X-Git-Url: http://www.git.cypherpunks.ru/?a=blobdiff_plain;f=r.go;h=5de2a455a4f5d0c315c406bd1d1aa3a56c3a375e;hb=ac9b31f3dfa1dbdf5c4571cfaf3c5b498faa0c0b;hp=2b917ae554c7ca20fc7cf43dfae462440de41063;hpb=29baa531948de1a2478a47d04bbab14012dc4c15;p=gorecfile.git

diff --git a/r.go b/r.go
index 2b917ae..5de2a45 100644
--- a/r.go
+++ b/r.go
@@ -1,6 +1,6 @@
 /*
 recfile -- GNU recutils'es recfiles parser on pure Go
-Copyright (C) 2020 Sergey Matveev <stargrave@stargrave.org>
+Copyright (C) 2020-2023 Sergey Matveev <stargrave@stargrave.org>
 
 This program is free software: you can redistribute it and/or modify
 it under the terms of the GNU General Public License as published by
@@ -15,19 +15,15 @@ You should have received a copy of the GNU General Public License
 along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */
 
-// GNU recutils'es recfiles parser on pure Go
 package recfile
 
 import (
 	"bufio"
 	"errors"
 	"io"
-	"regexp"
 	"strings"
 )
 
-var KeyValRe = regexp.MustCompile(`([a-zA-Z%][a-zA-Z0-9_]*):\s*(.*)$`)
-
 type Reader struct {
 	scanner *bufio.Scanner
 }
@@ -38,6 +34,31 @@ func NewReader(r io.Reader) *Reader {
 	return &Reader{bufio.NewScanner(r)}
 }
 
+func getKeyValue(text string) (string, string) {
+	cols := strings.SplitN(text, ":", 2)
+	if len(cols) != 2 {
+		return "", ""
+	}
+	k := cols[0]
+	if len(k) == 0 {
+		return "", ""
+	}
+	if !((k[0] == '%') ||
+		('a' <= k[0] && k[0] <= 'z') ||
+		('A' <= k[0] && k[0] <= 'Z')) {
+		return "", ""
+	}
+	for _, c := range k[1:] {
+		if !((c == '_') ||
+			('a' <= c && c <= 'z') ||
+			('A' <= c && c <= 'Z') ||
+			('0' <= c && c <= '9')) {
+			return "", ""
+		}
+	}
+	return k, strings.TrimPrefix(cols[1], " ")
+}
+
 // Get next record. Each record is just a collection of fields. io.EOF
 // is returned if there is nothing to read more.
 func (r *Reader) Next() ([]Field, error) {
@@ -59,7 +80,9 @@ func (r *Reader) Next() ([]Field, error) {
 		text = r.scanner.Text()
 
 		if continuation {
-			if text[len(text)-1] == '\\' {
+			if len(text) == 0 {
+				continuation = false
+			} else if text[len(text)-1] == '\\' {
 				lines = append(lines, text[:len(text)-1])
 			} else {
 				lines = append(lines, text)
@@ -95,12 +118,10 @@ func (r *Reader) Next() ([]Field, error) {
 			break
 		}
 
-		matches := KeyValRe.FindStringSubmatch(text)
-		if len(matches) == 0 {
+		name, line = getKeyValue(text)
+		if name == "" {
 			return fields, errors.New("invalid field format")
 		}
-		name = matches[1]
-		line = matches[2]
 
 		if len(line) > 0 && line[len(line)-1] == '\\' {
 			continuation = true
@@ -123,3 +144,29 @@ func (r *Reader) Next() ([]Field, error) {
 	}
 	return fields, nil
 }
+
+// Same as Next(), but with unique keys and last value.
+func (r *Reader) NextMap() (map[string]string, error) {
+	fields, err := r.Next()
+	if err != nil {
+		return nil, err
+	}
+	m := make(map[string]string, len(fields))
+	for _, f := range fields {
+		m[f.Name] = f.Value
+	}
+	return m, nil
+}
+
+// Same as Next(), but with unique keys and slices of values.
+func (r *Reader) NextMapWithSlice() (map[string][]string, error) {
+	fields, err := r.Next()
+	if err != nil {
+		return nil, err
+	}
+	m := make(map[string][]string)
+	for _, f := range fields {
+		m[f.Name] = append(m[f.Name], f.Value)
+	}
+	return m, nil
+}