Escape angle brackets in Android formatter, if the string contains a placeholder to enable users to use the official way to retrieve these strings. Improves #212.

5 files changed, 33 insertions, 12 deletions

diff --git a/CHANGELOG.md b/CHANGELOG.md
index f5b445b..a29a8bc 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,7 @@
+# next version
+
+- Improvement: Better support for placeholders in HTML styled Android strings (#212)
+
 # 1.0.1 (2017-10-17)
 
 - Bugfix: Always prefer the passed-in formatter (#221)
diff --git a/README.md b/README.md
index 8618a69..1e97111 100644
--- a/README.md
+++ b/README.md
@@ -78,7 +78,12 @@ Twine currently supports the following output formats:
 
 * [iOS and OS X String Resources][applestrings] (format: apple)
 * [Android String Resources][androidstrings] (format: android)
-    * Supports [basic styling][androidstyling] with \<b\>, \<i\>, \<u\> and \<a\> links. These tags will *not* be escaped. Use [`getText()`](https://developer.android.com/reference/android/content/res/Resources.html#getText(int)) to read these strings. Also tags inside `<![CDATA[` won't be escaped. See [\#212](https://github.com/scelis/twine/issues/212) for details.
+    * HTML tags will be escaped by replacing `<` with `&lt`
+    * Tags inside `<![CDATA[` won't be escaped.
+    * Supports [basic styling][androidstyling] with `<b>`, `<i>`, `<u>` and `<a>` links.
+	    * These tags will *not* be escaped, if the string doesn't contain placeholders so you can reference them directly in your layouts or use [`getText()`](https://developer.android.com/reference/android/content/res/Resources.html#getText(int)) to read them programatically. 
+	    * These tags *will* be escaped, if the string contains placeholders. You can use [`getString()`](https://developer.android.com/reference/android/content/res/Resources.html#getString(int,%20java.lang.Object...)) combined with [`fromHtml`](https://developer.android.com/reference/android/text/Html.html#fromHtml(java.lang.String)) as shown in the [documentation][androidstyling] to display them.
+	* See [\#212](https://github.com/scelis/twine/issues/212) for details.
 * [Gettext PO Files][gettextpo] (format: gettext)
 * [jquery-localize Language Files][jquerylocalize] (format: jquery)
 * [Django PO Files][djangopo] (format: django)
diff --git a/lib/twine/formatters/android.rb b/lib/twine/formatters/android.rb
index ced90b7..55c735c 100644
--- a/lib/twine/formatters/android.rb
+++ b/lib/twine/formatters/android.rb
@@ -116,10 +116,15 @@ module Twine
         value = gsub_unless(value, "'", "\\'") { |substring| substring =~ inside_cdata }
         value = gsub_unless(value, /&/, '&amp;') { |substring| substring =~ inside_cdata || substring =~ inside_opening_anchor_tag }
 
-        # escape opening angle brackes unless it's a supported styling tag
+        # if `value` contains a placeholder, escape all angle brackets
+        # if not, escape opening angle brackes unless it's a supported styling tag
         # https://github.com/scelis/twine/issues/212
         # https://stackoverflow.com/questions/3235131/#18199543
-        angle_bracket = /<(?!(\/?(b|u|i|a|\!\[CDATA)))/   # matches all `<` but <b>, <u>, <i>, <a> and <![CDATA
+        if number_of_twine_placeholders(value) > 0
+          angle_bracket = /<(?!(\/?(\!\[CDATA)))/           # matches all `<` but <![CDATA
+        else  
+          angle_bracket = /<(?!(\/?(b|u|i|a|\!\[CDATA)))/   # matches all `<` but <b>, <u>, <i>, <a> and <![CDATA
+        end
         value = gsub_unless(value, angle_bracket, '&lt;') { |substring| substring =~ inside_cdata }
 
         # escape non resource identifier @ signs (http://developer.android.com/guide/topics/resources/accessing-resources.html#ResourcesFromXml)
diff --git a/lib/twine/placeholders.rb b/lib/twine/placeholders.rb
index 1c7b200..bc44d05 100644
--- a/lib/twine/placeholders.rb
+++ b/lib/twine/placeholders.rb
@@ -6,6 +6,11 @@ module Twine
     PLACEHOLDER_FLAGS_WIDTH_PRECISION_LENGTH = '([-+0#])?(\d+|\*)?(\.(\d+|\*))?(hh?|ll?|L|z|j|t|q)?'
     PLACEHOLDER_PARAMETER_FLAGS_WIDTH_PRECISION_LENGTH = '(\d+\$)?' + PLACEHOLDER_FLAGS_WIDTH_PRECISION_LENGTH
     PLACEHOLDER_TYPES = '[diufFeEgGxXoscpaA]'
+    PLACEHOLDER_REGEX = /%#{PLACEHOLDER_PARAMETER_FLAGS_WIDTH_PRECISION_LENGTH}#{PLACEHOLDER_TYPES}/
+
+    def number_of_twine_placeholders(input)
+      input.scan(PLACEHOLDER_REGEX).size
+    end
 
