Do auto-unwrapping of multi-byte chars only for Chinese locale

Fixes #300.

doc/ox-hugo-manual.org

@@ -2828,6 +2828,29 @@ Here's an example of inlining an SVG:
 #+caption: An SVG with hyperlink
 #+attr_html: :inlined t
 [[file:../test/site/content-org/images/svg-with-hyperlinks.svg]]
+*** Chinese Support
+:PROPERTIES:
+:EXPORT_FILE_NAME: chinese-support
+:END:
+**** Auto-unwrapping of lines with multi-byte characters
+This issue came up on this [[https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547][emacs-china thread]] .. the issue was that
+when consecutive lines in Org source had Chinese characters, in HTML
+the last character on one line got separated from the first character
+on the next line by a space, which is not grammatically correct in
+Chinese.
+
+So in such cases, those lines must be unwrapped _without any spaces_
+to separate those characters across the lines.
+
+That of course would not be grammatically correct in English and even
+other languages with multi-byte characters (few examples: Hindi,
+Gujarati).
+
+So that line-unwrapping _with space removal_ is done *only if*,
+1. The /locale/ is [[https://www.gnu.org/software/gettext/manual/html_node/Locale-Environment-Variables.html][auto-detected]] to be Chinese via environment
+   variables /LANGUAGE/, /LC_ALL/ or /LANG/, or
+2. The /locale/ is *manually set* to Chinese by setting it to *zh*
+   using ~#+hugo_locale:~ keyword (or ~EXPORT_HUGO_LOCALE~ property).
 *** COMMENT Hugo Bundle
 :PROPERTIES:
 :EXPORT_FILE_NAME: hugo-bundle

ox-hugo.el

@@ -1635,6 +1635,35 @@ INFO is a plist used as a communication channel."
     ;; (message "[ox-hugo section-path DBG] section path: %S" section-path)
     section-path))
 
+;;;; Get Language
+(defun org-hugo--get-lang (info)
+  "Return the language used for the content.
+
+The returned value is a string that can consist of only English
+alphabets and an underscore.
+
+The first 2 characters of this string is a language codes as per
+ISO 639-1 standard.  See
+https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes.
+
+INFO is a plist used as a communication channel."
+  (let ((lang (plist-get info :lang-iso-code)))
+    (unless lang
+      (setq lang
+            (or (plist-get info :hugo-locale)
+                ;; https://www.gnu.org/software/gettext/manual/html_node/Locale-Environment-Variables.html
+                (getenv "LANGUAGE")
+                (getenv "LC_ALL")
+                (getenv "LANG")))
+      (when (stringp lang)
+        (setq lang
+              (replace-regexp-in-string "\\`\\([a-z]+_[A-Z]+\\).*\\'" "\\1" lang)))
+      (setq lang (org-string-nw-p lang))
+      (when lang
+        ;; (message "[org-hugo--get-lang DBG] language: %s" lang)
+        (plist-put info :lang-iso-code lang)))
+    lang))
+
 
 
 ;;; Transcode Functions
@@ -2434,15 +2463,18 @@ communication channel."
         ret)
 
     ;; (message "[org-hugo-paragraph DBG] para 1: %s" contents)
+
     ;; Join consecutive Chinese lines into a single long line without
     ;; unwanted space inbetween.
-    ;; https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5
-    ;; Example: 这是一个测试     -> 这是一个测试文本 ("This is a test text")
-    ;;          文本
-    (setq contents (replace-regexp-in-string
-                    "\\([[:multibyte:]]\\)[[:blank:]]*\n[[:blank:]]*\\([[:multibyte:]]\\)" "\\1\\2"
-                    contents))
-    ;; (message "[org-hugo-paragraph DBG] para 2: %s" contents)
+    (when (member (org-hugo--get-lang info) '("zh"))
+      ;; https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5
+      ;; Example: 这是一个测试     -> 这是一个测试文本 ("This is a test text")
+      ;;          文本
+      (setq contents (replace-regexp-in-string
+                      "\\([[:multibyte:]]\\)[[:blank:]]*\n[[:blank:]]*\\([[:multibyte:]]\\)" "\\1\\2"
+                      contents))
+      ;; (message "[org-hugo-paragraph DBG] para 2: %s" contents)
+      )
 
     (unless (org-hugo--plist-get-true-p info :hugo-preserve-filling)
       (setq contents (concat (mapconcat 'identity (split-string contents) " ") "\n")))
@@ -3154,14 +3186,7 @@ INFO is a plist used as a communication channel."
          (creator (and (plist-get info :with-creator)
                        (plist-get info :creator)))
          (locale (and (plist-get info :hugo-with-locale)
-                      (let* ((lang (or (plist-get info :hugo-locale)
-                                       ;; https://www.gnu.org/software/gettext/manual/html_node/Locale-Environment-Variables.html
-                                       (getenv "LANGUAGE")
-                                       (getenv "LC_ALL")
-                                       (getenv "LANG")))
-                             (lang (when (stringp lang)
-                                     (replace-regexp-in-string "\\`\\([a-z]+_[A-Z]+\\).*\\'" "\\1" lang))))
-                        lang)))
+                      (org-hugo--get-lang info)))
          (description (org-string-nw-p (plist-get info :description)))
          (aliases-raw (let ((aliases-raw-1 (org-string-nw-p (plist-get info :hugo-aliases))))
                         (when aliases-raw-1

