support plain text extract

cmd/parse.go

@@ -48,7 +48,7 @@ func simpleExtr(cmd *cobra.Command, args []string) {
 			defer wg.Done()
 			sem <- struct{}{}
 			defer func() { <-sem }()
-			defer fmt.Println(string(parseJSON(stdin)))
+			defer fmt.Println(string(*parseJSON(stdin)))
 		}()
 		RootClis.HelpFlags = false
 	}
@@ -70,15 +70,15 @@ func simpleExtr(cmd *cobra.Command, args []string) {
 					log.Warnln(err)
 					return
 				}
-				defer fmt.Println(string(parseJSON(input)))
+				defer fmt.Println(string(*parseJSON(input)))
 			}(v)
 		}
 		RootClis.HelpFlags = false
 	}
 	wg.Wait()
 }
 
-func parseJSON(dat []byte) []byte {
+func parseJSON(dat []byte) *[]byte {
 	var sraFields []*extract.SraFields
 	var pubMedFields []*extract.PubmedFields
 	var lock sync.Mutex
@@ -92,7 +92,7 @@ func parseJSON(dat []byte) []byte {
 			lock.Unlock()
 		}
 		dat, _ := json.MarshalIndent(pubMedFields, "", "    ")
-		return dat
+		return &dat
 	} else if RootClis.Mode == "sra" && len(dat) > 0 {
 		json.Unmarshal(dat, &sraJSON)
 		done := make(map[string]int)
@@ -103,13 +103,17 @@ func parseJSON(dat []byte) []byte {
 			lock.Unlock()
 		}
 		dat, _ := json.MarshalIndent(sraFields, "", "    ")
-		return dat
+		return &dat
+	} else if len(dat) > 0 {
+		obj, _ := extract.GetPlainFields("", &dat, &keyWords, RootClis.CallCor)
+		dat, _ := json.MarshalIndent(obj, "", "    ")
+		return &dat
 	}
-	return []byte{}
+	return nil
 }
 
 func init() {
 	RootCmd.Flags().StringVarP(&RootClis.Keywords, "keywords", "w", "algorithm, tool, model, pipleline, method, database, workflow, dataset, bioinformatics, sequencing, http, github.com, gitlab.com, bitbucket.org", "Keywords to extracted from abstract.")
 	RootCmd.Flags().BoolVarP(&RootClis.CallCor, "call-cor", "", false, "Wheather to calculate the corelated keywords, and return the sentence contains >=2 keywords.")
-	RootCmd.Flags().StringVarP(&RootClis.Mode, "mode", "", "", "mode to extract information (pubmed, sra).")
+	RootCmd.Flags().StringVarP(&RootClis.Mode, "mode", "", "", "mode to extract information: plain,pubmed, or sra.")
 }

go.mod

@@ -5,7 +5,7 @@ go 1.13
 require (
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
 	github.com/modern-go/reflect2 v1.0.1 // indirect
-	github.com/openbiox/ligo v0.0.0-20200423121108-3cfa60ff813d
+	github.com/openbiox/ligo v0.0.0-20200423134214-6abf93e36b96
 	github.com/sirupsen/logrus v1.5.0
 	github.com/spf13/cobra v1.0.0
 	github.com/spf13/pflag v1.0.5 // indirect

pdf2plain/go.mod

@@ -6,8 +6,17 @@ require (
 	code.sajari.com/docconv v1.1.0
 	github.com/JalfResi/justext v0.0.0-20170829062021-c0282dea7198 // indirect
 	github.com/advancedlogic/GoOse v0.0.0-20191112112754-e742535969c1 // indirect
+	github.com/araddon/dateparse v0.0.0-20200409225146-d820a6159ab1 // indirect
+	github.com/go-resty/resty/v2 v2.2.0 // indirect
+	github.com/jaytaylor/html2text v0.0.0-20200412013138-3577fbdbcff7 // indirect
 	github.com/levigross/exp-html v0.0.0-20120902181939-8df60c69a8f5 // indirect
-	github.com/openbiox/ligo v0.0.0-20200324043626-a1096f60db64
+	github.com/mattn/go-runewidth v0.0.9 // indirect
+	github.com/olekukonko/tablewriter v0.0.4 // indirect
+	github.com/openbiox/ligo v0.0.0-20200423121108-3cfa60ff813d
+	github.com/pkg/errors v0.9.1 // indirect
 	github.com/sirupsen/logrus v1.5.0
-	github.com/spf13/cobra v0.0.6
+	github.com/spf13/cobra v1.0.0
+	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/vbauerster/mpb/v4 v4.11.2 // indirect
+	golang.org/x/net v0.0.0-20200421231249-e086a090c8fd // indirect
 )