Add basic support for syntax highlighting in code blocks (#515)

* Add basic support for syntax highlighting in code blocks

* Work around additional new lines from chroma

* Make both formatter and style configurable

* Consolidate configs per review

* Fixed when PrefixContext and not ShowContextMulti

* Break out codeblock formatting to separate method per review

go.mod

@@ -25,7 +25,9 @@ require (
 )
 
 require (
+	github.com/alecthomas/chroma/v2 v2.4.0 // indirect
 	github.com/blang/semver v3.5.1+incompatible // indirect
+	github.com/dlclark/regexp2 v1.4.0 // indirect
 	github.com/dustin/go-humanize v1.0.0 // indirect
 	github.com/dyatlov/go-opengraph v0.0.0-20210112100619-dae8665a5b09 // indirect
 	github.com/francoispqt/gojay v1.2.13 // indirect

go.sum

@@ -154,6 +154,8 @@ github.com/advancedlogic/GoOse v0.0.0-20200830213114-1225d531e0ad/go.mod h1:f3HC
 github.com/advancedlogic/GoOse v0.0.0-20210820140952-9d5822d4a625/go.mod h1:f3HCSN1fBWjcpGtXyM119MJgeQl838v6so/PQOqvE1w=
 github.com/ajg/form v1.5.1/go.mod h1:uL1WgH+h2mgNtvBq0339dVnzXdBETtL2LeUXaIv25UY=
 github.com/ajstarks/svgo v0.0.0-20180226025133-644b8db467af/go.mod h1:K08gAheRH3/J6wwsYMMT4xOr94bZjxIelGM0+d/wbFw=
+github.com/alecthomas/chroma/v2 v2.4.0 h1:Loe2ZjT5x3q1bcWwemqyqEi8p11/IV/ncFCeLYDpWC4=
+github.com/alecthomas/chroma/v2 v2.4.0/go.mod h1:6kHzqF5O6FUSJzBXW7fXELjb+e+7OXW4UpoPqMO7IBQ=
 github.com/alecthomas/template v0.0.0-20160405071501-a0175ee3bccc/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc=
 github.com/alecthomas/template v0.0.0-20190718012654-fb15b899a751/go.mod h1:LOuyumcjzFXgccqObfd/Ljyb9UuFJ6TxHnclSeseNhc=
 github.com/alecthomas/units v0.0.0-20151022065526-2efee857e7cf/go.mod h1:ybxpYRFXyAe+OPACYpWeL0wqObRcbAqCMya13uyzqw0=
@@ -451,6 +453,8 @@ github.com/dhui/dktest v0.3.7/go.mod h1:nYMOkafiA07WchSwKnKFUSbGMb2hMm5DrCGiXYG6
 github.com/die-net/lrucache v0.0.0-20181227122439-19a39ef22a11/go.mod h1:ew0MSjCVDdtGMjF3kzLK9hwdgF5mOE8SbYVF3Rc7mkU=
 github.com/disintegration/imaging v1.6.0/go.mod h1:xuIt+sRxDFrHS0drzXUlCJthkJ8k7lkkUojDSR247MQ=
 github.com/disintegration/imaging v1.6.2/go.mod h1:44/5580QXChDfwIclfc/PCwrr44amcmDAg8hxG0Ewe4=
+github.com/dlclark/regexp2 v1.4.0 h1:F1rxgk7p4uKjwIQxBs9oAXe5CqrXlCduYEJvrF4u93E=
+github.com/dlclark/regexp2 v1.4.0/go.mod h1:2pZnwuY/m+8K6iRw6wQdMtk+rH5tNGR1i55kozfMjCc=
 github.com/dnaeon/go-vcr v1.0.1/go.mod h1:aBB1+wY4s93YsC3HHjMBMrwTj2R9FHDzUr9KyGc8n1E=
 github.com/docker/distribution v0.0.0-20190905152932-14b96e55d84c/go.mod h1:0+TTO4EOBfRPhZXAeF1Vu+W3hHZ8eLp8PgKVZlcvtFY=
 github.com/docker/distribution v2.7.1-0.20190205005809-0d3efadf0154+incompatible/go.mod h1:J2gT2udsDAN96Uj4KfcMRqY0/ypR+oyYUYmja8H+y+w=

matterircd.toml.example

@@ -150,6 +150,14 @@ ShowMentions = false
 # This disables that making them appear as normal PRIVMSGs.
 #DisableDefaultMentions = true
 
+# Enable syntax highlighting for code blocks.
+# Formatter and Style are passed through to the chroma v2 package.
+#   https://github.com/alecthomas/chroma/blob/master/formatters/tty_indexed.go#L262
+# terminal/terminal8 for 8-colors, terminal16, terminal256, terminal16m (16M true-colour).
+#   https://github.com/alecthomas/chroma/tree/master/styles
+# These are different colour schemes/styles. E.g. pygments, emacs, autumn, etc.
+SyntaxHighlighting = "terminal256:pygments"
+
 # Path to file to store last viewed information. This is useful for replying only
 # the messages missed.
 LastViewedSaveFile = "matterircd-lastsaved.db"

mm-go-irckit/userbridge.go

@@ -1,6 +1,7 @@
 package irckit
 
 import (
+	"bytes"
 	"encoding/binary"
 	"fmt"
 	"math/rand"
@@ -16,6 +17,7 @@ import (
 	"github.com/42wim/matterircd/bridge/mastodon"
 	"github.com/42wim/matterircd/bridge/mattermost"
 	"github.com/42wim/matterircd/bridge/slack"
+	"github.com/alecthomas/chroma/v2/quick"
 	"github.com/davecgh/go-spew/spew"
 	"github.com/mattermost/mattermost-server/v6/model"
 	"github.com/muesli/reflow/wordwrap"
@@ -139,22 +141,18 @@ func (u *User) handleDirectMessageEvent(event *bridge.DirectMessageEvent) {
 	}
 	text, prefix, suffix, showContext, maxlen := u.handleMessageThreadContext(prefixUser, event.MessageID, event.ParentID, event.Event, event.Text)
 
