Added almost all PubMed XML elements to parser

tripal_pub/includes/pubmed.inc

@@ -62,7 +62,7 @@ function tripal_pub_remote_search_pubmed_range($terms, $start = 0, $limit = 10)
      
   // repeat the search performed previously (using WebEnv & QueryKey) to retrieve
   // the PMID's within the range specied.  The PMIDs will be returned as a text list
-  $pmids_txt = tripal_pub_remote_search_pubmed_fetch($terms, $query_key, $web_env, 'uilist', $start, $limit);  
+  $pmids_txt = tripal_pub_remote_search_pubmed_fetch($terms, $query_key, $web_env, 'uilist', 'text', $start, $limit);  
   
   // iterate through each PMID and get the publication record. This requires a new search and new fetch
   $pmids = explode("\n", trim($pmids_txt));
@@ -74,7 +74,7 @@ function tripal_pub_remote_search_pubmed_range($terms, $start = 0, $limit = 10)
     $query = tripal_pub_remote_search_pubmed_query($terms); 
     
     // second retrieve the individual record
-    $pub_xml = tripal_pub_remote_search_pubmed_fetch($terms, $query['QueryKey'], $query['WebEnv'], 'xml', 0, 1);
+    $pub_xml = tripal_pub_remote_search_pubmed_fetch($terms, $query['QueryKey'], $query['WebEnv'], 'null', 'xml', 0, 1);
     $pub = tripal_pub_remote_search_pubmed_parse_pubxml($pub_xml);
     $pubs[] = $pub;    
     
@@ -133,11 +133,12 @@ function tripal_pub_remote_search_pubmed_query($terms){
 /*
  * 
  */
-function tripal_pub_remote_search_pubmed_fetch($terms, $query_key, $web_env, $rettype = 'xml', $start = 0, $limit = 10){
+function tripal_pub_remote_search_pubmed_fetch($terms, $query_key, $web_env, $rettype = 'null', 
+  $retmod = 'null', $start = 0, $limit = 10){
   
   // repeat the search performed previously (using WebEnv & QueryKey) to retrieve
   // the PMID's within the range specied.  The PMIDs will be returned as a text list
-  $fetch_url = "http://www.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?rettype=$rettype&retmode=text&retstart=$start&retmax=$limit&db=Pubmed&query_key=$query_key&WebEnv=$web_env";
+  $fetch_url = "http://www.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?rettype=$rettype&retmode=$retmod&retstart=$start&retmax=$limit&db=Pubmed&query_key=$query_key&WebEnv=$web_env";
   $rfh = fopen($fetch_url, "r");
   $results = '';
   while (!feof($rfh)) {
@@ -152,7 +153,13 @@ function tripal_pub_remote_search_pubmed_fetch($terms, $query_key, $web_env, $re
  * This function parses the XML containing details of a publication and
  * converts it into an associative array of where keys are Tripal Pub 
  * ontology terms and the values are extracted from the XML. The
- * XML should contain only a single publication record.
+ * XML should contain only a single publication record.  
+ * 
+ * Information about the valid elements in the PubMed XML can be found here:
+ * http://www.nlm.nih.gov/bsd/licensee/elements_descriptions.html
+ * 
+ * Information about PubMed's citation format can be found here
+ * http://www.nlm.nih.gov/bsd/policy/cit_format.html
  */
 function tripal_pub_remote_search_pubmed_parse_pubxml($pub_xml) {
   $pub = array();
@@ -171,8 +178,57 @@ function tripal_pub_remote_search_pubmed_parse_pubxml($pub_xml) {
           $pub['pub_database'] = 'PMID';
           break;        
         case 'Article':
+          $pub_model = $xml->getAttribute('PubModel');
+          $pub['publication_model'] = $pub_model;
           tripal_pub_remote_search_pubmed_parse_article($xml, $pub);
           break;
+        case 'MedlineJournalInfo':
+          tripal_pub_remote_search_pubmed_parse_medline_journal_info($xml, $pub);
+          break;
+        case 'ChemicalList':
+          // TODO: handle this
+          break;
+        case 'SupplMeshList':
+          // TODO: meant for protocol list
+          break;
+        case 'CitationSubset':
+          // TODO: not sure this is needed.         
+          break;
+        case 'CommentsCorrections':
+          // TODO: handle this
+          break;
+        case 'GeneSymbolList':
+          // TODO: handle this
+          break;
+        case 'MeshHeadingList':
+          // TODO: Medical subject headings
+          break;
+        case 'NumberOfReferences':
+          // TODO: not sure we should keep this as it changes frequently.
+          break;
+        case 'PersonalNameSubjectList':
+          // TODO: for works about an individual or with biographical note/obituary.
+          break;
+        case 'OtherID':
+          // TODO: ID's from another NLM partner.
+          break;
+        case 'OtherAbstract':
+          // TODO: when the journal does not contain an abstract for the publication.
+          break;
+        case 'KeywordList':
+          // TODO: handle this
+          break;
+        case 'InvestigatorList':
+          // TODO: personal names of individuals who are not authors (can be used with collection)
+          break;
+        case 'GeneralNote':
+          // TODO: handle this
+          break;
+        case 'DeleteCitation':
+          // TODO: need to know how to handle this
+          break;          
+        default:
+          break;
       }
     }
   }
@@ -180,6 +236,41 @@ function tripal_pub_remote_search_pubmed_parse_pubxml($pub_xml) {
   return $pub;
 }
 