test/site/content-org/all-posts.org

@@ -480,7 +480,7 @@ get escaped.. =foo_bar= must not become =foo\_bar=.
 :END:
 This post will be exported without =title= in the front-matter because
 it is explicitly set to /empty/ using =:EXPORT_TITLE:=.
-** En dash --, Em dash ---, Horizontal ellipsis ... in titles     :@upstream:
+** En dash --, Em dash ---, Horizontal ellipsis ... in titles          :@upstream:
 :PROPERTIES:
 :EXPORT_FILE_NAME: en-dash-em-dash-hellip-in-titles
 :END:
@@ -4938,8 +4938,10 @@ current subtree.
 abc
 def
 ghi
-这是一个测试
-文本
+
+{{{oxhugoissue(300)}}} --
+Тест
+Тест
 ** Filling is not preserved
 :PROPERTIES:
 :EXPORT_FILE_NAME: filling-is-not-preserved
@@ -4948,8 +4950,43 @@ ghi
 abc
 def
 ghi
+** Chinese locale                                            :chinese:locale:
+:PROPERTIES:
+:EXPORT_HUGO_LOCALE: zh
+:END:
+*** Filling automatically not preserved for Chinese characters (preserve filling on)
+:PROPERTIES:
+:EXPORT_FILE_NAME: filling-not-preserved-for-chinese-characters--preserve-filling-on
+:EXPORT_HUGO_PRESERVE_FILLING: t
+:END:
+#+begin_description
+Ensure that multi-byte characters are force-unwrapped if the locale is
+manually set or auto-detected as Chinese.
+#+end_description
+abc
+def
+ghi
 这是一个测试
 文本
+
+[[https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5][Reference]]
+*** Filling automatically not preserved for Chinese characters (preserve filling off)
+:PROPERTIES:
+:EXPORT_FILE_NAME: filling-not-preserved-for-chinese-characters--preserve-filling-off
+:EXPORT_HUGO_PRESERVE_FILLING:
+:END:
+#+begin_description
+Ensure that even when auto-unwrapping is enabled, no space is kept
+between the unwrapped multi-byte characters if the locale is manually
+set or auto-detected as Chinese.
+#+end_description
+abc
+def
+ghi
+这是一个测试
+文本
+
+[[https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5][Reference]]
 * Section Inheritance                                   :section_inheritance:
 ** Section A
 :PROPERTIES:
@@ -6360,6 +6397,7 @@ y slug para múltiples publicaciones, pero en diferentes idiomas.
 :EXPORT_AUTHOR: Feng Ruohang
 :EXPORT_HUGO_CUSTOM_FRONT_MATTER+: :source https://github.com/Vonng/Math/blob/master/nndl/nn-intro.md
 :EXPORT_DATE: 2017-11-29
+:EXPORT_HUGO_LOCALE: zh
 :END:
 - Disclaimer :: This post is from the [[https://github.com/Vonng/Math/blob/master/nndl/nn-intro.md][link]] posted by GitHub user
                 [[https://github.com/Vonng][*Vonng*]] in [[https://github.com/gohugoio/hugo/issues/234#issuecomment-347725391][this comment]]. All credit for this post

test/site/content/posts/filling-is-not-preserved.md

@@ -4,4 +4,4 @@ tags = ["filling"]
 draft = false
 +++
 
-abc def ghi 这是一个测试文本
+abc def ghi

test/site/content/posts/filling-is-preserved.md

@@ -7,4 +7,7 @@ draft = false
 abc
 def
 ghi
-这是一个测试文本
+
+`ox-hugo` Issue #[300](https://github.com/kaushalmodi/ox-hugo/issues/300) --
+Тест
+Тест

test/site/content/posts/filling-not-preserved-for-chinese-characters--preserve-filling-off.md

@@ -0,0 +1,10 @@
++++
+title = "Filling automatically not preserved for Chinese characters (preserve filling off)"
+description = "Ensure that even when auto-unwrapping is enabled, no space is kept between the unwrapped multi-byte characters if the locale is manually set or auto-detected as Chinese."
+tags = ["filling", "chinese", "locale"]
+draft = false
++++
+
+abc def ghi 这是一个测试文本
+
+[Reference](https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5)

test/site/content/posts/filling-not-preserved-for-chinese-characters--preserve-filling-on.md

@@ -0,0 +1,16 @@
++++
+title = "Filling automatically not preserved for Chinese characters (preserve filling on)"
+description = """
+  Ensure that multi-byte characters are force-unwrapped if the locale is
+  manually set or auto-detected as Chinese.
+  """
+tags = ["filling", "chinese", "locale"]
+draft = false
++++
+
+abc
+def
+ghi
+这是一个测试文本
+
+[Reference](https://emacs-china.org/t/ox-hugo-auto-fill-mode-markdown/9547/5)