+	lexer := ""
 	codeBlockBackTick := false
 	codeBlockTilde := false
 	text = wordwrap.String(text, maxlen)
 	lines := strings.Split(text, "\n")
 	for _, text := range lines {
-		if strings.HasPrefix(text, "```") && !codeBlockTilde {
-			codeBlockBackTick = !codeBlockBackTick
-		}
-		if strings.HasPrefix(text, "~~~") && !codeBlockBackTick {
-			codeBlockTilde = !codeBlockTilde
-		}
-		// skip empty lines for anything not part of a code block.
-		if !codeBlockBackTick && !codeBlockTilde && text == "" {
+
+		// TODO: Ideally, we want to read the whole code block and syntax highlight on that, but let's go with per-line for now.
+		text, codeBlockBackTick, codeBlockTilde, lexer = u.formatCodeBlockText(text, prefix, codeBlockBackTick, codeBlockTilde, lexer)
+
+		if text == "" {
 			continue
-		} else if text == "" {
-			text = " "
 		}
 
 		if showContext {
@@ -289,22 +287,18 @@ func (u *User) handleChannelMessageEvent(event *bridge.ChannelMessageEvent) {
 		text, prefix, suffix, showContext, maxlen = u.handleMessageThreadContext(event.ChannelID, event.MessageID, event.ParentID, event.Event, event.Text)
 	}
 
+	lexer := ""
 	codeBlockBackTick := false
 	codeBlockTilde := false
 	text = wordwrap.String(text, maxlen)
 	lines := strings.Split(text, "\n")
 	for _, text := range lines {
-		if strings.HasPrefix(text, "```") && !codeBlockTilde {
-			codeBlockBackTick = !codeBlockBackTick
-		}
-		if strings.HasPrefix(text, "~~~") && !codeBlockBackTick {
-			codeBlockTilde = !codeBlockTilde
-		}
-		// skip empty lines for anything not part of a code block.
-		if !codeBlockBackTick && !codeBlockTilde && text == "" {
+
+		// TODO: Ideally, we want to read the whole code block and syntax highlight on that, but let's go with per-line for now.
+		text, codeBlockBackTick, codeBlockTilde, lexer = u.formatCodeBlockText(text, prefix, codeBlockBackTick, codeBlockTilde, lexer)
+
+		if text == "" {
 			continue
-		} else if text == "" {
-			text = " "
 		}
 
 		if showContext {
@@ -1102,3 +1096,53 @@ func (u *User) handleMessageThreadContext(channelID, messageID, parentID, event,
 
 	return newText, prefix, suffix, showContext, maxlen
 }
+
+//nolint:gocyclo
+func (u *User) formatCodeBlockText(text string, prefix string, codeBlockBackTick bool, codeBlockTilde bool, lexer string) (string, bool, bool, string) {
+	// skip empty lines for anything not part of a code block.
+	if text == "" {
+		if codeBlockBackTick || codeBlockTilde {
+			return " ", codeBlockBackTick, codeBlockTilde, lexer
+		}
+		return "", codeBlockBackTick, codeBlockTilde, lexer
+	}
+
+	syntaxHighlighting := u.v.GetString(u.br.Protocol() + ".syntaxhighlighting")
+
+	if (strings.HasPrefix(text, "```") || strings.HasPrefix(text, prefix+"```")) && !codeBlockTilde {
+		codeBlockBackTick = !codeBlockBackTick
+		if codeBlockBackTick {
+			lexer = strings.TrimSpace(strings.TrimPrefix(strings.TrimPrefix(text, "```"), prefix+"```"))
+		}
+		return text, codeBlockBackTick, codeBlockTilde, lexer
+	}
+	if (strings.HasPrefix(text, "~~~") || strings.HasPrefix(text, prefix+"~~~")) && !codeBlockBackTick {
+		codeBlockTilde = !codeBlockTilde
+		if codeBlockTilde {
+			lexer = strings.TrimSpace(strings.TrimPrefix(strings.TrimPrefix(text, "~~~"), prefix+"~~~"))
+		}
+		return text, codeBlockBackTick, codeBlockTilde, lexer
+	}
+
+	if !(codeBlockBackTick || codeBlockTilde) || syntaxHighlighting == "" || lexer == "" {
+		return text, codeBlockBackTick, codeBlockTilde, lexer
+	}
+
+	formatter := "terminal256"
+	style := "pygments"
+	v := strings.SplitN(syntaxHighlighting, ":", 2)
+	if len(v) == 2 {
+		formatter = v[0]
+		style = v[1]
+	}
+
+	var b bytes.Buffer
+	err := quick.Highlight(&b, text, lexer, formatter, style)
+	if err == nil {
+		text = b.String()
+		// Work around https://github.com/alecthomas/chroma/issues/716
+		text = strings.ReplaceAll(text, "\n", "")
+	}
+
+	return text, codeBlockBackTick, codeBlockTilde, lexer
+}