gnu: Add ruby-latex-decode.

* gnu/packages/patches/ruby-latex-decode-fix-test.patch: New file.
* gnu/local.mk (dist_patch_DATA): Add it.
* gnu/packages/ruby.scm (ruby-latex-decode): New variable.

Signed-off-by: Ludovic Courtès <ludo@gnu.org>

2 files changed, 118 insertions, 0 deletions

diff --git a/gnu/packages/patches/ruby-latex-decode-fix-test.patch b/gnu/packages/patches/ruby-latex-decode-fix-test.patch
new file mode 100644
index 0000000000..58d7d16258
--- /dev/null
+++ b/gnu/packages/patches/ruby-latex-decode-fix-test.patch
@@ -0,0 +1,66 @@
+From 698e39a251d433e2cd2934586a2dfc5e6582c4f9 Mon Sep 17 00:00:00 2001
+From: Philip McGrath <philip@philipmcgrath.com>
+Date: Sun, 24 Jul 2022 09:12:55 -0400
+Subject: [PATCH] Improve readability of whitespace tests
+
+This change fixes the test for `~x` in some environments in which it
+was previously failing.
+
+Related to https://github.com/inukshuk/latex-decode/issues/13
+---
+ features/step_definitions/latex.rb |  5 +++++
+ features/symbols.feature           |  7 -------
+ features/whitespace.feature        | 13 +++++++++++++
+ 3 files changed, 18 insertions(+), 7 deletions(-)
+ create mode 100644 features/whitespace.feature
+
+diff --git a/features/step_definitions/latex.rb b/features/step_definitions/latex.rb
+index 4b1b595..fca63c0 100644
+--- a/features/step_definitions/latex.rb
++++ b/features/step_definitions/latex.rb
+@@ -5,3 +5,8 @@ end
+ Then /^the result should be ('|")(.*)\1$/ do |quote,value|
+   expect(@result).to eq(value)
+ end
++
++# For whitespace or other characters that are hard to read in source code:
++Then /^I should get '([^']*)' \+ U\+(\h{4,}) \+ '([^']*)'$/ do |pre,code,post|
++  expect(@result).to eq(pre + code.hex.chr(Encoding::UTF_8) + post)
++end
+diff --git a/features/symbols.feature b/features/symbols.feature
+index 46d9a51..f8a5e84 100644
+--- a/features/symbols.feature
++++ b/features/symbols.feature
+@@ -12,10 +12,3 @@ Feature: Decode LaTeX symbol directives
+     | \\uparrow            | ↑       |
+     | \\downarrow          | ↓       |
+     | \\rightarrow         | →       |
+-
+-  Scenarios: Whitespace
+-    | latex | unicode | description        |
+-    | x\\,x | x x     | small space        |
+-    | x~x   | x x     | non-breaking space |
+-    | ~x    |  x      | non-breaking space |
+-
+diff --git a/features/whitespace.feature b/features/whitespace.feature
+new file mode 100644
+index 0000000..fe5b208
+--- /dev/null
++++ b/features/whitespace.feature
+@@ -0,0 +1,13 @@
++Feature: Decode LaTeX whitespace directives
++  As a hacker who works with LaTeX
++  I want to be able to decode LaTeX whitespace
++
++  Scenario Outline: LaTeX to Unicode transformation
++    When I decode the string '<latex>'
++    Then I should get <pre> + U+<code> + <post>
++
++  Scenarios: Whitespace
++    | latex | pre | code | post | description                |
++    | x~x   | 'x' | 00A0 | 'x'  | non-breaking space         |
++    | ~y    | ''  | 00A0 | 'y'  | leading non-breaking space |
++    | z\\,z | 'z' | 2009 | 'z'  | small space                |
+-- 
+2.32.0
+
diff --git a/gnu/packages/ruby.scm b/gnu/packages/ruby.scm
index 746bb9b6ce..5111b9ae2b 100644
--- a/gnu/packages/ruby.scm
+++ b/gnu/packages/ruby.scm
@@ -13111,3 +13111,55 @@ adaptation of TeX math syntax for web display.  Ritex makes inserting math
 into HTML pages easy.  It supports most TeX math syntax as well as macros.")
     ;; doesn't clearly state -only vs -or-later
     (license license:gpl2)))
+
+(define-public ruby-latex-decode
+  (package
+    (name "ruby-latex-decode")
+    (version "0.4.0")
+    (source (origin
+              (method git-fetch)
+              (uri (git-reference
+                    (url "https://github.com/inukshuk/latex-decode")
+                    (commit (string-append "v" version))))
+              (sha256
+               (base32
+                "1f5j67ayd04pjkmzvn0hk7cr8yqvn0gyg9ns6a0vhzj2gwna9ihy"))
+              (file-name (git-file-name name version))))
+    (build-system ruby-build-system)
+    (native-inputs
+     (list ruby-cucumber
+           ruby-ritex
+           ruby-rspec))
+    (arguments
+     (list
+      #:test-target "cucumber"
+      #:phases
+      #~(modify-phases %standard-phases
+          (add-after 'extract-gemspec 'avoid-bundler
+            (lambda args
+              (substitute* "Rakefile"
+                (("require 'bundler" orig)
+                 (string-append "# " orig " # patched for Guix"))
+                (("Cucumber::Rake::Task\\.new[(]:cucumber[)]" orig)
+                 (string-append orig " do |c|\n"
+                                "  c.bundler = false # patched for Guix\n"
+                                "end"))
+                (("Bundler\\.setup" orig)
+                 (string-append "true # " orig " # patched for Guix")))
+              (substitute* "cucumber.yml"
+                ;; thanks to avoiding bundler, we can't use this option
+                ((" --publish-quiet")
+                 ""))))
+          (add-after 'replace-git-ls-files 'replace-another-git-ls-files
+            (lambda args
+              (substitute* "latex-decode.gemspec"
+                (("`git ls-files -- [{]test,spec,features[}]/\\*`")
+                 "`find {test,spec,features} -type f | sort`")))))))
+    (home-page "https://github.com/inukshuk/latex-decode")
+    (synopsis "Convert LaTeX to Unicode")
+    (description
+     "This package provides a gem to convert LaTeX input to Unicode.  Its
+original use was as an input filter for BibTeX-Ruby, but it can be used
+independently to decode LaTeX.  Many of the patterns used by this Ruby gem are
+based on François Charette's equivalent Perl module @code{LaTeX::Decode}.")
+    (license license:gpl3+)))