('tidy')) {
throw new \Exception('Missing extension [tidy].');
}
$tidy = new \tidy();
$html = $tidy->repairString($html, array(
'output-xhtml' => true,
'show-body-only' => true,
), 'utf8');
// fix: wrap up HTML into <div> tag to avoid error "Extra content at the end of the document"
// when loading into DOMDocument. Extra script was added in cca 2023 (?).
$html = "<div>{$html}</div>";
// convert whitespaces - inserted are either \n or
$html = strtr($html, [
' ' => ' ',
]);
$html = preg_replace('/\s+/u', ' ', $html);
$html = trim($html);
// load XHTML into DOM document
$xml = new \DOMDocument('1.0', 'utf-8');
$xml->loadXML($html);
$xpath = new \DOMXPath($xml);
$rows = $xpath->query('//table[@class="result-details"]/tbody');
$out = [];
if ($rows->length) {
foreach ($rows as $row) {
// Nazev / Commercial name
$nodeList = $xpath->query("./tr[1]/td[1]", $row);
if (!$nodeList->length) {
continue; // name is required
}
$name = $nodeList->item(0)->nodeValue;
$name = preg_replace('/\s+/u', ' ', $name); // viacnasobne inside spaces
// spisova znacka / register file No.
$nodeList = $xpath->query("./tr[2]/td[1]", $row);
$spisZnacka = $nodeList->length ? $nodeList->item(0)->nodeValue : '';
// ICO / company ID
$nodeList = $xpath->query("./tr[1]/td[2]", $row);
$ico = $nodeList->length ? $nodeList->item(0)->nodeValue : '';
// den zapisu / established date
$nodeList = $xpath->query("./tr[2]/td[2]", $row);
$denZapisuNum = $denZapisTxt = $nodeList->length ? $nodeList->item(0)->nodeValue : '';
if ($denZapisuNum && preg_match('/\d{4}/ui', $denZapisuNum, $match)) {
$parts = explode(' ', $denZapisuNum);
$parts = array_map('trim', $parts);
if (!empty($parts[2]) && $parts[2] > 1800 && !is_numeric($parts[1])) {
$month = self::numerizeMonth($parts[1]);
$day = trim($parts[0], ' .');
$denZapisuNum = $parts[2]
.'-'.str_pad($month, 2, '0', STR_PAD_LEFT)
.'-'.str_pad($day, 2, '0', STR_PAD_LEFT);
}
}
// sidlo/adresa - neda sa spolahnut na poradie prvkov :-(
$city = '';
$nodeList = $xpath->query("./tr[3]/td[1]", $row);
if ($nodeList->length) {
$addr = trim($nodeList->item(0)->nodeValue);
if (preg_match('/,\s*(\d{3} ?\d{2})\s+(.+)$/u', $addr, $match)) {
// Příborská 597, Místek, 738 01 Frýdek-Místek - nazov obce za PSC, prva je ulice a cislo
$city = $addr_city = $match[2];
list($addr_streetnr) = explode(',', $addr);
$addr_zip = $match[1];
} elseif (preg_match('/,\s*PSČ\s+(\d{3} ?\d{2})$/u', $addr, $match)) {
// Řevnice, ČSLA 118, okres Praha-západ, PSČ 25230 - PSC na konci, obec je prva, ulice a cislo druha
list($city, $addr_streetnr) = explode(',', $addr);
$addr_city = $city;
$addr_zip = $match[1];
} elseif(!preg_match('/\d{3} ?\d{2}/u', $addr, $match)) {
// Ústí nad Labem, Masarykova 74 - bez PSC - obec, ulice a cislo
$addr_streetnr = $addr_zip = '';
if (false !== strpos($addr, ',')) {
list($city, $addr_streetnr) = explode(',', $addr);
} else {
list($city) = explode(',', $addr);
}
$addr_city = $city;
}
// "Praha 10 - Dolní Měcholupy" -> Praha 10, pozn: Frydek-Mistek nema medzeru okolo pomlcky
