# ==============================================================================
# This file contains the following functions:
#       html__get_doctype
#       html_add_markup_brackets
#       html_convert_to_markup
#       html_ask_for_markup
#       html_make_A_NAME
#       html_make_A_HREF
#       html_entities
# ==============================================================================

NEDIT_require_macro_file("extensions.nm")
NEDIT_require_macro_file("array_utils.nm")

# ==============================================================================
# html__get_doctype(): looks in the current document for a <!DOCTYPE> directive,
#       stores the result in $html["doctype"][docname], and returns it. If this
#       has been done for the document, returns the previous result.
# ==============================================================================
define html__get_doctype
  {
  $html[""] = ""
  if ("doctype" in $html)
    doctype = $html["doctype"]
  else
    doctype = $empty_array

  full_name = $file_path $file_name
  if (full_name in doctype)
    return doctype[full_name]

  res = ""
  pos = search("(?n\\<!DOCTYPE>.*\\>", 0, "regex")
  if (pos > 0)
    res = get_range(pos, $search_end)

  doctype[full_name] = res
  $html["doctype"] = doctype
  return res
  }

# ==============================================================================
# html_add_markup_brackets(full_directive, isblock, beg, end): uses the string
#       parameter to either surround the current selection with markup (if there
#       is a selection) or to insert the markup as an open/close markup bracket
#       pair. The first word of the markup string is used to generate the markup
#       terminator.
# Parameters:
#       full_directive  The markup text (without <>).
#       isblock         True (non-zero) if newlines should be added.
#       beg             Position of selection start or cursor
#       end             Position of selection end
# ==============================================================================

define html_add_markup_brackets
  {
  full_directive = $1
  pos = search_string(full_directive, $sub_sep, 0)
  isblock = 0
  if (pos >= 0)
    {
    isblock = substring(full_directive, pos + 1)
    full_directive = substring(full_directive, 0, pos)
    }
  beg = $selection_start
  end = $selection_end
  if (beg >= 0 && $selection_left >= 0)
    return ""
  if (beg < 0)
    {
    beg = $cursor
    end = beg
    }
  if ($n_args > 1)
    isblock = $2
  if ($n_args > 2)
    {
    beg = $3
    end = beg
    }
  if ($n_args > 3)
    end = $4

  directive = replace_in_string(full_directive, "^\\s*(\\w\\S*).*", "\\1", \
                                "regex", "copy")

# directive_end = -1
# directive_start = search_string(full_directive, "[a-zA-Z0-1]", 0, "regex")
#
# if (directive_start >= 0)
#   directive_end = search_string(full_directive, ">", directive_start, "regex")
#
# if (directive_start >= 0 && directive_end >= 0)
#   directive = substring(full_directive, directive_start, directive_end)

  if (beg != end)
    {
    # we have a non-rectangular selection
    s = "<" full_directive ">" get_range(beg, end) "</" directive ">"
    replace_range(beg, end, s)
    end = beg + length(s)
    }
  else
    {
    # we do not have a selection
    if (!isblock)
      {
      replace_range(beg, end, "<" full_directive "></" directive ">")
      end = beg + length(full_directive) + 2
      }
    else
      {
      set_cursor_pos(beg)
      insert_string("<" full_directive ">")
      newline_and_indent()
      end = $cursor
      newline_and_indent()
      insert_string("</" directive ">")
      }
    }
  set_cursor_pos(end)
  return full_directive $sub_sep isblock
  }

# ==============================================================================
# html_convert_to_markup(): converts the selection "word1 word2 ..." to
#       "<word1 word2 ...></word1>", leaving the cursor between the two markup
#       brackets.
# ==============================================================================

define html_convert_to_markup
  {
  full_directive = get_selection()
  directive = full_directive
  start = $selection_start

  directive_end = -1
  directive_start = search_string(full_directive, "[a-zA-Z0-1]", 0, "regex")

  if (directive_start >= 0)
    directive_end = search_string(full_directive, ">", directive_start, "regex")

  if (directive_start >= 0 && directive_end >= 0)
    directive = substring(full_directive, directive_start, directive_end)

  replace_selection("<" full_directive "></" directive ">")
  deselect_all()
  set_cursor_pos(start + length(full_directive) + 2)
  }

