fix one warning in specific mysql configs; welcome 2022

gggeek · Feb 17, 2022 · 159abf0 · 159abf0
1 parent c4f4a83
commit 159abf0
Show file tree

Hide file tree

Showing 24 changed files with 121 additions and 25 deletions.
diff --git a/bin/php/checkattributes.php b/bin/php/checkattributes.php
@@ -3,7 +3,7 @@
  * A CLI script which checks problems with all object attributes of a given datatype in current database
  *
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/bin/php/checkschema.php b/bin/php/checkschema.php
@@ -3,7 +3,7 @@
  * A CLI script which checks problems with data in the current schema
  *
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/bin/php/checkstorage.php b/bin/php/checkstorage.php
@@ -3,7 +3,7 @@
  * A CLI script which checks for orphan storage files
  *
  * @author G. Giunta
- * @copyright (C) G. Giunta 2016-2021
+ * @copyright (C) G. Giunta 2016-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/bin/php/generatedefsfrompersistentobjects.php b/bin/php/generatedefsfrompersistentobjects.php
@@ -3,7 +3,7 @@
  * A CLI script which checks all persistent object defs for FKs and generates a file with definitions
  *
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezbinaryfilechecker.php b/classes/datatypecheckers/ezdbiezbinaryfilechecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezfloatchecker.php b/classes/datatypecheckers/ezdbiezfloatchecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezimagechecker.php b/classes/datatypecheckers/ezdbiezimagechecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezintegerchecker.php b/classes/datatypecheckers/ezdbiezintegerchecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezmediachecker.php b/classes/datatypecheckers/ezdbiezmediachecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2016-2021
+ * @copyright (C) G. Giunta 2016-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezstringchecker.php b/classes/datatypecheckers/ezdbiezstringchecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbiezuserchecker.php b/classes/datatypecheckers/ezdbiezuserchecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/datatypecheckers/ezdbinullabletypechecker.php b/classes/datatypecheckers/ezdbinullabletypechecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/ezdbibasechecker.php b/classes/ezdbibasechecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2016-2021
+ * @copyright (C) G. Giunta 2016-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/ezdbidatatypechecker.php b/classes/ezdbidatatypechecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/ezdbireportgenerator.php b/classes/ezdbireportgenerator.php
@@ -3,7 +3,7 @@
  * Formats results in various modes
  *
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/ezdbischemachecker.php b/classes/ezdbischemachecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 
@@ -229,7 +229,7 @@ public function checkFKDefinition( $childTable, $childCol, $parentTable, $parent
             if ($t1['type'] != $t2['type']) {
                 $diffs[] = "Column type mismatch: {$t1['type']} vs. {$t2['type']}";
             } else {
-                if ($t1['length'] != $t2['length']) {
+                if (isset($t1['length']) && isset($t2['length']) && $t1['length'] != $t2['length']) {
                     $diffs[] = "Column length mismatch: {$t1['length']} vs. {$t2['length']}";
                 }
             }

