New unicity policies for feeds with bad GUIDs (#4487)

New set of unicity criteria options.

New tolerance heuristic:

> `$invalidGuidsTolerance` (default 0.05) The maximum ratio (rounded) of invalid GUIDs to tolerate before degrading the unicity criteria.
> Example for 0.05 (5% rounded): tolerate 0 invalid GUIDs for up to 9 articles, 1 for 10, 2 for 30, 3 for 50, 4 for 70, 5 for 90, 6 for 110, etc.
> The default value of 5% rounded was chosen to allow 1 invalid GUID for feeds of 10 articles, which is a frequently observed amount of articles.

1 files changed, 4 insertions, 9 deletions

diff --git a/app/Models/Entry.php b/app/Models/Entry.php
index fe6702bcd..6e87fe5cd 100644
--- a/app/Models/Entry.php
+++ b/app/Models/Entry.php
@@ -456,7 +456,7 @@ HTML;
 	}
 
 	public function hash(): string {
-		if ($this->hash == '') {
+		if ($this->hash === '') {
 			//Do not include $this->date because it may be automatically generated when lacking
 			$this->hash = md5($this->link . $this->title . $this->authors(true) . $this->originalContent() . $this->tags(true));
 		}
@@ -481,16 +481,11 @@ HTML;
 			$this->date_added = $value;
 		}
 	}
+
 	public function _guid(string $value): void {
-		$value = trim($value);
-		if (empty($value)) {
-			$value = $this->link;
-			if (empty($value)) {
-				$value = $this->hash();
-			}
-		}
-		$this->guid = $value;
+		$this->guid = trim($value);
 	}
+
 	public function _title(string $value): void {
 		$this->hash = '';
 		$this->title = trim($value);