Improved robustness of the emoji feature

Changes:
- Use of `@emoji name` instead of `:name:`
- Support only GitHub emojis (i.e. without spaces or special characters in the name)
- Provided script to download images for LaTeX support.
- XML output now has <emoji> tag with name an unicode sequence.

doc/CMakeLists.txt

@@ -95,6 +95,7 @@ set(DOC_FILES
         language.tpl
         maintainers.txt
         translator.py
+        smile.png
 )
 file(GLOB LANG_FILES "${TOP}/src/translator_??.h")
 

doc/Doxyfile

@@ -55,3 +55,4 @@ ALIASES           = LaTeX="\f$\mbox{\LaTeX}\f$"
 ALIASES          += TeX="\f$\mbox{\TeX}\f$"
 LATEX_BATCHMODE   = YES
 LATEX_EXTRA_STYLESHEET = manual.sty
+LATEX_EMOJI_DIRECTORY  = ../doc

doc/commands.doc

@@ -75,6 +75,7 @@ documentation:
 \refitem cmdelse \\else
 \refitem cmdelseif \\elseif
 \refitem cmdem \\em
+\refitem cmdemoji \\emoji
 \refitem cmdendcode \\endcode
 \refitem cmdendcond \\endcond
 \refitem cmdenddocbookonly \\enddocbookonly
@@ -218,7 +219,6 @@ documentation:
 \refitem cmdperc \\\%
 \refitem cmdquot \\\"
 \refitem cmdchardot \\\.
-\refitem cmdcolon \:
 \refitem cmddcolon \::
 \refitem cmdpipe \\|
 \refitem cmdndash \\\--
@@ -2742,6 +2742,22 @@ class C {};
  */
 \endcode
 
+<hr>
+\section cmdemoji \\emoji "name"
+
+This command will produce an emoji character given its name.
+
+The supported names are the ones also supported by GitHub and listed here
+https://gist.github.com/rxaviers/7360908
+
+You can use the name with or without colons, i.e.
+`\emoji smile` is the same as writing `\emoji :smile:`.
+When an emoji is not supported the name with by places in the
+text with in between colons, i.e. `\emoji unsupported` will produce
+`:unsupported:` in the output. Doxygen will also give a warning message.
+
+See also the \ref emojisup "emoji support page" for details.
+
 <hr>
 \section cmdmsc \\msc ["caption"] [<sizeindication>=<size>]
 
@@ -3445,14 +3461,6 @@ class Receiver
   or to prevent starting a numbered list when the dot follows a number at
   the start of a line.
 
-<hr>
-\section cmdcolon \:
-
-  \addindex \\:
-  This command writes a single colon (\c \:) to the output. This
-  character sequence has to be escaped in some cases, because it is used
-  to define `emoji` see also \ref emojisup "Emoji support".
-
 <hr>
 \section cmddcolon \\::
 

doc/doxygen_manual.tex

@@ -72,6 +72,8 @@
 \setlength{\parindent}{0pt}
 \newcommand{\doxynormalparskip}{\setlength{\parskip}{0.2cm}}
 \newcommand{\doxytocparskip}{\setlength{\parskip}{0.2cm}}
+\newcommand{\doxygenemoji}[2]{%
+\IfFileExists{../doc/#2.png}{\raisebox{-0.1em}{\includegraphics[height=0.9em]{../doc/#2.png}}}{#1}}
 \doxynormalparskip
 \hbadness=750
 \tolerance=750

doc/emojisup.doc

@@ -18,21 +18,12 @@
 
 The [Unicode consortium](http://www.unicode.org/) has defined a set of
 [emoji](https://en.wikipedia.org/wiki/Emoji) with the corresponding unicode
-sequences and a so called "CLDR short name". The current version a v11.0 and can be found at
-[Full Emoji List, v11.0](https://unicode.org/emoji/charts/full-emoji-list.html) furthermore there is the list with 
-[Full Emoji Modifier Sequences, v11.0](http://www.unicode.org/emoji/charts/full-emoji-modifiers.html).
+sequences. Doxygen supports the subset of emoji characters as used by GitHub (based on the list
+https://api.github.com/emojis).
+An emoji is created using the \ref cmdemoji "\\emoji" command.
+For example `\emoji smile` (or `\emoji :smile:`) both produce \emoji smile.
 