# ==============================================================================
# html_ask_for_markup(): requests a markup directive in a dialog box: the user
#       should enter the leading directive such as "word1 word2 ...". This is
#       converted to "<word1 word2 ...>" and placed before the selected text.
#       The terminator, "</word1>", is added at the end of the selection,
#       leaving the cursor at the end of the terminator.
# ==============================================================================

define html_ask_for_markup
  {
  if ($selection_left >= 0 || $locked)
    {
    beep()
    return # disallow rectangular selections, or application to a locked file
    }

  if ($selection_start < 0)
    {
    beg = $cursor
    end = $cursor
    }
  else
    {
    beg = $selection_start
    end = $selection_end
    }

  set_locked(1)
  full_directive = string_dialog("Markup", "Block", "Inline", "Cancel")
  res = $string_dialog_button
  set_locked(0)
  if (full_directive == "" || res < 1 || res > 2)
    return ""

  res = html_add_markup_brackets(full_directive, res == 1, beg, end)
  deselect_all()
  return res
  }

# ==============================================================================
# html_make_A_NAME(): converts the selected text into an HTML anchor, such that
#       text like "hi there" becomes "<a name="hi_there">hi there</a>".
#       If there is a suspicion this is an xhtml file, use "id" instead of
#       "name".
# ==============================================================================

define html_make_A_NAME
  {
  s = ""

  if ($selection_start != -1)
    s = get_selection()

  while (!match_re(s, "^[A-Za-z][-A-Za-z0-9:_.]*$"))
    {
    u = replace_in_string(s, "(?n\\s)", "_", "regex", "copy")
    u = replace_in_string(u, "(?n[^-A-Za-z0-9:_.])", ":", "regex", "copy")
    u = replace_in_string(u, "^([^A-Za-z])", "name:-\\1", "regex", "copy")
    t = string_dialog("String\n    "s"\nDoes not conform to norms.\n" \
                      "It should match [A-Za-z][-A-Za-z0-9:_.]*\n" \
                      "Do you want to change it?\n" \
                      "(Suggestion:\n    "u"\nfor example)", \
                      "Change", "Keep", "Use suggested", "Cancel")
    r = $string_dialog_button
    if      (r == 1 && t != "") s = t   # Change
    else if (r == 2) break              # Keep
    else if (r == 3) s = u              # Use suggested
    else
      return
    }
  if (match_si(html__get_doctype(), "xhtml"))
    name_or_id = "id=\""s"\" name=\""s"\"" # XML uses ids, HTML uses names
  else
    name_or_id = "name=\""s"\""
  html_add_markup_brackets("a "name_or_id)
  }

# ==============================================================================
# html_make_A_HREF(): converts the selected text into an HTML local reference,
#       such that text like "hello" becomes "<a href="#hello">hello</a>".
# ==============================================================================

define html_make_A_HREF
  {
  s = ""
  if ($selection_start != -1)
    {
    s = get_selection()
    }
  markup = "a href=\"#"s"\""
  html_add_markup_brackets(markup)
  }

# ==============================================================================
# Information for the html_entities() function.
# ==============================================================================

# don't include &&amp; here: treat it separately since it must be done first
$html_simple_entity_list = \
"<&lt;"         ">&gt;"         "\"&quot;"

