From 5ca7345671490cbc9dfd3e122a4dfc617e977992 Mon Sep 17 00:00:00 2001
From: Guillaume Galy <ggaly@gatling.io>
Date: Wed, 11 Jan 2023 11:14:42 +0100
Subject: [PATCH] fix: inconsistent YAML multiline parsing (#475)

When the YAML block scalar (multiline string) syntax is used with only one line of text, the single line is not matched by the current regexp. It currently only works with at least two lines.

This fixes the regexp by adding a special case for a single line after the first indentation in the block.
---
 lexers/embedded/yaml.xml      |  2 +-
 lexers/testdata/yaml.actual   |  8 ++++++--
 lexers/testdata/yaml.expected | 12 ++++++++++++
 3 files changed, 19 insertions(+), 3 deletions(-)

diff --git a/lexers/embedded/yaml.xml b/lexers/embedded/yaml.xml
index baa7bd5de..97a0b6ed9 100644
--- a/lexers/embedded/yaml.xml
+++ b/lexers/embedded/yaml.xml
@@ -49,7 +49,7 @@
       </rule>
     </state>
     <state name="value">
-      <rule pattern="([&gt;|](?:[+-])?)(\n(^ {1,})(?:.*\n*(?:^\3 *).*)*)">
+      <rule pattern="([&gt;|](?:[+-])?)(\n(^ {1,})(?:(?:.*\n*(?:^\3 *).*)+|.*))">
         <bygroups>
           <token type="Punctuation"/>
           <token type="LiteralStringDoc"/>
diff --git a/lexers/testdata/yaml.actual b/lexers/testdata/yaml.actual
index 772e43fbe..ddb417e5c 100644
--- a/lexers/testdata/yaml.actual
+++ b/lexers/testdata/yaml.actual
@@ -105,7 +105,9 @@ e: |+
 f: >+
  multiline folded keep one space
  line 2
-g: test
+g: |
+  multiline literal with only one line
+h: test
 
 block_scalars_with_indent:
   a: |
@@ -129,7 +131,9 @@ block_scalars_with_indent:
   f: >+
     multiline folded keep
     line 2
-  g: test
+  g: |
+    multiline literal with only one line
+  h: test
 
 ####################
 # COLLECTION TYPES #
diff --git a/lexers/testdata/yaml.expected b/lexers/testdata/yaml.expected
index f054e3c07..a73b48a0a 100644
--- a/lexers/testdata/yaml.expected
+++ b/lexers/testdata/yaml.expected
@@ -251,6 +251,12 @@
   {"type":"NameTag","value":"g"},
   {"type":"Punctuation","value":":"},
   {"type":"TextWhitespace","value":" "},
+  {"type":"Punctuation","value":"|"},
+  {"type":"LiteralStringDoc","value":"\n  multiline literal with only one line"},
+  {"type":"TextWhitespace","value":"  \n"},
+  {"type":"NameTag","value":"h"},
+  {"type":"Punctuation","value":":"},
+  {"type":"TextWhitespace","value":" "},
   {"type":"Literal","value":"test"},
   {"type":"TextWhitespace","value":"\n\n"},
   {"type":"NameTag","value":"block_scalars_with_indent"},
@@ -297,6 +303,12 @@
   {"type":"NameTag","value":"g"},
   {"type":"Punctuation","value":":"},
   {"type":"TextWhitespace","value":" "},
+  {"type":"Punctuation","value":"|"},
+  {"type":"LiteralStringDoc","value":"\n    multiline literal with only one line"},
+  {"type":"TextWhitespace","value":"    \n  "},
+  {"type":"NameTag","value":"h"},
+  {"type":"Punctuation","value":":"},
+  {"type":"TextWhitespace","value":" "},
   {"type":"Literal","value":"test"},
   {"type":"TextWhitespace","value":"\n\n"},
   {"type":"Comment","value":"####################"},