+/*
+ * 
+ */
+function tripal_pub_remote_search_pubmed_parse_medline_journal_info($xml, &$pub) {
+  while ($xml->read()) {
+    // get this element name
+    $element = $xml->name;     
+        
+    // if we're at the </Article> element then we're done with the article...
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element == 'MedlineJournalInfo') {
+      return;  
+    }
+    if ($xml->nodeType == XMLReader::ELEMENT) { 
+      switch ($element) {
+        case 'Country':
+          // the place of publication of the journal
+          $xml->read();
+          $pub['journal_country'] = $xml->value;
+          break;
+        case 'MedlineTA':
+          // TODO: not sure how this is different from ISOAbbreviation
+          break;
+        case 'NlmUniqueID':
+          // TODO: the journal's unique ID in medline
+          break;
+        case 'ISSNLinking':
+          // TODO: not sure how this is different from ISSN
+          break;  
+        default:
+          break;
+      }
+    }
+  }
+}
+
 /*
  * 
  */
@@ -202,27 +293,160 @@ function tripal_pub_remote_search_pubmed_parse_article($xml, &$pub) {
           $xml->read();
           $pub['title'] = $xml->value;
           break;
-        case 'AbstractText':
-          $xml->read();
-          $pub['abstract'] = $xml->value;
+        case 'Abstract':
+          tripal_pub_remote_search_pubmed_parse_abstract($xml, $pub);
           break;
+        case 'Pagination':
+          tripal_pub_remote_search_pubmed_parse_pagination($xml, $pub);
+          break;  
+        case 'ELocationID':
+          $type = $xml->getAttribute('EIdType');
+          $valid = $xml->getAttribute('ValidYN');
+          $xml->read();
+          $elocation = $xml->value;
+          if ($type == 'doi' and $valid == 'Y') {
+            $pub['DOI'] = $elocation;
+          }
+          if ($type == 'pii' and $valid == 'Y') {
+            $pub['PII'] = $elocation;
+          }
+          $pub['elocation'] = $elocation;
+          break;        
         case 'Affiliation':
+          // the affiliation tag at this level is meant solely for the first author
+          $xml->read();
+          $pub['authors'][0]['affiliation'] = $xml->value;
           break;
         case 'AuthorList':
+          $complete = $xml->getAttribute('CompleteYN');
           tripal_pub_remote_search_pubmed_parse_authorlist($xml, $pub);
           break;
+        case 'InvestigatorList':
+          // TODO: perhaps handle this one day.  The investigator list is to list the names of people who 
+          // are members of a collective or corporate group that is an author in the paper. 
+          break;
         case 'Language':
           $xml->read();
-          $pub['language'] = $xml->value;
+          $lang_abbr = $xml->value;
+          // there may be multiple languages so we store these in an array
+          $pub['language'][] = tripal_pub_remote_search_get_language($lang_abbr);
+          $pub['language_abbr'][] = $lang_abbr;
+          break;
+        case 'DataBankList':
+          // TODO: handle this case
+          break;
+        case 'GrantList':
+          // TODO: handle this case
+          break;
+        case 'PublicationTypeList':
+          tripal_pub_remote_search_pubmed_parse_publication_type($xml, $pub);
+          break;
+        case 'VernacularTitle':
+          $xml->read();
+          $pub['vernacular_title'][] = $xml->value;;
+          break;
+        case 'ArticleDate': 
+          // TODO: figure out what to do with this element. We already have the
