/[drupal]/contributions/modules/import_html/modules/core.inc

Diff of /contributions/modules/import_html/modules/core.inc

Parent Directory | Revision Log | View Revision Graph Revision Graph | View Patch Patch

-revision 1.1.2.13,
Wed Sep  9 06:47:22 2009 UTC
+revision 1.1.2.14,
Tue Oct  6 14:09:02 2009 UTC
 Line 1
  <?php
- // $Id: core.inc,v 1.1.2.12 2009/07/01 02:40:56 dman Exp $
+ // $Id: core.inc,v 1.1.2.13 2009/09/09 06:47:22 dman Exp $
  /**
   * @file
   * Basic per-module functionality used by import_html.
   *
   * Each available module MAY implement a hook_import_html() function to add or
   * manipulate its own data as the node gets saved.
   *
   * This set defines the known core module attributes that can be added to a node
   * when importing. It sets the node (content & title) values, the menu and path
   * values, and possibly some taxonomy tags.
   *
   * Each of these (pretty integral) attributes is set in a modular way as an
   * example for extensions, and to keep the central import_html codebase lighter.
   *
   *
   * @ingroup import_html Import HTML
   * @author Dan Morrison http://coders.co.nz/
   */
  /**************************************************
   * NODE
   **************************************************/
  /**
   * Implementation of hook_import_html
   *
   * Add appropriate normal node attributes - title, body and teaser
   * - to the node
   */
  function node_import_html($profile, &$node, $datadoc = NULL) {
    // Setting a few basics will just avoid 'undefined' complaints when previewing
    $node->nid = isset($node->nid) ? $node->nid : 0;
    $node->sticky = isset($node->sticky) ? $node->sticky : FALSE;
    //
    //  Title
    // Note, internally the title should always be plaintext, encoded characters and no newlines
    // A meta 'title' may have been set,
    // but I'll use the real visible one by preference.
    $h1 = xml_get_element_content($datadoc, 'h1');
    if ($h1) {
      $node->title = $h1;
    }
    else {
      $title_elem = xml_get_element_content($datadoc, 'title');
      if ($title_elem) {
 Line 54 
 function node_import_html($profile, &$no
    //
    // BODY is the thing with id=content or whatever is set in the prefs
    //
    import_html_debug_code('XML DOM being scanned for XPATH data extraction to find the body content', xml_tostring($datadoc));
    $content_element = xml_getelementbyid($datadoc, 'content');
    if (!$content_element) {
      import_html_debug("Failed to find a body, anything with id='content' in this page", array(), WATCHDOG_NOTICE);
 Line 71 
 function node_import_html($profile, &$no
        $node->import_html_precision = $precision;
        $explanation = xml_getattribute($content_element, 'title');
        import_html_debug(
          "I'm %precision % confident that we found the right body element. !explanation",
          array('%precision' => 100*$precision, '!explanation' => $explanation),
          WATCHDOG_INFO
        );
 Line 87 
 function node_import_html($profile, &$no
    //
    //  Teaser
    //
    $teaser = xml_textcontent( xml_getelementbyid($datadoc, 'teaser') );
    $description = xml_textcontent( xml_getelementbyid($datadoc, 'description') );
    if (!$teaser) $teaser = $description ;
    if ($teaser) {
 Line 102 
 function node_import_html($profile, &$no
    /*
    // No custom teaser found, but
    // Imported HTML is highly likely to contain formatting that
    // will ruin teasers. Do our own cut-down version that assumes
    // FILTERED HTML
    if (!$teaser) {
      $teaser = filter_filter('process', 0, 1, node_teaser($node->body));
 Line 120 
 function node_import_html($profile, &$no
  /**************************************************
   * USER
   **************************************************/
  /**
   * Implementation of hook_import_html
   *
   * Add appropriate user identification to the new node.
   */
  function user_import_html($profile, &$node, $datadoc = NULL) {
    global $user;
    $node->uid = $user->uid;
 Line 145 
 function user_import_html($profile, &$no
   * its namespace. Miscellaneous class HTML import may create an attribute called
   * 'menu'. That would be bad.
   */
  function menu_import_html($profile, &$node, $datadoc = NULL) {
    unset($node->menu);
  }
    //
    // This is currently a freaking mess. I don't know what the menu router is trying to do these days.
    //