     def convert_twine_string_placeholder(input)
       # %@ -> %s
@@ -19,15 +24,13 @@ module Twine
       # %@ -> %s
       value = convert_twine_string_placeholder(input)
 
-      placeholder_syntax = PLACEHOLDER_PARAMETER_FLAGS_WIDTH_PRECISION_LENGTH + PLACEHOLDER_TYPES
-      placeholder_regex = /%#{placeholder_syntax}/
-
-      number_of_placeholders = value.scan(placeholder_regex).size
+      number_of_placeholders = number_of_twine_placeholders(input)
 
       return value if number_of_placeholders == 0
 
       # got placeholders -> need to double single percent signs
       # % -> %% (but %% -> %%, %d -> %d)
+      placeholder_syntax = PLACEHOLDER_PARAMETER_FLAGS_WIDTH_PRECISION_LENGTH + PLACEHOLDER_TYPES
       single_percent_regex = /([^%])(%)(?!(%|#{placeholder_syntax}))/
       value.gsub! single_percent_regex, '\1%%'
 
@@ -61,8 +64,7 @@ module Twine
     def convert_placeholders_from_twine_to_flash(input)
       value = convert_twine_string_placeholder(input)
 
-      placeholder_regex = /%#{PLACEHOLDER_PARAMETER_FLAGS_WIDTH_PRECISION_LENGTH}#{PLACEHOLDER_TYPES}/
-      value.gsub(placeholder_regex).each_with_index do |match, index|
+      value.gsub(PLACEHOLDER_REGEX).each_with_index do |match, index|
         "{#{index}}"
       end
     end
diff --git a/test/test_formatters.rb b/test/test_formatters.rb
index 87658b7..51d3c51 100644
--- a/test/test_formatters.rb
+++ b/test/test_formatters.rb
@@ -50,6 +50,10 @@ class TestAndroidFormatter < FormatterTest
       '<i>italic</i>'             => '<i>italic</i>',
       '<u>underline</u>'          => '<u>underline</u>',
 
+      '<b>%@</b>'                 => '&lt;b>%s&lt;/b>',
+      '<i>%@</i>'                 => '&lt;i>%s&lt;/i>',
+      '<u>%@</u>'                 => '&lt;u>%s&lt;/u>',
+
       '<span>inline</span>'       => '&lt;span>inline&lt;/span>',
       '<p>paragraph</p>'          => '&lt;p>paragraph&lt;/p>',
 
@@ -58,9 +62,10 @@ class TestAndroidFormatter < FormatterTest
       '<a href="target"></a>"out"'    => '<a href="target"></a>\"out\"',
       '<a href="http://url.com?param=1&param2=3&param3=%20">link</a>'   =>   '<a href="http://url.com?param=1&param2=3&param3=%20">link</a>',
 
-      '<p>escaped</p><![CDATA[]]>'    => '&lt;p>escaped&lt;/p><![CDATA[]]>',
-      '<![CDATA[]]><p>escaped</p>'    => '<![CDATA[]]>&lt;p>escaped&lt;/p>',
-      '<![CDATA[<p>unescaped</p>]]>'  => '<![CDATA[<p>unescaped</p>]]>',
+      '<p>escaped</p><![CDATA[]]>'                => '&lt;p>escaped&lt;/p><![CDATA[]]>',
+      '<![CDATA[]]><p>escaped</p>'                => '<![CDATA[]]>&lt;p>escaped&lt;/p>',
+      '<![CDATA[<p>unescaped</p>]]>'              => '<![CDATA[<p>unescaped</p>]]>',
+      '<![CDATA[<p>unescaped with %@</p>]]>'      => '<![CDATA[<p>unescaped with %s</p>]]>',
       '<![CDATA[]]><![CDATA[<p>unescaped</p>]]>'  => '<![CDATA[]]><![CDATA[<p>unescaped</p>]]>',
 
       '<![CDATA[&]]>'  => '<![CDATA[&]]>',