JarCastingCharacter and HTML Element Translations
Current release is 0.0.2. All releases are on Maven Central.
This blog post explains how to get UTF-8 support in the whole tool chain blog on UTF-8. Updates will be at here SKB Wiki on UTF-8.
The Problem
Today, as in 2016, UTF-8 should be standard when dealing with text. However, there are (still, and probably for a long time to come) many use cases where non-UTF-8 text needs to be processed. Some examples are
-
Text in databases or other sources might be using ASCII encoding and HTML entities,
-
Many text written for LaTeX is using 7-bit ASCII encoding with special commands for non-ASCII characters, and
-
Legacy HTML text might use HTML Entities rather than UTF-8 encoded characters.
When dealing with multiple targets (for instance LaTex, AsciiDoc, and HTML), character translation can become a nightmare. When defining the format of a normative source, all source representations must be translated to the required target representations.
Those translations can be very tricky, since they might require many target-specific exceptions.
This Solution
Use the SKB data for character maps and HTML Elements (SKB on Github) plus the SKB DataTool (DataTool on Github) to generate several translators that (hopefully) will ease the translation problems described above. The SKB data provides for character maps and HTML Element maps. Those maps are then used by the SKB DataTool to generate several Java classes with pre-defined mappings for this package.
There are three assumptions to translate a normative text as source to a target: characters, text formatting, and a combination of both.
Characters Translations
We assume that all characters are written in UTF-8. For instance, to write a German umlaut ö one would one simply write ö. Those UTF-8 characters will need then to be translated to a proper target representation.
For any other UTF-8 bases target, the example ö will just be the same: ö. If the target requires a different representation, we need to translate the ö to the target, e.g.:
-
for an ASCII 7-bit representation in LaTeX we need to translate it to
\"{o}. -
for an ASCII 7-bit representation in HTML we need to translate it to
ö.
This package provides character translators for doing exactly that.
Text Formatting Translations
Beside characters, the normative text source should also include standard formatting of text, such as bold and italic. Simple text markup languages (such as AsciiDoc) and LaTeX use tags that are very hard to parse. HTML however uses formatting tags that can be easily parsed and translated.
For instance, to mark text as bold in HTML one would use <b> and </b>. Using this HTML markup, we can write text for instance as follows:
-
for bold, write
<b>text in bold</b>and translate it to LaTeX as\textbf{text in bold}and to AsciiDoc as +text in bold+, -
for italic, write
<i>text in italic</i>and translate it to LaTeX as\textit{text in italic}and to AsciiDoc as +text in italic+.
This package provides translators for doing exactly that.
Combination of Character and Text Formatting Translations
When combining both, character and text formatting translations, a few special cases do apply. If we would simply translate all characters from source to target we would lose the text formatting.
For instance, the text
ä ö <b>ü</b>
would be translated to LaTeX as
\"{a} \"{o} \textlessb\textgreater\"{u}\textless/b\textgreater
To keep the text formatting, we first need to convert those formatting markups into a representation that is not picked up by the simple character translation (a temporary form), then realize the character translation, and then translate the temporary form of the formatting markup to the target representation. This package provides translators for doing exactly that, resulting in a translated text of:
\"{a} \"{o} \textbf{\"{u}}
Not everything will get translated
As mentioned above, the character and formatting translators are automatically generated. While the data source (SKB on Github) defines quite a few translations, it also might (will) miss some required translations. Over time, we hope that all required translations will be defined in the data source.
Features
This package provides three different types of translators, each providing different translation classes for different target:
-
Simple character translators,
-
Simple formatting translators (using HTML Elements), and
-
Combined translators.
Character Translators
Character translators all provide a method translateCharacters(String input) translating all source character representations found in input to a target representation. The translations provided currently are:
-
Text to AsciiDoc,
-
Text to HTML,
-
Text to LaTeX,
-
HTML to AsciiDoc, and
-
HTML to LaTeX.
HTML Element (text formatting) Translators
HTML Element translators all provide methods for:
-
Translating a text to a temporary representation -
text2tmp(String input), -
Translating a temporary representation to a target representation -
tmp2target(String input), and -
Directly translating from source to target -
translateHtmlElements(String input).
The translations provided currently are:
-
Text to AsciiDoc,
-
Text to HTML, and
-
Text to LaTeX.
Combined Translators
Combined translators provide all methods from the two above described translator interaces plus a method for a combined translation called translate(String input).
The translations provided currently are:
-
Text to AsciiDoc,
-
Text to HTML, and
-
Text to LaTeX.
Examples
Character Translations
The following code will take a given string with some UTF-8 characters and translate it to a number of targets. The first line creates a UTF-8 string. The following lines print out translations to AsciiDoc, HTML, and LaTeX.
String text = "ä ö ü Š β … € ™ ↔"; System.out.println(new Text2AsciiDoc().translateCharacters(text)); System.out.println(new Text2Html().translateCharacters(text)); System.out.println(new Text2Latex().translateCharacters(text));
The output of the example will be as follows. Line one below shows the translation to AsciiDoc. Line two shows the translation to HTML. Line three shows the translation to LaTeX.
ä ö ü Š β … € ™ ↔
ä ö ü Š β … € ™ ↔
\"{a} \"{o} \"{u} \v{S} \beta {\dots} {\euro} {\texttrademark} \(\leftrightarrow{}\)
HTML Element (text formatting) Translations
The following code will take a given string with some formatting (HTML Elements) and translate it to a number of targets. The first line creates a string with HTML Elements used for formatting. The following lines print out translations to AsciiDoc, HTML, and LaTeX.
String text = "<b>bold</b>, <i>italic</i>, H<sub>2</sub>O, x<sup>y</sup>"; System.out.println(new de.vandermeer.translation.helements.Text2AsciiDoc().translateHtmlElements(text)); System.out.println(new de.vandermeer.translation.helements.Text2Html().translateHtmlElements(text)); System.out.println(new de.vandermeer.translation.helements.Text2Latex().translateHtmlElements(text));
The output of the example will be as follows. Line one below shows the translation to AsciiDoc. Line two shows the translation to HTML. Line three shows the translation to LaTeX.
*bold*, _italic_, H_2O, x^y
<b>bold</b>, <i>italic</i>, H<sub>2</sub>O, x<sup>y</sup>
\textbf{bold}, \textit{italic}, H$_{2}$O, x$^{y}$
Combined Translators
The following example will take a given string with character and formatting and translate it to a number of targets. The first line has the actual string with combined elements. The following lines print out translations to AsciiDoc, HTML, and LaTeX.
String text = "<b>bold ä ö ü</b>, <i>italic Š β …</i>, €<sub>5</sub>O, ™<sup>↔</sup>"; System.out.println(new de.vandermeer.translation.combinations.Text2AsciiDoc().translate(text)); System.out.println(new de.vandermeer.translation.combinations.Text2Html().translate(text)); System.out.println(new de.vandermeer.translation.combinations.Text2Latex().translate(text));
The output of the example will be as follows. Line one below shows the translation to AsciiDoc. Line two shows the translation to HTML. Line three shows the translation to LaTeX.
*bold ä ö ü*, _italic Š β …_, €_5O, ™^↔
<b>bold ä ö ü</b>, <i>italic Š β …</i>, €<sub>5</sub>O, ™<sup>↔</sup>
\textbf{bold \"{a} \"{o} \"{u}}, \textit{italic \v{S} \beta {\dots}}, {\euro}$_{5}$O, {\texttrademark}$^{\(\leftrightarrow{}\)}$