+          // published date in the <PubDate> field, but this date should be in numeric
+          // form and may have more information.         
           break;
-        case 'ArticleDate':
-          break; 
         default:
           break;     
       }
     }
   }
 }
+/*
+ * A full list of publication types can be found here:
+ * http://www.nlm.nih.gov/mesh/pubtypes.html.
+ * 
+ * The Tripal Pub ontology doesn't yet have terms for all of the 
+ * publication types so we store the value in the 'publication_type' term.
+ */
+function tripal_pub_remote_search_pubmed_parse_publication_type($xml, &$pub) {
+  
+  while ($xml->read()) {
+    $element = $xml->name;    
+      
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element == 'PublicationTypeList') {
+      // we've reached the </Pagination> element so we're done.
+      return;
+    }
+    if ($xml->nodeType == XMLReader::ELEMENT) {
+      switch ($element) {
+        case 'PublicationType':  
+          $xml->read();          
+          $pub['publication_type'][] = $xml->value;
+          break;          
+        default:
+          break;
+      }
+    }
+  }  
+}
+/*
+ * 
+ */
+function tripal_pub_remote_search_pubmed_parse_abstract($xml, &$pub) {
+  $abstract = '';
+  
+  while ($xml->read()) {
+    $element = $xml->name;
+      
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element == 'Abstract') {
+      // we've reached the </Abstract> element so return   
+      $pub['abstract'] = $abstract;        
+      return;
+    }
+    // the abstract text can be just a singe paragraph or be broken into multiple
+    // abstract texts for structured abstracts.  Here we will just combine then 
+    // into a single element in the order that they arrive in HTML format
+    if ($xml->nodeType == XMLReader::ELEMENT) {
+      switch ($element) {
+        case 'AbstractText':
+          $label = $xml->getAttribute('Label');
+          $xml->read();
+          if ($label) {
+            $pub['structured_abstract_part'][][$label] = $xml->value;
+            $abstract .= "<p><b>$label</b></br>" . $xml->value . '</p>';
+          }
+          else {
+            $abstract .= '<p>' . $xml->value . '</p>';
+          }
+          break;
+        case 'CopyrightInformation':
+          $xml->read();
+          $pub['copyright'] = $xml->value;
+          break;          
+        default:
+          break;
+      }
+    }
+  }
+}
+/*
+ * 
+ */
+function tripal_pub_remote_search_pubmed_parse_pagination($xml, &$pub) {
+  while ($xml->read()) {
+    $element = $xml->name;
+      
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element == 'Pagination') {
+      // we've reached the </Pagination> element so we're done.
+      return;
+    }
+    if ($xml->nodeType == XMLReader::ELEMENT) {
+      switch ($element) {
+        case 'MedlinePgn':  
+          $xml->read();
+          $pub['pages'] = $xml->value;
+          break;          
+        default:
+          break;
+      }
+    }
+  }
+}
 /*
  * 
  */
@@ -238,60 +462,269 @@ function tripal_pub_remote_search_pubmed_parse_journal($xml, &$pub) {
       }
     }
     if ($xml->nodeType == XMLReader::ELEMENT) {
-      $xml->read();
-      $value = $xml->value;
       switch ($element) {
-        case 'ISSN':          
+        case 'ISSN':  
+          $issn_type = $xml->getAttribute('IssnType');          
+          $xml->read();
+          $issn = $xml->value;
+          $pub['ISSN'] = $issn;
+          if ($issn_type == 'Electronic') {            
+            $pub['eISSN'] = $issn;  
+          }
+          if ($issn_type == 'Print') {            
+            $pub['pISSN'] = $issn;  
