Should be all working, except that internal links no longer work, not…

… sure how to fix that
concretecms · Aug 25, 2011 · be476bd · be476bd
1 parent c430858
commit be476bd
Show file tree

Hide file tree

Showing 7 changed files with 98 additions and 24 deletions.
diff --git a/README b/README
@@ -21,3 +21,5 @@ Not so great:
 Would be nice:
   - Option to import just posts or just pages.
   - some kind of rudimentary support for image captions, like <div class="wp-import-image"> <img /> <span class="wp-import-caption">caption</span></div>
+  - Somehow keep post url structure, or update internal links in post content
+  - Better import of images, script as is tries to import flickr images that probably shouldn't be imported
diff --git a/controller.php b/controller.php
@@ -47,6 +47,9 @@ public function install() {
 
 	public function uninstall() {
 		$db = Loader::db();
+
+		$dashPage = Page::getByPath("/dashboard/wordpress_import");
+		$dashPage->delete();
 		$db->Execute('drop table WordpressItems');
 		parent::uninstall();
 	}

diff --git a/controllers/dashboard/wordpress_import/file.php b/controllers/dashboard/wordpress_import/file.php
@@ -74,7 +74,7 @@ public function import_wordpress_xml() {
 				$opt = new SelectAttributeTypeOption(0, $tags[$i]['tag_name'], $i);
 				$opt = $opt->saveOrCreate($akt);
 			}
-
+			
 		} else {
 			echo t("No file");
 			exit;

diff --git a/controllers/dashboard/wordpress_import/import.php b/controllers/dashboard/wordpress_import/import.php
@@ -28,11 +28,11 @@ public function get_root_pages() {
 		$json = Loader::helper('json');
 		$data = array();
 		$pageRoot = Page::getByID($this->post('new-root-pages'));
-		$data['page-title'] = $pageRoot->getCollectionName();
-		$data['page-url'] = $pageRoot->getCollectionPath();
+		$data['page_title'] = $pageRoot->getCollectionName();
+		$data['page_url'] = $pageRoot->getCollectionPath();
 		$postRoot = Page::getByID($this->post('new-root-posts'));
-		$data['post-title'] = $postRoot->getCollectionName();
-		$data['post-url'] = $postRoot->getCollectionPath();
+		$data['post_title'] = $postRoot->getCollectionName();
+		$data['post_url'] = $postRoot->getCollectionPath();
 		echo $json->encode($data);
 		exit;
 	}

diff --git a/libraries/wordpress_file_post_importer.php b/libraries/wordpress_file_post_importer.php
@@ -49,9 +49,68 @@ function importFile($fileUrl){
 		// itterate over each incoming URL adding if relevant
 		foreach($incoming_urls as $this_url) {
 			// try to D/L the provided file
-			$client = new Zend_Http_Client($this_url);
+// This all sets up the CURL actions to check the page
+$ch = curl_init();
+curl_setopt($ch, CURLOPT_URL, $this_url);
+curl_setopt($ch, CURLOPT_HEADER, true);
+curl_setopt($ch, CURLOPT_NOBODY, true);
+curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
+curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
+curl_setopt($ch, CURLOPT_MAXREDIRS, 10); //follow up to 10 redirections - avoids loops
+$data = curl_exec($ch);
+$http_code = curl_getinfo($ch, CURLINFO_HTTP_CODE); // Get the HTTP Code
+// Get final redirected URL, will be the same if URL is not redirected
+$new_url = curl_getinfo($ch, CURLINFO_EFFECTIVE_URL); 
+curl_close($ch);
+
+// Array of HTTP status codes. Trim down if you would like to.
+$codes = array(0=>'Domain Not Found',
+			   100=>'Continue',
+			   101=>'Switching Protocols',
+			   200=>'OK',
+			   201=>'Created',
+			   202=>'Accepted',
+			   203=>'Non-Authoritative Information',
+			   204=>'No Content',
+			   205=>'Reset Content',
+			   206=>'Partial Content',
+			   300=>'Multiple Choices',
+			   301=>'Moved Permanently',
+			   302=>'Found',
+			   303=>'See Other',
+			   304=>'Not Modified',
+			   305=>'Use Proxy',
+			   307=>'Temporary Redirect',
+			   400=>'Bad Request',
+			   401=>'Unauthorized',
+			   402=>'Payment Required',
+			   403=>'Forbidden',
+			   404=>'Not Found',
+			   405=>'Method Not Allowed',
+			   406=>'Not Acceptable',
+			   407=>'Proxy Authentication Required',
+			   408=>'Request Timeout',
+			   409=>'Conflict',
+			   410=>'Gone',
+			   411=>'Length Required',
+			   412=>'Precondition Failed',
+			   413=>'Request Entity Too Large',
+			   414=>'Request-URI Too Long',
+			   415=>'Unsupported Media Type',
+			   416=>'Requested Range Not Satisfiable',
+			   417=>'Expectation Failed',
+			   500=>'Internal Server Error',
+			   501=>'Not Implemented',
+			   502=>'Bad Gateway',
+			   503=>'Service Unavailable',
+			   504=>'Gateway Timeout',
+			   505=>'HTTP Version Not Supported');
+			if (isset($codes[$http_code])) {
+				if ($codes[$http_code] == "OK"){
+					$client = new Zend_Http_Client($this_url);
+
+
 			$response = $client->request();
-
 			if ($response->isSuccessful()) {
 				$uri = Zend_Uri_Http::fromString($this_url);
 				$fname = '';
@@ -106,6 +165,12 @@ function importFile($fileUrl){
 				// warn that we couldn't download the file
 				$errors[] = t('There was an error downloading ') . $this_url;
 			}
+				}
+			} else {
+				$errors[] = t("Error connecting to file's server, file skipped");
+			}
+
+
 		}
 	}
 	//print_r($errors);

