code.delx.au - gnu-emacs/blob - lisp/cedet/semantic/html.el

   1 ;;; semantic/html.el --- Semantic details for html files
   2
   3 ;;; Copyright (C) 2004, 2005, 2007, 2008 Free Software Foundation, Inc.
   4
   5 ;; Author: Eric M. Ludlam <zappo@gnu.org>
   6
   7 ;; This file is part of GNU Emacs.
   8
   9 ;; GNU Emacs is free software: you can redistribute it and/or modify
  10 ;; it under the terms of the GNU General Public License as published by
  11 ;; the Free Software Foundation, either version 3 of the License, or
  12 ;; (at your option) any later version.
  13
  14 ;; GNU Emacs is distributed in the hope that it will be useful,
  15 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  17 ;; GNU General Public License for more details.
  18
  19 ;; You should have received a copy of the GNU General Public License
  20 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  21
  22 ;;; Commentary:
  23 ;;
  24 ;; Parse HTML files and organize them in a nice way.
  25 ;; Pay attention to anchors, including them in the tag list.
  26 ;;
  27 ;; Copied from the original semantic-texi.el.
  28 ;;
  29 ;; ToDo: Find <script> tags, and parse the contents in other
  30 ;; parsers, such as javascript, php, shtml, or others.
  31
  32 (require 'semantic)
  33 (require 'semantic/format)
  34 (condition-case nil
  35     ;; This is not installed in all versions of Emacs.
  36     (require 'sgml-mode) ;; html-mode is in here.
  37   (error
  38    (require 'psgml-mode) ;; XEmacs uses psgml, and html-mode is in here.
  39    ))
  40
  41 ;;; Code:
  42 (eval-when-compile
  43   (require 'semantic/ctxt))
  44
  45 (defvar semantic-html-super-regex
  46   "<\\(h[1-9]\\|title\\|script\\|body\\|a +href\\)\\>"
  47   "Regular expression used to find special sections in an HTML file.")
  48
  49 (defvar semantic-html-section-list
  50   '(("title" 1)
  51     ("script" 1)
  52     ("body" 1)
  53     ("a" 11)
  54     ("h1" 2)
  55     ("h2" 3)
  56     ("h3" 4)
  57     ("h4" 5)
  58     ("h5" 6)
  59     ("h6" 7)
  60     ("h7" 8)
  61     ("h8" 9)
  62     ("h9" 10)
  63     )
  64   "Alist of sectioning commands and their relative level.")
  65
  66 (define-mode-local-override semantic-parse-region
  67   html-mode (&rest ignore)
  68   "Parse the current html buffer for semantic tags.
  69 INGNORE any arguments.  Always parse the whole buffer.
  70 Each tag returned is of the form:
  71  (\"NAME\" section (:members CHILDREN))
  72 or
  73  (\"NAME\" anchor)"
  74   (mapcar 'semantic-html-expand-tag
  75           (semantic-html-parse-headings)))
  76
  77 (define-mode-local-override semantic-parse-changes
  78   html-mode ()
  79   "We can't parse changes for HTML mode right now."
  80   (semantic-parse-tree-set-needs-rebuild))
  81
  82 (defun semantic-html-expand-tag (tag)
  83   "Expand the HTML tag TAG."
  84   (let ((chil (semantic-html-components tag)))
  85     (if chil
  86         (semantic-tag-put-attribute
  87          tag :members (mapcar 'semantic-html-expand-tag chil)))
  88     (car (semantic--tag-expand tag))))
  89
  90 (defun semantic-html-components (tag)
  91   "Return components belonging to TAG."
  92   (semantic-tag-get-attribute tag :members))
  93
  94 (defun semantic-html-parse-headings ()
  95   "Parse the current html buffer for all semantic tags."
  96   (let ((pass1 nil))
  97     ;; First search and snarf.
  98     (save-excursion
  99       (goto-char (point-min))
 100
 101       (let ((semantic--progress-reporter
 102              (make-progress-reporter
 103               (format "Parsing %s..."
 104                       (file-name-nondirectory buffer-file-name))
 105               (point-min) (point-max))))
 106         (while (re-search-forward semantic-html-super-regex nil t)
 107           (setq pass1 (cons (match-beginning 0) pass1))
 108           (progress-reporter-update semantic--progress-reporter (point)))
 109         (progress-reporter-done semantic--progress-reporter)))
 110
 111     (setq pass1 (nreverse pass1))
 112     ;; Now, make some tags while creating a set of children.
 113     (car (semantic-html-recursive-combobulate-list pass1 0))
 114     ))
 115
 116 (defun semantic-html-set-endpoint (metataglist pnt)
 117   "Set the end point of the first section tag in METATAGLIST to PNT.
 118 METATAGLIST is a list of tags in the intermediate tag format used by the
 119 html parser.  PNT is the new point to set."
 120   (let ((metatag nil))
 121     (while (and metataglist
 122                 (not (eq (semantic-tag-class (car metataglist)) 'section)))
 123       (setq metataglist (cdr metataglist)))
 124     (setq metatag (car metataglist))
 125     (when metatag
 126       (setcar (nthcdr (1- (length metatag)) metatag) pnt)
 127       metatag)))
 128
 129 (defsubst semantic-html-new-section-tag (name members level start end)
 130   "Create a semantic tag of class section.