  /**
   * Set alias and navigation for the given node
   * Helper for import_html_import_files()
 Line 170 
 function menu_import_html_after_save($pr
        import_html_debug("We will re-use the existing menu item, and keep its place in the sitemap");
      }
      else {
        // Check if a placeholder menu for the alias exists.
        // This can happen if the child items were created earlier and required a placeholder
        // Try both canonic and aliased paths
        if ( ! $menu = menu_get_item_by_path($normal_path) ) {
 Line 178 
 function menu_import_html_after_save($pr
        }
      }
      // $menu is a menu_link array
      if (! empty($menu)) {
        // Found a pre-existing menu, update it
        import_html_debug("Discovered that a placeholder menu item for %path already exists. Filling that one out with the full details now", array('%path' => $node->path));
 Line 189 
 function menu_import_html_after_save($pr
        $parent_menu = import_html_create_menu_path(dirname($node->path), $profile);
        #dpm(array('After creating the menu path, the parent menu is' => $parent_menu));
        if (empty($parent_menu)) {
          trigger_error("Code should not get here - failed to build menu - no parent menu returned to hang this node off.", E_USER_ERROR);
          return;
        }
 Line 204 
 function menu_import_html_after_save($pr
      }
      $node->menu = $menu;
      # dpm(array("about to save menu link " => $menu));
      menu_link_save($node->menu);
    }
  }
 Line 214 
 function menu_import_html_after_save($pr
   * Given a traditional URL path, return the parent menu item
   *
   * A helper function to menu_import_html_after_save()
   *
   * This is a recursive function that slices the given
   * (alias) path up until it finds a menu item it recognises.
   * Once found, it comes back down again building a path
 Line 225 
 function menu_import_html_after_save($pr
   * I'd like to have it fall back to system_admin_menu_block_page by default, but
   * can't see how to make that happen on the fly.
   *
   * @param $path
   *   the  path (aliased URL) we intend to construct
   * @param $title
   *   optional    Display title of the menu item at that path point, otherwist
   * it'll use the last fragment of the path as a display title
   * @return a menu object, the bottom-most, the one we asked for.
   * FALSE if it hit
   * the top unexpectedly
   *
   * @return a menu_link definition ARRAY
   *
   **/
  function import_html_create_menu_path($path, $profile = array()) {
 Line 261 
 function import_html_create_menu_path($p
    if ($p_path == dirname($p_path)) {
      import_html_debug("
          Hit topmost menu path '$p_path', this is either caused by creating a new menu tree
          from the bottom up and we just reached the top,
          or a recursion error (illegal paths).
      ", array(), WATCHDOG_INFO);
 Line 270 
 function import_html_create_menu_path($p
    }
    else {
      if (! $parent = menu_get_item_by_path($p_path)) {
        // Parent not found. Need to make it up.
        // Recurse...
        $parent = import_html_create_menu_path($p_path, $profile);
        // ... and welcome back again!
 Line 291 
 function import_html_create_menu_path($p
    import_html_debug("Creating a placeholder menu item for $path. Parent was found to be item {$parent['mlid']}", array(), WATCHDOG_INFO);
    $menu_link_item = import_html_create_placeholder_menu_item($path, $profile, $parent);
    if (empty($menu_link_item)) {
      trigger_error("Failed to create placeholder menu item ($path, $parent) ", E_USER_ERROR);
      return NULL;
 Line 302 
 function import_html_create_menu_path($p
  }
  /**
   * Create a dummy node to represent a menu container.
   *
   * This appears to be the only sane way to build a site from the bottom up.
   * I can, If I try, create menu trees that do not link to nodes ... but most
   * users will not be able to remove, over-write or edit them using the tools
   * they are used to.
   *
   * Instead, just make empty shells on-the-fly. You can kill them via content
   * management and things will bubble up and Drupal will not have a hernia.
   *
   * @param $parent is a parent link item array, containing (at least) a mlid and
   * menu_name
   */
 Line 323 
 function import_html_create_placeholder_
      'status'  => $profile['import_status'],