+          }                 
           break;
-        case 'Volume':          
+        case 'JournalIssue':   
+          // valid values of cited_medium are 'Internet' and 'Print'
+          $cited_medium = $xml->getAttribute('CitedMedium');          
+          tripal_pub_remote_search_pubmed_parse_journal_issue($xml, $pub);       
+          break;        
+        case 'Title': 
+          $xml->read();
+          $pub['journal_name'] = $xml->value;          
           break;
-        case 'Year':          
+        case 'ISOAbbreviation': 
+          $xml->read();
+          $pub['journal_iso_abbreviation'] = $xml->value;         
           break;
-        case 'Title':          
+        default:
           break;
-        case 'ISOAbbreviation':          
+      }
+    }
+  }
+}
+/*
+ * 
+ */
+function tripal_pub_remote_search_pubmed_parse_journal_issue($xml, &$pub) {
+  
+  while ($xml->read()) {
+    $element = $xml->name;
+      
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element = 'JournalIssue'){
+      // if we're at the </JournalIssue> element then we're done 
+      return;
+    }
+    if ($xml->nodeType == XMLReader::ELEMENT) {
+      switch ($element) {
+        case 'Volume': 
+          $xml->read();
+          $pub['volumn'] = $xml->value;         
+          break;
+        case 'Issue':
+          $xml->read();
+          $pub['issue'] = $xml->value;           
+          break;
+        case 'PubDate':
+          $date = tripal_pub_remote_search_pubmed_parse_date($xml, 'PubDate');
+          $year = $date['year'];
+          $month = $date['month'];
+          $day = $date['day']; 
+          $medline = $date['medline']; 
+            
+          if ($month and $day and $year) {
+            $pub['publication_date'] = "$year $month $day";
+          }
+          elseif ($month and !$day and $year) {
+            $pub['publication_date'] = "$year $month";
+          }
+          elseif (!$month and !$day and $year) {
+            $pub['publication_date'] = $year;
+          } 
+          elseif ($medline) {
+            $pub['publication_date'] = $medline;
+          }
+          else {
+            $pub['publication_date'] = "Unknown";
+          }             
           break;
         default:
           break;
       }
     }
-  }
+  }  
+}
+
+/*
+ * 
+ */
+function tripal_pub_remote_search_pubmed_parse_date ($xml, $element_name) {
+  $date = array();
+  
+  while ($xml->read()) {
+    $element = $xml->name;
+      
+    if ($xml->nodeType == XMLReader::END_ELEMENT and $element = $element_name){
+      // if we're at the </$element_name> then we're done 
+      return $date;
+    }
+    if ($xml->nodeType == XMLReader::ELEMENT) {           
+      switch ($element) {
+        case 'Year':
+          $xml->read(); 
+          $date['year'] = $xml->value;         
+          break;
+        case 'Month':
+          $xml->read(); 
+          $month = 
+          $date['month'] = $xml->value;          
+          break;
+        case 'Day':
+          $xml->read(); 
+          $date['day'] = $xml->value;          
+          break;
+        case 'MedlineDate':
+          // the medline date is when the date cannot be broken into distinct month day year.
+          $xml->read(); 
+          $date['medline'] = $xml->value;          
+          break;
+        default:
+          break;
+      }    
+    }
+  } 
 }
 /*
  * 
  */
 function tripal_pub_remote_search_pubmed_parse_authorlist($xml, &$pub) {
-  $authors = array();
-  $author = array();
-  $author_list = '';
+  $num_authors = 0;
   
