From d417c2a6ddcd2ed8b4f0157d96938b65c60d832b Mon Sep 17 00:00:00 2001
From: Niklas Fasching <niklas.fasching@gmail.com>
Date: Thu, 16 Apr 2020 15:49:32 +0200
Subject: [PATCH] Implement escaping in src (org only) and example blocks

inside src example blocks lines starting with `\s*,` are escaped - i.e. org
mode will not try to parse them as e.g. a headline. We don't want to render the
escape commata to html so let's take them out - and put them back in before
rendering to org. Doing it this way allows us to render them correctly even
when the input did not include them.

see https://orgmode.org/manual/Literal-Examples.html#Literal-Examples
---
 org/block.go                   |  4 ++++
 org/org_writer.go              |  9 ++++++++-
 org/testdata/blocks.html       | 18 ++++++++++++++++++
 org/testdata/blocks.org        | 17 +++++++++++++++++
 org/testdata/blocks.pretty_org | 19 ++++++++++++++++++-
 5 files changed, 65 insertions(+), 2 deletions(-)

diff --git a/org/block.go b/org/block.go
index 78ad9a7..4b88439 100644
--- a/org/block.go
+++ b/org/block.go
@@ -19,6 +19,7 @@ type Example struct {
 var exampleLineRegexp = regexp.MustCompile(`^(\s*):(\s(.*)|\s*$)`)
 var beginBlockRegexp = regexp.MustCompile(`(?i)^(\s*)#\+BEGIN_(\w+)(.*)`)
 var endBlockRegexp = regexp.MustCompile(`(?i)^(\s*)#\+END_(\w+)`)
+var exampleBlockEscapeRegexp = regexp.MustCompile(`(^|\n)([ \t]*),([ \t]*)(\*|,\*|#\+|,#\+)`)
 
 func lexBlock(line string) (token, bool) {
 	if m := beginBlockRegexp.FindStringSubmatch(line); m != nil {
@@ -51,6 +52,9 @@ func (d *Document) parseBlock(i int, parentStop stopFn) (int, Node) {
 		for ; !stop(d, i); i++ {
 			rawText += trim(d.tokens[i].matches[0]) + "\n"
 		}
+		if name == "EXAMPLE" || (name == "SRC" && len(parameters) >= 1 && parameters[0] == "org") {
+			rawText = exampleBlockEscapeRegexp.ReplaceAllString(rawText, "$1$2$3$4")
+		}
 		block.Children = d.parseRawInline(rawText)
 	} else {
 		consumed, nodes := d.parseMany(i, stop)
diff --git a/org/org_writer.go b/org/org_writer.go
index b562cf3..a59461d 100644
--- a/org/org_writer.go
+++ b/org/org_writer.go
@@ -2,6 +2,7 @@ package org
 
 import (
 	"fmt"
+	"regexp"
 	"strings"
 	"unicode"
 	"unicode/utf8"
@@ -16,6 +17,8 @@ type OrgWriter struct {
 	indent string
 }
 
+var exampleBlockUnescapeRegexp = regexp.MustCompile(`(^|\n)([ \t]*)(\*|,\*|#\+|,#\+)`)
+
 var emphasisOrgBorders = map[string][]string{
 	"_":   []string{"_", "_"},
 	"*":   []string{"*", "*"},
@@ -90,7 +93,11 @@ func (w *OrgWriter) WriteBlock(b Block) {
 	if isRawTextBlock(b.Name) {
 		w.WriteString(w.indent)
 	}
-	WriteNodes(w, b.Children...)
+	content := w.WriteNodesAsString(b.Children...)
+	if b.Name == "EXAMPLE" || (b.Name == "SRC" && len(b.Parameters) >= 1 && b.Parameters[0] == "org") {
+		content = exampleBlockUnescapeRegexp.ReplaceAllString(content, "$1$2,$3")
+	}
+	w.WriteString(content)
 	if !isRawTextBlock(b.Name) {
 		w.WriteString(w.indent)
 	}
diff --git a/org/testdata/blocks.html b/org/testdata/blocks.html
index a4da06a..48b7245 100644
--- a/org/testdata/blocks.html
+++ b/org/testdata/blocks.html
@@ -96,6 +96,24 @@ paragraphs
 </li>
 </ul>
 </blockquote>
+<div class="src src-org">
+<div class="highlight">
+<pre>
+  #+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  #+END_SRC
+
+,* I am not a real headline - commata escape characters aren&#39;t renderered
+</pre>
+</div>
+</div>
+<pre class="example">
+  #+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  #+END_SRC
+
+,* I am not a real headline - commata escape characters aren&#39;t renderered
+</pre>
 <script>
 console.log("Hello World!")
 </script>
diff --git a/org/testdata/blocks.org b/org/testdata/blocks.org
index b789401..a5b24f5 100644
--- a/org/testdata/blocks.org
+++ b/org/testdata/blocks.org
@@ -51,6 +51,23 @@ blocks like the quote block parse their content and can contain
         it can be made visible using css (e.g. =white-space: pre=).
 #+END_QUOTE
 
+#+BEGIN_SRC org
+  ,#+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  ,#+END_SRC
+
+,,* I am not a real headline - commata escape characters aren't renderered
+
+#+END_SRC
+
+#+BEGIN_EXAMPLE
+  ,#+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  ,#+END_SRC
+
+,,* I am not a real headline - commata escape characters aren't renderered
+#+END_EXAMPLE
+
 #+BEGIN_EXPORT html
 <script>
 console.log("Hello World!")
diff --git a/org/testdata/blocks.pretty_org b/org/testdata/blocks.pretty_org
index 40cbec8..8623988 100644
--- a/org/testdata/blocks.pretty_org
+++ b/org/testdata/blocks.pretty_org
@@ -51,6 +51,23 @@ blocks like the quote block parse their content and can contain
         it can be made visible using css (e.g. =white-space: pre=).
 #+END_QUOTE
 
+#+BEGIN_SRC org
+  ,#+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  ,#+END_SRC
+
+,,* I am not a real headline - commata escape characters aren't renderered
+
+#+END_SRC
+
+#+BEGIN_EXAMPLE
+  ,#+BEGIN_SRC bash
+  echo src (with language org) and example blocks support escaping using commata
+  ,#+END_SRC
+
+,,* I am not a real headline - commata escape characters aren't renderered
+#+END_EXAMPLE
+
 #+BEGIN_EXPORT html
 <script>
 console.log("Hello World!")
@@ -80,7 +97,7 @@ this unindented line is outside of the list item
 
   #+BEGIN_QUOTE
   #+BEGIN_EXAMPLE
-  #+END_QUOTE
+  ,#+END_QUOTE
   #+END_EXAMPLE
   #+END_QUOTE