      'promote' => FALSE,
      'body' => t("
        Placeholder page.
        Created by import_html.
        This is required to keep the menu in place during imports.
        Feel free to delete it!
      "),
    );
 Line 337 
 function import_html_create_placeholder_
      trigger_error("Failed to create placeholder node to hold the menus together. Not fatal, but some sections under $path will be unstructured", E_USER_WARNING);
      return NULL;
    }
    // Now make the menu for this dummy node
    $menu = array(
      'parent'     => $parent_menu['menu_name'] .':'. $parent_menu['mlid'],
 Line 346 
 function import_html_create_placeholder_
      'link_title' => import_html_guess_label('', $path),
      'link_path'  => 'node/'. $node->nid,
    );
    $mlid = menu_link_save($menu);
    if (! $mlid) {
 Line 363 
 function import_html_create_placeholder_
  /**
   * Return a menu item matching a given path or alias.
   * The alias for the requested path will be tried automatically.
   *
   * @return a menu link ARRAY. NULL if no valid menu item was found.
   */
  function menu_get_item_by_path($path) {
    // Caching works against me when adding lots of aliases in one go.
    // Cannot trust get_alias etc
    // Wipe it :(
    drupal_lookup_path('wipe'); // This is not working?
    // Use DB to fetch all aliases
    $aliases = array($path);
    $placeholders = array(' link_path = "%s" ');
 Line 410 
 function menu_get_item_by_path($path) {
  function path_import_html($profile, &$node, $datadoc = NULL) {
    drupal_lookup_path('wipe');
    // Path requires the nodes language to be set. Don't know why it's a PATH job, but
    // that's the module that complains about undefined property if it isn't
    $node->language = isset($node->language) ? $node->language : '';
  }
  /**
   * Add path alias support to the import_html process
   *
   * Needed to be run after save as we need to know the nid
   *
   * Adds an old-style (legacy) alias to the node path if required.
   */
  function path_import_html_after_save($profile, &$node, $datadoc = NULL) {
    #dpm(array('INVOKED path_import_html_after_save for '. $node->path));
    if ($profile['legacy_aliases'] && ($node->old_path != $node->path)) {
      import_html_debug("Setting up URL ALIASES for this item now. Directing {$node->old_path} to go to the system path ". drupal_get_normal_path($node->path), array(), WATCHDOG_DEBUG);
      $normal_path = "node/{$node->nid}";
      if ($normal_path == $node->path) {
        import_html_debug(
          "Failed to resolve %node_path into a system path. Cannot create alias at this time.",
          array('%node_path' => $node->path),
          WATCHDOG_NOTICE
        );
      }
      else {
        // path_set_alias is WRONG when it looks for existing items
        // drupal_lookup_path('wipe'); // DOES NOT WORK - to it ourselves.
 Line 446 
 function path_import_html_after_save($pr
          path_set_alias($normal_path, $node->old_path);
        }
        import_html_debug(
          "This document (known internally as '%normal_path' )
            should now be accessible via aliases as both
            '!main_alias' and '!legacy_alias' ",
          array(
            '%normal_path' => $normal_path,
 Line 458 
 function path_import_html_after_save($pr
        );
      }
    }
  }
  /**************************************************
 Line 468 
 function path_import_html_after_save($pr
  /**
   * Absorbs elements in the import document we recognise as being (probably)
   * taxonomy terms.
   *
   *
   * Any link rel="tag" is a start. The plaintext content of such a tag is
   * searched for as a term.
   *
   * both <a href rel='tag' > and <link rel="tag"> are good for me.
   *
   * @TODO this still needs to be expanded with real-world cases
   */
  function taxonomy_import_html($profile, &$node, $datadoc = NULL) {
 Line 482 
 function taxonomy_import_html($profile,
    $active_vocabs = taxonomy_get_vocabularies($node->type);
    if (!$active_vocabs) return;
    $freetag_vocab = $profile['freetag_vocab'];
    if ($freetag_vocab && !empty($node->keywords)) {
 Line 511 
 function taxonomy_import_html($profile,
      $reltypes = explode(' ', xml_getAttribute($link, 'rel'));
      $rellink = xml_getattribute($link, 'href');
      $label = trim(xml_textcontent($link));
      // Microformats may put their value into a 'title' attribute instead