   while ($xml->read()) {
     $element = $xml->name;
       
     if ($xml->nodeType == XMLReader::END_ELEMENT){
       // if we're at the </AuthorList> element then we're done with the article...
-      if($element = 'Article') {
+      if($element == 'AuthorList') {
+        // build the author list before returning
+        $author_list = '';
+        foreach ($pub['authors'] as $author) {
+          if ($author['valid'] == 'N') {
+            // skip non-valid entries.  A non-valid entry should have 
+            // a corresponding corrected entry so we can saftely skip it.
+            continue;
+          }
+          if ($author['collective']) {
+            $author_list .= $author['collective'] . ', ';
+          }
+          else {
+            $author_list .= $author['surname'] . ' ' . $author['first_initials'] . ', '; 
+          }             
+        }
+        $author_list = substr($author_list, 0, -2);
+        $pub['author_list'] = $author_list;
         return;
       }
       // if we're at the end </Author> element then we're done with the author and we can
       // start a new one.
-      if($element = 'Author') {
-        return;
+      if($element == 'Author') {
+        $num_authors++;
       }  
     }
     if ($xml->nodeType == XMLReader::ELEMENT) {
       switch ($element) {
-        case 'Author':          
+        case 'Author':
+          $valid = $xml->getAttribute('ValidYN');
+          $pub['authors'][$num_authors]['valid'] = $valid;          
+          break;
+        case 'LastName':  
+          $xml->read(); 
+          $pub['authors'][$num_authors]['surname'] = $xml->value;         
+          break;
+        case 'ForeName': 
+          $xml->read(); 
+          $pub['authors'][$num_authors]['given_name'] = $xml->value;          
+          break;
+        case 'Initials': 
+          $xml->read(); 
+          $pub['authors'][$num_authors]['first_initials'] = $xml->value;          
           break;
-        case 'LastName':          
+        case 'Suffix': 
+          $xml->read(); 
+          $pub['authors'][$num_authors]['suffix'] = $xml->value;          
           break;
-        case 'ForeName':          
+        case 'CollectiveName': 
+          $xml->read(); 
+          $pub['authors'][$num_authors]['collective'] = $xml->value;          
           break;
-        case 'Initials':          
+        case 'Identifier': 
+          // according to the specification, this element is not yet used.                  
           break;
         default:
           break;
       }
     }
   }
+}
+
+/*
+ * Language abbreviations were obtained here: 
+ * http://www.nlm.nih.gov/bsd/language_table.html
+ */
+function tripal_pub_remote_search_get_language($lang_abbr) {
+  $languages = array(
+    'afr' => 'Afrikaans',
+    'alb' => 'Albanian',
+    'amh' => 'Amharic',
+    'ara' => 'Arabic',
+    'arm' => 'Armenian',
+    'aze' => 'Azerbaijani',
+    'ben' => 'Bengali',
+    'bos' => 'Bosnian',
+    'bul' => 'Bulgarian',
+    'cat' => 'Catalan',
+    'chi' => 'Chinese',
+    'cze' => 'Czech',
+    'dan' => 'Danish',
+    'dut' => 'Dutch',
+    'eng' => 'English',
+    'epo' => 'Esperanto',
+    'est' => 'Estonian',
+    'fin' => 'Finnish',
+    'fre' => 'French',
+    'geo' => 'Georgian',
+    'ger' => 'German',
+    'gla' => 'Scottish Gaelic',
+    'gre' => 'Greek, Modern',
+    'heb' => 'Hebrew',
+    'hin' => 'Hindi',
+    'hrv' => 'Croatian',
+    'hun' => 'Hungarian',
+    'ice' => 'Icelandic',
+    'ind' => 'Indonesian',
+    'ita' => 'Italian',
+    'jpn' => 'Japanese',
+    'kin' => 'Kinyarwanda',
+    'kor' => 'Korean',
+    'lat' => 'Latin',
+    'lav' => 'Latvian',
+    'lit' => 'Lithuanian',
+    'mac' => 'Macedonian',
+    'mal' => 'Malayalam',
+    'mao' => 'Maori',
+    'may' => 'Malay',
+    'mul' => 'Multiple languages',
+    'nor' => 'Norwegian',
+    'per' => 'Persian',
+    'pol' => 'Polish',
+    'por' => 'Portuguese',
+    'pus' => 'Pushto',
+    'rum' => 'Romanian, Rumanian, Moldovan',
+    'rus' => 'Russian',
+    'san' => 'Sanskrit',
+    'slo' => 'Slovak',
+    'slv' => 'Slovenian',
+    'spa' => 'Spanish',
+    'srp' => 'Serbian',
+    'swe' => 'Swedish',
+    'tha' => 'Thai',
+    'tur' => 'Turkish',
+    'ukr' => 'Ukrainian',
+    'und' => 'Undetermined',
+    'urd' => 'Urdu',
+    'vie' => 'Vietnamese',
+    'wel' => 'Welsh',
+  );
+  return $languages[$lang_abbr];
 }