diff --git a/single_pages/dashboard/wordpress_import/file.php b/single_pages/dashboard/wordpress_import/file.php
@@ -21,8 +21,10 @@ function ok_submit() {
 <div class="ccm-dashboard-inner">
 <form id="import-wordpress-xml" method="post" action="<?php  echo $this->action('import_wordpress_xml') ?>">
  <div style="width:400px"> 
+	 <?php if ($records == 0){?>
  <h2> <?php echo  t("Please Upload/Select your Wordpress export xml file.") ?></h2>
 	  <?php   echo $al->file('wordpress-file','wp-file',t('Choose Wordpress File')); ?>
+ <?php } ?>
 		<?php  if ($records > 0) {
 			?><br/><h2>&nbsp;&nbsp;&nbsp;<?php echo  t("Ready to import ").$records.t(" records.");?></h2>
 		<?php }?>

diff --git a/single_pages/dashboard/wordpress_import/import.php b/single_pages/dashboard/wordpress_import/import.php
@@ -52,34 +52,34 @@
 					wpImport.continue_import(remain,done);
 				} else {
 					$("#progress h1").html("Import Finished.");
-					wpImport.importCategories();
 					$("#progress .titles").hide("slow");
+					wpImport.show_link();
 				}
 			},"json");
 		},
 		show_link: function() {
-			$.post('<?php echo $this->action('get_root_page') ?>', $("#import-wordpress-form").serialize(), function(data) {
-				if (data.post-url == data.page-url){
-					if(data.post-title && data.post-url) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported site: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.url+'">'+data.title+'</a>');
+			$.post('<?php echo $this->action('get_root_pages') ?>', $("#import-wordpress-form").serialize(), function(data) {
+				if (data.post_url == data.page_url){
+					if(data.post_title && data.post_url) {
+						$("#progress .parent-page").html('<?php echo t("Visit your imported site: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.page_url+'">'+data.page_title+'</a>');
 						$("#progress .parent-page").show("fast");
-					} else if(data.post-title) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported site: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.title+'</a>');
+					} else if(data.post_title) {
+						$("#progress .parent-page").html('<?php echo t("Visit your imported site: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.page_title+'</a>');
 						$("#progress .parent-page").show("fast");
 					}
 				} else {
-					if(data.post-title && data.post-url) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.url+'">'+data.title+'</a>');
-						$("#progress .parent-page").show("fast");
-					} else if(data.post-title) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.title+'</a>');
-						$("#progress .parent-page").show("fast");
+					if(data.post_title && data.post_url) {
+						$("#progress .parent-post").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.post_url+'">'+data.post_title+'</a>');
+						$("#progress .parent-post").show("fast");
+					} else if(data.post_title) {
+						$("#progress .parent-post").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.post_title+'</a>');
+						$("#progress .parent-post").show("fast");
 					}
-					if(data.post-title && data.post-url) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.url+'">'+data.title+'</a>');
+					if(data.page_title && data.page_url) {
+						$("#progress .parent-page").html('<?php echo t("Visit your imported pages: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>'+data.page_url+'">'+data.page_title+'</a>');
 						$("#progress .parent-page").show("fast");
-					} else if(data.post-title) {
-						$("#progress .parent-page").html('<?php echo t("Visit your imported posts: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.title+'</a>');
+					} else if(data.page_title) {
+						$("#progress .parent-page").html('<?php echo t("Visit your imported pages: ") ?><a href="<?php echo BASE_URL . DIR_REL ?>">'+data.page_title+'</a>');
 						$("#progress .parent-page").show("fast");
 					}
 				}
@@ -120,6 +120,8 @@
 			</div>
 			<p class="parent-page">
 			</p>
+			<p class="parent-post">
+			</p>
 			<ul class="titles" >
 			</ul>
 		</div>