From 67c7aa45c689b716753edf9cdd066aa5cb3ed5b3 Mon Sep 17 00:00:00 2001
From: Van De Velde <sven.van-de-velde@hpe.com>
Date: Thu, 5 Jul 2018 08:09:20 +0200
Subject: [PATCH] Updated <em> tag logic using _

In markdown, an _ is something special ... One cannot just scan for every _ in a text and emphasize every text between _ (underscores).
In many program languages, _ are used to represent a space in identifiers and functions etc.
Therefore, I've updated the markdown.lua logic to ensure that text between _ is only emphasized when the first _ starts with a space or any other non alpameric character, or the _ starts at the beginning of the text.

examples:

AI\_CARGO\_DISPATCHER gives AI\_CARGO\_DISPATCHER.
AI \_CARGO\_ DISPATCHER gives AI <em>CARGO</em> DISPATCHER.
\_AI\_CARGO\_DISPATCHER gives <em>AI</em>CARGO\_DISPATCHER
---
 markdown-tests.lua | 19 +++++++++++++++++++
 markdown.lua       | 16 +++++++++++++---
 2 files changed, 32 insertions(+), 3 deletions(-)

diff --git a/markdown-tests.lua b/markdown-tests.lua
index 187fe90..f2090f5 100644
--- a/markdown-tests.lua
+++ b/markdown-tests.lua
@@ -4392,6 +4392,25 @@ This is [an example] [id] reference-style link.
 <p><a href="kbs/wes-box-134.md.lp">Wesel - Boxtel [KBS 134]</a></p>
 ]==]
 
+tests.sven = [[
+TASK_CARGO_DISPATCHER
+~
+<p>TASK_CARGO_DISPATCHER</p>
+~
+TASK _CARGO_ DISPATCHER
+~
+<p>TASK<em>CARGO</em>DISPATCHER</p>
+~
+TASK_CARGO_ DISPATCHER
+~
+<p>TASK_CARGO_ DISPATCHER</p>
+~
+TASK*CARGO*DISPATCHER
+~
+<p>TASK<em>CARGO</em>DISPATCHER</p>
+]]
+
+
 -- Unhandled test: _M*A*S*H_
 
 local quiet_mode
diff --git a/markdown.lua b/markdown.lua
index 636e8f9..fd3349a 100644
--- a/markdown.lua
+++ b/markdown.lua
@@ -895,9 +895,19 @@ local function emphasis(text)
       text = text:gsub(s .. "([^%s][%*%_]?)" .. s, "<strong>%1</strong>")
       text = text:gsub(s .. "([^%s][^<>]-[^%s][%*%_]?)" .. s, "<strong>%1</strong>")
    end
-   for _, s in ipairs {"%*", "%_"} do
-      text = text:gsub(s .. "([^%s_])" .. s, "<em>%1</em>")
-      text = text:gsub(s .. "(<strong>[^%s_]</strong>)" .. s, "<em>%1</em>")
+   for _, s in ipairs {"%_"} do
+      text = text:gsub("([^%w]+)" .. s .. "([^%s_])" .. s, "%1<em>%2</em> ")
+      text = text:gsub("([^%w]+)" .. s .. "(<strong>[^%s_]</strong>)" ..s, "%1<em>%2</em>")
+      text = text:gsub("([^%w]+)" .. s .. "([^%s_][^<>_]-[^%s_])" .. s, "%1<em>%2</em>")
+      text = text:gsub("([^%w]+)" .. s .. "([^<>_]-<strong>[^<>_]-</strong>[^<>_]-)" .. s, "%1<em>%2</em>")
+      text = text:gsub("^" .. s .. "([^%s_])" .. s, "<em>%1</em> ")
+      text = text:gsub("^" .. s .. "(<strong>[^%s_]</strong>)" ..s, "<em>%1</em>")
+      text = text:gsub("^" .. s .. "([^%s_][^<>_]-[^%s_])" .. s, "<em>%1</em>")
+      text = text:gsub("^" .. s .. "([^<>_]-<strong>[^<>_]-</strong>[^<>_]-)" .. s, "<em>%1</em>")
+   end
+   for _, s in ipairs {"%*"} do
+      text = text:gsub(s .. "([^%s_])" .. s, "<em>%1</em> ")
+      text = text:gsub(s .. "(<strong>[^%s_]</strong>)" ..s, "<em>%1</em>")
       text = text:gsub(s .. "([^%s_][^<>_]-[^%s_])" .. s, "<em>%1</em>")
       text = text:gsub(s .. "([^<>_]-<strong>[^<>_]-</strong>[^<>_]-)" .. s, "<em>%1</em>")
    end