0.19.1

Added

• DOI on Zenodo to cite the Software itself

0.19.0

Added

• Utility function node_names_from_match for getting the node identifiers from the matches
• Tutorial for Python, Java and Rust on how to embedd graphANNIS in other programs
• Citation File Format (https://citation-file-format.github.io/) meta-data

Changed

• Renamed the "PartOfSubcorpus" component type to more general "PartOf"
• relANNIS import now takes the sub-corpus structure into account
• Quirks mode now also emulates the component search normalization behavior.
  Search nodes that where part of multiple dominance/pointing relation joins where duplicated and joined with the identity operator to work around the issue that nodes of different components could not be joined in relANNIS. This leads to additional output nodes in the find(...) query. See also the original JavaDoc for an explanation.
• The error_chain crate is no longer used for error reporting, instead a custom Error representation is used

Fixed

• "NULL" annotation namespaces where imported as "NULL" in relANNIS import
• Result ordering for "find(...)" function was not correct if token helper components where not loaded

0.18.1

Changed

• fixed issue where corpora which contain only tokens could not be queried for a subgraph with context

0.18.0

Added

• Release process is now using the cargo-release script

Changed

• Separate the update events in smaller chunks for relANNIS import to save memory

0.17.2

Fixed Bugs

• #70 get_all_components() returns all components with matching name if none with the same type exist

0.17.1

Fixed Bugs

• #69 relANNIS-Import: Subgraph query does not work if there is no coverage component.

0.17.0

Enhancements

• #68 Use applyUpdate() API to import legacy relANNIS files
• #67 Document the data model of graphANNIS
• #66 Automatic creation of inherited coverage edges
• #65 Add a new adjecency list based graph storage for dense components.

0.16.0

Fixed Bugs

• #62 Warn about missing coverage edges instead of failing the whole import

Enhancements

• #61 Implement the equal and not equal value operators

0.15.0

Fixed Bugs

• #59 Nodes are not deleted from graph storages via the "applyUpdate" API
• #55 Subgraph query does not work if there is no coverage component.
• #54 Check all existing matches when checking reflexivity

Enhancements

• #58 Implement ^ (near) operator
• #57 Implement ":arity" (number of outgoing edges) unary operator
• #52 Use CSV files for query set definition

0.14.2

Fixed Bugs

• #50 Non-reflexive operator join on "any token search" leads to non-empty result
• #48 Importing PCC 2.1 corpus hangs at "calculating statistics for component LeftToken/annis/"
• #46 Filter not applied for negated annotation search