      if (empty($label)) {
        $label = xml_getAttribute($link, 'title');
      }
      if (!empty($label)) {
 Line 521 
 function taxonomy_import_html($profile,
          if ($reltype == 'tag') {
            // Some of this magic should probably shift into absorb_metas()
            // Still thinking of the best way to encode tags into the raw HTML
            if (strstr($label, ":")) {
              // allow an extended format for this value
              // to support freetext vocab additions
 Line 535 
 function taxonomy_import_html($profile,
                $node->taxonomy['tags'][$vocabulary->vid] .= ",". $label;
              }
            }
            // Match against existing terms
            $terms = taxonomy_get_term_by_name($label);
            // Allow an optional 'S' on the end of terms when looking for a match. Just to be flexible.
            $terms = array_merge($terms, taxonomy_get_term_by_name($label ."s"));
            if (!is_array($node->taxonomy)) {
              $node->taxonomy = array();
            }
            if ($terms) {
              foreach ($terms as $term) {
                // If we did successfully identify the vocab earlier, filter on that. Otherwise let anything go
                if ($vocabulary && ($term->vid != $vocabulary->vid)) continue;
                $node->taxonomy[$term->tid] = $term;
              }
            }
 Line 562 
 function taxonomy_import_html($profile,
      } // has value
    } // each rel
    // Tag this new content if the profile has a global tag set
    // May have been set during the form process.
    //
    // Note that if the global tag is from a freetaging vocab, we insert it as a string
    // but if it was from a select category, we insert it by ID.
    $cats = $profile['import_category'];
    if (is_array($cats)) {
      foreach ($cats as $term_id) {
        if ($term = taxonomy_get_term($term_id)) {
          // Be careful not to add terms that are invalid or recently deleted
          // $node->taxonomy contains either strings or IDs.
          // Figure it out and insert into the object accordingly
          // Mixing ids and strings confuses taxonomy submit otherwise
          $term_vocab = taxonomy_vocabulary_load($term->vid);
 Line 594 
 function taxonomy_import_html($profile,
  }
  /**
-  * implimentation of hook_import_html_node_merge.
+  * Implimentation of hook_import_html_node_merge()
   *
   * HOOK_import_html_node_merge gets called after a new node has been created,
   * and is being used to update a pre-existing one.
   *
   * A per-module way of blending new values over old ones - conflict resolution
   * and checking for duplicates.
   *
   * If the values are done with, unset them from the $old node so they don't get
   * inadvertantly copied by the bulk operation that follows.
   *
   * Taxonomy needs this because: If I am MERGING with a pre-existing node that
   * already has terms (such as when doing a re-import) then the node taxonomy
   * already has a bunch of identified terms that are likely to be duplicates. I
 Line 612 
 function taxonomy_import_html($profile,
   * leaving the node with them set, and ALSO adding the freetagging list will
   * cause duplicate complaints. I need to investigate the existing terms and
   * remove them individually. Easiest way is to convert them into a flat list so
   * they get re-saved.
   *
   * taxonomy_preview_terms() didn't look like it will help.
   *
   * taxonomy_form_alter() seems to have the same problem. And solves it sorta.
   *
   * But not good enough for us without rewriting $node->taxonomy totally
   *
   * Taxonomy doesn't mind duplicates in the freetag string, so we will pull the
   * known terms out, just add them to the freetag string, and discard the oldies.
   *
   * It operates on both old and new by reference.
   */
  function taxonomy_import_html_node_merge(&$old_node, &$new_node, $profile = array()) {
 Line 638 
 function taxonomy_import_html_node_merge
            $new_node->taxonomy['tags'][$term->vid] = '';
          }
          $new_node->taxonomy['tags'][$term->vid] .= ', '. $term->name;
        }
      }
    }
  }

 Legend:



Removed from v.1.1.2.13
 


changed lines


 
Added in v.1.1.2.14
 Legend:



Removed from v.1.1.2.13
 


changed lines


 
Added in v.1.1.2.14
-Removed from v.1.1.2.13
+Added in v.1.1.2.14

	ViewVC Help
Powered by ViewVC 1.1.2