-A common way to denote an emoji is by means of `:<text>:`,
-doxygen supports the emoji as mentioned in the above mentioned unicode emoji lists in this way
-by means of the "CLDR short name" with the exception that in case a colon (`:`) is in the
-"CLDR short name" this colon has to be removed.
-Furthermore doxygen supports the list of emoji as used by github (based on the list 
-https://api.github.com/emojis). In this list also a reference is given to the unicode codes (just the
-first and last) and these unicodes are mapped onto the official unicode sequences.
-In case the "CLDR short name" and the "github name" are the same the reference from the 
-"CLDR short name" has precedence.
-
-Implementation
+\section emojirep Representation
 
 For the different doxygen output types there is an output defined:
 - Unicode code sequence, the actual representation is depending on the possibilities of the fonts loaded:
@@ -46,116 +37,79 @@ For the different doxygen output types there is an output defined:
   - man
   - perl
 
-\anchor emojiimage Emoji image retrieval
+\section emojiimage Emoji image retrieval
 
-In the  lists 
-[Full Emoji List, v11.0](https://unicode.org/emoji/charts/full-emoji-list.html) and
-[Full Emoji Modifier Sequences, v11.0](http://www.unicode.org/emoji/charts/full-emoji-modifiers.html).
-define images for the different vendors. These images can be retrieved by means of the following procedure (based on the code from Henning Pohl, https://github.com/henningpohl/latex-emoji):
+In the list of images can be downloaded via the following Python script:
 \code{.py}
-from bs4 import BeautifulSoup
-import base64
+# script to download the emoticons from GitHub and to produce a table for
+# inclusion in doxygen. Works with python 2.7+ and python 3.x
+import json
 import os
-import requests
-
-# http://www.unicode.org/emoji/charts/index.html
-# http://www.unicode.org/emoji/charts/full-emoji-list.html
-PAGE_URL = 'http://www.unicode.org/emoji/charts/full-emoji-list.html'
-PAGE_URL_SKIN = 'http://www.unicode.org/emoji/charts/full-emoji-modifiers.html'
-PAGE = 'full-emoji-list.html'
-PAGE_SKIN = 'full-emoji-modifiers.html'
-
-
-def get_header_names(header):
-    cols = header.find_all('th')
-    cols = [c.get_text() for c in cols]
-    cols = [c.replace('*','') for c in cols]
-    cols = [c.lower() for c in cols]
-    return cols
-
-def extract_image(column):
-    if 'miss' in column['class']:
-        return None
-
-    if 'miss7' in column['class']:
-        return None
-
-    data = column.img['src']
-    data_start = data.find("base64,")
-    if data_start == -1:
-        return None
-
-    data = base64.b64decode(data[data_start + len("base64,"):])
-    return data
-
-def save_image(folder, imgSrc, filename):
-    if os.path.exists(folder) is False:
-        os.mkdir(folder)
-
-    filename = os.path.join(folder, filename)
-    if os.path.exists(filename):
-        return
-
-    img = extract_image(imgSrc)
-    if img is not None:
-        with open(filename, 'wb') as out:
-            out.write(img)
+import argparse
+import re
+try:
+    import urllib.request as urlrequest
+except ImportError:
+    import urllib as urlrequest
+
+unicode_re = re.compile(r'.*?/unicode/(.*?).png\?.*')
+
+def get_emojis():
+    response  = urlrequest.urlopen('https://api.github.com/emojis')
+    raw_data  = response.read()
+    return json.loads(raw_data)
+
+def download_images(dir_name):
+    json_data = get_emojis()
+    num_items = len(json_data)
+    cur_item=0
+    for image,url in sorted(json_data.items()):
+        image_name = image+'.png'
+        cur_item=cur_item+1
+        if url.find('/unicode/')==-1 or not os.path.isfile(dir_name+'/'+image_name):
+            with open(dir_name+'/'+image_name,'wb') as file:
+                print('%s/%s: fetching %s' % (cur_item,num_items,image_name))
+                file.write(urlrequest.urlopen(url).read())
+        else:
+            print('%s/%s: skipping %s' % (cur_item,num_items,image_name))
+
+def produce_table():
+    json_data = get_emojis()
+    lines = []
+    for image,url in sorted(json_data.items()):
+        match = unicode_re.match(url)
+        if match:
+            unicodes = match.group(1).split('-')
+            unicodes_html = ''.join(["&#x"+x+";" for x in unicodes])
+            image_str = "\":"+image+":\","
+            unicode_str = "\""+unicodes_html+"\""
+            lines.append('  { %-42s %-38s }' % (image_str,unicode_str))
+    out_str = ',\n'.join(lines)
+    print("{")
+    print(out_str)
+    print("};")
+
+if __name__=="__main__":
+    parser = argparse.ArgumentParser()
+    group = parser.add_mutually_exclusive_group()
+    group.add_argument('-d','--dir',help='directory to place images in')
+    group.add_argument('-t','--table',help='generate code fragment',action='store_true')
+    args = parser.parse_args()
+    if args.table:
+        produce_table()
+    else:
+        download_images(args.dir)
 