diff --git a/classes/ezdbischemachecks.php b/classes/ezdbischemachecks.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/classes/ezdbistoragecker.php b/classes/ezdbistoragecker.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/doc/changelogs/changelog-0.26.0-to-0.26.1 b/doc/changelogs/changelog-0.26.0-to-0.26.1
@@ -0,0 +1,3 @@
+changelog from version 0.26.0 to 0.26.1
+
+. fixed: a warning when the db is set up in a way such tahtw e fail to get column definition length
diff --git a/doc/todo_ezimagefile b/doc/todo_ezimagefile
@@ -0,0 +1,93 @@
+NB: SOME/ALL OF THE STUFF BELOW FAILS BECAUSE PK FOR ezcoa IS ID,VERSION!
+
+a) check rows in ezcoa which do have a value for original file but miss a row in ezimagefile - matching based on attribute id
+  - found 0 when testing kc prod data
+  - remember that ezcoa<->ezimagefile is a many-to-one relationship, as ezimagefile misses version
+  - somewhat useful, but not sufficient: what if the matched ezimagefile row does not match any of the file paths in the ezcoa row?
+
+select count(*)
+from ezcontentobject_attribute a
+left join (select distinct(contentobject_attribute_id) as contentobject_attribute_id from ezimagefile) i on i.contentobject_attribute_id = a.id
+where data_type_string = 'ezimage'
+and (ExtractValue(data_text, '/ezimage/@filename') is not null and ExtractValue(data_text, '/ezimage/@filename') != '')
+and i.contentobject_attribute_id is null
+;
+
+b) check rows in ezcoa which miss a value for original file but do have rows in ezimagefile
+   - some of these do have instead '<original attribute_id="172" attribute_version="179"'
+   - the others hve NULL for ezdata_text value
+   - _but_, the rows in ezimagefile are probably referring to _other versions_ of the same attribute!
+   end result is: useless query...
+
+select count(*)
+from ezcontentobject_attribute a
+left join (select distinct(contentobject_attribute_id) as contentobject_attribute_id from ezimagefile) i on i.contentobject_attribute_id = a.id
+where data_type_string = 'ezimage'
+and (ExtractValue(data_text, '/ezimage/@filename') is null or ExtractValue(data_text, '/ezimage/@filename') = '')
+and i.contentobject_attribute_id is not null
+;
+
+c) check any ezimagefile which is not in the xml text in ezcoa.data_text (to retest...)
+
+- take care! in some cases '/ezimage/@filename' does contain colons and/or spaces,
+  eg: 'var/ovum_site/storage/images/24870301-1-eng-GB/C:\Users\lambertp\AppData\Local\Temp\F2 UK operators price comparison.jpg'
+  the fact is: there are few invalid chars in linux filenames... we moved to using | as path separator
+- take care 2: ExtractValue does not decode xml entities: we should replace &amp; to & => OK; to do: also replace &quot;
+- take care 3: we might be having problems when the match expression contains '\' chars... => OK: escaped them to \\
+- take care 4: we can not replace all spaces in filenames with the path !!! it will fail for files having spaces... => OK, we now explode using alias names, which should not contain spaces
+- take care 5: we should escape chars % and _ found in filepath when we use it for LIKE matching
+
+start with:
+
+select i.contentobject_attribute_id, group_concat(i.filepath)
+from ezimagefile i
+left join (
+  select
+    id,
+    concat(
+      '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@filename'), '&amp;', '&'),
+      '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@basename'), '&amp;', '&'), '_',
+      replace(
+        -- list of aliases, space separated
+        replace(ExtractValue(data_text, '/ezimage/alias/@name'), '&amp;', '&'),
+        ' ',
+        concat('.', ExtractValue(data_text, '/ezimage/@suffix'), '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@basename'), '&amp;', '&') , '_')
+      ),
+      '.', ExtractValue(data_text, '/ezimage/@suffix'), '|'
+    ) as all_filepaths
+  from ezcontentobject_attribute
+  where data_type_string = 'ezimage'
+    and ExtractValue(data_text, '/ezimage/@filename') != '' and ExtractValue(data_text, '/ezimage/@filename') is not null
+    -- and id = 27610921
+) a on i.contentobject_attribute_id = a.id and a.all_filepaths like concat('%|',replace(i.filepath, '\\', '\\\\'),'|%')
+where a.id is null
+group by i.contentobject_attribute_id
+
+d) check any ezcoa row of type ezimage which does have a value for original file but for which none of the files in the data_text field matches a row in ezimagefile
+
+- tested ok on kc prod data (0 results)
+- but some fixes still to be applied - see query above...
+- todo: add a test which would split the paths found in ecoa and check each of those separately
+
+select *
+from (
+  select
+    id,
+    concat(
+      '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@filename'), '&amp;', '&'),
+      '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@basename'), '&amp;', '&'), '_',
+      replace(
+        -- list of aliases, space separated
+        replace(ExtractValue(data_text, '/ezimage/alias/@name'), '&amp;', '&'),
+        ' ',
+        concat('.', ExtractValue(data_text, '/ezimage/@suffix'), '|', ExtractValue(data_text, '/ezimage/@dirpath'), '/', replace(ExtractValue(data_text, '/ezimage/@basename'), '&amp;', '&') , '_')
+      ),
+      '.', ExtractValue(data_text, '/ezimage/@suffix'), '|'
+    ) as all_filepaths
+  from ezcontentobject_attribute
+  where data_type_string = 'ezimage'
+    and ExtractValue(data_text, '/ezimage/@filename') != '' and ExtractValue(data_text, '/ezimage/@filename') is not null
+) a
+left join ezimagefile i on i.contentobject_attribute_id = a.id and a.all_filepaths like concat('%|',replace(i.filepath, '\\', '\\\\'),'|%')
+where i.id is null
+;
diff --git a/extension.xml b/extension.xml
@@ -2,8 +2,8 @@
 <software>
     <metadata>
         <name>eZ DB Integrity extension</name>
-        <version>0.26.0</version>
-        <copyright>Copyright (C) 2014-2021 Gaetano Giunta</copyright>
+        <version>0.26.1</version>
+        <copyright>Copyright (C) 2014-2022 Gaetano Giunta</copyright>
         <license>GNU General Public License v2.0</license>
     </metadata>
 

diff --git a/ezinfo.php b/ezinfo.php
@@ -6,8 +6,8 @@ static function info()
     {
         return array(
             'Name' => "<a href=\"https://github.com/gggeek/ezdbintegrity\">eZ DB Integrity extension</a>",
-            'Version' => "0.26.0",
-            'Copyright' => "Copyright (C) 2014-2021 Gaetano Giunta",
+            'Version' => "0.21.0",
+            'Copyright' => "Copyright (C) 2014-2022 Gaetano Giunta",
             'License' => "GNU General Public License v2.0"
         );
     }

diff --git a/interfaces/ezdbidatatypecheckerinterface.php b/interfaces/ezdbidatatypecheckerinterface.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
 

diff --git a/interfaces/ezdbischemafileformatinterface.php b/interfaces/ezdbischemafileformatinterface.php
@@ -1,7 +1,7 @@
 <?php
 /**
  * @author G. Giunta
- * @copyright (C) G. Giunta 2014-2021
+ * @copyright (C) G. Giunta 2014-2022
  * @license Licensed under GNU General Public License v2.0. See file license.txt
  */
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,3 @@
		changelog from version 0.26.0 to 0.26.1

		. fixed: a warning when the db is set up in a way such tahtw e fail to get column definition length