lisp/dom.el

   1 ;;; dom.el --- XML/HTML (etc.) DOM manipulation and searching functions
   2
   3 ;; Copyright (C) 2014-2016 Free Software Foundation, Inc.
   4
   5 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
   6 ;; Keywords: xml, html
   7
   8 ;; This file is part of GNU Emacs.
   9
  10 ;; GNU Emacs is free software: you can redistribute it and/or modify
  11 ;; it under the terms of the GNU General Public License as published by
  12 ;; the Free Software Foundation, either version 3 of the License, or
  13 ;; (at your option) any later version.
  14
  15 ;; GNU Emacs is distributed in the hope that it will be useful,
  16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 ;; GNU General Public License for more details.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
  22
  23 ;;; Commentary:
  24
  25 ;;; Code:
  26
  27 (require 'cl-lib)
  28 (eval-when-compile (require 'subr-x))
  29
  30 (defsubst dom-tag (node)
  31   "Return the NODE tag."
  32   ;; Called on a list of nodes.  Use the first.
  33   (if (consp (car node))
  34       (caar node)
  35     (car node)))
  36
  37 (defsubst dom-attributes (node)
  38   "Return the NODE attributes."
  39   ;; Called on a list of nodes.  Use the first.
  40   (if (consp (car node))
  41       (cadr (car node))
  42     (cadr node)))
  43
  44 (defsubst dom-children (node)
  45   "Return the NODE children."
  46   ;; Called on a list of nodes.  Use the first.
  47   (if (consp (car node))
  48       (cddr (car node))
  49     (cddr node)))
  50
  51 (defun dom-non-text-children (node)
  52   "Return all non-text-node children of NODE."
  53   (cl-loop for child in (dom-children node)
  54            unless (stringp child)
  55            collect child))
  56
  57 (defun dom-set-attributes (node attributes)
  58   "Set the attributes of NODE to ATTRIBUTES."
  59   (setq node (dom-ensure-node node))
  60   (setcar (cdr node) attributes))
  61
  62 (defun dom-set-attribute (node attribute value)
  63   "Set ATTRIBUTE in NODE to VALUE."
  64   (setq node (dom-ensure-node node))
  65   (let ((old (assoc attribute (cadr node))))
  66     (if old
  67         (setcdr old value)
  68       (setcar (cdr node) (nconc (cadr node) (list (cons attribute value)))))))
  69
  70 (defmacro dom-attr (node attr)
  71   "Return the attribute ATTR from NODE.
  72 A typical attribute is `href'."
  73   `(cdr (assq ,attr (dom-attributes ,node))))
  74
  75 (defun dom-text (node)
  76   "Return all the text bits in the current node concatenated."
  77   (mapconcat 'identity (cl-remove-if-not 'stringp (dom-children node)) " "))
  78
  79 (defun dom-texts (node &optional separator)
  80   "Return all textual data under NODE concatenated with SEPARATOR in-between."
  81   (mapconcat
  82    'identity
  83    (mapcar
  84     (lambda (elem)
  85       (if (stringp elem)
  86           elem
  87         (dom-texts elem separator)))
  88     (dom-children node))
  89    (or separator " ")))
  90
  91 (defun dom-child-by-tag (dom tag)
  92   "Return the first child of DOM that is of type TAG."
  93   (assoc tag (dom-children dom)))
  94
  95 (defun dom-by-tag (dom tag)
  96   "Return elements in DOM that is of type TAG.
  97 A name is a symbol like `td'."
  98   (let ((matches (cl-loop for child in (dom-children dom)
  99                           for matches = (and (not (stringp child))
 100                                              (dom-by-tag child tag))
 101                           when matches
 102                           append matches)))
 103     (if (equal (dom-tag dom) tag)
 104         (cons dom matches)
 105       matches)))
 106
 107 (defun dom-strings (dom)
 108   "Return elements in DOM that are strings."
 109   (cl-loop for child in (dom-children dom)
 110            if (stringp child)
 111            collect child
 112            else
 113            append (dom-strings child)))
 114
 115 (defun dom-by-class (dom match)
 116   "Return elements in DOM that have a class name that matches regexp MATCH."
 117   (dom-elements dom 'class match))
 118
 119 (defun dom-by-style (dom match)
 120   "Return elements in DOM that have a style that matches regexp MATCH."
 121   (dom-elements dom 'style match))
 122
 123 (defun dom-by-id (dom match)
 124   "Return elements in DOM that have an ID that matches regexp MATCH."
