1. Changed external links syntax in internal link syntax to repair link (dql-… file)

2. Replaced unicode "fi"-ligatur with "fi" to enable pdflatex-ing b/c utf8-inputenc-package does - for unknown reasons - not support unicode "fi"-ligatur

en/reference/annotations-reference.rst

@@ -525,7 +525,7 @@ Optional attributes:
 -  **mappedBy**: This option specifies the property name on the
    targetEntity that is the owning side of this relation. Its a
    required attribute for the inverse side of a relationship.
--  **inversedBy**: The inversedBy attribute designates the field in the
+-  **inversedBy**: The inversedBy attribute designates the field in the
    entity that is the inverse side of the relationship.
 -  **cascade**: Cascade Option
 -  **fetch**: One of LAZY or EAGER
@@ -599,7 +599,7 @@ Optional attributes:
 -  **orphanRemoval**: Boolean that specifies if orphans, inverse
    OneToOne entities that are not connected to any owning instance,
    should be removed by Doctrine. Defaults to false.
--  **inversedBy**: The inversedBy attribute designates the field in the
+-  **inversedBy**: The inversedBy attribute designates the field in the
    entity that is the inverse side of the relationship.
 
 Example:

en/reference/dql-doctrine-query-language.rst

@@ -1119,7 +1119,7 @@ There are situations when a query you want to execute returns a
 very large result-set that needs to be processed. All the
 previously described hydration modes completely load a result-set
 into memory which might not be feasible with large result sets. See
-the `Batch Processing <batch-processing>`_ section on details how
+the :doc:`Batch Processing <batch-processing>` section on details how
 to iterate large result sets.
 
 Functions