bionode-ncbi - npm Package Compare versions

Comparing version 1.0.4 to 1.1.0

lib/bionode-ncbi.js

@@ -227,3 +227,3 @@ // # bionode-ncbi
   if (srcUID) { stream.write(srcUID); stream.end() }
-  if (cb) { stream.pipe(concat(cb)) }
+  if (cb) { stream.on('data', cb) }
   else { return stream }
@@ -234,2 +234,3 @@ }
   var stream = through.obj(transform)
+  if (srcDB === 'tax') { srcDB = 'taxonomy' }
   return stream
@@ -261,21 +262,53 @@
     }
-    xml2js(obj.body, function(err, data) {
-      if (err) { self.emit('error', err) }
-      if (data.eLinkResult.LinkSet[0].LinkSetDb) {
-        data.eLinkResult.LinkSet[0].LinkSetDb.forEach(getMatch)
-        function getMatch(link) {
-          if (link.LinkName[0] === query.dbfrom + '_' + query.db) {
-            link.Link.forEach(pushLink)
-            function pushLink(link){
-              result.destUID = link.Id[0]
-              self.push(result)
-            }
-          }
-        }
-      }
+    xml2js(obj.body, gotParsed)
+    function gotParsed (err, data) {
+      if (err) { self.emit('error', err); return next() }
+      if (!data.eLinkResult.LinkSet[0].LinkSetDb) { return next() }
+      data.eLinkResult.LinkSet[0].LinkSetDb.forEach(getMatch)
+      self.push(result)
       next()
-    })
+    }
+    function getMatch(link) {
+      var linkName = query.dbfrom + '_' + query.db
+      if (link.LinkName[0] !== linkName) { return }
+      destUIDs = []
+      link.Link.forEach(getLink)
+      function getLink (link) { destUIDs.push(link.Id[0]) }
+      result.destUIDs = destUIDs
+    }
   }
 }
 
+// ## Property link (Plink)
+// Similar to Link but taked the srcID from a property of the Streamed object
+// and attached the result to a property with the name of the destination DB.
+//
+//     ncbi.search('genome', 'arthropoda')
+//     .pipe(ncbi.expand('tax'))
+//     .pipe(ncbi.plink('tax', 'sra')
+
+ncbi.plink = function (property, destDB) {
+  var srcDB = property.split('.').pop()
+  var destProperty = destDB+'id'
+  var stream = through.obj(transform)
+  return stream
+
+  function transform(obj, enc, next) {
+    var self = this
+    var id = tool.getValue(obj, property+'id')
+    if (!id) {
+      self.push(obj)
+      return next()
+    }
+    if (!obj[destProperty]) { obj[destProperty] = [] }
+    var link = ncbi.link(srcDB, destDB, id, gotData)
+    function gotData(data) {
+      // console.log(data)
+      if (data.destUIDs) { obj[destProperty] = data.destUIDs }
+      self.push(obj)
+      next()
+    }
+  }
+}
+
 // ## Download
@@ -493,1 +526,51 @@ // Takes a NCBI database string and a optional search term and downloads the datasets/sequence files.
 }
+
+
+// ## Expand
+// Takes a property (e.g., biosample) and optional destination property
+// (e.g., sample) and looks for a field named property+id (biosampleid)
+// in the Streamed object. Then it will do a ncbi.search for that id and save
+// the result under Streamed object.property.
+//
+//     ncbi.search('genome', 'arthropoda').pipe(ncbi.expand('assembly'))
+
+ncbi.expand = function(property, destProperty) {
+  var destProperty = destProperty || property
+  var db = property.split('.').pop()
+  if (db === 'tax') { db = 'taxonomy' }
+
+  var stream = through.obj(transform)
+  return stream
+
+  function transform(obj, enc, next) {
+    var self = this
+    var ids = tool.getValue(obj, property+'id')
+    if (!ids) {
+      self.push(obj)
+      return next()
+    }
+
+    // Taxonomy doesn't work just with ID number
+    if (db === 'taxonomy') { ids = ids+'[uid]' }
+
+    if (Array.isArray(ids)) {
+      async.map(ids, search, gotData)
+    } else {
+      search(ids, gotData)
+    }
+
+    function search(term, cb) {
+      var search = ncbi.search(db)
+      search.write(term)
+      search.on('data', function (data) { cb(null, data) })
+      search.on('end', next)
+    }
+
+    function gotData(err, data) {
+      if (err) { throw new Error(err) }
+      obj[destProperty] = data
+      self.push(obj)
+      next()
+    }
+  }
+}