-def scrape(page_url, page):
-    # Possibilities to obtain the basic data:
-    # - use request.get directly
-    soup = BeautifulSoup(requests.get(page_url).text, "html5lib")
-    # - download file (e.g. with wget http://www.unicode.org/emoji/charts/full-emoji-list.html)
-    # with open(page) as fp:
-    #     soup = BeautifulSoup(fp,"html5lib")
-
-    table = soup('table')[0]
-
-    # for version 11.0
-    # first row: smileys
-    # second row: face smileys
-    # third row: row with vendors, i.e. the one we want
-    header = table.find_all('tr')[2]
-    keys = get_header_names(header)
-
-    for row in header.find_next_siblings('tr'):
-        fields = {k:c for k, c in zip(keys, row.find_all('td')) }
-        if 'code' not in fields:
-            continue
-
-        codes = fields['code'].text.replace('U+', '').split(' ')
-        filename = "-".join(codes) + ".png"
-
-        save_image('ios', fields['appl'], filename)
-        save_image('android', fields['goog'], filename)
-        save_image('twitter', fields['twtr'], filename)
-        save_image('windows', fields['wind'], filename)
-        save_image('one', fields['one'], filename)
-        save_image('facebook', fields['fb'], filename)
-        save_image('samsung', fields['sams'], filename)
-        #save_image('gmail', fields['gmail'], filename)
-        #save_image('softbank', fields['sb'], filename)
-        #save_image('docomo', fields['dcm'], filename)
-        #save_image('kddi', fields['kddi'], filename)
-        #save_image('bw', fields['chart'], filename)
-
-if __name__ == '__main__':
-    scrape(PAGE_URL, PAGE)
-    scrape(PAGE_URL_SKIN, PAGE_SKIN)
 \endcode
-This results in a number of directories with the supported images. By means of the doxygen configuration parameter
+When invoking it with the `-d image_dir` option the images will by downloaded in the `image_dir` directory.
+By means of the doxygen configuration parameter
 \ref cfg_latex_emoji_directory "LATEX_EMOJI_DIRECTORY" the requested directory can be selected.
 
-It is also possible to use images from other sources or mix images from different sources, the only requirement is that the filename represents the unicode of the emoji. e.g. if we have the emoji <tt>\:grinning face with big eyes\:</tt> (also known as <tt>\:smiley\:</tt>) the coresponding unicode is `U+1F603` and the name of the file is `1F603.png`.<br>
-For a more complex emoji like <tt>\:keycap 1\:</tt> (also known as <tt>\:one\:</tt>) the coresponding unicode sequence is `U+0031U+FE0FU+20E3` and the name of the file is `0031-FE0F-20E3.png`.
-
-
-Note that when you want to use a colon (`:`) in your text it might be necessary to escape the colon (see \ref cmdcolon "\\:") as it might conflict with a, possible, emoji sequence.
-
+For convenience a zip with the result of running the script can also be downloaded from 
+http://www.doxygen.nl/dl/github_emojis.zip
 
 For a overview of the supported emoji one can issue the comand:<br>
-`doxygen.exe -f emoji <outputFileName>`
-
+`doxygen -f emoji <outputFileName>`
 
 \htmlonly
 Go to the <a href="langhowto.html">next</a> section or return to the

src/cmdmapper.cpp

@@ -120,7 +120,6 @@ CommandMap cmdMap[] =
   { ".",             CMD_PUNT },
   { "+",             CMD_PLUS },
   { "-",             CMD_MINUS },
-  { ":",             CMD_COLON },
   { "::",            CMD_DCOLON },
   { "\"",            CMD_QUOTE },
   { "_internalref",  CMD_INTERNALREF },
@@ -148,6 +147,7 @@ CommandMap cmdMap[] =
   { "--",            CMD_NDASH },
   { "---",           CMD_MDASH },
   { "_setscope",     CMD_SETSCOPE },
+  { "emoji",         CMD_EMOJI },
   { 0,               0 },
 };
 

src/cmdmapper.h

@@ -137,7 +137,7 @@ enum CommandType
   CMD_INCLUDEDOC   = 107,
   CMD_SNIPPETDOC   = 108,
   CMD_SNIPWITHLINES= 109,
-  CMD_COLON        = 110
+  CMD_EMOJI        = 110
 };
 
 enum HtmlTagType

src/docbookvisitor.cpp

@@ -182,14 +182,14 @@ void DocbookDocVisitor::visit(DocEmoji *s)
 {
 DB_VIS_C
   if (m_hide) return;
-  const char *res = EmojiEntityMapper::instance()->docbook(s->emoji());
+  const char *res = EmojiEntityMapper::instance()->unicode(s->index());
   if (res)
   {
     m_t << res;
   }
   else
   {
-    err("DocBook: non supported Emoji-entity found: %s\n",EmojiEntityMapper::instance()->html(s->emoji()));
+    m_t << s->name();
   }
 }