// whoops, avsak ani Ostrana-Hontice a dalsie .. :-( Pre city potrebujeme kratky nazov do 10-15 pismen
$city = explode('-', $city)[0];
// Praha 5 -> Praha
$city = preg_replace('/\d/u', '', $city);
// fix multiple spaces
$city = preg_replace('/\s+/u', ' ', $city);
}
$urlPlatnych = $urlUplny = $urlSbirkaListin = '';
/** @var \DOMNodeList - extract exactly 3 links */
$nodeList = $xpath->query("./../../ul[1]/li/a", $row);
if (3 == $nodeList->length ){
// e.g. "./rejstrik-firma.vysledky?subjektId=561565&typ=PLATNY"
$urlPlatnych = $nodeList->item(0)->getAttribute('href');
// e.g. "./rejstrik-firma.vysledky?subjektId=561565&typ=UPLNY"
$urlUplny = $nodeList->item(1)->getAttribute('href');
// e.g. "./vypis-sl-firma?subjektId=561565"
$urlSbirkaListin = $nodeList->item(2)->getAttribute('href');
}
// final data structure
$out[] = [
'name' => self::trimQuotes($name),
'ico' => preg_replace('/[^\d]/u', '', $ico),
'city' => self::trimQuotes($city), // shortened ie. "Praha" instead of "Praha-Barandov"
// address parts
'addr_city' => self::trimQuotes($addr_city), // full city e.g. "Praha-Barandov"
'addr_zip' => preg_replace('/[^\d]/u', '', $addr_zip),
'addr_streetnr' => self::trimQuotes($addr_streetnr),
// full original address
'addr_full' => self::trimQuotes($addr),
// date established
'den_zapisu_num' => $denZapisuNum,
'den_zapisu_txt' => $denZapisTxt,
// register file No.
'spis_znacka' => $spisZnacka,
// entity links - further data extraction ..
'urlPlatnych' => self::normalizeUrl($urlPlatnych),
'urlUplny' => self::normalizeUrl($urlUplny),
'urlSbirkaListin' => self::normalizeUrl($urlSbirkaListin),
];
}
}
return $out;
}
/**
* Vyhodi quotes z textu, aby neposkodilo HTML atributy
* @param string $txt
*/
protected static function trimQuotes($txt)
{
return trim(strtr($txt, ['"' => '', "'" => '']));
}
/**
* Return month as number, e.g. 30. ledna 2000 -> 2000-01-30
* @param text $month
*/
protected static function numerizeMonth($month)
{
$month = trim($month);
if (preg_match("/^(leden|ledn)/ui", $month)) {
return 1;
} elseif(preg_match("/^(únor)/ui", $month)) {
return 2;
} elseif(preg_match("/^(břez)/ui", $month)) {
return 3;
} elseif(preg_match("/^(dub)/ui", $month)) {
return 4;
} elseif(preg_match("/^(květ)/ui", $month)) {
return 5;
} elseif(preg_match("/^(červn)/ui", $month)) {
return 6;
} elseif(preg_match("/^(červen)/ui", $month)) {
return 7;
} elseif(preg_match("/^(srp)/ui", $month)) {
return 8;
} elseif(preg_match("/^(zář)/ui", $month)) {
return 9;
} elseif(preg_match("/^(říj)/ui", $month)) {
return 10;
} elseif(preg_match("/^(listop)/ui", $month)) {
return 11;
} elseif(preg_match("/^(prosin)/ui", $month)) {
return 12;
}
throw new \Exception("Failed converting month name [{$month}] to numeric.");
}
/**
* Return link converted to absolute without the session hash
* @param string $url
*/
protected static function normalizeUrl($url)
{
$url = dirname(self::URL_SERVER)."/".ltrim($url, ".\/");
$url = explode("&sp=", $url)[0]; // kill request hash, unclear purpose
return $url;
}
}