package.json

 {
   "name": "bionode-ncbi",
   "description": "Node.js module for working with the NCBI API (aka e-utils) using Streams.",
-  "version": "1.0.4",
+  "version": "1.1.0",
   "homepage": "http://github.com/bionode/bionode-ncbi",
@@ -6,0 +6,0 @@ "repository": {

test/bionode-ncbi.js

@@ -67,3 +67,3 @@ var fs = require('fs')
 test('Search', function(t) {
-  t.plan(5)
+  t.plan(3)
 
@@ -97,21 +97,22 @@ ncbi.search('assembly', 'Guillardia theta')
 
-  var start1 = Date.now()
-  ncbi.search({ db: 'sra', term: 'human', limit: 500, throughput: 500 })
-  .on('data', function(data) {})
-  .on('end', function() {
-    var msg = 'get 500 objects fast from sra using throughput of 500 per request'
-    var seconds = (Date.now() - start1) / 1000
-    var fast = seconds < 10
-    t.ok(fast, msg)
-  })
-
-  var start2 = Date.now()
-  ncbi.search({ db: 'sra', term: 'human', limit: 500, throughput: 5 })
-  .on('data', function(data) {})
-  .on('end', function() {
-    var msg = 'get 500 objects slowly from sra using throughput of 5 per request'
-    var seconds = (Date.now() - start2) / 1000
-    var slow = seconds > 10
-    t.ok(slow, msg)
-  })
+  // These tests fail randomly on Travis because of network speed
+  // var start1 = Date.now()
+  // ncbi.search({ db: 'sra', term: 'human', limit: 500, throughput: 500 })
+  // .on('data', function(data) {})
+  // .on('end', function() {
+  //   var msg = 'get 500 objects fast from sra using throughput of 500 per request'
+  //   var seconds = (Date.now() - start1) / 1000
+  //   var fast = seconds < 10
+  //   t.ok(fast, msg)
+  // })
+  //
+  // var start2 = Date.now()
+  // ncbi.search({ db: 'sra', term: 'human', limit: 500, throughput: 5 })
+  // .on('data', function(data) {})
+  // .on('end', function() {
+  //   var msg = 'get 500 objects slowly from sra using throughput of 5 per request'
+  //   var seconds = (Date.now() - start2) / 1000
+  //   var slow = seconds > 10
+  //   t.ok(slow, msg)
+  // })
 })
@@ -135,4 +136,4 @@
     var msg = 'same as previous, but doing bioproject->assembly instead of sra->assembly to try get same assembly UID as Search'
-    t.deepEqual(data.destUID, testData.assembly['guillardia-theta'].search.uid, msg)
+    t.deepEqual(data.destUIDs[0], testData.assembly['guillardia-theta'].search.uid, msg)
   })
 })

test/data.json

@@ -239,3 +239,3 @@ {
           "srcUID" : "35533",
-          "destUID" : "53577",
+          "destUIDs" : ["53577"],
           "destDB" : "bioproject"
@@ -250,3 +250,3 @@ }
           "srcUID" : "53577",
-          "destUID" : "503988",
+          "destUIDs" : ["503988"],
           "destDB" : "assembly"
@@ -253,0 +253,0 @@ }

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

1180215

increased by0.21%

Lines of code

4433

increased by1.67%

No dependency changes