From e97194bdb52446337ef1417d98eea49b70ce712a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Andreas=20Sch=C3=A4rtl?= <andreas.schaertl@fau.de>
Date: Thu, 30 Apr 2020 16:28:04 +0200
Subject: [PATCH] add fix-rdf-file.go
- It does the same as fix-rdf-file.py
- I was hoping that it would be faster, but it's actually
slower. I do use regex in the Python version so that's probably
it. For what it's worth, this confirms that the weird regular
expressoin in the Python version is actually correct.
---
ulo/fix-rdf-file.go | 66 +++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 66 insertions(+)
create mode 100644 ulo/fix-rdf-file.go
diff --git a/ulo/fix-rdf-file.go b/ulo/fix-rdf-file.go
new file mode 100644
index 0000000..288ddcb
--- /dev/null
+++ b/ulo/fix-rdf-file.go
@@ -0,0 +1,66 @@
+package main
+
+import (
+ "bufio"
+ "fmt"
+ "os"
+ "strings"
+)
+
+// characters that we want to escape
+var BadChars = []string{
+ "|", "\\", " ", "^", "<", ">",
+}
+
+// what we want to escape BadChars to; EscapedChars[i] should
+// containt he escaped version of BadChars[i]
+var EscapedChars = []string{}
+
+func init() {
+ // initialize EscapedChars
+
+ for _, bc := range BadChars {
+ escaped := fmt.Sprintf("%%%X", bc)
+ EscapedChars = append(EscapedChars, escaped)
+ }
+}
+
+func Fix(r rune) (fixed string) {
+ fixed = fmt.Sprintf("%c", r)
+
+ for i, bad := range BadChars {
+ escaped := EscapedChars[i]
+ fixed = strings.ReplaceAll(fixed, bad, escaped)
+ }
+
+ return fixed
+}
+
+func PrintFixed(line string) {
+ insideQuoted := false
+
+ out := strings.Builder{}
+
+ for _, r := range line {
+ if r == '"' {
+ insideQuoted = !insideQuoted
+ }
+
+ if insideQuoted {
+ fixed := Fix(r)
+ out.WriteString(fixed)
+ } else {
+ out.WriteRune(r)
+ }
+ }
+
+ fmt.Print(out.String())
+}
+
+func main() {
+ scanner := bufio.NewScanner(os.Stdin)
+ for scanner.Scan() {
+ PrintFixed(scanner.Text())
+ PrintFixed("\n")
+ }
+}
--
GitLab