Skip to content

Latest commit

 

History

History
85 lines (52 loc) · 2.71 KB

CHANGELOG.md

File metadata and controls

85 lines (52 loc) · 2.71 KB
Raw

0.9.3

  • Deprecating HL7.Query and HL7.Message and associated modules
  • Docs updates for HL7

0.9.2

  • Bug fix for internal format (caused issues between HL7 and HL7.Message)

0.9.1

  • Accidentally included nimble_parsec as a hard dependency, removing it.
  • Added keep_prefix_segments option to chunk_by_leading_segment/2

0.9.0

  • Planning to move all HL7.Query functionality to HL7 module with a map-based API.
  • HL7 module contains new map-based functions (experimentally).

0.8.0

  • Deprecates HL7.Query functions that take String in favor of funcitons that take HL7.Path
    • get_parts/2 -> find_all/2
    • get_part/2 -> find_first/2
    • replace_parts/3 -> update/3
    • Hard coded grammars may use ~p for compile time checks of grammars
    • Grammars coming from the outside can be changed into strings using HL7.Path.new/1
  • Deprecates HL7.FieldGrammar.to_indicies/1 use HL7.Path.new/1 instead

0.7.3

  • Fixes typespecs on HL7.Query.do_get_part

0.7.2

  • Fixes warnings for deprecations reporting bad locations

0.7.1

  • Fixes error when identifiers are used with String values in HL7.Query functions.

0.7.0

  • Precompiles hard coded HL7.Query strings
  • Deprecates passing Strings as arguments to HL7.Query funcitons
  • Adds sigil to create paths ~g{PID-3.1}

0.6.6

  • Added optional options to HL7.Message.new to support string validation, binary copies, and latin1 encodings.

0.6.5

  • Using binary copy while parsing to minimize shared binary references

0.6.4

  • Performance increase with new tokenizer and hot path for default message delimiters -- up to 6x faster parsing now!

0.6.1

Added wildcard selections, such as HL7.Query.select(message, "OBR*") to select segment types with all trailing segments until the lead segment is encountered again. Also includes !NTE to match any segment but an NTE segment.

0.6.0

Changes -- (looser parsing rules, next version will add warnings field)

  • HL7.Message will successfully parse messages without an HL7 Version (== nil)
  • HL7.Message now contains a fragments field for invalid segment data (empty list normally)

0.5.0

Bug Fixes

  • Hl7.Query.get_part was not returning multiple repetitions
  • Roadmap: will add an "invalid query" struct similar to invalid message

0.4.0

Bug Fixes (that could change expectations, thus version bump)

  • Messages with certain invalid headers will produce InvalidMessage and InvalidHeader structs instead of crashing.
  • Messages that do not specify encoding characters will use the HL7 default.
  • Messages that do not include Trigger Events will now act as "correct" HL7
  • Roadmap: will include a warnings list for parsed messages such that systems can decide on an appropriate strictness level.