 125   (dom-elements dom 'id match))
 126
 127 (defun dom-elements (dom attribute match)
 128   "Find elements matching MATCH (a regexp) in ATTRIBUTE.
 129 ATTRIBUTE would typically be `class', `id' or the like."
 130   (let ((matches (cl-loop for child in (dom-children dom)
 131                           for matches = (and (not (stringp child))
 132                                              (dom-elements child attribute
 133                                                            match))
 134                           when matches
 135                           append matches))
 136         (attr (dom-attr dom attribute)))
 137     (if (and attr
 138              (string-match match attr))
 139         (cons dom matches)
 140       matches)))
 141
 142 (defun dom-parent (dom node)
 143   "Return the parent of NODE in DOM."
 144   (if (memq node (dom-children dom))
 145       dom
 146     (let ((result nil))
 147       (dolist (elem (dom-children dom))
 148         (when (and (not result)
 149                    (not (stringp elem)))
 150           (setq result (dom-parent elem node))))
 151       result)))
 152
 153 (defun dom-previous-sibling (dom node)
 154   (when-let (parent (dom-parent dom node))
 155     (let ((siblings (dom-children parent))
 156           (previous nil))
 157       (while siblings
 158         (when (eq (cadr siblings) node)
 159           (setq previous (car siblings)))
 160         (pop siblings))
 161       previous)))
 162
 163 (defun dom-node (tag &optional attributes &rest children)
 164   "Return a DOM node with TAG and ATTRIBUTES."
 165   (if children
 166       `(,tag ,attributes ,@children)
 167     (list tag attributes)))
 168
 169 (defun dom-append-child (node child)
 170   "Append CHILD to the end of NODE's children."
 171   (setq node (dom-ensure-node node))
 172   (nconc node (list child)))
 173
 174 (defun dom-add-child-before (node child &optional before)
 175   "Add CHILD to NODE's children before child BEFORE.
 176 If BEFORE is nil, make CHILD NODE's first child."
 177   (setq node (dom-ensure-node node))
 178   (let ((children (dom-children node)))
 179     (when (and before
 180                (not (memq before children)))
 181       (error "%s does not exist as a child" before))
 182     (let ((pos (if before
 183                    (cl-position before children)
 184                  0)))
 185       (if (zerop pos)
 186           ;; First child.
 187           (setcdr (cdr node) (cons child (cddr node)))
 188         (setcdr (nthcdr (1- pos) children)
 189                 (cons child (nthcdr pos children))))))
 190   node)
 191
 192 (defun dom-ensure-node (node)
 193   "Ensure that NODE is a proper DOM node."
 194   ;; Add empty attributes, if none.
 195   (when (consp (car node))
 196     (setq node (car node)))
 197   (when (= (length node) 1)
 198     (setcdr node (list nil)))
 199   node)
 200
 201 (defun dom-pp (dom &optional remove-empty)
 202   "Pretty-print DOM at point.
 203 If REMOVE-EMPTY, ignore textual nodes that contain just
 204 white-space."
 205   (let ((column (current-column)))
 206     (insert (format "(%S " (dom-tag dom)))
 207     (let* ((attr (dom-attributes dom))
 208            (times (length attr))
 209            (column (1+ (current-column))))
 210       (if (null attr)
 211           (insert "nil")
 212         (insert "(")
 213         (dolist (elem attr)
 214           (insert (format "(%S . %S)" (car elem) (cdr elem)))
 215           (if (zerop (cl-decf times))
 216               (insert ")")
 217             (insert "\n" (make-string column ? ))))))
 218     (let* ((children (if remove-empty
 219                          (cl-remove-if
 220                           (lambda (child)
 221                             (and (stringp child)
 222                                  (string-match "\\`[\n\r\t  ]*\\'" child)))
 223                           (dom-children dom))
 224                        (dom-children dom)))
 225            (times (length children)))
 226       (if (null children)
 227           (insert ")")
 228         (insert "\n" (make-string (1+ column) ? ))
 229         (dolist (child children)
 230           (if (stringp child)
 231               (if (or (not remove-empty)
 232                       (not (string-match "\\`[\n\r\t  ]*\\'" child)))
 233                   (insert (format "%S" child)))
 234             (dom-pp child remove-empty))
 235           (if (zerop (cl-decf times))
 236               (insert ")")
 237             (insert "\n" (make-string (1+ column) ? ))))))))
 238
 239 (provide 'dom)
 240
 241 ;;; dom.el ends here