 131 NAME is the name of this section.
 132 MEMBERS is a list of semantic tags representing the elements that make
 133 up this section.
 134 LEVEL is the levelling level.
 135 START and END define the location of data described by the tag."
 136   (let ((anchorp (eq level 11)))
 137     (append (semantic-tag name
 138                           (cond (anchorp 'anchor)
 139                                 (t 'section))
 140                           :members members)
 141             (list start (if anchorp (point) end)) )))
 142
 143 (defun semantic-html-extract-section-name ()
 144   "Extract a section name from the current buffer and point.
 145 Assume the cursor is in the tag representing the section we
 146 need the name from."
 147   (save-excursion
 148     ; Skip over the HTML tag.
 149     (forward-sexp -1)
 150     (forward-char -1)
 151     (forward-sexp 1)
 152     (skip-chars-forward "\n\t ")
 153     (while (looking-at "<")
 154       (forward-sexp 1)
 155       (skip-chars-forward "\n\t ")
 156       )
 157     (let ((start (point))
 158           (end nil))
 159       (if (re-search-forward "</" nil t)
 160           (progn
 161             (goto-char (match-beginning 0))
 162             (skip-chars-backward " \n\t")
 163             (setq end (point))
 164             (buffer-substring-no-properties start end))
 165         ""))
 166     ))
 167
 168 (defun semantic-html-recursive-combobulate-list (sectionlist level)
 169   "Rearrange SECTIONLIST to be a hierarchical tag list starting at LEVEL.
 170 Return the rearranged new list, with all remaining tags from
 171 SECTIONLIST starting at ELT 2.  Sections not are not dealt with as soon as a
 172 tag with greater section value than LEVEL is found."
 173   (let ((newl nil)
 174         (oldl sectionlist)
 175         (case-fold-search t)
 176         tag
 177         )
 178     (save-excursion
 179       (catch 'level-jump
 180         (while oldl
 181           (goto-char (car oldl))
 182           (if (looking-at "<\\(\\w+\\)")
 183               (let* ((word (match-string 1))
 184                      (levelmatch (assoc-string
 185                                   word semantic-html-section-list t))
 186                      text begin tmp
 187                      )
 188                 (when (not levelmatch)
 189                   (error "Tag %s matched in regexp but is not in list"
 190                          word))
 191                 ;; Set begin to the right location
 192                 (setq begin (point))
 193                 ;; Get out of here if there if we made it that far.
 194                 (if (and levelmatch (<= (car (cdr levelmatch)) level))
 195                     (progn
 196                       (when newl
 197                         (semantic-html-set-endpoint newl begin))
 198                       (throw 'level-jump t)))
 199                 ;; When there is a match, the descriptive text
 200                 ;; consists of the rest of the line.
 201                 (goto-char (match-end 1))
 202                 (skip-chars-forward " \t")
 203                 (setq text (semantic-html-extract-section-name))
 204                 ;; Next, recurse into the body to find the end.
 205                 (setq tmp (semantic-html-recursive-combobulate-list
 206                            (cdr oldl) (car (cdr levelmatch))))
 207                 ;; Build a tag
 208                 (setq tag (semantic-html-new-section-tag
 209                            text (car tmp) (car (cdr levelmatch)) begin (point-max)))
 210                 ;; Before appending the newtag, update the previous tag
 211                 ;; if it is a section tag.
 212                 (when newl
 213                   (semantic-html-set-endpoint newl begin))
 214                 ;; Append new tag to our master list.
 215                 (setq newl (cons tag newl))
 216                 ;; continue
 217                 (setq oldl (cdr tmp))
 218                 )
 219             (error "Problem finding section in semantic/html parser"))
 220           ;; (setq oldl (cdr oldl))
 221           )))
 222     ;; Return the list
 223     (cons (nreverse newl) oldl)))
 224
 225 (define-mode-local-override semantic-sb-tag-children-to-expand
 226   html-mode (tag)
 227   "The children TAG expands to."
 228   (semantic-html-components tag))
 229
 230 ;;;###autoload
 231 (defun semantic-default-html-setup ()
 232   "Set up a buffer for parsing of HTML files."
 233   ;; This will use our parser.
 234   (setq semantic-parser-name "HTML"
 235         semantic--parse-table t
 236         imenu-create-index-function 'semantic-create-imenu-index
 237         semantic-command-separation-character ">"
 238         semantic-type-relation-separator-character '(":")
 239         semantic-symbol->name-assoc-list '((section . "Section")
 240
 241                                            )
 242         semantic-imenu-expandable-tag-classes '(section)
 243         semantic-imenu-bucketize-file nil
 244         semantic-imenu-bucketize-type-members nil
 245         senator-step-at-start-end-tag-classes '(section)
 246         semantic-stickyfunc-sticky-classes '(section)
 247         )
 248   (semantic-install-function-overrides
 249    '((tag-components . semantic-html-components)
 250      )
 251    t)
 252   )
 253
 254 (define-child-mode html-helper-mode html-mode
 255   "`html-helper-mode' needs the same semantic support as `html-mode'.")
 256
 257 (provide 'semantic/html)
 258
 259 ;; Local variables:
 260 ;; generated-autoload-file: "loaddefs.el"
 261 ;; generated-autoload-feature: semantic/loaddefs
 262 ;; generated-autoload-load-name: "semantic/html"
 263 ;; End:
 264
 265 ;;; semantic/html.el ends here