$html_entity_list = \
" &nbsp;"       "¡&iexcl;"      "¢&cent;"       "£&pound;"      \
"¤&curren;"     "¥&yen;"        "¦&brvbar;"     "§&sect;"       \
"¨&uml;"        "©&copy;"       "ª&ordf;"       "«&laquo;"      \
"¬&not;"        "­&shy;"        "®&reg;"        "¯&macr;"       \
"°&deg;"        "±&plusmn;"     "²&sup2;"       "³&sup3;"       \
"´&acute;"      "µ&micro;"      "¶&para;"       "·&middot;"     \
"¸&cedil;"      "¹&sup1;"       "º&ordm;"       "»&raquo;"      \
"¼&frac14;"     "½&frac12;"     "¾&frac34;"     "¿&iquest;"     \
"À&Agrave;"     "Á&Aacute;"     "Â&Acirc;"      "Ã&Atilde;"     \
"Ä&Auml;"       "Å&Aring;"      "Æ&AElig;"      "Ç&Ccedil;"     \
"È&Egrave;"     "É&Eacute;"     "Ê&Ecirc;"      "Ë&Euml;"       \
"Ì&Igrave;"     "Í&Iacute;"     "Î&Icirc;"      "Ï&Iuml;"       \
"Ð&ETH;"        "Ñ&Ntilde;"     "Ò&Ograve;"     "Ó&Oacute;"     \
"Ô&Ocirc;"      "Õ&Otilde;"     "Ö&Ouml;"       "×&times;"      \
"Ø&Oslash;"     "Ù&Ugrave;"     "Ú&Uacute;"     "Û&Ucirc;"      \
"Ü&Uuml;"       "Ý&Yacute;"     "Þ&THORN;"      "ß&szlig;"      \
"à&agrave;"     "á&aacute;"     "â&acirc;"      "ã&atilde;"     \
"ä&auml;"       "å&aring;"      "æ&aelig;"      "ç&ccedil;"     \
"è&egrave;"     "é&eacute;"     "ê&ecirc;"      "ë&euml;"       \
"ì&igrave;"     "í&iacute;"     "î&icirc;"      "ï&iuml;"       \
"ð÷&eth;"       "ñ&ntilde;"     "ò&ograve;"     "ó&oacute;"     \
"ô&ocirc;"      "õ&otilde;"     "ö&ouml;"       "÷&divide;"     \
"ø&oslash;"     "ù&ugrave;"     "ú&uacute;"     "û&ucirc;"      \
"ü&uuml;"       "ý&yacute;"     "þ&thorn;"      "ÿ&yuml;"       \
""

$html_entities = $empty_array

# ==============================================================================
# html_entities([string, [, "all"]]): converts invalid characters in the string
#       passed (or in the selected region if no arguments given, or the string
#       is empty) to the equivalent character entities. If the second argument
#       "all" is given, all known entities are substituted; otherwise those
#       which may be HTML markup will be skipped ('&', '<', '>', '"').
# ==============================================================================

define html_entities
  {
  $html_entities[""] = ""

 if (!("simple" in $html_entities))
    {
    entity_list = $empty_array
    split_list = split($html_simple_entity_list, "(?<=;)", "regex")
    for (ind in split_list)
      {
      ent = split_list[ind]
      if (ent != "")
        entity_list[substring(ent, 0, 1)] = substring(ent, 1)
      }
    $html_entities["simple"] = entity_list
    }

 if (!("other" in $html_entities))
    {
    entity_list = $empty_array
    split_list = split($html_entity_list, "(?<=;)", "regex")
    for (ind in split_list)
      {
      ent = split_list[ind]
      if (ent != "")
        entity_list[substring(ent, 0, 1)] = substring(ent, 1)
      }
    $html_entities["other"] = entity_list
    }

  s = ""
  all = 0
  if ($n_args > 0)
    s = $1
  if ($n_args > 1 && $2 == "all")
    all = 1

  if (s == "" && $selection_start != -1)
    {
    s = get_selection()
    }

  len = length(s)

  if (len > $html_entities["other"][])
    {
    if (all)
      {
      # start with & -> &amp;
      if (search_string(s, "&", 0, "case") != -1)
        s = replace_in_string(s, "&", "&amp;", "case")

      # go through the $html_entities["simple"] table, trying to substitute the
      # character for the entity within the string each time
      entity_list = $html_entities["simple"]
      for (c in entity_list)
        {
        entity = entity_list[c]
        if (search_string(s, c, 0, "case") != -1)
          s = replace_in_string(s, c, entity, "case")
        }
      }
    # go through the $html_entities["other"] table, trying to substitute the
    # character for the entity within the string each time
    entity_list = $html_entities["other"]
    for (c in entity_list)
      {
      entity = entity_list[c]
      if (search_string(s, c, 0, "case") != -1)
        s = replace_in_string(s, c, entity, "case")
      }
    }
  else
    {
    # go through the string to be altered, checking each character to
    # see if it should be substituted
    ss = ""
    for (j = 0; j < len; j++)
      {
      c = substring(s, j, j+1)
      if (c in $html_entities["other"])
        ss = ss $html_entities["other"][c]
      else if (all)
        {
        if (c in $html_entities["simple"])
          ss = ss $html_entities["simple"][c]
        else
          ss = ss c
        }
      else
        ss = ss c
      }
    s = ss
    }

  return s
  }