PASTA: will not retrieve the full DOI

[isangil]

Per report from J.L (woods Hole)

Before I dig deeper I am wondering if you have ever seen where the DOI is only partially retrieved. I used VBO to retrieve the DOI and get only 10.6073/pasta/2ae057a80d when it should have been 10.6073/pasta/2ae057a80dfc191ba3638ee41aa6a729

[isangil]

Dave Reid encoded this helper in the PastaAPI

public function fetchDOI() {
list($scope, $identifier, $revision) = $this->dataSet->getPackageIDParts();
$url = static::getApiUrl("package/doi/eml/{$scope}/{$identifier}/{$revision}");
$request = drupal_http_request($url, array('timeout' => 10));
if (empty($request->error) && $request->code == 200 && !empty($request->data))
if (preg_match('/doi:([\d.]+/pasta/[a-e0-9]+)/', $request->data, $matches)) {
return $matches[1];
}
...

[isangil]

The JL followed up, spotting the problem

"After adding an Exception message before the preg_match I can see that it was getting the DOI.
But with (preg_match('/doi:([\d.]+/pasta/[a-e0-9]+)/', $request->data, $matches)) it would cut out anything above 'e' so put in 'z' and it works. e.g. (preg_match('/doi:([\d.]+/pasta/[a-z0-9]+)/', $request->data, $matches))

Was there a reason for only only a-e?"

[isangil]

And No, there is no reason to stop at "e".... this was a typo.

[davereid]

Oh my gosh, no, it should include 'f' as well.

[isangil]

:)

[isangil]

fix with this very elaborate and arduous commit
235c86c