Initial commit

1 files changed, 119 insertions, 0 deletions

diff --git a/tags.go b/tags.go
new file mode 100644
index 0000000..43a2833
--- /dev/null
+++ b/tags.go
@@ -0,0 +1,119 @@
+// Almost everything in this file is adapted from Ergo IRCd
+// This is probably considered a derived work for copyright purposes
+//
+// SPDX-License-Identifier: MIT
+
+package main
+
+import (
+	"strings"
+	"unicode/utf8"
+)
+
+func parseTags(tagsString string) (tags map[string]string, err error) {
+	tags = make(map[string]string)
+	for 0 < len(tagsString) {
+		tagEnd := strings.IndexByte(tagsString, ';')
+		endPos := tagEnd
+		nextPos := tagEnd + 1
+		if tagEnd == -1 {
+			endPos = len(tagsString)
+			nextPos = len(tagsString)
+		}
+		tagPair := tagsString[:endPos]
+		equalsIndex := strings.IndexByte(tagPair, '=')
+		var tagName, tagValue string
+		if equalsIndex == -1 {
+			// Tag with no value
+			tagName = tagPair
+		} else {
+			tagName, tagValue = tagPair[:equalsIndex], tagPair[equalsIndex+1:]
+		}
+		// "Implementations [...] MUST NOT perform any validation that would
+		//  reject the message if an invalid tag key name is used."
+		if validateTagName(tagName) {
+			if !validateTagValue(tagValue) {
+				err = ErrInvalidTagContent
+				return
+			}
+			tags[tagName] = UnescapeTagValue(tagValue)
+		}
+		tagsString = tagsString[nextPos:]
+	}
+	return
+}
+
+func UnescapeTagValue(inString string) string {
+	// buf.Len() == 0 is the fastpath where we have not needed to unescape any chars
+	var buf strings.Builder
+	remainder := inString
+	for {
+		backslashPos := strings.IndexByte(remainder, '\\')
+
+		if backslashPos == -1 {
+			if buf.Len() == 0 {
+				return inString
+			} else {
+				buf.WriteString(remainder)
+				break
+			}
+		} else if backslashPos == len(remainder)-1 {
+			// trailing backslash, which we strip
+			if buf.Len() == 0 {
+				return inString[:len(inString)-1]
+			} else {
+				buf.WriteString(remainder[:len(remainder)-1])
+				break
+			}
+		}
+
+		// Non-trailing backslash detected; we're now on the slowpath
+		// where we modify the string
+		if buf.Len() < len(inString) {
+			buf.Grow(len(inString))
+		}
+		buf.WriteString(remainder[:backslashPos])
+		buf.WriteByte(escapedCharLookupTable[remainder[backslashPos+1]])
+		remainder = remainder[backslashPos+2:]
+	}
+
+	return buf.String()
+}
+
+var escapedCharLookupTable [256]byte
+
+func init() {
+	for i := 0; i < 256; i += 1 {
+		escapedCharLookupTable[i] = byte(i)
+	}
+	escapedCharLookupTable[':'] = ';'
+	escapedCharLookupTable['s'] = ' '
+	escapedCharLookupTable['r'] = '\r'
+	escapedCharLookupTable['n'] = '\n'
+}
+
+// https://ircv3.net/specs/extensions/message-tags.html#rules-for-naming-message-tags
+func validateTagName(name string) bool {
+	if len(name) == 0 {
+		return false
+	}
+	if name[0] == '+' {
+		name = name[1:]
+	}
+	if len(name) == 0 {
+		return false
+	}
+	// Let's err on the side of leniency here; allow -./ (45-47) in any position
+	for i := 0; i < len(name); i++ {
+		c := name[i]
+		if !(('-' <= c && c <= '/') || ('0' <= c && c <= '9') || ('A' <= c && c <= 'Z') || ('a' <= c && c <= 'z')) {
+			return false
+		}
+	}
+	return true
+}
+
+// "Tag values MUST be encoded as UTF8."
+func validateTagValue(value string) bool {
+	return utf8.ValidString(value)
+}